Article Title: The Remote Assessment of Parkinsonism Supporting Ongoing Development of Interventions in Gaucher Disease – Study Protocol Short Running Title: The RAPSODI Study

Mutations in GBA which are causative of Gaucher disease in their biallelic form, are the most common genetic risk factor for Parkinson disease. The diagnosis of Parkinson disease relies upon clinically defined motor features which appear after irreversible neurodegeneration. Prodromal symptoms of Parkinson disease may provide a means to predict latent pathology, years before the onset of motor features. Previous work has reported prodromal features of Parkinson disease in GBA mutation carriers, however this has been insufficiently sensitive to identify those that will develop Parkinson disease. The Remote Assessment of Parkinsonism Supporting Ongoing Development of Interventions in Gaucher Disease (RAPSODI) study assesses a large cohort of GBA mutation carriers, to aid development of procedures for earlier diagnosis of Parkinson disease. Summary Points  The therapeutic focus for Parkinson disease is currently symptomatic, rather than disease-modifying or neuroprotective.  Non-motor prodromal features of Parkinson disease may precede the motor symptoms required for clinical diagnosis by several years.  GBA mutations, which are associated with Gaucher disease in their biallelic form, have emerged as the most common genetic risk factor for Parkinson Disease.  GBA associated PD displays a slightly exaggerated phenotype, including earlier age of onset, enhanced cognitive decline, more severe affective disturbances, and a greater likelihood of manifesting REM Sleep Behaviour Disorder, hyposmia and autonomic dysfunction. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 22, 2021. ; https://doi.org/10.1101/2021.07.21.21260533 doi: medRxiv preprint

a means to predict latent pathology, years before the onset of motor features. Previous work has reported prodromal features of Parkinson disease in GBA mutation carriers, however this has been insufficiently sensitive to identify those that will develop Parkinson disease. The Remote Assessment of Parkinsonism Supporting Ongoing Development of Interventions in Gaucher Disease (RAPSODI) study assesses a large cohort of GBA mutation carriers, to aid development of procedures for earlier diagnosis of Parkinson disease.

Summary Points
 The therapeutic focus for Parkinson disease is currently symptomatic, rather than disease-modifying or neuroprotective.
 Non-motor prodromal features of Parkinson disease may precede the motor symptoms required for clinical diagnosis by several years.
 GBA mutations, which are associated with Gaucher disease in their biallelic form, have emerged as the most common genetic risk factor for Parkinson Disease.
 GBA associated PD displays a slightly exaggerated phenotype, including earlier age of onset, enhanced cognitive decline, more severe affective disturbances, and a greater likelihood of manifesting REM Sleep Behaviour Disorder, hyposmia and autonomic dysfunction.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 22, 2021. ; https://doi.org/10.1101/2021.07.21.21260533 doi: medRxiv preprint  The RAPSODI GD study is an annual remote assessment which aims to define the clinical prodrome of PD in a large cohort of GBA mutation carriers, to accurately predict clinical diagnosis.
 The secondary objectives of RAPSODI GD are to risk stratify homozygote and heterozygote GBA mutation carriers for PD risk, understand the variable penetrance of the GBA associated PD phenotype, and create a platform for the future evaluation of biomarkers of disease.
 The putative implication of the RAPSODI GD study is to contribute towards earlier diagnosis of GBA associated PD, to provide a timeframe for delivering neuroprotective interventions.
 Defining prodromal PD in GBA carriers may have wider implications for sporadic PD.

