p.L1795F LRRK2 variant is a common cause of Parkinson’s disease in Central Europe

Pathogenic variants in LRRK2 are one of the most common genetic risk factors for Parkinson’s disease (PD). Recently, the lesser-known p.L1795F variant was proposed as a strong genetic risk factor for PD, however, further families are currently lacking in literature. A multicentre young onset and familial PD cohort (n = 220) from 9 movement disorder centres across Central Europe within the CEGEMOD consortium was screened for rare LRRK2 variants using whole exome sequencing data. We identified 4 PD cases with heterozygous p.L1795F variant. All 4 cases were characterised by akinetic-rigid PD phenotype with early onset of severe motor fluctuations, 2 receiving LCIG therapy and 2 implanted with STN DBS; all 4 cases showed unsatisfactory effect of advanced therapies on motor fluctuations. Our data also suggest that p.L1795F may represent the most common currently known pathogenic LRRK2 variant in Central Europe compared to the more studied p.G2019S, being present in 1.81% of PD cases within the Central European cohort and 3.23% of familial PD cases. Together with the ongoing clinical trials for LRRK2 inhibitors, this finding emphasises the urgent need for more ethnic diversity in PD genetic research.


Introduction
Twenty years ago, the leucine-rich repeat kinase 2 (LRRK2) gene was identi ed as a key driver of Parkinson's disease (PD) pathophysiology [ 1 , 2 ].Since then, it has received recognition as one of the most common genetic causes of autosomal dominant PD, as well as one of the strongest genetic risk factors in sporadic PD with promising potential for gene-speci c disease-modifying therapy [ 3 ].
The 51-exon LRRK2 gene encodes a multidomain enzyme of the same name, containing a leucine-rich repeat (LRR) domain, followed by Ras-like GTPase domain (ROC), C-terminal of ROC (COR) domain and kinase (KIN) domain, that represent the LRRK2's catalytic core [ 4 ].Additional domains described are an armadillo repeat domain (ARM), an ankyrin repeat domain (ANK) in the N-terminal non-catalytic region and a WD40-repeat at the C-terminal region [ 5 ].Several missense mutations within the catalytic core have been well-characterised as pathogenic for PD, such as p.G2019S and p.I2020T located in the kinase domain, p.N1437D/H/S and p.R1441G/C/H/S located in the ROC domain, p.Y1699C or p.F1700L in the COR domain [ 6 , 7 , 8 ].The functional consequence of these mutations is a dysregulation of LRRK2 enzymatic activity, including increased phosphorylation of a subset of Rab proteins [ 4 ].(Fig. 1).Additionally, several variants were shown to moderately increase the risk for PD, such as p.Q416* and p.A419V located in the ARM domain [ 9 , 10 ], p.R1628P and p.M1646T within the COR domain [ 10 , 11 , 12 ] and p.G2385R within the WD40 domain [ 13 ] Interestingly, a non-coding LRRK2 variant c.622C > T was nominated by Genome-Wide Association Studies (GWAS) to also affect the LRRK2 expression and mediate PD risk [ 14 ].
Lately, the p.L1795F variant (rs111910483) was proposed as a strong genetic risk factor for PD, with an estimated OR of 2.5 [ 15 ].It was also previously shown to have a functional effect via enhanced kinase activity, providing more evidence for its pathogenicity [ 7 ].The p.L1795F variant was previously identi ed in 2 siblings with PD, within a family with several family members affected in each generation.However, no segregation was shown due to the unavailability of additional family members [ 16 ], and further reports are lacking in the literature.
The aim of this study was to screen a multicentre early onset and familial PD cohort from 9 movement disorder centres across Central Europe within the Central European Group on Genetics of Movement Disorders (CEGEMOD) consortium [ 17 ] for rare LRRK2 variants using whole exome sequencing (WES) data to access the possible contribution of proposed p.L1795F LRRK2 variant to PD in Central Europe.

