Human loss-of-function variants suggest that partial LRRK2 ... · across 141,456 individuals...

Human loss-of-function variants suggest that partial LRRK2 inhibition is a safe therapeuticstrategyforParkinson’sdiseaseNicolaWhiffin1,2,3*,IrinaM.Armean3,4*,AaronKleinman5*,JamieL.Marshall3,EricV.Minikel3,Julia K. Goodrich3,4, Nicholas M Quaife1,2, Joanne B Cole3,6,7,8, Qingbo Wang3,4,9, Konrad J.Karczewski3,4, Beryl B. Cummings3,4,10, Laurent Francioli3,4, Kristen Laricchia3,4, Anna Guan5,Babak Alipanahi5†, Peter Morrison5, Marco A.S. Baptista11, Kalpana M. Merchant11, GenomeAggregationDatabaseProductionTeam,GenomeAggregationDatabaseConsortium,JamesS.Ware1,2,3, Aki S. Havulinna12,13, Bozenna Iliadou14, Jung-Jin Lee15, Girish N. Nadkarni16,17, ColeWhiteman18, the 23andMe Research Team, Mark Daly3,4,19, Tõnu Esko3,20, ChristinaHultman14,17,RuthJ.F.Loos16,21,LiliMilani20,AarnoPalotie4,12,19,CarlosPato18,MichelePato18,Danish Saleheen15,22,23, Patrick F. Sullivan14,24, Jessica Alföldi3,4, Paul Cannon5**, Daniel G.MacArthur3,4**1National Heart & Lung Institute and MRC London Institute of Medical Sciences, Imperial CollegeLondon,London,UK2CardiovascularResearchCentre,RoyalBrompton&HarefieldHospitalsNHSTrust,London,UK3Program in Medical and Population Genetics, Broad Institute of MIT and Harvard, Cambridge, MA02142,USA4AnalyticandTranslationalGeneticsUnit,MassachusettsGeneralHospital,Boston,MA02114,USA523andMe,Inc.6PrograminMetabolism,BroadInstituteofMITandHarvard,Cambridge,MA02142,USA7CenterforGenomicMedicine,MassachusettsGeneralHospital,Boston,MA02114,USA8Divisionof Endocrinology andCenter forBasic andTranslationalObesityResearch,BostonChildren’sHospital,Boston,MA,USA9PrograminBioinformaticsandIntegrativeGenomics,HarvardMedicalSchool,Boston,MA02115,USA10PrograminBiologicalandBiomedicalSciences,HarvardMedicalSchool,Boston,MA,02115,USA11MichaelJ.FoxFoundation12InstituteforMolecularMedicineFinland(FIMM),HiLIFE,UniversityofHelsinki,Helsinki,Finland13National Institute for Health and Welfare, 00271, Helsinki, Finland. 14DepartmentofMedicalEpidemiologyandBiostatistics,KarolinskaInstitutet,Stockholm,Sweden15Department of Biostatistics and Epidemiology, Perelman School of Medicine at the University ofPennsylvania,Philadelphia,PA,USA16TheCharlesBronfman Institute forPersonalizedMedicine, IcahnSchoolofMedicineatMountSinai,NewYork,NY,USA17DepartmentofMedicine,IcahnSchoolofMedicineatMountSinai,NewYork,NY,USA18Department of Psychiatry and the Behavioral Sciences, State University of New York, Downstate Medical Center, New York, NY,USA 19Stanley Center for Psychiatric Research, Broad Institute of MIT and Harvard, Cambridge,Massachusetts02142,USA

.CC-BY 4.0 International licenseunder anot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available

The copyright holder for this preprint (which wasthis version posted September 18, 2019. ; https://doi.org/10.1101/561472doi: bioRxiv preprint

https://doi.org/10.1101/561472

http://creativecommons.org/licenses/by/4.0/

20EstonianGenomeCenter,InstituteofGenomics,UniversityofTartu,Tartu,Estonia21TheMindichChildHealthandDevelopment Institute, IcahnSchoolofMedicineatMountSinai,NewYork,NY,USA22DepartmentofMedicine,PerelmanSchoolofMedicineattheUniversityofPennsylvania,Philadelphia,PA,USA23CenterforNon-CommunicableDiseases,Karachi,Pakistan24DepartmentsofGeneticsandPsychiatry,UniversityofNorthCarolina,ChapelHill,NC,USA*Theseauthorscontributedequallytothiswork**Theseauthorscontributedequallytothiswork†Currentaffiliation:Google,Inc.Correspondence should be addressed to Daniel G MacArthur [email protected] or [email protected] genetic variants predicted to cause loss-of-function of protein-coding genes (pLoFvariants) provide natural in vivomodels of human gene inactivation, and can be valuableindicatorsofgenefunctionandthepotential toxicityof therapeutic inhibitors targetingthesegenes1,2.Gain-of-kinase-functionvariants inLRRK2areknowntosignificantly increasetheriskof Parkinson’s disease3,4, suggesting that inhibition of LRRK2 kinase activity is a promisingtherapeuticstrategy.Whilepreclinicalstudiesinmodelorganismshaveraisedsomeon-targettoxicity concerns5–8, the biological consequences of LRRK2 inhibition have not been well-characterized in humans. Here we systematically analyse pLoF variants in LRRK2 observedacross141,456individualssequencedintheGenomeAggregationDatabase(gnomAD)9,49,960exome sequenced individuals from the UK Biobank, and over 4 million participants in the23andMegenotypeddataset.Afterstringentvariantcuration,weidentify1,455individualswithhigh-confidence pLoF variants in LRRK2, 82.5% with experimental validation. We show thatheterozygous pLoF variants in LRRK2 reduce LRRK2 protein levels but are not stronglyassociatedwithreducedlifeexpectancy,orwithanyspecificphenotypeordiseasestate.Thesedata suggest that therapeutics thatpartiallydownregulate LRRK2 levelsor kinaseactivityareunlikely tohavemajoron-targetsafety liabilities.Our resultsdemonstrate thevalueof large-scalegenomicdatabasesandphenotypingofhumanLoFcarriers for targetvalidation indrugdiscovery.



https://doi.org/10.1101/561472


MaintextNovel therapeutic strategies are desperately needed in Parkinson’s disease (PD), one of themostcommonage-relatedneurologicaldiseases,whichaffectsabout1%ofpeopleovertheageof6010,11.Around30%offamilialand3-5%ofsporadicPDcaseshavebeenlinkedtoageneticcause12. LRRK2 missense variants account for a significant fraction of cases, including high-penetrance variants13, moderately penetrant variants such as G2019S14, and risk factorsidentified in genome-wide association studies15. Although the precise mechanism by whichLRRK2 variants mediate their pathogenicity remains unclear, a common feature isaugmentationofkinaseactivityassociatedwithdisease-relevantalterationsincellmodels3,16,17.Discovery of RabGTPases as substrates of LRRK218 has also highlighted the role of LRRK2 inregulation of the endolysosomal and vesicular trafficking pathways implicated in Parkinson’sdisease18,19.LRRK2kinaseactivity isalsoupregulatedmoregenerally inPDpatients (withandwithoutLRRK2variants)20.LRRK2hasthereforebecomeaprominentdrugtarget,withmultipleLRRK2kinaseinhibitorsandsuppressors21indevelopmentasdisease-modifyingtreatmentsforPD20,22,23.Despite thesepromising indications, thereareconcernsabout thepotential toxicityofLRRK2inhibitors.Thesemainlyarisefrompreclinicalstudies,wherehomozygousknockoutsofLRRK2inmice,andhighdosetoxicologystudiesofLRRK2kinaseinhibitorsinratsandprimates,haveshown abnormal phenotypes in lung, kidney and liver5–8.Whilemodel organisms have beeninvaluable in understanding the function of LRRK2, they also have important limitations asexemplifiedby the inconsistencies in thephenotypicconsequencesof reducedLRRK2activityseen among yeast, fruit flies,worms,mice, rats and non-human primates24. Complementarydatafromnaturalhumanknockoutsarecriticalforunderstandingbothgenefunctionandthepotentialconsequencesofalong-termreductioninLRRK2kinaseactivityinhumans.Large-scale human genetics is an increasingly powerful source of data for the discovery andvalidation of therapeutic targets in humans1. Predicted loss-of-function (pLoF) variants,predicted to largely or entirely abolish the function of affected alleles, are a particularlyinformativeclassofgeneticvariation.Suchvariantsarenaturalmodelsfor life-longorganism-wide inhibition of the target gene, and can provide information about both the efficacy andsafety of a candidate target2,25–28. However, pLoF variants tend to be rare in humanpopulations29,andarealsoenrichedforbothsequencingandannotationartefacts30.Assuch,leveragingpLoFvariation indrugtargetassessmenttypicallyrequiresvery largecollectionsofgenetically andphenotypically characterized individuals, combinedwith deep curationof thetargetgeneandcandidatevariants31.AlthoughpreviousstudiesofpLoFvariantsinLRRK2have



https://doi.org/10.1101/561472


found no association with PD risk32, no study has assessed the broader phenotypicconsequencesofsuchvariants.We identifiedLRRK2pLoFvariants,andassessedtheassociatedphenotypicchanges, in threelargecohortsofgeneticallycharacterizedindividuals.Firstly,weannotatedLRRK2pLoFvariantsintwolargesequencingcohorts:thegnomADv2.1.1dataset,whichcontains125,748exomesand15,708genomesfromunrelatedindividuals9,andthe46,062exome-sequencedunrelatedEuropeanindividualsfromtheUKBiobank33.Weidentified633individualsingnomADand258individualsintheUKBiobankwith150uniquecandidateLRRK2LoFvariants,acombinedcarrierfrequencyof0.48%.All variantswereobservedonly in theheterozygous state.Compared tothespectrumobservedacrossallgenes,LRRK2 isnotsignificantlydepletedforLoFvariants ingnomAD(loss-of-functionobserved/expectedupperboundfraction(LOEUF)9=0.64).To ensure that the 150 identified variants are indeed expected to cause LoF, we manuallycurated them to remove those of low-quality or with annotation errors (Figure 1a;SupplementaryTables1and2).Weremoved16variantsidentifiedeitheraslow-confidencebythe Loss-of-function Transcript Effect Estimator (LOFTEE; 6 variants in 409 individuals)9, ormanuallycuratedaslow-qualityorunlikelytocausetrueLoF(10variantsin129individuals;seemethods).OneadditionalindividualwasexcludedfromtheUKBiobankcohortastheycarriedbothanpLoFvariantandtheG2019Sriskallele.Thisanalysisresultsinafinaldatasetof255gnomADindividualsand97UKbiobankindividualswith 134 unique high-confidence pLoF variants (Figure 1a), an overall carrier frequency of0.19%; less than half the frequency estimated from uncurated variants, reaffirming theimportanceofthoroughcurationofcandidateLoFvariants31.Asubsetof25gnomADsampleswith 19 unique LRRK2 pLoF variants with DNA available were all successfully validated bySangersequencing(SupplementaryTable3).Secondly, we examined LRRK2 pLoF variants in over four million consented researchparticipants from the personal genetics company 23andMe, Inc. Participants have beengenotypedona varietyofplatforms, and imputedagainst a referencepanel comprising56.5millionvariantsfrom1000Genomesphase334,andUK10K35.EightputativeLRRK2LoFvariantswere identified in the 23andMe cohort (LOFTEE high-confidence). After manual curation allputativecarriersofeachvariantweresubmittedforvalidationbySangersequencing.Variantswith<5confirmedcarrierswereexcludedfromtheanalysis.Theresultingcohortconsistedof749 individuals, each of whom is a Sanger-confirmed carrier for one of three pLoF variants(Figure 1a; Supplementary Table 4). The high rate of Sanger confirmation for rs183902574(>98%)allowedconfidentadditionof354putativecarriersofrs183902574,fromexpansionof



https://doi.org/10.1101/561472


the 23andMe dataset, without Sanger confirmation. Analyses with and without thesegenotyped only carriers were not significantly different (Supplementary Table 6). Across thetwo most frequent pLoF alleles we observed an extremely small number (<5) of sequence-confirmedhomozygotes;however,giventheverysmallnumberofobservations,wecanmakeno robust inferenceexcept thathomozygous inactivationof LRRK2appears tobecompatiblewithlife.FortheremainderofthismanuscriptwefocusonheterozygouspLoFcarriers.

