Functional monoamine oxidase B gene intron 13 polymorphism predicts putaminal dopamine turnover in de novo Parkinson's disease

Objective: The objective of this study was to evaluate the effects of common functional polymorphisms in genes involved in dopamine metabolism on striatal dopamine turnover in de novo Parkinson's disease (PD).


Supporting Data
Additional Supporting Information may be found in the online version of this article at the publisher's web-site.  Parkinson's disease (PD) is characterized by a progressive loss of dopaminergic neurons of the substantia nigra pars compacta projecting to the striatum. Motor symptoms only emerge after a substantial loss of striatal dopaminergic nerve terminals has already occurred, 1 which implies that a variety of compensatory mechanisms take place in the preclinical phase of PD. 2 One of these compensatory mechanisms is an increase of putaminal dopamine turnover, which can be measured with 18 Ffluorodopa PET even earlier than changes in dopamine synthesis and storage. [3][4][5][6] Moreover, an extensively elevated dopamine turnover in early PD has been discussed as intrinsic risk factor for the development of motor complications. 7,8 Conversely, mild elevation of putaminal dopamine turnover in de novo PD has been shown to be associated with a lower risk for motor complications. 9 Although previous studies have demonstrated that age and levodopa treatment as known clinical risk factors for motor complications correlate with dopamine turnover in early PD, 8,10 it remains unclear whether functional polymorphisms in genes involved in dopamine metabolism intrinsically contribute to individual changes of dopamine turnover. Thus, we aimed to investigate whether common functional polymorphisms in the dopa decarboxylase (DDC, rs921451), 11 monoamine oxidase B (MAOB, rs1799836), 12 catechol-O-methyltransferase (COMT, rs4680) 13 and dopamine transporter/solute carrier family 6 member 3 (DAT/SLC6A3, variable number tandem repeats) [14][15][16] genes were predictive for striatal 18 F-fluorodopa uptake and dopamine turnover measured by extended 18 F-fluorodopa PET imaging 3,10,17,18 in de novo PD.

Participants and Methods
This observer-blinded cohort study analyzed participants with diagnostic criteria defined PD from a previous randomized clinical trial 10,19 who had received extended 18 F-fluorodopa PET imaging prior to any antiparkinsonian treatment. Participants underwent genetic testing of selected functional polymorphisms during subsequent follow-up. We included all remaining participants from the PET study for whom blood collection was possible. From the original cohort of 40 patients, 6 had died (2 from pulmonary embolism, 4 from unknown causes), 4 were lost to follow-up, and 2 had been diagnosed with essential tremor (both had normal putaminal K occ values).
Motor symptoms and activities of daily living were estimated using the Unified PD Rating Scale (UPDRS), 20 disease stage was assessed with the Hoehn and Yahr scale. 21 Clinical evaluations were carried out in drugnaïve patients by movement disorder trained physicians blinded to PET data (M.L., M.W.). Clinical and genetic testing had approval from the ethics committee at Technische Universität Dresden, Dresden, Germany Dresden (EK91052003), the clinical trial was registered with ClinicalTrials.gov (NCT00153972). Written informed consent was obtained from all participants.
PET measurements were carried out in drug-naïve participants using a 3D ECAT EXACT HR + system (Siemens/CTI, Knoxville, Tennessee) as described previously 3,10 (refer to Supplementary Methods for details). In short, we applied a tissue reference model in which the activity curve of the occipital cortex served as input function for calculation of the tissue input uptake rate constant K occ and effective distribution volume ratio (EDVR). 17,18 K occ and (EDVR) from the left and right sides were averaged for each scan. 10 Genotyping of the common functional polymorphisms DDC (rs921451), 11 MAOB (rs1799836), 12 COMT (rs4680), 13 and DAT/SLC6A3 (variable number tandem repeats, VNTR) 14-16 was performed as described previously (refer to Supplementary Methods for details). In brief, genomic DNA was extracted by standard methods from venous blood samples and polymorphisms were analyzed by an allelic discrimination Taqman assay (Applied Biosystems, Foster City, California). Positive control DNA template and negative control (DNA/RNA-free water) were included in each genotyping panel. DAT-VNTR polymorphism was determined by standard polymerase chain reaction  Table S1). [11][12][13][14][15][16] Statistical analyses were performed with IBM SPSS statistics, version 23.0 (IBM Corporation, Armonk, New York). Comparisons of demographic, clinical, genetic, and PET data between the groups were made with unpaired t-test or Mann-Whitney U test, or χ 2 test, as appropriate. To account for potentially confounding factors on the comparison of PET measures, we additionally performed an analysis of covariance (ANCOVA), in which age, UPDRS part III score, and symptom duration were entered as continuous covariates and gender as fixed factor. If not mentioned otherwise, data are displayed as means ± standard deviation or numbers and percentages. The significance level (2-tailed) was set at P < .0125 to correct for multiple comparisons using the Bonferroni method.