Results
The mean age of included PD patients from Central Europe (n = 220) was 53.5 ± 12.9 years of whom 136 (61.8%) were men.Family history was positive in 93 patients (42.3%) and 117 patients (53.2%) developed PD before the age of 40 years.The basic characteristics of the patients included are summarized in Suppl.Table 1.
Pedigrees for all 4 index cases are provided in Fig. 2. Most families report additional members with PD; however, the majority of these individuals were deceased, and biological material was not available to test if the identi ed LRRK2 variant segregated with the disease in these cases.Interestingly, all patients were from the same region close to the east Slovak-Hungarian border, with 1 patient (P1) of Hungarian origin and 3 patients (P2-4) from Slovakia (Table 1).In all 4 cases, the p.L1795F variant was validated by Sanger sequencing (Suppl.Figure 1).

Genotype-phenotype correlation
The clinical features of PD patients harbouring the heterozygous p.L1795F variant are shown in Table 2 and Suppl.Structure and function of the p.L1795F variant Previous biochemical analysis of the p.L1795F variant reported an approximately ve-fold increase in phosphorylation of Rab10 at p.T73, a well-characterised substrate for LRRK2, compared to the wild-type protein, as well as an approximately two-fold increase in autophosphorylation at p.S1292, and a halving of phosphorylation at p.S935 [ 7 ].These data are similar to the functional impact of pathogenic mutations in the ROC and COR domain, including p.N1437H, p.R1441C, and p.Y1699C.Notably, the p.L1795F residue is located in close proximity to pathogenic mutations in the ROC and COR domain (Fig. 4).Taken together these data are consistent with, but do not demonstrate, a pathogenic role for the p.L1795F variant.

Discussion
In this study, we screened for the p.L1795F LRRK2 variant in a multicentre early onset and familial PD cohort within the CEGEMOD consortium [ 17 ] to determine its possible contribution to PD in Central Europe.In contrast to other European populations, the genetic background of PD in Central Europe remains mostly unknown as genetic reports are scarce.Leveraging WES data, we identi ed 3 heterozygous PD carriers and 1 heterozygous PD carrier was additionally identi ed through standard genetic testing.The minor allele frequency (MAF) in our PD cohort was 0.00909, compared to the MAF of 0.0004084 reported previously within the largest PD-associated rare variant meta-analysis so far [ 15 ].Similarly, the p.L1795F prevalence in Central European PD patients seems to be much higher compared to the other already-known LRRK2

Genetic Studies
Peripheral blood was obtained from all individuals and genomic DNA was extracted from 0.5ml of frozen EDTA whole blood samples using QIAamp DNA Blood Mini Kit (Qiagen, UK), based on the manufacturer's instructions, after obtained informed consent from all participants.Samples were then quanti ed and checked for purity using a NanoDrop spectrophotometer (Thermo Scienti c) and whole exome sequencing (WES) was performed by Novogene (Cambridge, UK) according to their protocol [ 74 ].The libraries were sequenced with Illumina's NovaSeq 6000.All samples passed the internal QC steps with an average of 6.7 million total reads per sample and an average sequence of depth of 33.7 ± 4.2 x.

Figures Figure 1
Figures

Table 1
Demographics and clinical characteristics of LRRK2 p.L1795F positive PD patients.

Table 2
Detailed phenotype of identi ed LRRK2 p.L1795F positive PD patients.
PD = Parkinson's Disease; MDS-UPDRS -Movement Disorder Society-Uni ed Parkinsons Disease Scale; H&Y = Hoehn and Yahr; MoCA -Montreal Cognitive Assessment; GCase = Betaglucocerebrosidase; L-dopa = Levodopa; LCIG = Levodopa-Carbidopa Intestinal Gel; DBS = Deep Brain Stimulation; STN = Subthalamic Nucleus; y = years; NA = not available; pathogenic variants, which are quite rare in Central Europe[ 20 ].The p.L1795F variant was present in 1.81% of Central European PD cases and 3.23% of familial PD cases, compared to the prevalence of the most common p.G2019S, which was estimated to be around 0.33% as reported previously[ 20 ].Population heterogeneity seems to be a prominent factor in the LRRK2 allelic distribution.The variants' prevalence varies geographically, suggesting the locations of founder events and the dispersion of founders' descendants.The most-studied mutation in LRRK2, p.G2019S, was demonstrated to arise independently at least twice in humans[ 21 ], with one of the founding mutational events occurring in the Middle East approximately 4000 years ago [ 16 ,21 , 22, 23 ], leading to its higher frequencies in North African Polish PD family was reported carrying the p.N1437H variant that is mostly described in Scandinavians[ 40 ].The LRRK2 protein possesses two distinct enzymatic activities via its kinase and ROC domains, the second being intimately linked to the COR domain.Interestingly, throughout evolution ROC domains are always accompanied by a COR domain[ 41 ], therefore they are considered functionally inseparable and described as a ROC-COR supradomain.The disease mechanism underlying LRRK2-associated PD is proposed via increased kinase activity with pathogenic mutations leading to gain-of-function and hyperphosphorylation of substrate proteins[ 5 ].The GTPase activity has received less attention than the kinase, however, several pathogenic mutations located within the ROC-COR domains indicate that the LRRK2 GTPase activity also plays a role in PD pathophysiology.[ 42ciated within the ROC-COR supradomain were described to reduce the GTPase activity, which in turn elevates the kinase activity [ 5 ,31 , ,44]. Acposed p.L1795F LRRK2 variant is located in close proximity to pathogenic mutations in the ROC and COR domain.It has been previously in vitro demonstrated to enhance the kinase activity and computationally predicted as likely pathogenic or damaging (AlphaMissense: Noted as likely pathogenic with a score of 0.744[ 45 ], REVEL score = 0.638; Conservation Previous studies suggest that sleep disturbances and urinary dysfunction are among the commonly reported NMS by LRRK2-PD patients and asymptomatic carriers[ 49 ].Certain NMS, especially REM sleep behaviour disorder and cognitive decline, may occur at slightly reduced frequencies compared to sporadic PD [50 , 51 ].Different LRRK2 variants also appear to be linked with speci c clinical phenotypes.For instance, the p.R1441G PD carriers may be more likely to develop excessive tremor[ 2 ].Motor uctuations were more likely to develop in p.R1441G/C/H/S compared to the individuals carrying the p.G2019S variant[ 52 ].Similarly, the p.G2385R carriers tend to have a more rapidly progressive parkinsonism and motor decline with more motor uctuations compared to p.G2019S [53 , 54 ].Atypical features, such as dementia and amyotrophy were described in individuals carrying the p.Y1699C variant[ 1 ]. 2 siblings diagnosed as late onset PD (at 60y and 66y respectively) [ 16 ].Interestingly, the patient with signi cantly younger onset at 25y also carried rare heterozygous MAPT p.R538P (c.1613G > C; p.Arg538Pro) variant whose clinical signi cance is currently unknown (CADD score 25.1; Polyphen: probably damaging; SIFT: deleterious, carol: deleterious) as it lacks reports in the literature.Several MAPT loci variants have already been reported to interact with the LRRK2 gene [ 55 ] and might increase the susceptibility to PD, such as MAPT IVS1 + 124 C > G variant that seems to modify the PD risk in LRRK2 p.G2385R carriers in East Asians, as well as slightly decrease the age of onset [ 56 ].Similarly, several polymorphic variations in the MAPT gene have already been shown to possibly in uence the age of onset in LRRK2-associated PD [ 57 , 58 ], though the results are inconclusive and the exact mechanism remains unknown.Family history resembled an autosomal-dominant mode of inheritance in 3 cases (75%) as suggested previously [ 16 ], though additional family members affected were not available for segregation analysis.A striking feature of this variant in all identi ed cases in our study seems to be the lack of tremor in clinical presentation, as well as very early onset of severe dyskinesia and motor uctuations, with a narrow therapeutic window and unsatisfactory response to advanced treatment options such as LCIG therapy or DBS.Similarly, in the family carrying p.L1795F variant