Figure 1:Annotation and curationof candidate LRRK2 pLoF variants. a) Flow chart showing thevariant filtering and curation of candidate LRRK2 LoF variants in the gnomAD, UK Biobank and23andMe cohorts. Of the 1,103 carriers identified in 23andMe, 749 were confirmed by Sangersequencingwiththeremainderuntested.b)TheancestrydistributionofLRRK2pLoFvariantcarriersingnomAD(AFR-African/AfricanAmerican;AMR-American/Latino;ASJ-AshkenaziJewish;EAS-EastAsian;FIN-Finnish;NFE-Non-FinnishEuropean;SAS-SouthAsian).pLoFvariantsseenmorethan 10 times appear in colour with remaining variants in grey. LRRK2 pLoF variants aremostlyindividually extremely rare (less than 1/10,000 carrier frequency), with the exception of twononsense variants almost exclusively restricted to the admixed American/Latino population(Cys1313TerandArg1725Ter),andtwolargelyNon-FinnishEuropean-specificvariants(Leu2063Ter



https://doi.org/10.1101/561472


and Arg772Ter). All variant protein descriptions are with respect to ENSP00000298910.7. c)SchematicoftheLRRK2genewithpLoFvariantsmarkedbyposition,withtheheightofthemarkercorrespondingtoallelecountingnomAD(greybars)andUKBiobank(bluebars).The51exonsareshown as rectangles coloured by protein domain, with the remaining exons in grey. The threevariantsgenotypedinthe23andMecohortareannotatedwiththeirsamplecountinblacktext.Thethreecombineddatasetsprovideatotalof1,455carrierindividualswith134uniqueLRRK2pLoF variants. These variants are found across all major continental populations (Figure 1b;SupplementaryFigure1)andshowneitheranyobviousclusteringalongthelengthoftheLRRK2protein,norspecificenrichmentordepletioninanyoftheknownannotatedproteindomains(Chi-squareP=0.22;Figure1c),consistentwithsignaturesoftrueLoF31.ToconfirmthatLRRK2pLoFvariants leadtoareduction inLRRK2protein levels,weanalysedtotal protein lysates from cell lineswith three unique pLoF variants by immunoblotting.Weobtained lymphoblastoid cell lines (LCLs) from two families with naturally occuringheterozygous LoF variants, and a third variant was CRISPR/Cas9 engineered into embryonicstem cells (Supplementary Figure 2), which were differentiated into cardiomyocytes. In allinstances,LRRK2protein levelswerevisiblyreducedcomparedtonon-carriercontrols (Figure2).TheseresultsagreewithapreviousstudywhichassessedthreeseparatepLoFvariants,andfoundsignificantreductionsinLRRK2proteinlevels32. Together,thesesixfunctionallyvalidatedvariants represent 82.5% of total pLoF carriers in this study (1,201/1,455). AlthoughheterozygouspLoFcarriershaveLRRK2proteinremaining,webelievethatthisstaterepresentsthe closest genetic model to the partial kinase inactivation expected to be generated bytherapeuticLRRK2inhibitors.



https://doi.org/10.1101/561472


Figure2:LRRK2pLoFheterozygoteshavereducedLRRK2proteincomparedtocellsharboringnoLoFvariants.a)ImmunoblotofLRRK2andloadingcontrolGAPDHonlymphoblastoidcelllinesfromfive individuals harboring no pLoF variants (LRRK2-WT) and three individuals harboring aheterozygous pLoF variant (Cys1313Ter - 12-40699748-T-A; Arg1483Ter - 12-40704362-C-T). b)Immunoblot of LRRK2, α-actinin (specific tomuscle), and GAPDH on three control lines and oneCRISPR/Cas9 engineered LRRK2 heterozygous line of cardiomyocytes differentiated from ESCs(Arg1693Ter - 12-40714897-C-T). All variant protein descriptions are with respect toENSP00000298910.7.Wenext sought to determinewhether lifelong loweringof LRRK2protein levels through LoFresults in an apparent reduction in lifespan.We foundno significant differencebetween theage distribution of LRRK2 pLoF variant carriers and non-carriers in either the gnomAD or23andMedatasets (two-sidedKolmogorov-SmirnovP=0.085and0.46respectively;Figure3a),suggestingnomajorimpactonlongevity,thoughwenotethatatcurrentsamplesizesweareonlypoweredtodetectastrongeffect(seeMethods).



https://doi.org/10.1101/561472


Figure 3: LRRK2 pLoF variants are not associatedwith reduced lifespan and are not associatedwithanyadversephenotypes.(a)TheagedistributionsofLRRK2pLoFcarriersarenotsignificantlydifferent from those of non-carriers in both gnomAD and 23andMe. (b) Manhattan plot ofphenome-wide association study results for carriers of three LRRK2 pLoF variants against non-carriersinthe23andMecohort.Eachpointrepresentsadistinctphenotypewiththesegroupedintorelated categories (delineated by alternating black and grey points). The dotted horizontal linerepresentsaBonferonicorrectedPvaluethresholdfor368tests.For a subset of studies within gnomAD, phenotype data is available from study or nationalbiobank questionnaires or from linked electronic health records (see online methods). Wemanuallyreviewedtheserecordsforall60ofthe255gnomADLRRK2pLoFcarriersforwhomdata were available, and recorded any phenotypes affecting the lung, liver, kidney,cardiovascularsystem,nervoussystem,immunity,andcancer(SupplementaryTable5).Wedidnot findanover-representationofanyparticularphenotypeorphenotypecategory inLRRK2pLoFcarriers.



https://doi.org/10.1101/561472


The23andMedatasetincludesself-reporteddataforthousandsofphenotypesacrossadiverserangeofcategories.Weperformedaphenome-wideassociationstudyonLRRK2pLoFcarrierscomparedtonon-carriersfor368health-relatedtraits(seemethods)andfoundnosignificantassociationbetweenany individualphenotypeandcarrierstatus(Figure3b). Inparticular,wefound no significant associations with any lung, liver or kidney phenotypes (SupplementaryTables6and7).The UK Biobank resource includes measurements for 30 blood serum and four urinebiomarkers.We found no difference in any of these biomarkers between pLoF carriers andcontrols (Supplementary Table 8; Supplementary Figure 3). In particular, there was nodifference between carriers and non-carriers for urine biomarkers transformed into clinicalmeasures of kidney function (Figure 4a; see onlinemethods) and no difference in six bloodbiomarkerscommonlyusedtoassessliverfunction(Figure4c).Wealsoobservednodifferenceinspirometrymeasurementsoflungfunction(Figure4b).We grouped self-reported disease diagnoses in UK Biobank individuals into categoriescorresponding to organ system and/or mechanism (Supplementary Table 9; see onlinemethods). We observed no enrichment for any of these phenotype groups in LRRK2 pLoFcarrierswhencomparedtonon-carriers(SupplementaryTable10).WealsominedICD10codesfromhospitaladmissionsanddeathrecordsforanyepisodesrelatingtolung,liverandkidneyphenotypes, removing any with a likely infectious or other external causes (SupplementaryTable 11; see onlinemethods), and identified six pLoF carriers with ICD10 codes relating totheseorgansystems(6.19%),comparedto4,536non-carriers(9.87%;SupplementaryTables12and13).



https://doi.org/10.1101/561472


Figure 4: LRRK2 pLoF carriers donot have impaired lung, liver or kidney function. For all plots,pLoF carriers are shown in teal andnon-carriers in grey. Themean and1 standarddeviation arerepresentedbytheblackcircleandline.(a)UrinebiomarkersAlbuminandCreatininetransformedinto twoclinicalmarkersofkidney function (seeonlinemethods).NopLoFcarriers showsignsofseverely impaired kidney function. (b) Z-scores of age, sex and height corrected spirometrymeasures of lung function36. (c) Blood serum biomarkers of liver function. The plots for alkalinephosphatase,alanineaminotransferase,aspartateaminotransferase,bilirubinandcreatininehavebeen top truncated removing 47, 29, 92, 8 and 27 non-carriers respectively. The violin plots andsummarystatisticswerecalculatedonthefulldataset.AllpLoFcarriersarewithineachplotarea.Our results indicate that approximately one in every 500humans is heterozygous for a pLoFvariantinLRRK2,resultinginasystemiclifelongdecreaseinLRRK2proteinlevels,andthatthispartialinhibitionhasnodiscernibleeffectonsurvivalorhealth.Theseresultssuggestthatdrugsthat reduce LRRK2 kinase activity are unlikely to show toxicity due to on-target effects. Thisobservation is particularly important given that individuals at risk of PD will need to taketherapeutics chronically, likely soon after clinical diagnosis and extending for decades.Whileinitialphase I studiesof the safetyof LRRK2kinase inhibitorshave shownapromising safetyprofile23,37, these are not yet able to address the question of long-term, on-targetpharmacology-relatedsafetyprofiles.



https://doi.org/10.1101/561472


The rarity of pLoF variants in LRRK2, combined with the relatively low prevalence of PD,preventsusfromdirectlyassessingwhetherLRRK2 inhibitionmayreducethe incidenceofPDwithcurrentsamplesizes(SupplementaryTable6).Futurecohortswithmuchlargernumbersofsequencedandphenotypedindividuals(likelyinthemillionsofsamples)willberequiredtoanswer this question.As such, our study focuses entirely on thequestionofwhether partialgenetic LRRK2 inactivation has broader phenotypic consequences that might correspond toadverseeffectsofchronicinhibitionofLRRK2inhibitors.We acknowledge multiple limitations to the work described here. Firstly, we relied onheterogeneous phenotype data mostly derived from self-reported questionnaires. Both23andMe and gnomAD record only age at recruitment, and participants are of a relativelyyoung age compared to the typical age of onset for PD. Our ascertainment of LRRK2 pLoFvariantsin23andMewasnecessarilyincomplete,duetotheavailabilityoftargetedgenotypingratherthansequencingdatainthiscohort;thismeansthatasubsetofthe23andMeindividualstreatedasnon-carrierscouldbecarriersofLRRK2pLoFvariantsnotgenotypedor imputed inthisdataset.Finally,thelow-frequencyofnaturallyoccuringpLoFvariantsinLRRK2resultsinarelativelysmallnumberofcarrierswhichcouldbeassessedforeachbiomarkerandphenotype,meaning that we are not well-powered to detect subtle or infrequent clinical phenotypesarising from LRRK2 haploinsufficiency. However, our study strongly suggests that any clinicalphenotypearising frompartial inactivationof LRRK2 is likely tobe substantiallymorebenignthanearly-onsetPD.Thisstudyprovidesan importantproofofprincipleforthevalueofvery largegeneticallyandphenotypically characterized cohorts, combined with thorough variant curation, in exploringthesafetyprofileofcandidatedrugtargets.Overthecomingyears,theavailabilityofcompleteexome or genome sequence data for hundreds of thousands of individuals who are deeplyphenotyped and/or available for genotype-based recontact studies, combined with deepcuration and experimental validation of candidate pLoF variants, will provide powerfulresourcesfortherapeutictargetvalidationaswellasbroaderstudiesofthebiologyofhumangenes.DataandcodeavailabilityThedatapresentedinthismanuscriptandthecodeusedtomakethefiguresareavailable inhttps://github.com/macarthur-lab/gnomad_lrrk2.



