UBC Faculty Research and Publications

Phosphorylated α-synuclein in Parkinson’s disease: correlation depends on disease severity Stewart, Tessandra; Sossi, Vesna; Aasly, Jan O; Wszolek, Zbigniew K; Uitti, Ryan J; Hasegawa, Kazuko; Yokoyama, Teruo; Zabetian, Cyrus P; Leverenz, James B; Stoessl, Alexander J; Wang, Yu; Ginghina, Carmen; Liu, Changqin; Cain, Kevin C; Auinger, Peggy; Kang, Un J; Jensen, Poul H; Shi, Min; Zhang, Jing Jan 31, 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-40478_2015_Article_185.pdf [ 895.09kB ]
JSON: 52383-1.0132629.json
JSON-LD: 52383-1.0132629-ld.json
RDF/XML (Pretty): 52383-1.0132629-rdf.xml
RDF/JSON: 52383-1.0132629-rdf.json
Turtle: 52383-1.0132629-turtle.txt
N-Triples: 52383-1.0132629-rdf-ntriples.txt
Original Record: 52383-1.0132629-source.json
Full Text

Full Text

RESEARCHPhosphorylated α-synuclenKz7gydegenerative disease, with a complicated etiology featuring PD [5,6], as α-syn in LBs from patients with sporadicStewart et al. Acta Neuropathologica Communications  (2015) 3:7 DOI 10.1186/s40478-015-0185-3pS129 in human disease imperative for more accurate9th Avenue, HMC Box 359635, Seattle, WA 98104, USAFull list of author information is available at the end of the articleboth genetic and environmental components [1,2]. Theabundant, membrane-associated protein α-synuclein (α-syn) is the primary component of Lewy bodies (LBs), a de-fining feature of PD, although whether these inclusionsare themselves detrimental, or are a protective mechanismfor sequestering toxic soluble oligomeric forms of α-syn,remains controversial [3,4].[5-8] and genetic [9] forms of PD is overwhelminglyphosphorylated at S129. Further, pS129 alters character-istics of α-syn such as its propensity for aggregation[6,10,11], toxicity [10-12], and protein associations [13,14].However, considerable controversy exists over the effectsof pS129 on neurodegeneration in PD, with contradictoryfindings in transgenic fly [10] and mammalian modelsstudying effects of phosphorylation on α-syn toxicity[11,15] and neurodegeneration [16]. These contrasts sug-gest that the effects of pS129 may be highly model-dependent, making assessment of the natural course of* Correspondence: zhangj@uw.edu1Department of Pathology, University of Washington School of Medicine, 325Parkinson’s disease (PD) is a common age-related neuro-Introduction: α-Synuclein (α-syn) is a key protein in Parkinson’s disease (PD), and one of its phosphorylated forms,pS129, is higher in PD patients than healthy controls. However, few studies have examined its levels in longitudinallycollected cerebrospinal fluid (CSF) or in preclinical cases. In this study, CSF and clinical data were contributed by >300subjects from three cohorts (the longitudinal DATATOP cohort, a large cross-sectional cohort, and a cohort of LRRK2mutation carriers).Results: Consistent with our previous observation that CSF pS129 positively correlated with Unified Parkinson’s DiseaseRating Scale (UPDRS) scores, CSF pS129 in the DATATOP cohort increased over approximately two years of diseaseprogression (mean change 5.60 pg/ml, p = 0.050). Intriguingly, in the DATATOP cohort, pS129 negatively correlatedwith UPDRS scores at the baseline (R = −0.244, p = 0.017), but not final point, suggesting that this association maydepend on disease stage. Reanalysis of our previous cohort with stratification by PD stage, and addition of a cohort ofLRRK2 mutation carriers with very early/preclinical PD, supported the idea that the relationship between CSF pS129 anddisease severity over a wider range of PD stages might be represented with a U-shaped curve, in which lower pS129levels correlated with worse clinical condition at early stages, but better condition at later stages.Conclusion: The observation of a negative-to-positive transition of correlation of pS129 to disease severity as PDprogresses could have profound impact on how pS129 is used as a biomarker clinically as well as in modeling PDexperimentally.Keywords: Parkinson’s disease, Cerebrospinal fluid, Biomarker, α-synuclein, Phosphorylation, DATATOPIntroduction Posttranslational modifications of α-syn, particularlyphosphorylation at serine 129 (pS129), may be critical indisease: correlation depeTessandra Stewart1, Vesna Sossi2, Jan O Aasly3, ZbigniewTeruo Yokoyama5, Cyrus P Zabetian6,7,8, James B LeverenCarmen Ginghina1, Changqin Liu1,13, Kevin C Cain14, PegMin Shi1 and Jing Zhang1*Abstract© 2015 Stewart et al.; licensee BioMed CentraCommons Attribution License (http://creativecreproduction in any medium, provided the orDedication waiver (http://creativecommons.orunless otherwise stated.Open Accessin in Parkinson’sds on disease severityWszolek4, Ryan J Uitti4, Kazuko Hasegawa5,,8,9,10, Alexander Jon Stoessl11, Yu Wang12,Auinger15, Un Jung Kang16, Poul Henning Jensen17,l. This is an Open Access article distributed under the terms of the Creativeommons.org/licenses/by/4.0), which permits unrestricted use, distribution, andiginal work is properly credited. The Creative Commons Public Domaing/publicdomain/zero/1.0/) applies to the data made available in this article,Stewart et al. Acta Neuropathologica Communications  (2015) 3:7 Page 2 of 11modeling of human PD, as well as to assist future transla-tional investigations.Our recent study in a cross-sectional cohort foundthat, unlike total α-syn, CSF pS129 is higher in PD thanhealthy controls [17]. Remarkably, it also significantlycorrelated with PD severity, suggesting that, in additionto being mechanistically important in the pathology ofPD, pS129 may serve as a biomarker for PD progression.However, this hypothesis must be interrogated in longi-tudinally collected samples, particularly in subjects with-out confounding factors presented by medication. Yet,longitudinal studies of PD face a number of challenges,including repeated collections of CSF over a prolongedtime, allowing enough time to detect biochemical alter-ations in this slowly progressive disease.We studied pS129 in the DATATOP (the deprenyl andtocopherol antioxidative therapy of Parkinsonism) co-hort, in which recently diagnosed