reported previously, severe dyskinesia was present in 1 PD carrier, though resting tremor was also present at PD onset in 1 sibling contrary to our cohort [ 16 ].All 4 patients reported urinary dysfunction and orthostatic hypotension as the most common NMS.Sleep diturbances and constipation were also present in 50% of the carriers identi ed.None of the p.L1795F carriers were diagnosed with RBD and only one patient was diagnosed with level 1 PD-dementia based on MoCA score (13pts), being in his early eighties and after 28 years of In conclusion, the genetic analysis presented herein, taken together with functional and structural data, supports a pathogenic role for the p.L1795F variant.Therefore, we propose that LRRK2 p.L1795F variant should be included in the standard genetic testing in PD patients from Central Europe as it seems to represent one of the major contributors to autosomal-dominant PD in this region.It was previously shown to have a high OR and have a functional effect via enhanced kinase activity providing more evidence for its pathogenicity.Our data suggests that p.L1795F may represent the most common currently known pathogenic LRRK2 variant in Central Europe compared to more studied p.G2019S and therefore should be prioritized.Together with the ongoing clinical trials for LRRK2 inhibitors, this nding emphasizes the urgent need for more ethnic diversity in PD genetic research.We screened Whole Exome Sequencing (WES) data of 219 early onset (below the age of 50 years) or familial PD patients (EOPD = 179, FPD = 93; males 136, mean age = 53.5 ± 12.9 years, mean age of onset = 41.7 ± 11.2 years) recruited from 9 movement disorder centres in Czech Republic, Hungary, Poland and Slovakia within the CEGEMOD consortium as described elsewhere[ 17 ].The research protocol was approved by the ethics committees from all participating centres.All patients provided informed consent.Each individual with PD was diagnosed in accordance with the Movement Disorder Society (MDS) clinical diagnostic criteria [ 60 ].Demographic and clinical data were collected using standardised protocol. The Moveent Disorder Society-Uni ed Parkinson's Disease Rating (MDS-UPDRS) Part III [ 61 ] and Hoehn and Yahr (H&Y) stage systems [ 62 ] were used to evaluate motor severity.Cognitive functions were evaluated using the Montreal Cognitive Assessment (MoCA) [ 63 ] and the Parkinson's Disease -Cognitive Rating Scale (PD-CRS) [ 64 ].Non-Motor Symptoms Scale (NMSS) [ 65 ] was used for the assessment of non-motor symptoms in PD and autonomic dysfunction was reported using the Scale for Outcomes in Parkinson's disease for Autonomic symptoms (SCOPA-AUT) [ 66 ].The REM Sleep Behaviour Disorder Screening Questionnaire (RBDSQ) [ 67 ] was used as a self-screening tool for RBD.The general level of daytime sleepiness using The Epworth Sleepiness Scale (ESS) [ 68 ].Fatigue was measured with the Multidimensional Fatigue Inventory (MFI) [ 69 ].The Beck Depression Inventory II (BDI-II) [ 70 within PD cohorts, with the p.G2019S variant's combined prevalence being reported as only 0.33% (MAF = 0.00164)[ 20 ].The studied p.R1441H, p.N1437H, p.Y1699C, p.I2020T, p. M1869V, p.P755L, p.I1122V, p.L1114=, p. G2385R and p.R1428P were not identi ed at all.Interestingly, the risk factor p.A419V, which was linked to PD in East and Central Asian populations, was also identi ed in the Hungarian PD cohort[ 20 ]and one LRRK2-associated PD is characterised by features consistent with idiopathic PD[ 27 ].Initial motor features typically include slowly progressive asymmetric tremor at rest and/or bradykinesia, rigidity and gait abnormalities.The range of disease onset is typically in the 50s and 60s but varies, even within the same family, as early onset (in the 20s) and late onset (in the 90s) have also been described [ 46 ].Disease progression also varies signi cantly among individuals and is related to age of onset [ 25 , 47 ].Nonmotor symptoms (NMS) may appear prior to movement disorder or emerge with motor disease progression, such as hyposmia/anosmia, constipation, depression, anxiety and other neuropsychiatric features [ 48 ].The novel p.L1795F is consistent with the LRRK2-associated PD phenotype described in the literature, although it is associated with more rapidly progressive parkinsonism and earlier onset of severe motor uctuations.The age of onset varies from early (25y) to late (69y) age, similarly as in the rst family reported with Conclusion Methods Study and participants 0.4.0; https://www.broadinstitute.org/gatk/)was used for base quality score recalibration, variant calling and variant quality score recalibration.Variants were annotated using VEP Variant Effect Predictor(https://genomebiology.biomedcentral.com/articles/10.1186/s13059-016-0974-4). cDNA and protein sequence variants are described in accordance with the recommendations of Human Genome Variation Society.All LRRK2 variants were harmonized with the canonical Ensemble feature