https://doi.org/10.1101/561472


AcknowledgmentsWe thank the research participants and employees of 23andMe, Inc., and the researchparticipants ingnomAD, formakingthisworkpossible.N.W. issupportedbyaRosetreesandStoneygateImperialCollegeResearchFellowship.E.V.M.issupportedbyNIHF31AI22592.KJKwas supported by NIGMS F32 GM115208. This work was supported by the Michael J. FoxFoundation for Parkinson’s Research grant 12868, NIDDK U54DK105566, NIGMSR01GM104371,WellcomeTrust [107469/Z/15/Z];Medical ResearchCouncil (UK);NIHRRoyalBrompton Biomedical Research Unit; NIHR Imperial Biomedical Research Centre. T.E issupportedbyEstonianResearchCouncilgrantPUT1660.L.M.issupportedbyEstonianResearchCouncil grant PRG184. This research has been conducted using the UK Biobank Resource(https://doi.org/10.1101/572347)underApplicationNumber42890.ConflictsofinterestA.K.,B.A.,A.G.,P.M.,P.C.,andmembersofthe23andMeResearchTeamarecurrentorformeremployeesof23andMe,Inc.,andholdstockorstockoptionsin23andMe.DGMisafounderwithequityinGoldfinchBio,andhasreceivedresearchsupportfromAbbVie,Astellas,Biogen,BioMarin,Eisai,Merck,Pfizer,andSanofi-Genzyme.EVMhasreceivedresearchsupportintheformofcharitablecontributionsfromCharlesRiverLaboratoriesandIonisPharmaceuticals,andhasconsultedforDeerfieldManagement.KJKownsstockinPersonalis.MJDisafounderofMazeTherapeutics.



https://doi.org/10.1101/561472


References

1. Nelson,M.R.etal.Thesupportofhumangeneticevidenceforapproveddrugindications.

Nat.Genet.47,856–860(2015).

2. Plenge,R.M.,Scolnick,E.M.&Altshuler,D.Validatingtherapeutictargetsthroughhuman

genetics.Nat.Rev.DrugDiscov.12,581–594(2013).

3. Greggio,E.etal.KinaseactivityisrequiredforthetoxiceffectsofmutantLRRK2/dardarin.

Neurobiol.Dis.23,329–341(2006).

4. West,A.B.etal.Parkinson’sdisease-associatedmutationsinleucine-richrepeatkinase2

augmentkinaseactivity.Proc.Natl.Acad.Sci.U.S.A.102,16842–16847(2005).

5. Andersen,M.A.etal.PFE-360-inducedLRRK2inhibitioninducesreversible,non-adverse

renalchangesinrats.Toxicology395,15–22(2018).

6. Fuji,R.N.etal.EffectofselectiveLRRK2kinaseinhibitiononnonhumanprimatelung.Sci.

Transl.Med.7,273ra15(2015).

7. Baptista,M.A.S.etal.Lossofleucine-richrepeatkinase2(LRRK2)inratsleadsto

progressiveabnormalphenotypesinperipheralorgans.PLoSOne8,e80705(2013).

8. Hinkle,K.M.etal.LRRK2knockoutmicehaveanintactdopaminergicsystembutdisplay

alterationsinexploratoryandmotorco-ordinationbehaviors.Mol.Neurodegener.7,25

(2012).

9. Karczewski,K.J.etal.Variationacross141,456humanexomesandgenomesrevealsthe

spectrumofloss-of-functionintoleranceacrosshumanprotein-codinggenes.bioRxiv

531210(2019).doi:10.1101/531210

10. deLau,L.M.L.&Breteler,M.M.B.EpidemiologyofParkinson’sdisease.LancetNeurol.5,



https://doi.org/10.1101/561472


525–535(2006).

11. Polymeropoulos,M.H.etal.MappingofageneforParkinson’sdiseasetochromosome

4q21-q23.Science274,1197–1199(1996).

12. Klein,C.&Westenberger,A.GeneticsofParkinson’sdisease.ColdSpringHarb.Perspect.

Med.2,a008888(2012).

13. Zimprich,A.etal.MutationsinLRRK2causeautosomal-dominantparkinsonismwith

pleomorphicpathology.Neuron44,601–607(2004).

14. Goldwurm,S.etal.EvaluationofLRRK2G2019Spenetrance:relevanceforgenetic

counselinginParkinsondisease.Neurology68,1141–1143(2007).

15. Do,C.B.etal.Web-basedgenome-wideassociationstudyidentifiestwonovellocianda

substantialgeneticcomponentforParkinson’sdisease.PLoSGenet.7,e1002141(2011).

16. MacLeod,D.etal.ThefamilialParkinsonismgeneLRRK2regulatesneuriteprocess

morphology.Neuron52,587–593(2006).

17. West,A.B.etal.Parkinson’sdisease-associatedmutationsinLRRK2linkenhancedGTP-

bindingandkinaseactivitiestoneuronaltoxicity.Hum.Mol.Genet.16,223–232(2007).

18. Steger,M.etal.PhosphoproteomicsrevealsthatParkinson’sdiseasekinaseLRRK2

regulatesasubsetofRabGTPases.Elife5,(2016).

19. Roosen,D.A.&Cookson,M.R.LRRK2attheinterfaceofautophagosomes,endosomes

andlysosomes.Mol.Neurodegener.11,73(2016).

20. DiMaio,R.etal.LRRK2activationinidiopathicParkinson’sdisease.Sci.Transl.Med.10,

(2018).

21. Zhao,H.T.etal.LRRK2AntisenseOligonucleotidesAmeliorateα-SynucleinInclusion



https://doi.org/10.1101/561472


FormationinaParkinson’sDiseaseMouseModel.Mol.Ther.NucleicAcids8,508–519

(2017).

22. Chen,Z.-C.etal.PhosphorylationofamyloidprecursorproteinbymutantLRRK2promotes

AICDactivityandneurotoxicityinParkinson’sdisease.Sci.Signal.10,(2017).

23. Chen,J.,Chen,Y.&Pu,J.Leucine-RichRepeatKinase2inParkinson’sDisease:Updated

fromPathogenesistoPotentialTherapeuticTarget.Eur.Neurol.79,256–265(2018).

24. Daniel,G.&Moore,D.J.ModelingLRRK2PathobiologyinParkinson’sDisease:FromYeast

toRodents.inBehavioralNeurobiologyofHuntington’sDiseaseandParkinson'sDisease

(eds.Nguyen,H.H.P.&Cenci,M.A.)331–368(SpringerBerlinHeidelberg,2015).

25. Cohen,J.C.,Boerwinkle,E.,Mosley,T.H.,Jr&Hobbs,H.H.SequencevariationsinPCSK9,

lowLDL,andprotectionagainstcoronaryheartdisease.N.Engl.J.Med.354,1264–1272

(2006).

26. TGandHDLWorkingGroupoftheExomeSequencingProject,NationalHeart,Lung,and

BloodInstituteetal.Loss-of-functionmutationsinAPOC3,triglycerides,andcoronary

disease.N.Engl.J.Med.371,22–31(2014).

27. MyocardialInfarctionGeneticsConsortiumInvestigatorsetal.Inactivatingmutationsin

NPC1L1andprotectionfromcoronaryheartdisease.N.Engl.J.Med.371,2072–2082

(2014).

28. Minikel,E.V.etal.Quantifyingpriondiseasepenetranceusinglargepopulationcontrol

cohorts.Sci.Transl.Med.8,322ra9(2016).

29. Lek,M.etal.Analysisofprotein-codinggeneticvariationin60,706humans.Nature536,

285–291(2016).



https://doi.org/10.1101/561472


30. MacArthur,D.G.etal.Asystematicsurveyofloss-of-functionvariantsinhumanprotein-

codinggenes.Science335,823–828(2012).

31. Minikel,E.V.etal.Evaluatingpotentialdrugtargetsthroughhumanloss-of-function

geneticvariation.bioRxiv530881(2019).doi:10.1101/530881

32. Blauwendraat,C.etal.FrequencyofLossofFunctionVariantsinLRRK2inParkinson

Disease.JAMANeurol.75,1416–1422(2018).

33. VanHout,C.V.etal.Wholeexomesequencingandcharacterizationofcodingvariationin

49,960individualsintheUKBiobank.bioRxiv572347(2019).doi:10.1101/572347

34. 1000GenomesProjectConsortiumetal.Aglobalreferenceforhumangeneticvariation.

Nature526,68–74(2015).

35. UK10KConsortiumetal.TheUK10Kprojectidentifiesrarevariantsinhealthanddisease.

Nature526,82–90(2015).

36. Gupta,R.P.&Strachan,D.P.Ventilatoryfunctionasapredictorofmortalityinlifelong

non-smokers:evidencefromlargeBritishcohortstudies.BMJOpen7,e015381(2017).

37. Troyer,M.Safety,TolerabilityandTargetEngagementDemonstratedinPhase1Studyof

LRRK2InhibitorDNL201inHealthyYoungandElderlyAdults.in

https://denalitherapeutics.com/uploads/documents/events/DNL201_MJFF_PDTx_Presenta

tion.pdf(MJFFParkinson’sDiseaseTherapeuticsConference,2018).

38. McLaren,W.etal.TheEnsemblVariantEffectPredictor.GenomeBiol.17,122(2016).

39. Cummings,B.B.etal.Transcriptexpression-awareannotationimprovesrarevariant

discoveryandinterpretation.bioRxiv554444(2019).doi:10.1101/554444

40. Pato,M.T.etal.Thegenomicpsychiatrycohort:partnersindiscovery.Am.J.Med.Genet.



https://doi.org/10.1101/561472


BNeuropsychiatr.Genet.162B,306–312(2013).

41. Saleheen,D.etal.Humanknockoutsandphenotypicanalysisinacohortwithahighrate

ofconsanguinity.Nature544,235–239(2017).

42. Ripke,S.etal.Genome-wideassociationanalysisidentifies13newrisklocifor

schizophrenia.Nat.Genet.45,1150–1159(2013).

43. BipolarDisorderandSchizophreniaWorkingGroupofthePsychiatricGenomics

Consortium.Electronicaddress:[email protected]&BipolarDisorderand

SchizophreniaWorkingGroupofthePsychiatricGenomicsConsortium.Genomic

DissectionofBipolarDisorderandSchizophrenia,Including28Subphenotypes.Cell173,

1705–1715.e16(2018).

44. Genovese,G.etal.Increasedburdenofultra-rareprotein-alteringvariantsamong4,877

individualswithschizophrenia.Nat.Neurosci.19,1433–1441(2016).