PD patients (that is,subjects with early, unmedicated clinical PD) underwentrepeated CSF sampling, two years apart. To extend ourfindings to a wider range of disease stages, we furtherexamined CSF samples obtained from subjects withLRRK2 mutations (which lead to late-onset PD similarto the idiopathic disease), including patients at preclinicalstages [18,19], and results published previously [17,20]from a large multi-center collaborative cohort includingsubjects with later stage PD. With these cohorts, we iden-tified a potential non-linear relationship between CSFpS129 and disease severity.Materials and methodsStudy subjectsCohorts used in this study were approved by the Institu-tional Review Boards of all participating institutions.Although cohorts were recruited and samples werecollected independently, similar protocols were usedthroughout, including collection between 6 and 10 am,freezing immediately after collection, use of polypropyl-ene tubes, addition of protease inhibitor cocktail, andthawing immediately prior to the assay. Samples werecentrifuged for 10 minutes at 15,000 rpm before all as-says (α-syn, pS129, or hemoglobin). Two longitudinalsamples were collected from the DATATOP cohort sub-jects, and one cross-sectional sample was collected fromeach Multi-center collaborative cohort and LRRK2 co-hort subject.DATATOP cohortDetailed description of the DATATOP cohort, recruitedin 1987, has been previously published, [21] as have theadditional criteria used in our analyses [20,22]. To elim-inate potential confounding effects, only those assignedto the placebo group were included in this study. Sub-jects were followed for ~24 months, with the primaryoutcome of time to “endpoint,” defined as the develop-ment of symptoms severe enough to require treatmentwith levodopa, as determined by a blinded clinician.Clinical data, such as Unified Parkinson’s Disease RatingScale (UPDRS) score, was evaluated approximately everythree months, and CSF was collected twice, once each atthe baseline and final time points. We included subjectswhose follow-up time was >6 months, resulting in a co-hort of 95 PD patients.Multi-center collaborative cohortCSF samples were collected from a total of 209 PD pa-tients, recruited at nine collaborative centers, using similarcollection and quality control protocols. Patients under-went extensive clinical evaluation; a detailed description ofthis cohort, including pS129 scores, has been previouslypublished [17]. Of note, the range of MMSE scores islarge, and some of our subjects, especially those enrolledin Multicenter-collaborative cohort, have MMSE scoresmeeting the criteria of cognitive impairment or even de-mentia. However, as >75% typical PD patients show cogni-tive dysfunction eventually [23,24], we did not excludethese subjects, because it would be impossible to also ex-clude the equivalent patients (that is, those who wouldeventually become demented) from the cohorts with sub-jects at earlier stages.LRRK2 cohortCSF samples from 23 asymptomatic and seven symp-tomatic LRRK2 mutation carriers at early stages (sub-jects with UPDRS > 20 were excluded to allow exclusivefocus on early PD) were assessed. Note that disease dur-ation, used to define “early” PD in the other cohorts, isnot available, as most of these subjects have not yet beendiagnosed with PD. Further details concerning recruit-ment, demographics, and other CSF markers in a subsetof this cohort have been previously published [18,19].PET studiesWithin one year of CSF sample collection, each LRRK2cohort subject traveled to the Pacific Parkinson’s ResearchCentre (Vancouver, BC, Canada) for positron emissiontomography (PET) scans. Any antiparkinsonian medica-tions were stopped at least 12 h before assessment. Sub-jects were scanned with 11C-(±)-α-dihydrotetrabenazine(TBZ), a ligand of vesicular monoamine transporter 2, aspreviously described [18,19]. All PET data were normal-ized to age-matched control values. Because the asymp-tomatic group included both subjects with markedreductions in TBZ scores (likely at an early PD stage) andsubjects with normal TBZ scores (who may be truly un-affected at the time of sampling), we limited some analysesto subjects with age-normalized TBZ scores <1.0.Stewart et al. Acta Neuropathologica Communications  (2015) 3:7 Page 3 of 11Luminex assaysCSF pS129 was measured according to our previouslypublished protocol [17]. Briefly, MicroPlex microspheres(Luminex, Austin, TX, USA) were coupled with anti-α-synuclein antibody (ASY-1 [25,26]). For each subject,75 μl CSF was diluted with 25 μl assay diluent (0 · 1%bovine serum albumin [Sigma, St Louis, MO] in PBS).Analytes were detected using biotinylated anti-humanpS129 antibody [27] and streptavidin-R-PE (Prozyme,Hayward, CA, USA). Plates were read on a LiquiChipLuminex 200 Workstation (Qiagen, Valencia, CA, USA).CSF α-syn was measured as previously described; [28]results for the entire DATATOP cohort have been pub-lished separately [20]. In all assays, samples were ran-domly distributed across plates to avoid confoundingdue to the plate location, and assay operators were blindto the disease status or time point of samples.Although frozen DATATOP CSF samples underwentprolonged storage, concentrations of CSF markers (Aβ,tau, α-syn), reported elsewhere [20,22], were comparable torecently collected samples in our previous studies [28,29].Hemoglobin measurementsBlood contamination was assessed using Human Hemo-globin ELISA Quantitation Kit (Bethyl Lab Inc., Mont-gomery, TX, USA) according to the manufacturer’sinstructions.Statistical analysisAll analyses were performed using PASW Statistics 19(IBM, Chicago, IL). Our α-syn and pS129 assays typicallyresult in inter-plate variability <15% and intra-plate vari-ability <10% [17,28]; no corrections were made for as-says that performed within this range. Within eachcohort, interassay variability was ensured by includingidentical reference samples on each assay plate, andinter-plate variability controlled by correcting for thesereferences. No corrections were made between cohorts.Values are reported as mean ± standard deviation (SD)unless otherwise noted. Change in biomarker levels wasanalyzed using a paired t-test on the difference (final-baseline) of CSF sample values following Shapiro-Wilktest for normality. Because a previous study demon-strated that blood contamination influences levels oftotal α-syn [28], 25 subjects with at least 1 sample con-taining >200 ng/ml hemoglobin, the level at which bloodα-syn confounds CSF α-syn levels, were excluded fromanalysis comparing longitudinal changes in pS129, α-synor pS129/α-syn ratio. Associations between CSF levelsand UPDRS scores are reported as Pearson correlation.Association with age-normalized TBZ scores is reportedas Spearman correlation, due to the highly non-normaldistribution of TBZ scores. Subjects diagnosed with PD≤2 years before sample collection (the median diseaseduration for DATATOP subjects at baseline) were con-sidered “early” stage cases for the purpose of segregationof cohorts into “early” vs “late” PD.ResultsDemographic distribution of DATATOP cohortClinical evaluations and CSF were assessed in 95 subjectsassigned to the original DATATOP placebo group. Demo-graphic characteristics, clinical scores, and CSF levels ofprotein markers of all cohorts are presented in Table 1.Endpoint, defined as development of symptoms of suffi-cient severity to require administration of dopamine sup-plementing drugs, was reached by 72 DATATOP subjects.Longitudinal alteration of phosphorylated α-syn in CSFWe tested the change in pS129 and the pS129/α-synratio over the two-year DATATOP follow-up period.Despite a negative trend, suggesting a small decrease,no significant difference was observed between baselineand final total α-syn (mean difference −190.50 pg/ml,95% CI −382.57, 1.56, p = 0.052), in subjects meeting Hgbcutoff. In contrast, both pS129 and pS129/α-syn ratio in-creased (5.60 pg/ml, 95% CI [−0.01, 11.21], p = 0.050 and0.034; 95% CI [0.007, 0.061], p = 0.014, respectively). Weexamined separately the subset of patients who reachedendpoint, and found that the differences between finaland baseline scores for pS129 and pS129/α-syn ratio weresomewhat larger, indicating that the changes are morepronounced in the subset of subjects that progressed torequiring levodopa therapy (Table 2; Figure 1). Examiningthe relationship between total and phosphorylated α-syn,we found that at baseline, pS129 was significantly cor-related with α-syn (R = −0.420; p < 0.001), but not at thefinal time point (R = −0.109; p = 0.335).This finding of increasing pS129 is quite consistentwith previous studies; however, the DATATOP cohortdoes not include neurologically normal control subjectsin which to examine age-related changes in CSF pS129,so any changes due only to aging would not be detectedin this set. In order to approximate this measurement,we studied the cross-sectional relationship between ageand pS129 at the baseline and final time points, andfound no correlation (p > 0.05; Pearson correlation), ineither the full cohort or only the subset that reachedendpoint. In contrast to a previous study [28], there wasno significant correlation between age and total α-syn atbaseline (R = 0.176; p = 0.107); however, a significant cor-relation was observed between age and total α-syn at thefinal time point (R = 0.267; p = 0.017).CSF pS129 and disease severity in cohorts at differingstages of disease progressionHaving observed a longitudinal increase in pS129 in re-cently diagnosed PD patients, we next sought to examineTable 1 Demographic characteristics of cohortsBaseline(N = 95)Final(N = 95)UW-collaborative(N = 209)LRRK2(N = 30)Age, y (mean ± SD) 61.08 ± 8.91 63.08 ± 8.68 65.94 ± 10.54 53.13 ± 13.88Sex F/M (% Male) 35/60 (63%) 35/60 (63%) 52/157 (75%) 15/15 (50%)Duration of disease, yMean ± SD 1.9 ± 1.41 4.02 ± 1.55 7.9 ± 6.48 –Range 0 – 6 1 – 9 0 – 42± 10Stewart et al. Acta Neuropathologica Communications  (2015) 3:7 Page 4 of 11MMSEMean ± SD 28.85 ± 1.57 28.81Range 23 – 30 22 – 3H&Y†Median 2.0 2.0the association between CSF pS129 and disease severityacross a wider range of PD stages than has previously beenexamined, by studying early clinical PD in recently diag-nosed subjects (DATATOP), moderate to severe clinicalPD (multi-center collaborative cohort) and very early/pre-clinical PD (LRRK2 cohort).Range 1.0 – 2.5 1.0 – 4.0UPDRS‡ TotalMean ± SD 24.87 ± 12.50 40.73 ± 1Range 0 – 61 9 – 88UPDRS MotorMean ± SD 16.90 ± 9.52 27.36 ± 1Range 0 – 50 5 – 62Ps129 (pg/ml)Mean ± SD 114.66 ± 17.14 117.89 ±Range 57.11 – 151.63 68.66 – 1Total α-syn (pg/ml)Mean ± SD 630.83 ± 703.08 639.48 ±Range 186.3 – 6343.5 173.6 – 4RatioMean ± SD 0.2530 ± 0.1232 0.2688 ±Range 0.01 – 0.77 0.02 – 0.8†Hoehn and Yahr.‡Unified Parkinson’s Disease Rating Scale3.Table 2 Paired t-test, change in marker: (Hgb > 200 excludedidentical subject groupsSubjects reaching endpoint (N = 51) Subjects not rMean ST dev Sig MeanPs129 (pg/ml) 8.73 24.43 0.014 −2.78Ps129/Syn 0.040 0.123 0.024 0.018Syn (pg/ml) −169.93 877.37 0.173 −245.73Bold font indicates significance at the p=0.05 level..48 28.47 ± 3.07 –11 – 302.0 –DATATOP cohort: early clinical PDWe examined the relationship of CSF pS129 with PD se-verity. A negative correlation was observed betweenpS129 and UPDRS motor scores at baseline (R = −0.244,p = 0.017), but not at the final time point (Table 3), afterapproximately two years of disease progression, when0.0 – 5.06.63 – –2.13 23.72 ± 11.87 5.20 ± 4.730 – 71 0 – 2017.92 74.01 ± 26.67 63.79 ± 22.7370.78 0.00 – 203.17 35.2 – 145.1677.95 553.61 ± 409.22 744.78 ± 1169.94523.0 140.0 – 2990.0 170.98 – 5735.080.1433 0.177 ± 0.098 0.172 ± 0.1326 0.00 – 0.53 0.01 – 0.56for both α-syn and pS129 to allow direct comparison ofeaching endpoint (N = 19) All subjects w/Hgb ≤ 200 (N = 70)ST dev Sig Mean ST dev Sig19.09 0.534 5.60 23.53 0.0500.079 0.347 0.034 0.113 0.014134.65 0.085 −190.50 805.51 0.052l αregeStewart et al. Acta Neuropathologica Communications  (2015) 3:7 Page 5 of 11motor scores were significantly worse. The pS129/α-synratio was not correlated with UPDRS scores at any timepoint. Controlling for age and gender did not signifi-cantly