Brain Microglial Activation Increased in Glucocerebrosidase (GBA) Mutation Carriers without Parkinson's disease

ABSTRACT Background Glucocerebrosidase gene mutations are a common genetic risk factor for Parkinson's disease. They exhibit incomplete penetrance. The objective of the present study was to measure microglial activation and dopamine integrity in glucocerebrosidase gene mutation carriers without Parkinson's disease compared to controls. Methods We performed PET scans on 9 glucocerebrosidase gene mutation carriers without Parkinson's disease and 29 age‐matched controls. We measured microglial activation as 11C‐(R)‐PK11195 binding potentials, and dopamine terminal integrity with 18F‐dopa influx constants. Results The 11C‐(R)‐PK11195 binding potential was increased in the substantia nigra of glucocerebrosidase gene carriers compared with controls (Student t test; right, t = −4.45, P = 0.0001). Statistical parametric mapping also localized significantly increased 11C‐(R)‐PK11195 binding potential in the occipital and temporal lobes, cerebellum, hippocampus, and mesencephalon. The degree of hyposmia correlated with nigral 11C‐(R)‐PK11195 regional binding potentials (Spearman's rank, P = 0.0066). Mean striatal 18F‐dopa uptake was similar to healthy controls. Conclusions In vivo 11C‐(R)‐PK11195 PET imaging detects neuroinflammation in brain regions susceptible to Lewy pathology in glucocerebrosidase gene mutation carriers without Parkinson's. © 2020 The Authors. Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson and Movement Disorder Society

A B S T R AC T : Background: Glucocerebrosidase gene mutations are a common genetic risk factor for Parkinson's disease. They exhibit incomplete penetrance. The objective of the present study was to measure microglial activation and dopamine integrity in glucocerebrosidase gene mutation carriers without Parkinson's disease compared to controls.

Methods:
We performed PET scans on 9 glucocerebrosidase gene mutation carriers without Parkinson's disease and 29 age-matched controls. We measured microglial activation as 11 C-(R)-PK11195 binding potentials, and dopamine terminal integrity with 18 F-dopa influx constants. Results: The 11 C-(R)-PK11195 binding potential was increased in the substantia nigra of glucocerebrosidase gene carriers compared with controls (Student t test; right, t = −4.45, P = 0.0001). Statistical parametric mapping also localized significantly increased 11 C-(R)-PK11195 binding potential in the occipital and temporal lobes, cerebellum, hippocampus, and mesencephalon. The degree of hyposmia correlated with nigral 11 C-(R)-PK11195 regional binding potentials (Spearman's rank, P = 0.0066). Mean striatal 18 F-dopa uptake was similar to healthy controls. Conclusions: In vivo 11 C-(R)-PK11195 PET imaging detects neuroinflammation in brain regions susceptible to Lewy  The glucocerebrosidase gene (GBA) encodes the lysosomal hydrolase glucocerebrosidase. In the biallelic (homozygous or compound heterozygous) state, GBA mutations may cause Gaucher disease (GD) which leads to glucosylceramide accumulation in visceral organs and, in a minority of cases, the central nervous system (neuronopathic GD). GBA mutations are the most significant genetic risk factor for Parkinson's disease (PD) and dementia with Lewy bodies (DLB) 1-3 ; however, penetrance is only 10%-30%. 4-6 PD patients carrying a GBA mutation have an earlier disease onset and a higher risk of dementia. 7 At postmortem, α-synuclein aggregations identical to those found in idiopathic PD 1 and DLB 8 are present in GBA-PD subjects. Asymmetrically reduced striatal 18 Fdopa uptake, 9,10 striatal dopamine transporter binding, 11,12 and an altered striatal asymmetry index 13 have been reported in PD patients with GBA mutations. Conversely 123 I-isoflupane dopamine transporter uptake has been demonstrated to be upregulated in non-PD GBA carriers compared with controls and is higher in GBA PD compared to idiopathic PD cases. 14,15 GBA mutation carriers without PD exhibit prodromal PD features, [16][17][18][19] which progress with time. 20 Glial activation has been demonstrated in postmortem PD brains. 21,22 Nigral microglial activation along with reduced striatal 18 F-Dopa uptake is present in idiopathic rapid eye movement sleep behavior disorder (RBD). 23 It is also a feature of neuronopathic GD at postmortem 8 and in GD mouse models. 24 No studies have investigated in vivo the presence of brain microglial activation in GBA mutation carriers and related this to the presence of striatal dopaminergic dysfunction. We therefore measured 11 C-(R)-PK11195 regional binding potentials (BP ND ) and 18 F-dopa K i in GBA mutation carriers without evidence of Parkinson's disease.