45. Borodulin,K.etal.CohortProfile:TheNationalFINRISKStudy.Int.J.Epidemiol.(2017).

doi:10.1093/ije/dyx239

46. Leitsalu,L.etal.CohortProfile:EstonianBiobankoftheEstonianGenomeCenter,

UniversityofTartu.Int.J.Epidemiol.44,1137–1147(2015).

47. Bellenguez,C.etal.Arobustclusteringalgorithmforidentifyingproblematicsamplesin

genome-wideassociationstudies.Bioinformatics28,134–135(2012).

48. Levey,A.S.&Stevens,L.A.EstimatingGFRusingtheCKDEpidemiologyCollaboration

(CKD-EPI)creatinineequation:moreaccurateGFRestimates,lowerCKDprevalence

estimates,andbetterriskpredictions.Americanjournalofkidneydiseases:theofficial

journaloftheNationalKidneyFoundation55,622–627(2010).



https://doi.org/10.1101/561472


49. Freidlin,B.,Zheng,G.,Li,Z.&Gastwirth,J.L.Trendtestsforcase-controlstudiesofgenetic

markers:power,samplesizeandrobustness.Hum.Hered.53,146–152(2002).

50. McKenna,A.etal.TheGenomeAnalysisToolkit:aMapReduceframeworkforanalyzing

next-generationDNAsequencingdata.GenomeRes.20,1297–1303(2010).

51. Hsu,P.D.etal.DNAtargetingspecificityofRNA-guidedCas9nucleases.Nat.Biotechnol.

31,827–832(2013).

52. Wang,Q.&Ui-Tei,K.ComputationalPredictionofCRISPR/Cas9TargetSitesReveals

PotentialOff-TargetRisksinHumanandMouse.inGenomeEditinginAnimals:Methods

andProtocols(ed.Hatada,I.)43–53(SpringerNewYork,2017).

53. Dale,R.K.,Pedersen,B.S.&Quinlan,A.R.Pybedtools:aflexiblePythonlibraryfor

manipulatinggenomicdatasetsandannotations.Bioinformatics27,3423–3424(2011).

54. Naito,Y.,Hino,K.,Bono,H.&Ui-Tei,K.CRISPRdirect:softwarefordesigningCRISPR/Cas

guideRNAwithreducedoff-targetsites.Bioinformatics31,1120–1123(2015).



https://doi.org/10.1101/561472


SupplementaryTablesandFiguresSupplementaryTable1:CuratedLRRK2LoFvariantsidentifiedingnomAD.AllpositionsarewithrespecttoGRCh37.Supplementary Table 2: Curated LRRK2 LoF variants identified in 46,062 unrelated Europeanindividuals in the UK Biobank. All positions are shown with respect to both GRCh37 andGRCh38.Supplementary Table 3: Details of variants and primers for Sanger validation of LRRK2 LoFcarriers.AllpositionsarewithrespecttoGRCh37.Supplementary Table 4. Details of LRRK2 variants identified in the 23andMe cohort. Allpositions are with respect to GRCh37. All variant protein descriptions are annotated onENSP00000298910.Supplementary Table 5. Summary of phenotype data from gnomAD LRRK2 LoF carriers. AllpositionsarewithrespecttoGRCh37.SupplementaryTable6:Associationstatisticsforphenotypesspecificto lung,kidneyand liver(alsoplotted inFigure3c).Forprivacyreasons,countsof less thanfivearerounded.Noneofthephenotypesassociatedwithcarrierstatus,evenwithoutcorrection(p<0.05).SupplementaryTable7:23andMephenotypeassociationstatisticsforallphenotypesplottedinFigure3b.Supplementary Table 8: Details of 30 blood serum and 4 urine biomarkers tested in the UKBiobankdataset.ShownarethecountofpLoFcarriersandnon-carriersaswellasthemeanandstandard-deviationwithineachcohort.P-valuescorrespondtoa2-sideWilcoxontest.Supplementary Table 9: Details of phenotype groupings for diseases coded within the field'2002Non-cancer illnesscode,self-reported'.ShownarethetreepositionswithinUKBiobankalongwithprimaryandsecondarygroupingsusedinthisstudy.SupplementaryTable10:Associationstatisticsforallphenotypegroupings intheUKBiobank.pLoFcarriersandnon-carrierswerecountedonlyoncecorrespondingtowhetherornottheyhadanyof theself-reportedphenotypeswithinagrouping.P-valuescorrespond toaFisher'sexacttest.



https://doi.org/10.1101/561472


SupplementaryTable11:ICD10codestakenfromUKBiobankfields'41270-Diagnoses-ICD10','40001 - Underlying (primary) cause of death: ICD10' and '40002 - Contributory (secondary)causesofdeath:ICD10'todetectlung,liverandkidneydiseaseepisodes.Supplementary Table 12: ICD10 codes relating to lung, liver and kidney disease reported inLRRK2pLoFcarriers.Supplementary Table 13: Details of ICD10 codes relating to lung, liver and kidney diseasereportedineachofthesixLRRK2pLoFcarrierswithatleastonereported.AllpositionsarewithrespecttoGRCh37.SupplementaryTable14:Detailsof30blood serumand4urinebiomarkers tested in theUKBiobankdataset.ShownarethecountofpLoFcarriersandG2019Sriskallelecarriersaswellasthemeanandstandard-deviationwithineachcohort.P-valuescorrespondtoa2-sideWilcoxontest.SupplementaryTable15:Thepower todetectanageeffectof thesamemagnitudeasAPOErs429358inthe23andMedataset.



https://doi.org/10.1101/561472


SupplementaryFigure1EthnicitydistributionofLRRK2LoFcarriersinthe23andMecohort.



https://doi.org/10.1101/561472


SupplementaryFigure2a)Sangersequenceof isogenichESCengineeredcolonyforheterozygousLRRK2 variant 10, clone 13 (GRCh37:12-40714897-C-T). The engineered cell line maintains b) anormalkaryotype,c)normalcolonymorphologyandexpressionofOCT4,andd)differentiatesintothecardiomyocytelineage.Thebrightfieldimageofcardiomyocyteswascapturedatday17ofthedifferentiationprotocol.



https://doi.org/10.1101/561472




https://doi.org/10.1101/561472


Supplementary Figure3:Comparisonof 30blood serumand4urinebiomarkersbetweenLRRK2pLoFcarriers(teal),G2019Sriskallelecarriers(blue)andcarriersofneither(None;grey).Themeanandstandarddeviation ineachcohortareshownbyblackcirclesand linesrespectively.Forsomebiomarkersplotshavebeentoptruncatedtoremoveoutlinesinthenon-carriercohort.ValuesforallpLoFandG2019Scarriersareshownwithineachplotarea.Ineachcase,summarystatisticsarecalculatedonthefulldataset.

SupplementaryFigure4 IGVvisualizationof the splicedonor variantGRCh37:12-40626187-T-C intheGTExLRRK2pLoFcarrierexomesequencingdataandlungtissueRNA-seqdatacomparedtoacontrolGTExlungRNA-seqsample.ThepLoFvariantisobservedonreadscontainingananchoringmissense variant, GRCh37:12-40626185-A-G (A green and G orange), and these reads arepresentingnormalsplicingasseeninthecontrolRNA-seqsample.



https://doi.org/10.1101/561472


SupplementaryFigure5:ProportionofLRRK2pLoFcarriers,G2019Sriskallelecarriersandnon-carriersintheUKBiobankthataremale(darkgrey)andfemale(lightgrey).

SupplementaryFigure6:AgedistributionofLRRK2pLoFcarriers,G2019Sriskallelecarriersandnon-carriersintheUKBiobank.



https://doi.org/10.1101/561472


OnlineMethodsgnomADvariantannotationandcurationThe gnomAD resource, including both sample and variant quality control (including sampleancestry assignment), is fully described in our companionpaper9. gnomADversion2.1.1wasusedforthisanalysis.PutativeLoFvariantsweredefinedasstopgained,frameshiftoressentialsplice site (splice donor or splice acceptor) as annotated by the Ensembl Variant EffectPredictor38.Variants were included if they were annotated as LoF on any of the three high-confidenceGENCODEannotatedprotein-codingtranscriptsthatareexpressedinthelung,liverorkidney.All variants also underwent transcript expression-aware annotation which evaluates thecumulative expression status of the transcripts harboring a variant in theGTEx dataset39. Allhigh-confidence variants were found in exons with high evidence of expression across allrelevanttissuesinGTEx.Inaddition,allwerehigh-confidencepLoFonthecanonicaltranscript,whichistheonlytranscripttoincludethekinasedomain.Variants were filtered out if they were flagged as low-confidence by the Loss-of-functionTranscript Effect Estimator (LOFTEE)9. For the remaining variants, a manual curation wasperformed,includinginspectionofvariantqualitymetrics,readdistributionandthepresenceofnearbyvariantsusingtheIntegrativegenomeviewer(IGV)andsplicesitepredictionalgorithmsusingAlamut.Asinglesplice-sitevariant(12-40626187-T-C),foundin77gnomADcarriers,wasidentifiedinanindividual with RNAseq data in the Genotype-Tissue Expression (GTEx) project. The RNA-seqreadsweremanuallyinspectedtolookforanyeffectonsplicing.Assessingthereaddistributionofalinkedheterozygousvariantinthisindividualshowedconvincinglythatthevarianthasnodiscernible effect on transcript splicing (Supplementary Figure 4). All available tissues wereassessedwith reads from lung tissue shown inSupplementaryFigure4. Thevariantwasalsoidentified in 8 UK Biobank carriers and in 23andMe and was similarly excluded from thesecohorts.Wehavecompliedwithallrelevantethicalregulations.ThisstudywasoverseenbytheBroadInstitute’sOfficeofResearchSubjectProtectionandthePartnersHumanResearchCommittee.Informedconsentwasobtainedfromallparticipants.SangervalidationofgnomADvariantcarriers



https://doi.org/10.1101/561472


Sanger validationwas performed on genomic DNA derived from peripheral blood under thefollowingPCRconditions:98°C2min;30cycles20sec98°C,20sec54°C,1min72°C;3min72°Cusing the Herculase II Fusion DNA polymerase (Agilent, 600679). PCR products (5ul) wereanalyzed on a 2% agarose gel and the remaining productwas purifiedwith theQiagen PCRPurificationkit.SequenceanalysiswasperformedwithbothPCRprimersatQuintarabio.DetailsofvariantsandPCRprimersusedforeachcanbefoundinSupplementaryTable3.gnomADphenotypecurationandcohortdescriptionsThebelowdescribedstudieswithLRRK2pLoFcarriershadavailablephenotypedata.Foreachstudy,allavailablerecordsweremanuallyreviewedtoidentifyanyreportsofhealthproblemswhichwerecategorised into the followingclasses: lung, liver,kidney, cardiovascular,nervoussystem,immuneandcancer.TheGenomicPsychiatryCohort(GPC)projectThe GPC cohort is a longitudinal resource with the aim of making population-based dataavailablethroughtheNationalInstituteofMentalHealth(NIMH).TheNIMHrepositorycontainswhole-genome sequencing data and detailed clinical and demographic data particularlyfocussed on schizophrenia and bipolar disorders. A large fraction of participants, 88%, haveconsented for re-contact40.Thescreeningquestionnaireconsistedof32yes/noquestionsaboutmentalhealth issuesand23 yes/no questions covering other medical problems including liver, digestive andcardiovascular problems. There were no specific questions relating to lung or kidneyphenotypesalthoughparticipantswereaskedtoansweryes/notohavinganyadditionalhealthproblems. Ifaparticipantansweredyes to thisquestion,wemarked theexistenceof lungorkidney disease as ‘unknown’. One sample was excluded due to conflicting questionnaireanswers.Theageofthe25LRRK2carriersrangedfrom19to67.Twocarriers,aged55and60,reportedhaving had liver problems with 4 participants over 60 years reporting no liver problems.ThePakistanRiskofMyocardialInfarctionStudy(PROMIS)The PROMIS cohort comprises 10,503 individuals characterized using a phenotypequestionnairewith>350 items coveringdemographic anddietary characteristics andover80