Challenges facing early diagnosis and intervention in Parkinson disease.
Parkinson Disease (PD) is the second most common neurodegenerative disease, with increasing global prevalence which doubled between 1990 and 2016 [1,2]. The clinicopathology of PD is characterised by motor symptoms including bradykinesia, tremor and rigidity, attributed to the progressive degeneration of dopaminergic neurons in the substantia nigra pars compacta, [3,4]. An important neuropathological hallmark of PD is the presence of filamentous intracellular inclusions of α-synuclein, known as Lewy Bodies [5].
Current treatments of PD focus primarily on symptom management, rather than disease-modification [6]. With an aging population, the putative continuation of the . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 22, 2021. ; https://doi.org/10.1101/2021.07.21.21260533 doi: medRxiv preprint upward trend in cases of PD calls for the development of preventative therapies, of which there currently are none [7,8]. One of the main challenges for developing neuroprotective therapies for PD is that clinical diagnosis relies upon the presentation of defined motor deficits, which manifest after the onset of neurodegeneration [9]. As such, there is increasing interest in characterising the prodromal features of PD, which may precede motor symptoms by several years [10]. Several latent features of PD's clinical prodrome have been validated thus far, including REM sleep behaviour disorder (RBD), hyposmia, depression and anxiety, global cognitive impairment, constipation and autonomic dysfunction [11]. In recent years, studies such as PREDICT-PD, have optimised the validation of prodromal features of PD through remote assessment methods, enhancing recruitment and power, and allowing for repeated measures without inter-observer variation [12].
Gaucher disease and GBA.
The GBA gene located on chromosome 1q21, encodes for the lysosomal hydrolase glucocerebrosidase (GCase), and has 12 exons (exon 1 is noncoding) [13,14]. Gaucher disease (GD) is a rare autosomal recessive disorder caused by biallelic homozygous or compound heterozygous GBA mutations, of which there have been over 400 pathogenic mutations identified [15,16,17]. In GD, a loss of function of GBA results in deficient GCase activity, and the subsequent accumulation of its substrate, a sphingolipid called glucosylceramide primarily in macrophages [18]. Literature concerning the epidemiology of GD is highly variable, with birth incidence of GD in the general population falling between 0.39 to 5.80 per 100,000, and prevalence ranging between 0.70 and 1.75 per 100,000 [19]. The prevalence of GD varies with ethnicity and is markedly higher in Ashkenazi Jewish populations, who also have a higher GBA mutation carrier frequency [19]. GD has three clinical subtypes: 1, 2 and 3 [20]. Type 1 GD is the mildest and most common phenotype, accounting for 94% of cases in western . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 22, 2021. ; https://doi.org/10.1101/2021.07.21.21260533 doi: medRxiv preprint populations [19]. Type 1 is characterised by clinicopathological features including bone lesions, hepatosplenomegaly, thrombocytopenia and anaemia, in the absence of primary neurological disease [18]. Types 2 and 3 are neuronopathic forms, with mean lethality at < 4 years old and < 40 years old respectively [21]. GBA mutations vary in pathogenicity therefore those that cause type 1 GD (such as N370S) are considered mild, whereas those that also cause type 2 and 3 GD (such as L444P) are denoted as severe [22,23].

The link Between GBA mutations and PD.
Within the last three decades, clinicians have reported the presentation of Parkinsonian motor features in a subset of patients with GD [24,25]. An increased incidence of PD within GD kindreds was observed, and PD patients with GBA mutations were more likely to report a first or second degree relative with PD, than those without GBA mutations [26,27,28]. Parkinsonism has presented in GD patients with both neuronopathic and non-neuronopathic GBA variants [29]. Subsequently, in an international multi-centre analysis a strong association was detected between GBA mutations and PD, with an odds ratio of 5.43 for carrying heterozygous GBA mutations in PD patients vs controls [28]. This work established heterozygous GBA mutations as the most common genetic risk factor for PD. Whilst an increased lifetime risk for developing PD has been established for both biallelic and monoallelic carriers of GBA mutations, it is still only a minority of individuals with these mutations that will go on to develop PD [30,31]. Heterozygous GBA mutation carrier status confers a 3-15% risk of PD by age 80, contingent upon the severity of the variant, however the literature concerning the penetrance of these mutations in PD is variable [23,30,32,33].
Penetrance of GBA mutations in PD has been estimated between 10-30% and can be stratified by genetic and nongenetic factors, including age-specificity and severity of mutation [33,34].
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The prodrome of GBA associated PD.
There is an overrepresentation of literature concerning the natural history of PD associated with the two most frequent GBA pathogenic variants, L444P and N370S, compared to the role of many of the other hundreds of variants [13,22,35,36,37]. The phenotype of GBA associated PD (GBA PD), differs from sporadic PD (sPD), with a slightly earlier age of onset of approximately 5 years, with a dose effect for severe vs mild mutations, [38,23]. It has also emerged that GBA PD is associated with more nonmotor symptoms than its idiopathic counterpart, including a greater likelihood of manifesting cognitive dysfunction concomitant with more aggressive cognitive decline, more severe affective disorder, and an increased risk of experiencing RBD, hyposmia and autonomic dysfunction [39,40,30,41,42].

