- Email: [email protected]
Validation of a commercially available enzyme-linked immunoabsorbent assay for the quantification of human α-Synuclein in cerebrospinal fluid
Journal of Immunological Methods 426 (2015) 70–75
Contents lists available at ScienceDirect
Journal of Immunological Methods journal homepage: www.elsevier.com/locate/jim
Research paper
Validation of a commercially available enzyme-linked immunoabsorbent assay for the quantification of human α-Synuclein in cerebrospinal fluid Niels Kruse a,⁎, Brit Mollenhauer a,b a b
Department of Neuropathology, University Medical Center, Robert-Koch-Str. 40, 37075 Göttingen, Germany Paracelsus-Elena-Klinik, Klinikstr. 16, 34128 Kassel, Germany
a r t i c l e
i n f o
Article history: Received 15 May 2015 Received in revised form 10 July 2015 Accepted 4 August 2015 Available online 10 August 2015 Keywords: alpha-Synuclein ELISA Validation
a b s t r a c t The quantification of α-Synuclein in cerebrospinal fluid (CSF) as a biomarker has gained tremendous interest in the last years. Several commercially available immunoassays are emerging. We here describe the full validation of one commercially available ELISA assay for the quantification of α-Synuclein in human CSF (Covance alpha-Synuclein ELISA kit). The study was conducted within the BIOMARKAPD project in the European initiative Joint Program for Neurodegenerative Diseases (JPND). We investigated the effect of several pre-analytical and analytical confounders: i.e. (1) need for centrifugation of freshly drawn CSF, (2) sample stability, (3) delay of freezing, (4) volume of storage aliquots, (5) freeze/thaw cycles, (6) thawing conditions, (7) dilution linearity, (8) parallelism, (9) spike recovery, and (10) precision. None of these confounders influenced the levels of α-Synuclein in CSF significantly. We found a very high intra-assay precision. The inter-assay precision was lower than expected due to different performances of kit lots used. Overall the validated immunoassay is useful for the quantification of α-Synuclein in human CSF. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Levels of α-Synuclein (aSyn) in cerebrospinal fluid (CSF) in Parkinson's disease and related synucleinopathies have been analyzed in several studies (for review see Sako et al., 2014). However, not all of these studies reached the same conclusions. The reason for variable results relies largely on the use of home-made ELISA assays using different antibody combinations (Mollenhauer et al., 2010). In a recent ring trial we investigated the performance of a newly released commercially available human specific ELISA (Covance alpha-Synuclein ELISA kit) to quantify aSyn in cerebrospinal fluid in 18 different laboratories (Kruse et al., 2015). In this study pre-analytical confounders were largely excluded through the centrally provided samples. In the current manuscript we extended this initial validation study at one center and analyzed the effect of various preanalytical and analytical manipulations on the aSyn quantification with this assay. This study was conducted in accordance to the standard operating procedure developed within the BIOMARKAPD consortium (Andreasson et al., in press). In particular we analyzed (1) the need for centrifugation of freshly drawn CSF, (2) sample stability under various conditions, (3) delay of freezing, (4) volume of storage aliquots, (5) freeze/thaw cycles, (6) thawing conditions, (7) dilution linearity, (8) parallelism, (9) spike recovery, and (10) precision of replicate ⁎ Corresponding author. E-mail addresses: [email protected] (N. Kruse), [email protected] (B. Mollenhauer).
http://dx.doi.org/10.1016/j.jim.2015.08.003 0022-1759/© 2015 Elsevier B.V. All rights reserved.
aSyn quantification in independent experiments. The most important outcome is that aSyn seems to be very stable under all conditions applied. 2. Material and methods 2.1. Collection of CSF samples For validation experiments randomly selected CSF samples were used. Details on sample processing and diagnostic criteria were published recently (Mollenhauer et al., 2011). Our study was done in accordance with the Declaration of Helsinki and with informed written consent provided by all patients or by their next of kin in the case of cognitive impairment (mini-mental status examination score 23 of 25 or below). Our study was approved by the ethics committees at the University of Göttingen and the Landesärztekammer Hessen (Germany). CSF samples had no elevation in red or white blood cell counts and semiquantitative hemoglobin levels. 2.2. Preparation of CSF samples Preparation of CSF samples has been described recently (Mollenhauer et al., 2011). Depending on the CSF volume available after performing routine diagnostics CSF aliquots (except for Aliquot volume samples (see Section 2.2.4 below)) were prepared at 375 μl or 60 μl in 0.5 ml polypropylene reaction tubes (Sarstedt, Nümbrecht, Germany; Ref.# 72.699). For most of the variables to be analyzed
N. Kruse, B. Mollenhauer / Journal of Immunological Methods 426 (2015) 70–75
71
Fig. 3. Delay of freezing has no effect on quantification of CSF aSyn.