alter the results in any case (not shown).The cross-sectional correlation with UPDRS (lower insubjects with worse UPDRS scores at baseline) and thelongitudinal change in pS129 levels (increasing withprogression of disease severity) appear contradictory.Additionally, the DATATOP baseline correlation resultsseemed to conflict with our previous study [17], inwhich pS129 was significantly positively correlated withUPDRS motor score in a large cohort. This apparentparadox could be explained if the negative correlation isunique to the baseline DATATOP cohort, rather than afeature of PD progression; however, we wished to inves-tigate whether it might be a feature of a multi-stage re-sponse, reflecting a non-linear (U-shaped) time courseFigure 1 Longitudinal changes in pS129 and relationship with totanot reach endpoint. Squares: subjects that reached endpoint. Solid line:between total α-syn and pS129 levels by time point. Circles: baseline timtotal α-syn and pS129 for each time point.of pS129 expression. Because a second, comparable lon-gitudinal cohort was not available to study this question,we instead examined two additional cross-sectional co-horts, one with more advanced PD subjects [17] and onewith preclinical/very early PD subjects [18,19], to at-tempt to estimate the course of CSF pS129 over a longerperiod of PD progression.Table 3 Comparison of correlation between UPDRS motor scordurationDATATOP baseline DATATON Corr P N CorpS129Whole cohort 95 −0.244 0.017 95 0.14Disease duration ≤2 years 69 −0.289 0.016 18 −0.Disease duration >2 years 23 −0.067 0.778 74 −0.B: baseline time point. F: final time point. Most LRRK2 cohort subjects have not yetand 3 multi-center collaborative cohort subjects.Bold font indicates significance at the p=0.05 level.Comparison of DATATOP and multi-center collaborativecohorts: early vs moderate/severe clinical PDTo determine whether the difference in the direction ofcorrelation between CSF pS129 and UPDRS motorscores in DATATOP and our previous multi-center col-laborative cohort could be explained by the much earlierdisease status of the DATATOP cohort, we re-analyzedboth cohorts after stratifying by disease stage (Table 3;Figure 2 B-E), defining subjects with a disease durationless than or equal to the median for the baseline DATA-TOP cohort (2 years) as having early stage PD. In theDATATOP cohort, the significant negative correlationbetween UPDRS and pS129 was observed in the earlystage subjects at the baseline time point (disease dur-ation ≤ 2 years; n = 69), but not in the smaller group withearly PD at the final time point (n = 18). No significanttrend was observed at either time point in the subset-syn. A) Change in pS129 by baseline level. Triangles: subjects that didression line generated from all subjects combined. B) Associationpoint. Squares: final time point. Lines represent correlation betweenwith more advanced PD (disease duration >2 years; n =23 and 74 for baseline and final, respectively). Next, westratified the multi-center collaborative cohort using thesame criteria, although this group included many moresubjects with longer disease durations. The trend of in-creasing UPDRS motor scores with higher pS129 CSFvalues was significant in both the whole group (n = 199)es and CSF pS129 levels, overall and stratified by diseaseP final Multi-center collaborative LRRK2r P N Corr P N Corr P7 0.834 199 0.147 0.038 30 −0.050 0.794093 0.712 32 0.043 0.817 –050 0.674 164 0.173 0.027 –received PD diagnoses. Disease durations not available for 3 DATATOP subjectsann sStewart et al. Acta Neuropathologica Communications  (2015) 3:7 Page 6 of 11Figure 2 The relationship between pS129 and motor symptoms chduration A) Relationship between CSF pS129 and UPDRS motor scores iand late stage PD subset (n = 164), but not in those withearlier stage PD. When the DATATOP baseline and finaltime point data (early PD) were analyzed together withthe multi-center cross-sectional data (moderate to latePD), a U-shaped relationship between CSF pS129 andUPDRS motor scores appeared, i.e., a negative associ-ation during early PD (higher pS129 was associated withlower UPDRS and lesser disease severity), followed by anincreasingly positive association with progressing disease(Figure 2A). Therefore, this data supports the idea thatthe direction of the association between pS129 andUPDRS is non-linear and dependent on disease stage,and there may be a transition between negative andpositive correlation between CSF pS129 and disease se-verity, leading to a U-shaped curve over long periods offollow-up or a wide range of severities included in across-sectional study.LRRK2 cohort: very early and preclinical PDIf the hypothesis that the correlation between CSFpS129 and disease severity is dependent on PD progres-sion is true, then it would be expected that subjects withvery early or preclinical PD would show a negative cor-relation (similar to DATATOP baseline). However, pre-clinical/very early PD patients have very low (or 0)combined datasets including multi-center collaborative cohort and DATATOand UPDRS in cohort subsets divided by disease stage. B) DATATOP DD less tless than 2 years. D) Cross-sectional DD greater than 2 years. MC: Multi-centerDB: DATATOP baseline time point. DF: DATATOP final time point. Diamonds: mcircles: DATATOP final.ges with disease severity. pS129 and UPDRS motor scores by diseaseubjects in DATATOP and cross-sectional cohorts. Line is Loess curve forUPDRS scores, making analysis of the correlation diffi-cult. Therefore, we sought to test this hypothesis using acohort of LRRK2 mutation carriers, in which preclinicalPD signs were assessed by neuroimaging (Table 1). Asexpected, no correlation was observed between pS129and UPDRS (Table 3) in this small cohort of mostly clin-ically asymptomatic subjects, as the UPDRS values forasymptomatic subjects were near or equal to 0. However,LRRK2 subjects underwent PET imaging, and those withclinical PD show markedly reduced age-normalized TBZscores (Figure 3A); therefore, TBZ scores, which appearto undergo changes at an earlier stage than UPDRS,were