Results
Our cohort included 28 patients (19 [68%] men, 9 [32%] women; mean age 60.3 ± 9.4 years), who had been diagnosed 1.8 ± 2.0 years after the onset of motor symptoms. Mean age at disease onset was 57.3 ± 9.2 years, and mean UPDRS motor score was 19.4 ± 7.6. Genotyping revealed allele frequencies  Table S1). All genotypes were in the Hardy-Weinberg equilibrium. We did not detect significant differences of demographic and clinical variables after stratification for genotypes in the selected gene variants (Supplementary Table S2).
Patients carrying the MAOB CC/(C)/CT (rs1799836) genotype, which is associated with low/intermediate enzyme activity in the brain, 12 had a higher EDVR (lower dopamine turnover) in the whole putamen compared to patients carrying the MAOB TT/(T) genotype that has been linked to high enzyme activity (Fig. 1, Table 1). We also observed slightly higher 18 Ffluorodopa uptake (K occ ) in the MAOB CC/(C)/CT (rs1799836) genotype, which was not significant after correction for multiple comparisons. No differences were detected between genotypes of the remaining genes ( Fig. 1, Supplementary Tables S3-S5). The difference in EDVR between MAOB (rs1799836) genotype groups was confirmed in separated analysis of anterior and posterior portions of the putamen with similar distribution between the 2 putaminal regions. In contrast, we found no differences for K occ and EDVR between groups in the caudate nucleus for all genotypes (Table 1).
Because PET measures had been reported to potentially correlate with age, gender, and disease severity, 8,9 we additionally calculated ANCOVAs to control for the candidate covariates age, UPDRS motor score, symptom duration, and gender. Adjustment for these potential confounders largely confirmed the differences of putaminal EDVR between the MAOB (rs1799836) genotype groups (Table 1). No significant differences of PET measures were observed for the remaining genotype comparisons (Supplementary Tables S3-S5). Moreover, we did not identify significant interactions between genotype groups when calculating ancillary ANCOVAs for K occ and EDVR with Bonferroni correction.