Fig. 1. No effect of different centrifugation conditions on CSF aSyn quantification results.
(centrifugation, precision, stability, freeze/thaw cycles, delay of freezing, and thawing conditions), multiple 60 μl volumes of CSF were sufficient, while larger amounts were necessary of other experiments (aliquot volumes, dilution linearity, and parallelism). After preparation all CSF aliquots were stored at − 80 °C until analysis. 2.2.1. Centrifugation samples Three different aliquots were prepared from CSF obtained from six donors. (1) One aliquot was left after lumbar puncture (LP) without centrifugation. (2) One aliquot was centrifuged at 1600 ×g and room temperature (RT) for 10 min to obtain the RT samples and (3) the remaining CSF was centrifuged at 1600 ×g at 4 °C.
2.2.4. Aliquot volume samples CSF samples (n = 7 donors) were prepared at 125 μl, 250 μl, and 375 μl in 500 μl tubes (corresponding to 25%, 50%, and 75% of the storage tube volumes) and left for one year at −80 °C before analysis. 2.2.5. Freeze/thaw cycle samples Three aliquots were thawed at RT for 60 min and refrozen at −80 °C. After at least 12 h at − 80 °C this procedure was repeated twice on two aliquots and another two times on one aliquot, respectively. CSF samples from eight donors were used for this purpose and generated samples being thawed up to five times before final thawing for analysis. 2.2.6. Thawing conditions Aliquots from four different CSF samples were thawed under different conditions: either rapidly by keeping them in the hand or thawing them for up to 3 h on ice, at 5–8 °C in a refrigerator or at RT before performing the assay.
2.2.2. Stability samples Nine aliquots were prepared to test CSF stability under different conditions. CSF from seven donors was used for this purpose. Aliquot 1 was left at − 80 °C until analysis. Aliquots 2 were left at 5–8 °C for 24 h before being frozen at − 80 °C. Aliquots 3 were left at 5–8 °C for one week before being frozen at −80 °C. Aliquots 4 were left at RT for 24 h before being frozen at −80 °C. Aliquots 5 were left at −20 °C for one month before being frozen at −80 °C.
2.2.7. Dilution linearity samples CSF samples (n = 6 donors) were spiked with approximately 100-times the endogenous concentration of aSyn and then serially diluted two-fold before analysis.
2.2.3. Delay of freezing samples After lumbar puncture CSF aliquots (n = 5 donors) were left for 2 h, 24 h, 48 h, and five days at RT or 4 °C before being frozen at −80 °C.
2.2.9. Spike recovery samples Aliquots of CSF samples (n = 5 donors) were spiked with aSyn standards at final concentrations of 1000 pg/ml, 500 pg/ml, and
Fig. 2. High reproducibility of CSF aSyn quantification under various storage conditions.
Fig. 4. No effect of storage aliquot volume on CSF aSyn quantification.
2.2.8. Parallelism samples CSF samples (n = 5 donors) were serially diluted two-fold for analysis starting with non-diluted CSF as the highest concentration.
72
N. Kruse, B. Mollenhauer / Journal of Immunological Methods 426 (2015) 70–75
version (Mollenhauer et al., 2008). In brief, CSF samples were diluted 1:10 with 1 × Reagent Diluent for analysis. Plates were washed four times with 300 μl wash buffer. aSyn standards and CSF samples were applied at 200 μl volumes to reaction wells. The reaction plate was incubated overnight at 4 °C without shaking. The next day the content of the wells was discarded and plates were washed as described above. 50 μl biotinylated primary antibody was added at 1000-fold dilution of the stock solution in 1 × Reagent Diluent. Plates were then incubated for 2 h at RT with shaking at 300 rpm. After washing again, 200 μl streptavidin-coupled horse-radish peroxidase, diluted 15,000-fold in 1 × Reagent Diluent, was added and incubated for 1 h at RT with shaking at 300 rpm. Plates were washed again, 100 μl chemiluminescence substrate was added and plates were read in a Tecan Infinite 200Pro electrochemiluminescence reader. Experiments were performed using kit lots #D13IC02685 and D14IK01932. Fig. 5. No effect of multiple freeze/thaw cycles on CSF aSyn quantification.