used to measure severity of the pathology in thesesubjects. In LRRK2 mutation carriers, a trend of decreasedpS129 with greater disease severity was observed, and wassignificant in the subset of subjects with early PD path-ology, indicated by TBZ < 1 (n = 20, Spearman’s rho 0.456,p = 0.043; Figure 3B). This result is consistent with thetrend seen in early PD cases of DATATOP cohort, i.e.higher ps129 values were associated with higher TBZscores and lesser disease severity (note that high TBZscores and low UPDRS scores both reflect milder disease),suggesting that the finding of a negative relationship atearly disease stages is not unique to the DATATOPcohort.P baseline and final time points. B)-E) Relationship between pS129han 2 years. C) DATATOP DD greater than 2 years. D) Cross-sectional DDcollaborative cohort. DD: Disease duration, time since diagnosis in years.ulti-center collaborative cohort. Black circles: DATATOP baseline. GrayStewart et al. Acta Neuropathologica Communications  (2015) 3:7 Page 7 of 11DiscussionSeveral key findings are presented in this study: 1) pS129and pS129/α-syn ratio increased in longitudinally col-lected CSF from untreated early PD patients, particularlythose who progressed to requiring dopamine therapy;and 2) the relationship between pS129 and PD severitydiffered at earlier and later disease stages (high pS129reflected less severe symptoms at earlier stages, but notat later stages). We examined this possibility using re-analysis of a previously studied cohort, after stratificationby disease duration, and found that this progression ap-peared as a U-shaped curve when the two cohorts werecombined. Moreover, this interpretation was supportedFigure 3 LRRK2 subjects with greater brain pathology have lower CSFdiagnoses. B. Relationship between TBZ scores and pS129 in LRRK2 mutatioUPDRS measurements in Figure 2. Dotted regression lines represent whole≥1. Black triangles: Subjects without PD diagnosis and TBZ < 1. Open triangdiagnosis and UPDRS motor scores < 20.by examination of a cohort of LRRK2 carriers with earlyor preclinical PD.Longitudinal changes in pS129 in PD progressionsα-Syn in LBs is highly phosphorylated at S129 [5,6,10],particularly in more severe stages of Lewy pathology[7,20], suggesting a progressive increase in pS129 as PDadvances. Additionally, cross-sectional studies in braintissue and CSF indicate increased phosphorylated α-synin PD [6,17], suggesting that it may be useful as a PDbiomarker. However, whether it increases longitudinallyin the CSF of individual subjects has been less well ex-amined. Here, we found that pS129 increased over thepS129. A) Significant decrease in TBZ scores in subjects with PDn carriers. Note that x-axis is inverted to facilitate comparison withcohort; solid lines represent cohort excluding subjects with TBZ scoresles: Subjects with no diagnosis and TBZ > 1. Circles: Subjects with PDStewart et al. Acta Neuropathologica Communications  (2015) 3:7 Page 8 of 11approximately two-year DATATOP study, while α-syn,though it trended to be lower in more advanced stages,did not achieve statistical significance in the placebogroup. The ratio of pS129/α-syn showed a greater in-crease, as anticipated given their opposite expecteddirections of change. This investigation is, to our know-ledge, the first study of pS129 in living subjects withCSF collected longitudinally over a reasonably longperiod of time (about two years). The observation is en-tirely consistent with our previous cross-sectional studydemonstrating that pS129 and the pS129/syn werehigher in PD than in controls, and pS129 significantlypositively correlated with severity of motor symptoms[17]. Our data, however, contrasted with another recentstudy of post-mortem ventricular fluid, which observedno difference in pS129 between PD patients and controls[30]. The lack of significance between advanced PD andcontrols, when assessed using ventricular fluids collectedat autopsy, could be attributed to a number of factors,including cohort size, antibodies used, or use of autopsyventricular fluid, which has a different protein compos-ition from lumbar CSF [31]. However, later studies bythe same group found increased plasma pS129 in PD,but no longitudinal increase in follow-up periods be-tween 3 months and up to 4 years [32,33]. In light ofour CSF results, the lack of further increase in plasmapS129 in longitudinal samples could be explained by acombination of different sample types, as well as by thefollow-up period, particularly considering the modestmagnitude of the change observed in CSF here, evenover approximately two years. Based on our CSF investi-gations, it is likely that a much longer period of follow-up is needed to detect biochemical alterations in slowneurodegenerative processes [20,22].pS129 and disease severity: evidence for a non-linearpattern in PD progression?The most intriguing observation in the current study isthat pS129 negatively correlated with PD severity atbaseline, i.e. at a relatively early stage of the disease, inthe DATATOP cohort. We hypothesized that this para-doxical observation might depend on disease stage, andseparation of two cohorts by stage showed a trend, inboth cohorts, toward more positive correlation with in-creasing severity. The effect of this change would bereflected in clinical trials as contradictory associationsbetween early and late cohorts, as was observed whencomparing the baseline DATATOP and multi-center col-laborative cohorts. We considered that this relationshipcould be explained if pS129 decreases in the initial orpreclinical stages of PD (leading to the negative correl-ation at early but post-diagnosis time points such asDATATOP baseline), then progressively increases atlater stages (leading first to a loss of any correlation,followed by the introduction of a positive correlation, aswell as longitudinally observed increase in concentra-tion). We therefore sought to determine if a non-linearrelationship appeared when a wider range of diseasestages are studied together, particularly by including sub-jects at very early stages. Remarkably, the