https://doi.org/10.1101/561472


blood biomarker measurements41. The predominant focus of the questionnaire was cardiacfunctionandphenotype.Whisttheparticipantswerespecificallyaskedtoreportsufferingfromasthma, no other lung, liver, kidney, nervous system or immune phenotypes were directlyassayedsotheseweremarkedas‘unknown’fortheseindividuals.The12LRRK2LoFcarriersinPROMISdidnotdifferintermsofage,sexandmyocardialinfarctionstatuswhencomparedtotheentirecohort.TheSwedishSchizophreniaandBipolarStudiesCases with schizophrenia or bipolar disorder were identified from Swedish nationalhospitalization registers42,43. Controlswere selected at random frompopulation registers. Allindividuals had whole exome sequencing data44. All available ICD codes from inpatienthospitalizationsandoutpatientspecialisttreatmentcontactswereprovidedforeachpatient.FINRISK:theNationalFINRISKstudyThe FINRISK survey has been carried out for 40 years since 1972 every five years usingindependent, randomand representativepopulation samples fromdifferentpartsof Finland.Forthiswork,weusedsequencingandhealthregisterdatafromFINRISKsurveys1992—200745.

Fullhealthrecords including ICD10codeswerereviewedbystudycoordinatorswhoprovideduswithyes/noanswersforeachofourphenotypeclasses.TheBioMebiobankatTheCharlesBronfmanInstituteforPersonalizedMedicineatMountSinaiTheMountSinaiBioMeBiobank,foundedinSeptember2007,isanongoing,broadly-consentedelectronic health record (EHR)-linked bio- and data-repository that enrolls participants non-selectively from theMount SinaiMedical Center patient population (New York City). BioMeparticipants represent a broad racial, ethnic and socioeconomic diversitywith a distinct andpopulation-specific disease burden, characteristic of the communities served byMount SinaiHospital. Currently comprised of over 47,000 participants, BioMe participants are of African(AA,24%),Hispanic/Latino(HL,35%),European(EA,32%ofwhom40%AshkenaziJewish)andother/mixedancestry.BioMe is linked to Mount Sinai’s system-wide Epic EHR, which captures a full spectrum ofbiomedical phenotypes, including clinical outcomes, covariate and exposure data from past,presentandfuturehealthcareencounters.Themediannumberofoutpatientencountersis21per participant, reflecting predominant enrollment of participants with common chronic



https://doi.org/10.1101/561472


conditions from primary care facilities. Clinical phenotype data has been meticulouslyharmonizedandvalidated.Genome-wide genotype data and whole exome sequencing data is available for >30,000participants. Inaddition,wholegenomesequencingdata isavailablefor>11,000participants.ThefullelectronicrecordsofthreeBioMeLRRK2pLoFcarrierswerereviewedbylocalcliniciansandwewereprovidedwithdetailedsummaries.EstonianBiobankoftheEstonianGenomeCenter,UniversityofTartu The Estonian Biobank cohort is composed of volunteers from the general Estonian residentadultpopulation46.Thecurrentnumberofparticipants -closeto165,000--represents15%,oftheEstonianadultpopulation,makingitideallysuitedtopopulation-basedstudies.Participantswere recruited throughout Estonia by medical personnel, and participants receive astandardized health examination, donate blood, and fill out a 16-module questionnaire onhealth-related topics such as lifestyle, diet and clinical diagnoses. A detailed phenotypesummary fromahealth surveyand linkeddata including ICD10 codes, clinical lab values andtreatment and medication information is annually updated through linkage with nationalelectronichealthdatabasesandregistries.UKBiobankvariantdetectionandcurationThe49,960exomesequenced individuals fromtheUKBiobankwererestrictedtoasubsetof46,062 unrelated individuals of European ancestry. Relatedness was determined using KINGkinship coefficient estimates from the genotype relatedness file with a cutoff of 0.0884 toincludepairsof individualswithgreater than3rd-degree relatedness. Europeanancestrywasdetermined by projecting individuals on to 1000 Genomes phase 334 principal componentanalysiscoordinatespace,followedbyAberrantRpackage47clusteringtoretainonlyindividualsfallingwithin1000GenomesprojectEURPC1andPC2limits(lambda=4.5).Wefurtherremovedindividualsthatself-reportedasnon-Europeanethnicity.WeidentifiedallindividualswithputativeLoFvariantsdetectedintheFEanalysispipelineusingGATK 3.0 for variant calling and filtering33. We did not use the SPB pipeline calls due toadvertisederrors in theRegeneronGeneticsCenterpipelineat the timewewereconductingthese analyses. Variants were included if they were annotated as LoF on any transcriptexpressedinthelung,liverorkidney.AswiththegnomADanalysis,variantswerefilteredoutifthey were flagged as low-confidence by LOFTEE, before manual curation of the remaining



https://doi.org/10.1101/561472


variants.Thiscurationincludedinspectionofvariantqualitymetrics,readdistributionandthepresenceofnearbyvariantsusingIGVandsplicesitepredictionalgorithmsusingAlamut.Inaddition,266individualsinthefullgenotypedcohortof488,288sampleswhowerecarriersof theG2019S risk allelewere identified.One individualwhowas a carrier for both a LRRK2pLoFvariantandG2019Swasexcludedfromallanalyses.CarriersofG2019Swerenotincludedinthe‘non-carrier’cohortinanyoftheanalyses.LRRK2pLoFcarriers,G2019Sriskallelecarriersandnon-carriersarewellmatchedforbothsex(SupplementaryFigure5)andage(SupplementaryFigure6).UKBiobankphenotypeanalysisBloodserumandurinebiomarkersThefirstrecordedvalueofallfieldsrelatingto‘Bloodbiochemistry’(fieldcodes30600-30890)and‘Urineassays’(fieldcodes30510-30535)wasextractedforall individuals.Thedistributionofvalues forallbiomarkerswasplotted (SupplementaryFigure2),anda two-sidedWilcoxontestwasusedtotestforadifferencebetweenLRRK2pLoFcarriersandnon-carriers.These data were also extracted for G2019S risk allele carriers and these individuals werecompared to pLoF carriers. There was no significant difference in any of the 34 biomarkersbetweenpLoFandG2019Scarriersafteraccountingformultipletesting(SupplementaryTable14).ClinicalmeasuresofkidneyfunctionAlbumintocreatinineratio(ACR)wascalculatedbydividingtheurinemicroalbuminvalue(field30500; mg/L) by the urine creatinine value (field 30510; umol/L) multiplied by a factor of0.0001131222. Glomerular filtration rate (eGFR) was calculated using the CKD EpidemiologyCollaboration (CKD-EPI) creatinine equation48. Normal range values for both ACR and eGFRwere taken from the National Kidney Foundation website(https://www.kidney.org/kidneydisease/).SpirometrymeasuresoflungfunctionToassess lungfunctionweusedGlobalLung Initiative(GLI)2012referenceequationZ-scoresstandardisedforage,sexandheight forForcedexpiratoryvolumein1-second(FEV1),Forced



https://doi.org/10.1101/561472


vital capacity (FVC) and FEV1/ FVC ratio measured using Spirometry. These calculations areavailableinfieldcodes20256,20257and20258andweredescribedpreviously36.GroupedphenotypeanalysisThelistofallcodingswithinthefield‘20002Non-cancerillnesscode,self-reported’,weretakenfrom the UK Biobank showcase (http://biobank.ctsu.ox.ac.uk/crystal/coding.cgi?id=6). Allselectablecodingsweregivenaprimarygroupingpertainingtothemainsystemrelatingtothatdisease. In rare instanceswheremore thanonegrouping couldbeassigned, the secondwasincludedasasecondarygrouping.Diseaseswithanautoimmunebasisweregivenasecondarygrouping to reflect a similar underlying mechanism. Due to the opposing effects of somerespiratorydiseases,whereappropriate,phenotypes in this categoryweregivena secondarygrouping of Airway, Interstitial or Pleural. Any codings reflecting symptoms, trauma/injury,benign cancer, mental health phenotypes or diseases arising as a result of infection wereexcluded.Allphenotypecodingsandassignedgroupingsare listed inSupplementaryTable9.Any coding within the field ‘20001 Cancer code, self-reported’ was assigned a grouping of‘Cancer’.To test foranassociationbetweenanyphenotypegroupandLRRK2 pLoF carrier statuseachindividual was counted once as either having self-reported any of the phenotypes within agroup,orhavingreportednone.AFisher’sexacttestwasusedtotestforanassociation.AnalysisofICD10codesAll ICD10codes relating todiseasesof the liver (K70-K77),diseasesof the respiratory systemnotspecific to theupper respiratory tract (J20-J22, J40-J47, J80-J99)orkidneydiseases (N00-N29)werecuratedtoexcludeanywithaprimaryinfectiousorexternalcause(SupplementaryTable11).Asthmawasexcludedfromallanalysestoavoidanyissuescausedbythedeliberateascertainmentoftheexomesequencedportionofthecohortbasedonasthmastatus.For each individual, we extracted all ICD10 codes from the fields ‘41270 Diagnoses - ICD10’(recordedfromepisodes inhospital), ‘40001Underlying (primary)causeofdeath: ICD10’and‘40002 Contributory (secondary) causes of death: ICD10’. The number of carriers and non-carrierswithanyICD10coderelatingtolung(5pLoFcarriers;2,378non-carriers),liver(0pLoFcarriers;652non-carriers)orkidneydisease(3pLoFcarriers;2,272non-carriers)werecounted.For J43 (Emphysema), J44 (Other chronic obstructive pulmonary disease) and J47(Bronchiectasis),ICD10codeswerenotcountediftheywerereportedalongsideexposuretoorhistoryoftobaccouse(Z77.22,P96.81,Z87.891,Z57.31,F17orZ72.0).