250 pg/ml. Spiking was performed by adding one volume of standards to nine volumes of CSF. Endogenous protein concentrations were determined by spiking one volume Reagent Diluent to nine volumes CSF to account for the dilution factor. Furthermore, nine volumes of Reagent Diluent were spiked with one volume of protein standards to confirm spiked protein concentrations. 2.2.10. Precision samples 24 samples were prepared from CSF obtained from seven donors for analysis of intra-assay and inter-assay precision. 12 aliquots from individual CSF samples were measured in one experiment followed by quantification of aSyn in three additional aliquots in four independent experiments. Intra-assay precision refers to the comparability of results performed on the same day and inter-assay variability refers to the comparability of results from independent experiments performed on different days.
2.4. Data analysis Standards and samples were analyzed in duplicates. Mean and standard deviations were calculated for each duplicate. Background values were subtracted from standard and sample readings. Standard curves were plotted using Microsoft Excel 2003. Resulting curve parameters of linear regression curves were used to calculate aSyn concentrations in individual samples, as is considered acceptable according to the kit insert. Concentrations were adjusted for dilution factors. Results from duplicates with b15% SD were considered acceptable. Reliable results were defined as those with CVs for replicate determinations of analyte concentration of b20%. In some analysis aSyn concentrations were normalized for clarity. Normalization was either done for the dilution factor (Figs. 7 and 8) or to samples thawed quickly (Fig. 6). 2.5. Calculations All calculations were done using Microsoft Excel software.
2.3. Quantification of aSyn
3. Results
A general outline of the validation strategy including documentation of results will be presented elsewhere (Andreasson et al., in press). Determination of aSyn concentrations in CSF was done with one commercially available alpha-Synuclein ELISA kit according to the manufacturer's instruction (Covance, Dedham, MA, USA; Cat.-# SIG38974) (Mollenhauer et al., 2013). This assay is based on a former
In this study we performed a full validation of the Covance Alpha-Synuclein ELISA kit. According to the kit instructions aSyn concentrations down to 6.1 pg/ml can be quantified. In a previous European-wide study (Kruse et al., 2015) reliable quantification over the entire detection range of the standard curve was confirmed. Our data agreed to those reported by Kang et al. (2013) and were
Fig. 6. Different thawing conditions do not influence quantification results of aSyn in CSF.
N. Kruse, B. Mollenhauer / Journal of Immunological Methods 426 (2015) 70–75
73
reproduced in this study. All aSyn concentrations measured in this study were significantly higher than the lowest concentration of the standard curve and we therefore conclude that these were reliable. 3.1. Centrifugation Results of our different centrifugation conditions are shown in Fig. 1 and demonstrate that quantification under all conditions gave similar results. When concentrations were normalized to the concentration of the reference treatment (centrifugation at 4 °C) none of the samples deviated by more than 20% (range: 1.7–13.5%). When mean concentrations were calculated for all three conditions CVs varied between 1.23% and 6.41%. We therefore conclude that centrifugation has no effect on the aSyn concentration determined in these samples without artificial blood contamination. Fig. 8. Parallelism of serially diluted CSF samples and standards.
3.2. Stability Stability of aSyn in CSF was determined under various conditions as described in the Materials and methods section. These analyses clearly demonstrate that nearly all samples were quantified with similar results for individual CSF samples. With the exception of two aliquots (sample 1, 24 h room temperature and one month − 20 °C) all quantification results were in the expected range (Fig. 2). These results indicate that CSF may be stored for extended time periods at temperatures other than −80 °C, what is usually recommended. 3.3. Delay of freezing After having demonstrated that CSF samples may be stored for extended time period at room temperature or 5–8 °C we looked at further time points for storage at either room temperature and 5–8 °C before they were frozen at − 80 °C. Under all conditions we found highly consistent aSyn concentrations indicating that aSyn in CSF is very stable. CVs were 10% or less for all CSF samples (Fig. 3). 3.4. Aliquot volume Due to a potential increased concentration of analyte by evaporation of solvent we investigated if this occurs with aSyn in CSF by preparing three different aliquot volumes (25%, 50%, and 75% of the maximal vial volume) of CSF. This procedure also allowed us to investigate if absorption is an issue when variable storage volumes are used. For 7 different CSF samples analyzed after storage for approximately one year at − 80 °C we did not observe significant changes in aSyn
Fig. 7. Dilution linearity of serially diluted CSF samples after spiking with high concentration of aSyn.