trend of anegative correlation between pS129 and PD severity atearly stages was maintained in a cohort consisting ofLRRK2 mutation carriers exhibiting early signs of PD.Although UPDRS scores were typically very low in thesesubjects, TBZ imaging, which reflects nigrostriatal dam-age robustly [34], showed that higher pS129 levels wereassociated with better disease states.Of four groups (DATATOP baseline and final, multi-center collaborative, and LRRK2), only the DATATOPfinal cohort, at intermediate disease stage, showed no as-sociation between pS129 and UPDRS, in contrast to thelate multi-center collaborative cohort, where a clearpositive correlation is shown. This lack of correlation ina portion of DATATOP cohort may be explained by amixture of the limited number of the early stage subjectsat the final time point and the comparatively earlier sta-tus of even the “late” subjects, most of which justreached the point of requiring PD medication. This leadsto the hypothesis that these subjects are at the stageswhere the pS129/severity relationship is undergoing anegative-to-positive transition (i.e., near the vertex of aU-shaped curve). It should also be noted that, althoughthe reported severity (UPDRS scores) in the multi-centercollaborative cohort overlapped substantially with that ofthe DATATOP final group, in the former, all subjectswere on anti-parkinsonism medications, partially mask-ing the true severity of their symptoms. In other words,subjects in this cohort can be considered to have moresevere PD than DATATOP subjects with equivalentscores. For the reasons discussed above, it is not surpris-ing that neither the levels of the markers at baseline, northeir changes over the duration of the study, predictedlongitudinal progression in UPDRS of the DATATOPcohort (data not shown).Taken together, these data suggest the possibility thatthe relationship between pS129 and disease severity mayalter with progression. The mechanisms by which such aphenomenon occurs, and what it means for the role ofpS129 in PD pathogenesis, remain to be examined. Onepossible interpretation of the current study is that phos-phorylation increases as a compensatory mechanism,explaining both its increasing levels and apparentlycontradictory association with less severe symptoms atearly-mid stages, but that the benefits are eventuallyovercome by accumulating negative effects of this orother PD-related changes, explaining how pS129 couldcontinue to increase, and become associated with worseoutcomes in patients with more advanced disease.to consider that biochemical and clinical measures of dis-samples in addition to two cross-sectional cohorts, weCompliance with ethical standardsStewart et al. Acta Neuropathologica Communications  (2015) 3:7 Page 9 of 11Regardless of the mechanisms involved, however, theseobservations might provide some insight to the conflict-ing results obtained in various animal models of PD. Forexample, experiments in fly [10] and rat models [11,15]found opposite effects of phosphomimetic mutant α-synon neuronal degeneration, and differing effects of phos-phorylation on the propensity of α-syn to form aggre-gates have also been reported [6,10,11,16,35]. Further,differences have been observed between the phosphomi-metic and genuine phosphorylated proteins [16,35],complicating interpretation of these studies. If alter-ations in pS129 depend on the stage of the disease, it islikely that other disease-related molecular interactionsinfluence the effects of pS129 in the cell, and, for futurestudies, the effects of pS129 must be investigated inmodel systems that recapitulate the human condition asclosely as possible. An important caveat that must beconsidered is that, while patients in the early/preclinicalcohorts (DATATOP and LRRK2) were untreated, thosein the multi-center collaborative cohort, in addition tobeing at more advanced stages, are also undergoingtherapeutic treatment. With these datasets, one couldargue that the differing relationship between disease se-verity and pS129 with progression is driven by effects ofdrug treatments (for example, treatment could result inincreased pS129 levels, such that those with worsesymptoms requiring increased therapy exhibit the high-est levels, and masking a negative relationship such aswas observed in both the DATATOP and LRRK2 co-horts). While this remains a theoretical possibility, thefact that a further increase in pS129 occurred in the lon-gitudinal DATATOP cohort (Table 2), where all subjectsare unmedicated, indicates PD therapy is unlikely theprimary cause of this observation.Additional considerationsSeveral additional caveats must also be considered. Oneobvious concern is the lack of neurologically normalcontrols in the DATATOP study, meaning that no groupexists for comparison of PD-related changes in pS129with its natural course with aging. Therefore, whether theincreasing levels of pS129 observed in the DATATOP co-hort are due to PD progression or aging cannot be defini-tively determined by this dataset alone. However, somedata exists to suggest that aging is not the primary factor.First, we previously examined this question in our multi-center collaborative cohort, and found no relationshipbetween age and pS129 in older PD patients or controlsubjects [31]. Further, we examined the cross-sectional re-lationship between pS129 and age in the DATATOP co-hort, and again found no association. Together, these datasuggest that CSF pS129 is not dependent on age in theolder patients included here. However, this must be con-firmed in longitudinally collected control subjects in laterThe authors report no conflicts of interest. Protocols in-cluding human subjects were approved by all participat-ing institutions, and performed in accordance with theethical standards of the institutional research committeeand with the 1964 Helsinki declaration and its lateramendments or comparable ethical standards. Informedconsent was obtained from all subjects prior to anyprocedures.Competing interestsThe authors declare that they have no competing interests.Authors’ contributionsJZ conceived and supervised the project, and drafted the manuscript withTS, MS, and KCC TS assisted in experimental design and execution,performed analyses, and prepared figures. VS and AJS, were responsible forPET analysis. JOA, KKJ, ZKW, RJU, KH, TY, CPZ, and JBL were responsible forpatient characterization and sample collection. MS and CL participated instudy design and execution. CG and YW handled samples and performedLuminex and hemoglobin assays. PHJ provided reagents and contributed tomanuscript preparation. PA and UJK participated in data analysis. All authorscritically reviewed the manuscript. All authors read and approved the finalmanuscript.AcknowledgementsWe deeply appreciate the donation of CSF and participation by subjects inthis study. This work was supported by the Michael J Fox Foundation, theParkinson Study Group, and the NIH (NIA: R01 AG033398; NIEHS: P42gression. These observations, though obtained in a largecohort, need to be further validated in an independentinvestigation, e.g. in prospective studies like the ongoingParkinson’s Progression Markers Initiative. The signifi-cance of this study also goes beyond a biomarker re-search, informing future mechanistic studies of thedisease, which should consider this natural course ofpS129 in humans when modeling PD experimentally.provide evidence for pathological alterations of pS129,along with total α-syn, in the natural course of PD pro-ease reflect different things: CSF α-syn and pS129 aremeasures of brain-wide pathology while UPDRS motorscore reflects largely the degeneration of nigrostriatal sys-tem. Despite these caveats, sensitive and objective detec-tion of alterations in CSF pS129 and total α-syn, whichboth alter over two years of progression, [20] wouldgreatly aid clinical trials of novel, disease-modifying treat-ments targeting α-syn-related systems.In summary, in this study of pS129 using longitudinalstudies. Additionally, when using UPDRS to assess theutility of CSF pS129 and/or pS129/α-syn in monitoringthe central nervous systems of PD patients, it is importantES004696-5897, P30 ES007033-6364, R01 ES016873, R01 ES019277 and T32ES015459; NINDS: R01 NS057567, R01 NS065070, P50 NS062684-6221, P50NS072187, and U01 NS082137).Stewart et al. Acta Neuropathologica Communications  (2015) 3:7 Page 10 of 11Author details1Department of Pathology, University of Washington School of Medicine, 3259th Avenue, HMC Box 359635, Seattle, WA 98104, USA. 2Department of Physicsand Astronomy, University of British Columbia, Vancouver Hospital and HealthSciences Centre, Vancouver, BC, Canada. 3Department of Neurology, St. OlavsHospital, Trondheim, Norway. 4Department of Neurology, Mayo Clinic Florida,Jacksonville, FL, USA. 5Department of Neurology, National Hospital Organization,Sagamihara National Hospital, Kanagawa, Japan. 6Geriatric Research, Educationand Clinical Center, Veterans Affairs Puget Sound Health Care System, Seattle,WA, USA. 7Parkinson’s Disease Research, Education and Clinical Center, VeteransAffairs Puget Sound Health Care System, Seattle, WA, USA. 8Department ofNeurology, University of Washington School of Medicine, Seattle, WA, USA.9Department of Psychiatry and Behavioral Sciences, University of WashingtonSchool of Medicine, Seattle, WA, USA. 10Mental Illness Research, Education andClinical Center, Veterans Affairs Puget Sound Health Care System, Seattle, WA,USA. 11Pacific Parkinson’s Research Centre, University of British Columbia andVancouver Coastal Health, Vancouver, BC, Canada. 12Department ofNeurosurgery, Tongji Hospital, Tongji Medical College, Huazhong University ofScience and Technology, Wuhan, Hubei 430030, China. 13Department ofEndocrinology and Metabolism and Xiamen Diabetes Institute, the FirstAffiliated Hospital of Xiamen University, Xiamen, China. 14Department ofBiostatistics, University of Washington School of Public Health, Seattle, WA, USA.15Department of Neurology, Center for Human Experimental Therapeutics,University of Rochester School of Medicine and Dentistry, Rochester, NY, USA.16Department of Neurology, University of Chicago Medicine and BiologicalSciences, Chicago, IL, USA. 17Department of Biomedicine, Aarhus University, OleWorms alle 1170, DK-8000 Aarhus-C, Denmark.Received: 26 November 2014 Accepted: 6 January 2015References1. Burbulla LF, Kruger R. Converging environmental and genetic pathways inthe pathogenesis of Parkinson’s disease. J Neurol Sci. 2011;306:1–8.doi:10.1016/j.jns.2011.04.005.2. Valente EM, Arena G, Torosantucci L, Gelmetti V. Molecular pathwaysin sporadic PD. Parkinsonism Relat Disord. 2012;18 Suppl 1:S71–3.doi:10.1016/s1353-8020(11)70023-2.3. Karpinar DP, Balija MB, Kugler S, Opazo F, Rezaei-Ghaleh N, Wender N, et al.Pre-fibrillar alpha-synuclein variants with impaired beta-structure increaseneurotoxicity in Parkinson’s disease models. EMBO J. 2009;28:3256–68.doi:10.1038/emboj.2009.257.4. Winner B, Jappelli R, Maji SK, Desplats PA, Boyer L, Aigner S, et al. In vivodemonstration that alpha-synuclein oligomers are toxic. Proc Natl Acad SciU S A. 2011;108:4194–9. doi:10.1073/pnas.1100976108.5. Anderson JP, Walker DE, Goldstein JM, de Laat R, Banducci K, Caccavello RJ,et al. Phosphorylation of Ser-129 is the dominant pathological modificationof alpha-synuclein in familial and sporadic Lewy body disease. J Biol Chem.2006;281:29739–52. doi:10.1074/jbc.M600933200.6. Fujiwara H, Hasegawa M, Dohmae N, Kawashima A, Masliah E, Goldberg MS,et al. alpha-Synuclein is phosphorylated in synucleinopathy lesions. Nat CellBiol. 2002;4:160–4. doi:10.1038/ncb748.7. Walker DG, Lue LF, Adler CH, Shill HA, Caviness JN, Sabbagh MN, et al.Changes in properties of serine 129 phosphorylated alpha-synuclein withprogression of Lewy-type histopathology in human brains. Exp Neurol.2013;240:190–204. doi:10.1016/j.expneurol.2012.11.020.8. Saito Y, Kawashima A, Ruberu NN, Fujiwara H, Koyama S, Sawabe M, et al.Accumulation of phosphorylated alpha-synuclein in aging human brain.J Neuropathol Exp Neurol. 2003;62:644–54.9. Covy JP, Yuan W, Waxman EA, Hurtig HI, Van Deerlin VM, Giasson BI. Clinicaland pathological characteristics of patients with leucine-rich repeat kinase-2mutations. Mov Disord. 