https://doi.org/10.1101/561472


23andMevariantannotationandcurationPutative LRRK2 LoF variants were defined as those classified as high-confidence by LOFTEE.Variantsweremanuallyassessedforcallrate,genotypingandimputationqualityandmanuallycuratedtoensuretheywereexpectedtocausetrueLoF.For each of the two genotyped LRRK2 pLoF, we determined carrier status by manuallyinspectingandcustomcallingtheprobeintensityplots.Fortheimputedvariants,carrierstatuswasdetermined fromtheMinimac-Imputeddosage.As thesecallingmethodsmightproducefalse positives, we confirmed the participants’ genotypes through Sanger sequencing.Individuals with discordant genotypes were excluded. This resulted in a cohort of 749individuals,eachofwhomisaSangersequence-confirmedcarrierforoneofthreepLoFvariants(SupplementaryTable4).Duringinitialselectionandsequencing,expansionofthedatabaseledtoinclusionofanumberofadditional individualsgenotypedforoneofthepLoFvariants,rs183902574.Weperformedcustom-calling on these individuals and found 354 deemed high-confidence carriers(Supplementary Table 4). As these individuals were not Sanger sequenced, all subsequentanalyseswereperformedbothincludingandexcludingtheseindividuals.Participants provided informed consent and participated in the research online, under aprotocol approved by the external AAHRPP-accredited IRB, Ethical & Independent ReviewServices(E&IReview).Testingthepowertodetectanageeffectin23andMeAs a positive control for the age analysis,we tested theAPOEAlzheimer’s disease risk allelers429358,whichhasaknowneffectonlifespan.Thiseffect ishighlysignificantinthisdataset(P=1.2x10-211).Giventhatthecarriercountforrs429358ismuchhigherthanforLRRK2pLoF,weassessedthepowerofthe23andMedatasettodetectanageeffectassociatedwithLRRK2pLoFvariantsthatisofthesameeffectsizeastheknowneffectoftheAPOEallelers429358bysamplingcarriersof this variant. We randomly selected N carriers of rs429358 from the 23andMe dataset,performedaKStestontheagedistributionofthosecarriersversus4,000,000non-carriers,andconsideredtheresultingp-value.Werepeatedthisprocess100times,andthencomputedtheproportion of these simulations with p < 0.05. This tells us our power to reject the null



https://doi.org/10.1101/561472


hypothesisthatAPOEdoesnothaveanaffectonageatalpha=0.05,ifwehadNcarriersinthedataset.WerepeatedthisfordifferentvaluesofN,between1,000and20,000(SupplementaryTable15).Associationtestinginthe23andMedatasetPhenotypeselectionThe 23andMe dataset includes self-reported phenotype data for thousands of phenotypesacross a diverse range of categories. These phenotypes have different sample sizes andprevalences,sothepowertodetectassociationsvarieswidely.Webeganwithacuratedsetof748 disease phenotypes. We then applied a liberal filter based on our power to detect anassociationwith carrier status.More specifically, assuming aMAF of 2E-05,we restricted tophenotypeswherewehadpower0.1todetectanassociationeffectwithOR>1.3(forbinarytraits)orbeta>0.2(forquantitativetraits)atalpha=0.0001significance.This leftuswith460binaryand14quantitativephenotypes.AssociationtestingFor thesubsetof368health-relatedphenotypes (i.e.excludingany related todiet,druguse,lifestyleandpersonality),wefirstrestrictedto individuals forwhomwehadphenotypicdata.Wecalculatedpairwiseidentitybydescent(IBD)overallindividualsusingamodifiedversionofIBD64,andtheniterativelyremovedindividualsuntilwewereleftwithasetofparticipants,notwoofwhomshared>700cMinIBD.Wethentestedtheassociationbetweenthephenotypeand carrier status, controlling for age, sex, genotyping platform, and the first ten geneticprinciplecomponents.Weusedlogisticregressionforbinaryphenotypes,andlinearregressionforquantitativephenotypes.To control for population structure we restricted our analyses to participants with > 97%European ancestry, but the results did not qualitatively change when we dropped thisrestriction. We also tested associations using only individuals whose carrier status wasconfirmedbySangersequencing,butthisalsodidnotresultinanymeaningfuldifference.ABonferroni correctedP-value threshold for368 independent testsof1.36x10-4wasused toassessstatisticalsignificance.Poweranalysis



https://doi.org/10.1101/561472


Foreachphenotype,wecomputethetheoreticaloddsratiowearepoweredtodetect(giveninSupplementaryTables6and7)as follows:Letmbetheproportionof individualsused in theassociation study of that phenotypewho are LRRK2 pLoF carriers, and let n0 and n1 be thenumberofcontrolsandcases,respectively.ForeachORintheinterval[1,10]atstepsof0.02,wecomputethepoweroftheCochran-Armitagetrendtesttodetectanassociationbetweenavariant withMAFm and odds ratio OR at alpha=0.05, with n0 controls and n1 cases49.WereportthesmallestORsuchthatthispoweris>=0.8.AnalysisofLRRK2proteinlevelsCellcultureAll cell lines tested negative for mycoplasma contamination on a monthly basis with theMycoAlert™Detection kit (Lonza, LT07-118) andMycoAlert™AssayControl Set (Lonza, LT07-518).Cellsweregrownat37°Cwith5%CO2.HumanembryonicstemcellcultureAll pluripotent stem cellswere approved byHarvard ESCRO protocol #E00052 and #E00067.Humanembryonicstemcells(hESCs)wereobtainedfromWiCellResearchInstitute(WA01,H1)under an MTA. Cell lines were authenticated by visual inspection of cell morphology withbrightfield microscopy, staining with anti-Oct4 antibody to determine maintenance ofpluripotency (SantaCruz, sc-5279, datanot shown), sent toWiCell Research Institute after 6monthsofpassagingorafter isogenic cell linegeneration forkaryotyping,and in somecasesPCA of RNA sequencing data to confirm clustering with other pluripotent stem cell lines.PluripotentstemcellswereplatedontohESCqualifiedmatrigel(VWR,BD354277)coated6wellplates, mTeSR1 media was changed daily (StemCell Technologies, 85850), and cells werepassagedevery5-7dayswith0.5mMEDTA.LymphoblastoidcellcultureLymphoblastoidcelllines(LCLs)wereobtainedfromCoriellBiorepository(GM18500,GM18501,GM18502,HG01345,HG01346)andapprovedbytheBroadInstituteOfficeofResearchSubjectProtectionprotocol#3639.Celllineswereauthenticatedbyvisualinspectionofcellmorphologywith brightfield microscopy and in some cases PCA of RNA sequencing data to confirmclustering with GTEx LCLs. LCL media was changed every other day with RPMI 1640 (LifeTechnologies),2mML-glutamine(LifeTechnologies),and15%FBS(Sigma).



https://doi.org/10.1101/561472


CardiomyocytedifferentiationCardiomyocyte differentiation of the control and engineered H1 hESC lines was performedaccording to the protocol by Lian et al., 2012. Briefly, 500,000 cells were plated on hESCqualified matrigel (VWR, BD354277), grown in mTeSR1 media for four days (StemCellTechnologies, 85850), and switched to RPMImedia (Life Technologies)with B27 supplement(LifeTechnologies),switchingtoB27withinsulinatday7,fortheremainderoftheprotocol.Onday0ofdifferentiation,12µMCHIR99021(Tocris)wasappliedfor24hours.Atday3,cultureswere treated with 5µM IWP2 (Tocris) for 24 hours. Bright field images and movies wereacquired at day 17 and cells were harvested for protein/RNA extraction at day 19.IsogeniccelllineengineeringThefollowingguideandHRtemplateweredeliveredintosinglecellH1hESCsvianucleofection(Lonza 4D-Nucleofector X unit) using theP3PrimaryCell Kit (V4XP-3024), pulse codeCA137,and pX459 (Addgene): AATAAGGCATTTCATATAGT andACAGGCCTGTGATAGAGCTTCCCCATTGTGAGAACTCTGAAATTATCATCTGACTATATGAAATGCCTTATTTTCCAATGGGATTTTGGTCAAGATTAA.Cells were allowed to recover from nucleofection in mTeSR supplemented with 10µM RockInhibitor (Y-27632, Tocris) overnight. The following three days the cells were treated with0.25µg/mlpuromycin(VWR)inmTeSR.CellswerethenculturedinmTeSRuntilcolonieswerereadytobesplit.Engineeredcellsweresplitintosinglecellsandplatedinmatrigelcoated96-wellplatesatadensityof0.5cellsperwell.Plateswerescreenedforcolonies8-10daysafterplatingandgrownuntilcolonieswerereadytobesplit.Colonieswerethensplitwith0.5mMEDTA into two identical 96-well plates, one for DNA extraction/PCR/sequencing and one forfreezingcells.Oncecolonieswerereadytobesplit,96-wellplateswerefrozeninmFreSR(StemCellTechnologies)andstoredat -80°CuntilHRpositivewellswere identified.HReditedcellswere then thawed and expanded for 4 generations, validated by Sanger sequencing,karyotyping,andOCT4stainingpriortoproceedingwithcardiomyocytedifferentiation.Off-targetanalysisofCRISPR/Cas9engineeringTo detect any potential off-target effects caused by CRISPR/Cas9 genome editing, whole-genomesequencing(WGS)wasconductedforboththeengineeredandcontrolcell lines.DNAextraction, Quality Control (QC) and 30x PCR-Free WGS were performed by the GenomicsPlatformattheBroadInstitute.AllPrepDNA/RNAextractionkitwasused,followingitsprotocol.Alignment, marking of duplicates, recalibration of quality scores and variant calling are allperformedusingGATKbestpractices50.



https://doi.org/10.1101/561472


Weidentified157,230variantsintheengineeredcelllinethatwerenotfoundinthecontrolcellline as candidate variants. For the guide RNA (gRNA) used, we defined potential off-targetregions as those with <4bps mismatch against the full 20bp gRNA sequence (334 regions),and/or <2bpmismatch against the seed (=PAMproximal) 12bpof the gRNA sequence (5,780regions),eachfollowedbytheNGGPAM.Welookedforanycandidatevariantthatfallintothepotentialoff-targetregion,resultingindetectionofonlyonevariant(chr8-65084564-A-AT)thatfallsontoaregionwith1mismatchagainsttheseed12bpofgRNAsequence(chr8:65084560-65084575). No variantswith <4bpmismatch against the full 20bp gRNA sequence or perfectmatch at the seed regionwere detected. Since amismatch at the seed region decreases thelikelihoodofoff-targetvariants51,52,andalsobecausethesinglevariantwedetectedisaknownvariant (rs1161563412) observed in the populationwithout apparent phenotypic association,we concludednomajor off-target effect exists at the level of violating themain steps of ourresearch. All the analysis for the detection of potential off-target were conducted usingpybedtools53andCRISPRdirect54software.WesternblotanalysisCell pellets were snap-frozen in liquid nitrogen and stored at -80°C. Cells were douncehomogenized in ice-cold radioimmunoprecipitation assay buffer (RIPA) (89901; ThermoScientific) containing protease inhibitors (HaltTM Protease Inhibitor, Thermo Scientific).Homogenateswererotatedat4°Cfor30min,followedbycentrifugationat15,000gfor20minat4°C.Equalamountsofprotein(50µg)wereelectrophoresedon4-20%SDS-PAGE(Biorad)andtransferred to nitrocellulose membranes. The following antibodies were used forimmunoblotting: LRRK2 (75-253, UC Davis/NIH NeuroMab Facility), α-actinin (A7811, Sigma),GAPDH (sc-25778, SantaCruz), anti-rabbit IgGHRP (7074,Cell Signaling), andanti-mouse IgGHRP (7076,Cell Signaling). Immunoblotsweredevelopedusingenhanced chemiluminescence(SuperSignal West Pico Chemiluminescent Substrate, Thermo Scientific) on an AmershamImager600.