Discussion
We provide first evidence that a common functional variation in the MAOB gene leads to differences in putaminal dopamine metabolism as measured by extended 18 F-fluorodopa PET imaging in patients with early untreated PD. We found that the MAOB CC/(C)/CT rs1799836 genotype (associated with low/intermediate enzyme activity in the brain) resulted in lower putaminal dopamine turnover when compared with the MAOB TT/(T) genotype (high enzyme activity). 18 F-fluorodopa uptake (K occ from standard PET scanning 10-60 minutes after tracer administration) is assumed to mainly reflect tracer influx and central DDC activity. 22 Previous studies have indicated that common DDC polymorphisms (including rs921451) modulate the therapeutic response to levodopa without changing levodopa metabolism in the periphery, suggesting that DDC polymorphisms and resulting variations in DDC activity influence bioavailability of dopamine in the CNS. 11 Indeed, subsequent PET studies in healthy volunteers provided evidence that 1 of 4 common DDC haplotypes is predictive of 18 Ffluorodopa uptake in healthy humans. 23 Although this haplotype contains a major allele that is in linkage disequilibrium with rs921451, 23 we did not find any difference of 18 F-fluorodopa uptake with respect to DDC (rs921451) genotypes in our PD cohort. However, this haplotype is not uniquely defined by this single polymorphism. 23 Further studies are needed to assess the relationship between DDC genetics, central DDC activity, and levodopa response in PD.
In contrast to DDC, COMT and MAO-B as catabolizing enzymes in the dopamine metabolism are proposed to particularly influence dopamine turnover as measured by extended 18 F-fluorodopa PET protocols. 3,[24][25][26] Accordingly, PD patients carrying the functional Data are means ± standard deviations. ANCOVA, analysis of covariance; EDVR, Effective distribution volume ratio from 18 F-fluorodopa positron emission tomography; MAOB, monoamine oxidase B. K occ , 18 F-fluorodopa uptake from positron emission tomography. a P values are from unpaired 2-sided t-tests comparing MAOB C(C)/CT and MAOB TT alleles. P values of < .0125 (2-sided) were considered significant to correct for multiple comparisons using the Bonferroni method and have been marked in bold. b F and P values are from ANCOVA comparing MAOB CC/(C)/CT and MAOB TT/(T) alleles and controlling for the candidate covariates age, UPDRS part III motor score as a measure of disease severity and symptom duration as well as gender as fixed factor. P values of < .0125 (2-sided) were considered significant to correct for multiple comparisons using the Bonferroni method and have been marked in bold.
COMT GG/GG (rs4680) genotype associated with low enzyme activity 13 showed lower late 18 F-fluorodopa uptake (measure of dopamine turnover) when compared with COMT AA/AA carriers in the frontal cortex but not the striatum. 24 In agreement, COMT (rs4680) genotypes in our cohort did not predict striatal dopamine turnover in early PD patients. This observation is in contrast to our findings on MAO-B as the other main dopamine catabolizing enzyme: in close agreement with biochemical analyses showing a correlation of MAOB CC/(C)/CT rs1799836 genotype with low/intermediate enzyme activity in postmortem human brain tissue, 12 our study demonstrated that the MAOB CC/(C)/ CT genotype is associated with lower putaminal dopamine turnover. We also observed a nonsignificant increase of 18 F-fluorodopa uptake with this genotype, which is in agreement with studies showing that K occ is to some extent sensitive to dopamine turnover even if calculated in the first 90 minutes after tracer injection. 18 Although previous 18 F-fluorodopa PET studies reported that decreased DAT levels are directly associated with higher dopamine turnover, 27 we did not observe any effects of the DAT/SLC6A3 (VNTR) genotype known to influence DAT expression 14-16 on dopamine turnover in our cohort. The reasons for this discrepancy remain unclear but might be related to differences in DAT level estimation between studies (direct quantification by PET imaging by Sossi and colleagues, 27 only indirect assumption in our approach).
Although we present data on one of the largest PD cohorts with systematic EDVR measurement by PET imaging, our study has several limitations. First, the cohort size and consequently the genotype group sizes are relatively small, which generated the need for statistical adjustment to confounding variables and may limit generalizability to a less homogenous population. Second, our data were generated from white patients selected at one university-based movement disorder outpatient center and is thus primarily reflective for patients of similar background and setting. Third, although patient groups were clinically indistinguishable at the time of PET imaging, we cannot completely exclude that the compositions of the various genotype groups may relevantly differ, particularly because genotype group sizes in some cases were relatively small after dichotomization. On the contrary, statistical adjustment for major demographic and clinical covariates largely confirmed our unadjusted analyses.
Together, our study suggests that the functional MAOB gene intron 13 (rs1799836) polymorphism predicts putaminal dopamine metabolism in early untreated PD with the MAOB TT allele linked to high enzyme activity leading to higher intrinsic dopamine turnover, which has been shown to be associated with a higher subsequent risk for motor complications. 9 Future studies are warranted to investigate whether this polymorphism modulates the risk or the time of onset of motor complications in PD and whether individual assessment of this gene polymorphism might be helpful for early risk stratification in PD patients.