GCase in sporadic PD.
Whilst the molecular mechanisms of pathogenesis have not been fully elucidated for GBA PD, much like in GD there is a significant reduction in GCase activity [43].
Histopathological analyses of post-mortem PD brain tissue have identified reduced GCase activity and protein levels in regions with elevated expression of α-synuclein [43], and this was subsequently reproduced [44,45]. The most significant reduction in GCase activity was noted in the substantia nigra, and notably these observations were found in both GBA PD tissue and tissue from individuals with sPD [43]. Work in iPSC models has suggested that the involvement of GCase in both GBA PD and sPD, may result from a bidirectional relationship between reduced lysosomal GCase activity and cytotoxic αsynuclein aggregation [46,47,48]. Due to the implication of deficient GCase activity in both GBA PD and sPD, GCase has become an attractive therapeutic target for PD [49,50,51]. Murine and iPSC models of GBA PD and sPD have demonstrated the promise of gene therapies and small-molecule chaperone modulation, to increase . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 22, 2021. ; https://doi.org/10.1101/2021.07.21.21260533 doi: medRxiv preprint GCase activity and subsequently reduce α-synuclein accumulation with neuroprotective outcomes [52,53]. Understanding the mechanisms of GCase-mediated PD pathogenesis is therefore a salient research focus for not only GBA PD but more widely sPD.

Objectives and Rationale
The increasing global burden of PD highlights the relevance of utilising PD's clinical prodrome for earlier diagnosis, and intervention [2]. As the most common genetic risk factor for PD, with premotor features and putative implications for sPD, defining prodromal PD features associated with GBA is an important target for enhancing detection of disease and earlier intervention timepoints [28,39,49,3].
The RAPSODI GD study aims to define prodromal symptoms of PD in individuals who carry GBA mutations, both homozygote and heterozygote, and to further elucidate the natural history of GBA PD. It will test the hypothesis that the presentation of

Sampling.
Purposive sampling has been used to recruit participants from the following strata: 1. Heterozygous GBA mutation carriers, 2. Homozygous GBA mutation carriers, is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 22, 2021. ; https://doi.org/10.1101/2021.07.21.21260533 doi: medRxiv preprint Manchester Children's Hospital). From these PICs GD patients with homozygous GBA mutations are recruited. These participants act as index cases, to recruit first degree relatives from. These relatives once consented are subsequently screened and determined to be homozygous or heterozygous for GBA mutations, or controls. Our sister study, PD FRONTLINE, is a newly established offshoot which screens idiopathic PD cases for GBA mutations. All idiopathic PD cases will initially be recruited and screened via the PD FRONTLINE platform to ensure GBA mutation status. These GBA mutation positive individuals will then be asked to reconsent and reassign from PD FRONTLINE to RAPSODI GD and will act as index cases to contact first degree relatives through. Relatives will be determined to be homozygous or heterozygous for GBA mutations, or controls. Family tracing from index cases is only pursued with permission from the participant. Due to the descriptive nature of RAPSODI GD, we envision that recruitment will remain active throughout the full course of the study period, up to the limit of 10,000 participants. Individuals with other neurological disorders, 4. Individuals on drugs known to be associated with Parkinsonism.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Remote annual and biennial assessments.
Annual remote assessments within RAPSODI GD take place online, via the RAPSODI GD portal (www.rapsodistudy.com). The battery of assessments takes approximately one hour to complete.