2009;24:32–9. doi:10.1002/mds.22096.10. Chen L, Feany MB. Alpha-synuclein phosphorylation controls neurotoxicityand inclusion formation in a Drosophila model of Parkinson disease. NatNeurosci. 2005;8:657–63. doi:10.1038/nn1443.11. Gorbatyuk OS, Li S, Sullivan LF, Chen W, Kondrikova G, Manfredsson FP,et al. The phosphorylation state of Ser-129 in human alpha-synucleindetermines neurodegeneration in a rat model of Parkinson disease. ProcNatl Acad Sci U S A. 2008;105:763–8. doi:10.1073/pnas.0711053105.12. Kragh CL, Lund LB, Febbraro F, Hansen HD, Gai WP, El-Agnaf O, et al.Alpha-synuclein aggregation and Ser-129 phosphorylation-dependent celldeath in oligodendroglial cells. J Biol Chem. 2009;284:10211–22.doi:10.1074/jbc.M809671200.13. Lou H, Montoya SE, Alerte TN, Wang J, Wu J, Peng X, et al. Serine 129phosphorylation reduces the ability of alpha-synuclein to regulate tyrosinehydroxylase and protein phosphatase 2A in vitro and in vivo. J Biol Chem.2010;285:17648–61. doi:10.1074/jbc.M110.100867.14. Pronin AN, Morris AJ, Surguchov A, Benovic JL. Synucleins are a novel classof substrates for G protein-coupled receptor kinases. J Biol Chem.2000;275:26515–22. doi:10.1074/jbc.M003542200.15. McFarland NR, Fan Z, Xu K, Schwarzschild MA, Feany MB, Hyman BT, et al.Alpha-synuclein S129 phosphorylation mutants do not alter nigrostriataltoxicity in a rat model of Parkinson disease. J Neuropathol Exp Neurol.2009;68:515–24. doi:10.1097/NEN.0b013e3181a24b53.16. Sato H, Arawaka S, Hara S, Fukushima S, Koga K, Koyama S, et al.Authentically phosphorylated alpha-synuclein at Ser129 acceleratesneurodegeneration in a rat model of familial Parkinson’s disease. J Neurosci.2011;31:16884–94. doi:10.1523/jneurosci.3967-11.2011.17. Wang Y, Shi M, Chung KA, Zabetian CP, Leverenz JB, Berg D, et al.Phosphorylated alpha-synuclein in Parkinson’s disease. Sci Transl Med.2012;4:121ra120. doi:10.1126/scitranslmed.3002566.18. Aasly JO, Shi M, Sossi V, Stewart T, Johansen KK, Wszolek ZK, et al.Cerebrospinal fluid amyloid beta and tau in LRRK2 mutation carriers.Neurology. 2012;78:55–61. doi:10.1212/WNL.0b013e31823ed101.19. Shi M, Furay AR, Sossi V, Aasly JO, Armaly J, Wang Y, et al. DJ-1 andalphaSYN in LRRK2 CSF do not correlate with striatal dopaminergicfunction. Neurobiol Aging. 2012;33(836):e835–7.doi:10.1016/j.neurobiolaging.2011.09.015.20. Stewart T, Liu C, Ginghina C, Cain KC, Auinger P, Cholerton B, et al.Cerebrospinal fluid alpha-synuclein predicts cognitive decline in Parkinsondisease progression in the DATATOP cohort. Am J Pathol. 2014;184:966–75.doi:10.1016/j.ajpath.2013.12.007.21. Investigators TPSGD. DATATOP: a multicenter controlled clinical trial in earlyParkinson’s disease. Parkinson Study Group Archives of neurology.1989;46:1052–60.22. Zhang J, Mattison HA, Liu C, Ginghina C, Auinger P, McDermott MP, et al.Longitudinal assessment of tau and amyloid beta in cerebrospinalfluid of Parkinson disease. Acta Neuropathol. 2013;126:671–82.doi:10.1007/s00401-013-1121-x.23. Aarsland D, Andersen K, Larsen JP, Lolk A, Kragh-Sorensen P. Prevalence andcharacteristics of dementia in Parkinson disease: an 8-year prospectivestudy. Arch Neurol. 2003;60:387–92.24. Hely MA, Reid WG, Adena MA, Halliday GM, Morris JG. The Sydneymulticenter study of Parkinson’s disease: the inevitability of dementia at20 years. Mov Disord. 2008;23:837–44. doi:10.1002/mds.21956.25. Fjorback AW, Varming K, Jensen PH. Determination of alpha-synucleinconcentration in human plasma using ELISA. Scand J Clin Lab Invest.2007;67:431–5. doi:10.1080/00365510601161497.26. Lindersson E, Beedholm R, Hojrup P, Moos T, Gai W, Hendil KB, et al.Proteasomal inhibition by alpha-synuclein filaments and oligomers. J BiolChem. 2004;279:12924–34. doi:10.1074/jbc.M306390200.27. Waxman EA, Giasson BI. Specificity and regulation of casein kinase-mediatedphosphorylation of alpha-synuclein. J Neuropathol Exp Neurol. 2008;67:402–16.doi:10.1097/NEN.0b013e31816fc995.28. Hong Z, Shi M, Chung KA, Quinn JF, Peskind ER, Galasko D, et al. DJ-1 andalpha-synuclein in human cerebrospinal fluid as biomarkers of Parkinson’sdisease. Brain. 2010;133:713–26. doi:10.1093/brain/awq008.29. Shi M, Bradner J, Hancock AM, Chung KA, Quinn JF, Peskind ER, et al.Cerebrospinal fluid biomarkers for Parkinson disease diagnosis andprogression. Ann Neurol. 2011;69:570–80. doi:10.1002/ana.22311.30. Foulds PG, Yokota O, Thurston A, Davidson Y, Ahmed Z, Holton J, et al. Postmortem cerebrospinal fluid alpha-synuclein levels are raised in multiplesystem atrophy and distinguish this from the other alpha-synucleinopathies,Parkinson’s disease and Dementia with Lewy bodies. Neurobiol Dis.2012;45:188–95. doi:10.1016/j.nbd.2011.08.003.31. Zetterberg H, Smith DH, Blennow K. Biomarkers of mild traumatic braininjury in cerebrospinal fluid and blood. Nat Rev Neurol. 2013;9:201–10.doi:10.1038/nrneurol.2013.9.32. Foulds PG, Mitchell JD, Parker A, Turner R, Green G, Diggle P, et al.Phosphorylated alpha-synuclein can be detected in blood plasma and ispotentially a useful biomarker for Parkinson’s disease. FASEB J. 2011;25:4127–37.doi:10.1096/fj.10-179192.33. Foulds PG, Diggle P, Mitchell JD, Parker A, Hasegawa M, Masuda-SuzukakeM, et al. A longitudinal study on alpha-synuclein in blood plasmaas a biomarker for Parkinson’s disease. Sci Rep. 2013;3:2540.doi:10.1038/srep02540.34. Hsiao IT, Weng YH, Hsieh CJ, Lin WY, Wey SP, Kung MP, et al. Correlation ofParkinson disease severity and 18 F-DTBZ positron emission tomography.JAMA Neurol. 2014;71:758–66. doi:10.1001/jamaneurol.2014.290.35. Paleologou KE, Schmid AW, Rospigliosi CC, Kim HY, Lamberto GR,Fredenburg RA, et al. Phosphorylation at Ser-129 but not the phosphomimicsS129E/D inhibits the fibrillation of alpha-synuclein. J Biol Chem.2008;283:16895–905. doi:10.1074/jbc.M800747200.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionStewart et al. Acta Neuropathologica Communications  (2015) 3:7 Page 11 of 11Submit your manuscript at www.biomedcentral.com/submit


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items