https://doi.org/10.1101/561472


GROUPAUTHORS

GenomeAggregationDatabase Production Team: Jessica Alföldi1,2, IrinaM. Armean1,2,3, EricBanks4, Louis Bergelson4, Kristian Cibulskis4, Ryan L Collins1,5,6, Kristen M. Connolly7, MiguelCovarrubias4, Beryl Cummings1,2,8, Mark J. Daly1,2,9, Stacey Donnelly1, Yossi Farjoun4, StevenFerriera10, Laurent Francioli1,2, Stacey Gabriel10, Laura D. Gauthier4, Jeff Gentry4, NamrataGupta10,1, Thibault Jeandet4, Diane Kaplan4, Konrad J. Karczewski1,2, Kristen M. Laricchia1,2,Christopher Llanwarne4, Eric V. Minikel1, Ruchi Munshi4, BenjaminM Neale1,2, Sam Novod4,Anne H. O'Donnell-Luria1,11,12, Nikelle Petrillo4, Timothy Poterba9,2,1, David Roazen4, ValentinRuano-Rubio4, Andrea Saltzman1, Kaitlin E. Samocha13, Molly Schleicher1, Cotton Seed9,2,Matthew Solomonson1,2, Jose Soto4, Grace Tiao1,2, Kathleen Tibbetts4, Charlotte Tolonen4,ChristopherVittal9,2,GordonWade4,ArcturusWang9,2,1,QingboWang1,2,6,JamesSWare14,15,1,NicholasAWatts1,2,BenWeisburd4,NicolaWhiffin14,15,11.PrograminMedicalandPopulationGenetics,BroadInstituteofMITandHarvard,Cambridge,Massachusetts02142,USA2.AnalyticandTranslationalGeneticsUnit,MassachusettsGeneralHospital,Boston,Massachusetts02114,USA3. European Molecular Biology Laboratory, European Bioinformatics Institute, Wellcome Genome Campus, Hinxton,Cambridge,CB101SD,UnitedKingdom4.DataSciencesPlatform,BroadInstituteofMITandHarvard,Cambridge,Massachusetts02142,USA5.CenterforGenomicMedicine,MassachusettsGeneralHospital,Boston,MA02114,USA6.PrograminBioinformaticsandIntegrativeGenomics,HarvardMedicalSchool,Boston,MA02115,USA7.GenomicsPlatform,BroadInstituteofMITandHarvard,Cambridge,Massachusetts02142,USA8.PrograminBiologicalandBiomedicalSciences,HarvardMedicalSchool,Boston,MA,02115,USA9.StanleyCenterforPsychiatricResearch,BroadInstituteofMITandHarvard,Cambridge,Massachusetts02142,USA10.BroadGenomics,BroadInstituteofMITandHarvard,Cambridge,Massachusetts02142,USA11.DivisionofGeneticsandGenomics,BostonChildren'sHospital,Boston,Massachusetts02115,USA12.DepartmentofPediatrics,HarvardMedicalSchool,Boston,Massachusetts02115,USA13.WellcomeSangerInstitute,WellcomeGenomeCampus,Hinxton,CambridgeCB101SA,UK14.NationalHeart&LungInstituteandMRCLondonInstituteofMedicalSciences,ImperialCollegeLondon,LondonUK15.CardiovascularResearchCentre,RoyalBrompton&HarefieldHospitalsNHSTrust,LondonUK

GenomeAggregationDatabaseConsortium:CarlosAAguilarSalinas1,TariqAhmad2,ChristineM. Albert3,4, Diego Ardissino5, Gil Atzmon6,7, John Barnard8, Laurent Beaugerie9, Emelia J.Benjamin10,11,12, Michael Boehnke13, Lori L. Bonnycastle14, Erwin P. Bottinger15, Donald WBowden16,17,18, Matthew J Bown19,20, John C Chambers21,22,23, Juliana C. Chan24, DanielChasman3,25, JudyCho15,MinaK.Chung26,BruceCohen27,25,AdolfoCorrea28,DanaDabelea29,Mark J. Daly30,31,32, Dawood Darbar33, Ravindranath Duggirala34, Josée Dupuis35,36, Patrick T.Ellinor30,37,RobertoElosua38,39,40,JeanetteErdmann41,42,43,TõnuEsko30,44,MarttiFärkkilä45,JoseFlorez46, Andre Franke47, Gad Getz48,49,25, Benjamin Glaser50, Stephen J. Glatt51, DavidGoldstein52,53,ClicerioGonzalez54,LeifGroop55,56,ChristopherHaiman57,CraigHanis58,MatthewHarms59,60,MikkoHiltunen61,MattiM.Holi62,ChristinaM.Hultman63,64,MikkoKallela65,JaakkoKaprio56,66, Sekar Kathiresan67,68,25, Bong-Jo Kim69, Young Jin Kim69, George Kirov70, Jaspal



https://doi.org/10.1101/561472


Kooner23,22,71,SeppoKoskinen72,HarlanM.Krumholz73,SubraKugathasan74,SooHeonKwak75,Markku Laakso76,77, Terho Lehtimäki78, Ruth J.F. Loos15,79, Steven A. Lubitz30,37, Ronald C.W.Ma24,80,81, Daniel G. MacArthur31,30, Jaume Marrugat82,39, Kari M. Mattila78, StevenMcCarroll32,83, Mark I McCarthy84,85,86, Dermot McGovern87, Ruth McPherson88, James B.Meigs89,25,90, Olle Melander91, Andres Metspalu44, Benjamin M Neale30,31, Peter M Nilsson92,Michael C O'Donovan70, Dost Ongur27,25, Lorena Orozco93, Michael J Owen70, Colin N.A.Palmer94, Aarno Palotie56,32,31, Kyong Soo Park75,95, Carlos Pato96, Ann E. Pulver97, NazneenRahman98, AnneM. Remes99, John D. Rioux100,101, Samuli Ripatti56,66,102, DanM. Roden103,104,Danish Saleheen105,106,107, Veikko Salomaa108, Nilesh J. Samani19,20, Jeremiah Scharf30,32,67,Heribert Schunkert109,110, Moore B. Shoemaker111, Pamela Sklar*112,113,114, Hilkka Soininen115,HarrySokol9,TimSpector116,PatrickF.Sullivan63,117,JaanaSuvisaari108,EShyongTai118,119,120,YikYing Teo118,121,122, Tuomi Tiinamaija56,123,124, Ming Tsuang125,126, Dan Turner127, Teresa Tusie-Luna128,129, Erkki Vartiainen66, James S Ware130,131,30, Hugh Watkins132, Rinse K Weersma133,MaijaWessman123,56,JamesG.Wilson134,RamnikJ.Xavier135,1361.UnidaddeInvestigaciondeEnfermedadesMetabolicas.InstitutoNacionaldeCienciasMedicasyNutricion.MexicoCity2.PeninsulaCollegeofMedicineandDentistry,Exeter,UK3.DivisionofPreventiveMedicine,BrighamandWomen'sHospital,Boston,Massachusetts,USA.4.DivisionofCardiovascularMedicine,BrighamandWomen'sHospitalandHarvardMedicalSchool,Boston,Massachusetts,USA.5.DepartmentofCardiology,UniversityHospital,43100Parma,Italy6.DepartmentofBiology,FacultyofNaturalSciences,UniversityofHaifa,Haifa,Israel7.DepartmentsofMedicineandGenetics,AlbertEinsteinCollegeofMedicine,Bronx,NY,USA,104618.DepartmentofQuantitativeHealthSciences,LernerResearchInstitute,ClevelandClinic,Cleveland,OH44122,USA9.SorbonneUniversité,APHP,GastroenterologyDepartment,SaintAntoineHospital,Paris,France10.NHLBIandBostonUniversity'sFraminghamHeartStudy,Framingham,Massachusetts,USA.11.DepartmentofMedicine,BostonUniversitySchoolofMedicine,Boston,Massachusetts,USA.12.DepartmentofEpidemiology,BostonUniversitySchoolofPublicHealth,Boston,Massachusetts,USA.13.DepartmentofBiostatisticsandCenterforStatisticalGenetics,UniversityofMichigan,AnnArbor,Michigan4810914.NationalHumanGenomeResearchInstitute,NationalInstitutesofHealth,Bethesda,MD,USA15.TheCharlesBronfmanInstituteforPersonalizedMedicine,IcahnSchoolofMedicineatMountSinai,NewYork,NY16.DepartmentofBiochemistry,WakeForestSchoolofMedicine,Winston-Salem,NC,USA17.CenterforGenomicsandPersonalizedMedicineResearch,WakeForestSchoolofMedicine,Winston-Salem,NC,USA18.CenterforDiabetesResearch,WakeForestSchoolofMedicine,Winston-Salem,NC,USA19.DepartmentofCardiovascularSciences,UniversityofLeicester,Leicester,UK20.NIHRLeicesterBiomedicalResearchCentre,GlenfieldHospital,Leicester,UK21.DepartmentofEpidemiologyandBiostatistics,ImperialCollegeLondon,London,UK22.DepartmentofCardiology,EalingHospitalNHSTrust,Southall,UK23.ImperialCollegeHealthcareNHSTrust,ImperialCollegeLondon,London,UK24.DepartmentofMedicineandTherapeutics,TheChineseUniversityofHongKong,HongKong,China.25.DepartmentofMedicine,HarvardMedicalSchool,Boston,MA26.DepartmentsofCardiovascularMedicine,CellularandMolecularMedicine,MolecularCardiology,andQuantitativeHealthSciences,ClevelandClinic,Cleveland,Ohio,USA.27.McLeanHospital,Belmont,MA28.DepartmentofMedicine,UniversityofMississippiMedicalCenter,Jackson,Mississippi,USA29.DepartmentofEpidemiology,ColoradoSchoolofPublicHealth,Aurora,Colorado,USA.