Recruitment and Clinical Assessments
Between 2015 and 2016, 9 biallelic (homozygous or compound heterozygous) or heterozygous carriers of GBA mutations were recruited from University College London, UK (see Table 1 for characteristics). All subjects had exons 1-11 of the GBA gene sequenced (Table 1). Biallelic carriers had type 1 GD, whereas heterozygous carriers were drawn from GD kindreds. No subjects met PD (UK Brain Bank) diagnostic criteria, and none were genetically related. Two of 5 GD patients were receiving enzyme replacement therapy (ERT; velaglucerase 800 IU weekly and 4000 IU monthly) and 3 of 5 substrate reduction therapy (SRT: eligustat 84 IU twice daily in 2 of 3, miglustat 300 mg once daily in 1 of 3). Both SRT and ERT were administered throughout the duration of the study. Ethical approval was obtained from London, UK (10/H0720/21), and Midtjylland, Denmark (M-2014-397-14), research ethics committees. Each GBA carrier had 11 C-(R)-PK11195 and 18 Fdopa PET, an MRI, and neurological examination. Prodromal PD features were rated with the University of Pennsylvania Smell Identification Test (UPSIT), Montreal cognitive assessment, RBD questionnaire (RBDSQ), PD Non-Motor Symptoms Scale, the Movement Disorder Society Unified Parkinson's Disease Rating Scale (MDS-UPDRS) parts II and III, and Beck's Depression Inventory.
All scans and examinations were performed at Aarhus University Hospital, Denmark. GBA carrier PET findings were compared with in-house PET data from 29 age-matched healthy controls (20 had 11 C-[R]-PK11195 BP ND PET, and 9 had 18 F-dopa PET) recruited for a previously published study. 25 Assessments of control prodromal PD features were not available.

PET and MRI
We performed prespecified region-of-interest (ROI) analyses comparing GBA mutation carriers with controls. Selected ROIs were the substantia nigra (SN), putamen, and caudate for 11 C-(R)-PK11195 BP ND and the putamen and caudate for 18 F-dopa K i . We performed statistical parametric mapping (SPM) of 11 C-(R)-PK11195 uptake across all brain voxels. Technical details of the PET and MRI scanning and analysis procedures are available in the supplementary materials.

Statistics
For the ROI analyses, statistical calculations and graphs were produced with Stata v14.2 software (StataCorp., College Station, TX). The 18 F-dopa K i and 11 C-(R)-PK11195 BP ND values from specified ROIs were compared in carrier and control groups using the Student t test (P < 0.05). When there was a significant difference in 11 C-(R)-PK11195 BP ND between the GBA and control groups, secondary analyses correlating PD prodromal features with 11 C-(R)-PK11195BP ND were undertaken (Spearman's rank: all clinical scales were non normally distributed, P < 0.05). A Bonferroni correction was applied to all significant results.

Participants
Participant characteristics are listed in Table 1. Nine GBA mutation carriers (5 biallelic and 4 heterozygous) were selected on the basis of their genotype and the absence of PD features. Two age-matched control groups (20 for 11 C-(R)-PK11195 BP ND PET and 9 for 18 F-dopa PET) were included in the final GBA analysis. Some GD patients had musculoskeletal problems typical of GD reflected in raised MDS UPDRS III scores, but these were not specific for PD. This reflects the limitations of the MDS UPDRS when used in the context of non-PD comorbidities and applied to subjects without diagnosed PD. No participants had a bradykinetic or rigid syndrome on expert examination. There were no missing data.
Upregulated Cortical, Hippocampal, and Mesencephalon 11C-(R)-PK11195 BP ND in GBA+ Group SPM-localized clusters of voxels with significantly increased 11 C-(R)-PK11195 BP ND in GBA carriers bilaterally in the occipital and temporal cortices, cerebellum, left hippocampus, and central and anterior mesencephalon (Table S3 and Fig. 1B,C). No brain regions showed reduced 11 C-(R)-PK11195 BP ND compared with controls.

No Difference in Mean 18 F-Dopa K i Between GBA+ and Control Participants
The GBA carriers showed no significant decreases in mean 18 F-dopa K i across striatal ROIs compared with controls (Tables S1 and S2, Fig. S1). Two participants had putamen and/or caudate 18 F-dopa K i more than 2 SDs below the control mean (Table S4). Greater variance in 18 F-dopa K i (see Table S1) was seen in the GBA group (SD of 0.002 in the putamen and caudate compared with SD of 0.001 in controls). Post hoc analysis (Student t test) comparing the anterior, medial, and posterior putamen did not show any significant mean differences between GBA mutation carriers and controls.

No Correlation Between Nigral 11 C-(R)-PK11195 BP ND and 18 F-Dopa K i in GBA+ Group
There was no association between the SN 11 C-(R)-PK11195 BP ND and putamen or caudate (Table S2) 18 F-dopa K i in the GBA group.