Informed Consent.
Participant consent is gained upon enrolment within the online RAPSODI GD portal (www.rapsodistudy.com), or the PD FRONTLINE portal for participants with idiopathic PD (www.pdfrontline.com).

Self-report rating scales.
The RAPSODI GD study uses three self-administered scales to assess: 1.

Bradykinesia akinesia incoordination test.
Motor dysfunction is assessed via the Bradykinesia Akinesia Incoordination (BRAIN) Test [57]. On a standard physical computer/laptop keyboard, participants must . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
Incoordination, akinesia, dysmetria and kinesia are assigned scores based on the variance in time travelling between the keys, the time spent on keys, the accuracy of pressing the correct keys, and the total number of key presses, respectively.

Cognitive testing.
Several aspects of cognitive function are assessed via a battery of precision cognitive tests. This was historically delivered via the online platform "CogTrack TM " [58], until the platform expired, and cognitive testing moved to a new platform "NeurOn" in November 2020 (www.neuropsychology.online). The battery includes tasks of image encoding, simple reaction time, digit span backwards, spatial working memory and image recall. Within this battery participants are scored on dysfunction in working and episodic memory, disturbances in attention focus and sustainment, and memory retrieval rate.

Hyposmia assessment.
Olfactory function is assessed on a biennial basis via the University of Pennsylvania Smell Identification Test (UPSIT) [59]. The UPSIT is delivered remotely through the post, and scores are uploaded directly onto the RAPSODI GD portal by participants. The UPSIT utilises a series of booklets, each containing 10 microencapsulated odorant panels. Participants apply mechanical pressure to the panels to release the odorants, and then select which of four options they believe each odorant to represent. The participants are assigned a score of olfaction sensitivity from 0-40 depending on the accuracy of their responses.

Genotyping.
In their baseline year with RAPSODI GD, participants are posted a saliva sample collection kit (DNA Genotek Oragene.DNA OG-500), with biological substance category . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 22, 2021. ; https://doi.org/10.1101/2021.07.21.21260533 doi: medRxiv preprint B compliant return packaging. Genomic DNA is extracted via the protocol provided in the DNA purification handbook published by DNA Genotek, for use with the Oragene.DNA OG-500 kits [60]. The full 8.9kb amplicon of the GBA gene is sequenced for all known deleterious variants, on the Oxford Nanopore MinION, via a protocol validated by Leija-Salazar et al. [61]. Due to RAPSODI GD's active selection of individuals with GBA mutations, there is a likelihood that individuals from an Ashkenazi Jewish population will be overrepresented within the cohort [23]. As such, genomic DNA samples are also MinION sequenced for the most common LRRK2 mutation, G2019S, as this is a salient and potentially confounding PD associated risk factor for Ashkenazi Jewish individuals [62].
Upon detection of mutation positive cases, DNA samples will be sent for external confirmatory Sanger sequencing in an NHS accredited lab.

Follow up.
The self-reporting scales, Bradykinesia akinesia incoordination test, and cognitive testing will be repeated once a year from baseline. The UPSIT will be repeated once every other year from baseline.

Future Perspective
The development and validation of clearly defined features of prodromal PD for GBA mutation carriers, could eventually allow for the accurate detection of at-risk individuals. Clinical detection of disease in this latent phase, would lend itself to the identification of earlier therapeutic timeframes to deliver neuroprotective interventions before irreversible neurodegeneration can take place. As the mechanisms of pathogenesis in the GCase-mediated pathway of PD are elucidated, biomarkers and therapeutic targets characterized for GBA PD may also prove salient in developing preventative measures within sPD as well.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 22, 2021. ; https://doi.org/10.1101/2021.07.21.21260533 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 22, 2021. ; https://doi.org/10.1101/2021.07.21.21260533 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 22, 2021. ; https://doi.org/10.1101/2021.07.21.21260533 doi: medRxiv preprint