https://doi.org/10.1101/561472


30.PrograminMedicalandPopulationGenetics,BroadInstituteofMITandHarvard,Cambridge,MA,USA31.AnalyticandTranslationalGeneticsUnit,MassachusettsGeneralHospital,Boston,Massachusetts02114,USA32.StanleyCenterforPsychiatricResearch,BroadInstituteofMITandHarvard,Cambridge,MA,USA33.DepartmentofMedicineandPharmacology,UniversityofIllinoisatChicago34.DepartmentofGenetics,TexasBiomedicalResearchInstitute,SanAntonio,TX,USA35.DepartmentofBiostatistics,BostonUniversitySchoolofPublicHealth,Boston,MA02118,USA36.NationalHeart,Lung,andBloodInstitute'sFraminghamHeartStudy,Framingham,MA01702,USA37.CardiacArrhythmiaServiceandCardiovascularResearchCenter,MassachusettsGeneralHospital,Boston,MA38.CardiovascularEpidemiologyandGenetics,HospitaldelMarMedicalResearchInstitute(IMIM).Barcelona,Catalonia,Spain39.CIBERCV,Barcelona,Catalonia,Spain40.DepartamentofMedicine,MedicalSchool,UniversityofVic-CentralUniversityofCatalonia.Vic,Catalonia,Spain41.InstituteforCardiogenetics,UniversityofLübeck,Lübeck,Germany42. 1. DZHK (German Research Centre for Cardiovascular Research), partner site Hamburg/Lübeck/Kiel, 23562 Lübeck,Germany43.UniversityHeartCenterLübeck,23562Lübeck,Germany44.EstonianGenomeCenter,InstituteofGenomics,UniversityofTartu,Tartu,Estonia45.HelsinkiUniversityandHelsinkiUniversityHospital,ClinicofGastroenterology,Helsinki,Finland.46.DiabetesUnitandCenterforGenomicMedicine,MassachusettsGeneralHospital;ProgramsinMetabolismandMedical&PopulationGenetics,BroadInstitute;DepartmentofMedicine,HarvardMedicalSchool47.InstituteofClinicalMolecularBiology(IKMB),Christian-Albrechts-UniversityofKiel,Kiel,Germany48.BioinformaticsProgram,MGHCancerCenterandDepartmentofPathology49.CancerGenomeComputationalAnalysis,BroadInstitute.50.EndocrinologyandMetabolismDepartment,Hadassah-HebrewUniversityMedicalCenter,Jerusalem,Israel51.DepartmentofPsychiatryandBehavioralSciences;SUNYUpstateMedicalUniversity52. Institute for GenomicMedicine, ColumbiaUniversityMedical Center, HammerHealth Sciences, 1408, 701West 168thStreet,NewYork,NewYork10032,USA.53. DepartmentofGenetics&Development,ColumbiaUniversityMedicalCenter,HammerHealthSciences,1602,701West168thStreet,NewYork,NewYork10032,USA.54.CentrodeInvestigacionenSaludPoblacional.InstitutoNacionaldeSaludPublicaMEXICO55.LundUniversity,Sweden56.InstituteforMolecularMedicineFinland(FIMM),HiLIFE,UniversityofHelsinki,Helsinki,Finland57.LundUniversityDiabetesCentre58.HumanGeneticsCenter,UniversityofTexasHealthScienceCenteratHouston,Houston,TX7703059.DepartmentofNeurology,ColumbiaUniversity60.InstituteofGenomicMedicine,ColumbiaUniversity61.InstituteofBiomedicine,UniversityofEasternFinland,Kuopio,Finland62.DepartmentofPsychiatry,PL320,HelsinkiUniversityCentralHospital,Lapinlahdentie,00180Helsinki,Finland63.DepartmentofMedicalEpidemiologyandBiostatistics,KarolinskaInstitutet,Stockholm,Sweden64.IcahnSchoolofMedicineatMountSinai,NewYork,NY,USA65.DepartmentofNeurology,HelsinkiUniversityCentralHospital,Helsinki,Finland.66.DepartmentofPublicHealth,FacultyofMedicine,UniversityofHelsinki,Finland67.CenterforGenomicMedicine,MassachusettsGeneralHospital,Boston,Massachusetts02114,USA68. CardiovascularDisease Initiative andProgram inMedical andPopulationGenetics, Broad InstituteofMITandHarvard,Cambridge,Massachusetts02142,USA69.CenterforGenomeScience,KoreaNationalInstituteofHealth,Chungcheongbuk-do,RepublicofKorea.70.MRCCentreforNeuropsychiatricGenetics&Genomics,CardiffUniversitySchoolofMedicine,HadynEllisBuilding,MaindyRoad,CardiffCF244HQ71.NationalHeartandLungInstitute,CardiovascularSciences,HammersmithCampus,ImperialCollegeLondon,London,UK.72.DepartmentofHealth,THL-NationalInstituteforHealthandWelfare,00271Helsinki,Finland.73. Section of Cardiovascular Medicine, Department of Internal Medicine, Yale School of Medicine, New Haven,Connecticut3CenterforOutcomesResearchandEvaluation,Yale-NewHavenHospital,NewHaven,Connecticut.



https://doi.org/10.1101/561472


74.DivisionofPediatricGastroenterology,EmoryUniversitySchoolofMedicine,Atlanta,Georgia,USA.75.DepartmentofInternalMedicine,SeoulNationalUniversityHospital,Seoul,RepublicofKorea76.TheUniversityofEasternFinland,InstituteofClinicalMedicine,Kuopio,Finland77.KuopioUniversityHospital,Kuopio,Finland78. Department of Clinical Chemistry, Fimlab Laboratories and Finnish Cardiovascular ResearchCenter-Tampere, Faculty ofMedicineandHealthTechnology,TampereUniversity,Finland79.TheMindichChildHealthandDevelopmentInstitute,IcahnSchoolofMedicineatMountSinai,NewYork,NY80.LiKaShingInstituteofHealthSciences,TheChineseUniversityofHongKong,HongKong,China.81.HongKongInstituteofDiabetesandObesity,TheChineseUniversityofHongKong,HongKong,China.82.CardiovascularResearchREGICORGroup,HospitaldelMarMedicalResearchInstitute(IMIM).Barcelona,Catalonia.83.DepartmentofGenetics,HarvardMedicalSchool,Boston,MA,USA84.OxfordCentreforDiabetes,EndocrinologyandMetabolism,UniversityofOxford,ChurchillHospital,OldRoad,Headington,Oxford,OX37LJUK85.WellcomeCentreforHumanGenetics,UniversityofOxford,RooseveltDrive,OxfordOX37BN,UK86. Oxford NIHR Biomedical Research Centre, Oxford University Hospitals NHS Foundation Trust, John Radcliffe Hospital,OxfordOX39DU,UK87. F Widjaja Foundation Inflammatory Bowel and Immunobiology Research Institute, Cedars-Sinai Medical Center, LosAngeles,CA,USA.88.AtherogenomicsLaboratory,UniversityofOttawaHeartInstitute,Ottawa,Canada89.DivisionofGeneralInternalMedicine,MassachusettsGeneralHospital,Boston,MA,0211490.PrograminPopulationandMedicalGenetics,BroadInstitute,Cambridge,MA91.DepartmentofClinicalSciences,UniversityHospitalMalmoClinicalResearchCenter,LundUniversity,Malmo,Sweden.92.LundUniversity,Dept.ClinicalSciences,SkaneUniversityHospital,Malmo,Sweden93.InstitutoNacionaldeMedicinaGenómica(INMEGEN),MexicoCity,14610,Mexico94.MedicalResearchInstitute,NinewellsHospitalandMedicalSchool,UniversityofDundee,Dundee,UK.95. Department of Molecular Medicine and Biopharmaceutical Sciences, Graduate School of Convergence Science andTechnology,SeoulNationalUniversity,Seoul,RepublicofKorea96.DepartmentofPsychiatry,KeckSchoolofMedicineattheUniversityofSouthernCalifornia,LosAngeles,California,USA.97.DepartmentofPsychiatryandBehavioralSciences,JohnsHopkinsUniversitySchoolofMedicine,Baltimore,Maryland,USA98.DivisionofGeneticsandEpidemiology,InstituteofCancerResearch,LondonSM25NG99. MedicalResearchCenter,OuluUniversityHospital,Oulu,FinlandandResearchUnitofClinicalNeuroscience,Neurology,UniversityofOulu,Oulu,Finland.100.ResearchCenter,MontrealHeartInstitute,Montreal,Quebec,Canada,H1T1C8101.DepartmentofMedicine,FacultyofMedicine,UniversitédeMontréal,Québec,Canada102.BroadInstituteofMITandHarvard,CambridgeMA,USA103.DepartmentofBiomedicalInformatics,VanderbiltUniversityMedicalCenter,Nashville,Tennessee,USA.104.DepartmentofMedicine,VanderbiltUniversityMedicalCenter,Nashville,Tennessee,USA.105. Department of Biostatistics and Epidemiology, Perelman School of Medicine at the University of Pennsylvania,Philadelphia,PA,USA106.DepartmentofMedicine,PerelmanSchoolofMedicineattheUniversityofPennsylvania,Philadelphia,PA,USA107.CenterforNon-CommunicableDiseases,Karachi,Pakistan108.NationalInstituteforHealthandWelfare,Helsinki,Finland109.DeutschesHerzzentrumMünchen,Germany110.TechnischeUniversitätMünchen111.DivisionofCardiovascularMedicine,NashvilleVAMedicalCenterandVanderbiltUniversity,SchoolofMedicine,Nashville,TN37232-8802,USA.112.DepartmentofPsychiatry,IcahnSchoolofMedicineatMountSinai,NewYork,NY,USA113.DepartmentofGeneticsandGenomicSciences,IcahnSchoolofMedicineatMountSinai,NewYork,NY,USA114.InstituteforGenomicsandMultiscaleBiology,IcahnSchoolofMedicineatMountSinai,NewYork,NY,USA115.InstituteofClinicalMedicine,neurology,UniversityofEasternFinad,Kuopio,Finland116.DepartmentofTwinResearchandGeneticEpidemiology,King'sCollegeLondon,LondonUK



https://doi.org/10.1101/561472


117.DepartmentsofGeneticsandPsychiatry,UniversityofNorthCarolina,ChapelHill,NC,USA118.SawSweeHockSchoolofPublicHealth,NationalUniversityofSingapore,NationalUniversityHealthSystem,Singapore119.DepartmentofMedicine,YongLooLinSchoolofMedicine,NationalUniversityofSingapore,Singapore120.Duke-NUSGraduateMedicalSchool,Singapore121.LifeSciencesInstitute,NationalUniversityofSingapore,Singapore.122.DepartmentofStatisticsandAppliedProbability,NationalUniversityofSingapore,Singapore.123.FolkhälsanInstituteofGenetics,FolkhälsanResearchCenter,Helsinki,Finland124.HUCHAbdominalCenter,HelsinkiUniversityHospital,Helsinki,Finland125.CenterforBehavioralGenomics,DepartmentofPsychiatry,UniversityofCalifornia,SanDiego126.InstituteofGenomicMedicine,UniversityofCalifornia,SanDiego127. JulietKeidan InstituteofPediatricGastroenterology,ShaareZedekMedicalCenter,TheHebrewUniversityof Jerusalem,Israel128.InstitutodeInvestigacionesBiomédicasUNAMMexicoCity129.InstitutoNacionaldeCienciasMédicasyNutriciónSalvadorZubiránMexicoCity130.NationalHeart&LungInstitute&MRCLondonInstituteofMedicalSciences,ImperialCollegeLondon,LondonUK131.CardiovascularResearchCentre,RoyalBrompton&HarefieldHospitalsNHSTrust,LondonUK132.RadcliffeDepartmentofMedicine,UniversityofOxford,OxfordUK133. Department of Gastroenterology and Hepatology, University of Groningen and University Medical Center Groningen,Groningen,theNetherlands134.DepartmentofPhysiologyandBiophysics,UniversityofMississippiMedicalCenter,Jackson,MS39216,USA135.PrograminInfectiousDiseaseandMicrobiome,BroadInstituteofMITandHarvard,Cambridge,MA,USA136.CenterforComputationalandIntegrativeBiology,MassachusettsGeneralHospital

23andMeResearchTeam:MichelleAgee,AdamAuton,RobertK.Bell,KatarzynaBryc,SarahL.Elson,PierreFontanillas,NicholasA.Furlotte,BarryHicks,DavidA.Hinds,KarenE.Huber,EthanM.Jewett,YunxuanJiang,Keng-HanLin,NadiaK.Litterman,MatthewH.McIntyre,KimberlyF.McManus, Joanna L.Mountain, ElizabethS.Noblin,CarrieA.M.Northover, Steven J. Pitts,G.DavidPoznik,J.FahSathirapongsasuti,JanieF.Shelton,SuyashShringarpure,ChaoTian,JoyceY.Tung,VladimirVacic,XinWang,CatherineH.WilsonAffiliation:23andMe,Inc.



https://doi.org/10.1101/561472


Human loss-of-function variants suggest that partial LRRK2 ... · across 141,456 individuals...

Documents

Transcript of Human loss-of-function variants suggest that partial LRRK2 ... · across 141,456 individuals...