Discussion
Our data indicate that both heterozygous and biallelic GBA mutation carriers can have increased 11 C-(R)-PK11195 BP ND in brain regions susceptible to Lewy body formation. 26 It is unclear whether this is a cytotoxic or neuroprotective process. Only 10%-30% of GBA mutation carriers will develop PD. It is therefore unlikely that all the participants in this study will convert. Which GBA carriers are likely to progress to PD and the mechanisms underlying this conversion are of particular interest. 11 C-(R)-PK11195 BP ND values in the SN correlated with UPSIT scores, suggesting that those GBA carriers who have reduced olfactory function have higher nigral inflammation. Correlation of striatal 11 C-(R)-PK11195 FIG. 1. (A) Top left, box and dot plots of 11 C-PK11195 binding potential (BP ND ) in the substantia nigra of GBA+ heterozygous carriers (white circles), biallelic GBA+ carriers (black circles), and controls (hollow black diamonds). Please note data points are offset across x axis for ease of interpretation. Middle line is median, box is interquartile range. (B) Top right, 11 C-PK11195 binding potential (BP ND ) in GBA carriers > controls. Colored areas depicted on the single-subject brain template illustrate clusters of voxels of 11 C-PK11195 binding potential (BP ND ) surviving P < 0.05 with family-wise error rate (FWE) correction in the brain stem region of GBA+ carriers compared with control subjects. Non-brain stem clusters are masked. GBA, n = 9; controls, n = 20. (C) Bottom left, 11 C PK11195 binding potential (BP ND ) in GBA carriers > controls. Red areas depicted on the brain surface template illustrate clusters of voxels of 11 C-PK11195 BP ND surviving P < 0.05 with FWE correction in cortical regions of GBA+ carriers compared with control subjects. GBA+, n = 9; controls, n = 20. (D) Bottom right, scatterplots of 11 C-PK11195 BP ND in the substantia nigra of GBA+ carriers against University of Pennsylvania Smell Identification Test (UPSIT) score. GBA+ heterozygous carriers (white), biallelic GBA+ carriers (black). [Color figure can be viewed at wileyonlinelibrary.com] BP ND with age and MDS UPDRS III score has also been shown in early PD cases. 27 Despite mean nigral 11 C-(R)-PK11195 BP ND being increased in the GBA group, no significant reduction in mean putamen 18 F-dopa uptake was seen. It is known that 18 F-dopa lacks the sensitivity to detect early dopaminergic dysfunction because of compensatory upregulation of dopa decarboxylase in the remaining terminals. Early reductions may be better detected with dopamine transporter markers. 28,29 Our finding of normal striatal F-dopa uptake in GBA carriers may not necessarily equate to normal dopamine terminal function, although no GBA carrier exhibited clinical features of PD.
Interestingly 18 F-dopa Ki was more variable in the GBA group compared with controls. Recently, 184 nonmanifesting GBA carriers were reported to have increased dopamine transporter binding across striatal regions. 15 This is in line with an increase in striatal 18 Fdopa K i found in a portion of our GBA+ cases. It has been reported that 11 C-(R)-PK11195 binding to microglia "burns out" as amyloidosis in early Alzheimer's disease advances 30 but increases again as tau tangles form. 31,32 A biphasic trajectory could explain the lack of correlation between 18 F-dopa K i and 11 C-(R)-PK11195 BP ND in our data set.

Limitations
The relatively small sample size, its cross-sectional design, and the unknown future disease status of GBA mutation carriers are limitations. We acknowledge that GBA mutations exhibit a variable penetrance and phenotype, in terms of both PD and GD. Reproducing these results in larger (ideally prospective) and more genotypically and phenotypically homogenous cohorts is needed. Nevertheless, we believe these are important and highly relevant pilot data that will inform the design of future studies.
The 11 C-PK11195 BP ND has high nonspecific binding, which provides a lower specific-to-background PET signal ratio than newer markers of activated microglia; therefore, our results may underestimate glial activation. This study used 11 C-(R)-PK11195 BP ND as a marker of the translocator protein (TSPO) expressed by the mitochondria of activated microglia, and, in contrast to newer TSPO tracers available, the binding is not influenced by the polymorphism of the TSPO expressed by individuals. The limitations of supervised cluster analysis in conditions with possible widespread microglial activation should also be acknowledged, as it could lead to an underestimation of 11 C-(R)-PK11195 BP ND , particularly in small ROIs.
Three of 5 and 2 of 5 subjects were taking substrate reduction therapy or enzyme replacement therapy (ERT), respectively. The former is under evaluation as a PD neuroprotective agent (clinicaltrials.gov, NCT02906020).
ERT is not thought to cross the blood-brain barrier, although 1 report suggests a portion may. 33 We cannot exclude the possibility that the reduced nigral and putamen 11 C-(R)-PK11195 BP ND in biallelic compared with heterozygous cases could represent suppression of glial activation by these drugs.

Conclusions
Our findings indicate that GBA mutations are associated with microglial activation in Lewy-susceptible brain regions in subjects without either a prodromal or clinical diagnosis of PD. Further studies are required to assess whether 11 C-(R)-PK11195 BP ND PET, (with or without additional biomarkers) can predict GBA carrier conversion to PD and striatal dopamine loss.