concentration (Fig. 4). Mean values calculated from all three volumes from each CSF samples exhibited CV values below 20% (range: 1.6–14.3%). 3.5. Freeze/thaw cycles The effect of multiple freeze/thaw cycles was negligible: with up to five freeze/thaw cycles eight CSF samples showed highly comparable aSyn concentrations among themselves (Fig. 5). CVs from comparison of three aliquots per CSF sample after increasing numbers of freeze/thaw cycles were 1.2–11.0%. 3.6. Thawing conditions Different thawing procedures did not impair quantification of aSyn as compared to quickly frozen aliquots used for normalization. Even after leaving CSF samples for up to 3 h at room temperature all concentrations were determined to be within 20% of the reference aliquots (Fig. 6). 3.7. Dilution linearity Fig. 7 shows the results from experiments performed to determine the dilution linearity. aSyn concentrations were normalized for the results of the aliquots diluted 1:8 as this dilution closely matched the 1:10 dilution recommended in the assay instructions. We found that aliquots diluted four-fold to 32-fold gave comparable results. Most concentrations were within 20% of that of the reference aliquot. Only
Fig. 9. Reproducibility of CSF aSyn quantification in independent experiments.
74
N. Kruse, B. Mollenhauer / Journal of Immunological Methods 426 (2015) 70–75
4. Discussion
Table 1 Intra-assay and inter-assay precision. CSF sample
Intra-assay precision
Inter-assay precision
CSF 2 CSF 3 CSF 4a CSF 5a CSF 9 CSF 10a CSF 11a
4.2% 5.3% 4.2% 5.0% 4.5% 2.6% 3.7%
13.0% 10.8% 24.0% 29.3% 20.3% 64.9% 29.8%
a
Experiments on these CSF samples were performed with different kit lots.
one CSF sample had slightly larger variability (CSF 5 with 78% at four-fold dilution and 123% at 32-fold dilution). 3.8. Parallelism Parallelism of serially diluted CSF samples is demonstrated in Fig. 8. Normalization of aSyn concentrations was again relative to those obtained for the 1:8 dilutions. We found that CSF diluted 1:4 to 1:16 in most cases gave reliable results. 3.9. Spike recovery aSyn concentrations were measured in CSF aliquots spiked with three different concentrations of aSyn standards (i.e. 1000 pg/ml, 500 pg/ml and 250 pg/ml). The concentration of the spike solutions and the endogenous concentration of the CSF samples were measured in parallel. After subtraction of the endogenous aSyn concentration mean recovery rates were calculated to be 70.1 ± 4.1% (range: 66.6–78.7%) (data not shown). 3.10. Precision Fig. 9 and Table 1 show information for intra-assay precision and inter-assay precision from 24 aliquots of seven different samples in five independent experiments. Intra-assay precision was very robust with CVs between 2.6 and 5.3%. Inter-assay precision was more inconsistent with variation between 10.8 and 20.3% for one lot and 24.0–64.9% for two different lots (Table 1). The reason for this observation is that these analyses were performed on ELISA kits from two different lots (Fig. 10). Compared to ELISA kit lot 1 much higher signal values were detected for individual standard concentrations. At similar signal readings for individual samples this shift of the standard curve therefore lead to lower concentration assignments.
Knowledge on pre-analytical and analytical confounders on quantification of CSF biomarkers is of importance as large multicenter biomarker studies need to compare measurements across different centers with expected variation and to minimize further avoidable source of variability. In this study we performed a comprehensive validation of one ELISA assay for the quantification of human aSyn in CSF. For other biomarkers, mainly in the field of Alzheimer's disease (AD) similar studies have been enormous helpful to design biomarker studies and to compare studies form different centers (Mattsson et al., 2013). The biomarker field in PD is lagging behind. Some characteristics of CSF aSyn are similar to beta amyloid (such as the lipophilic nature and in consequence the propensity to stick to certain tube and pipette tip material) but some characteristics are particular in CSF aSyn from other analytes, e.g. the high abundance on erythrocytes. To our knowledge this is the first publication analyzing various pre-analytical and analytical confounders related to aSyn quantification. We found that none of the confounders analyzed in this study did negatively influence the quantitative results. This indicates that aSyn is a very stable protein and robust for multicenter assessments. None of the potential sources of variation, i.e. the need for centrifugation of freshly drawn CSF, sample stability under various conditions, delay of freezing at − 80 °C, volume of storage aliquots, repeated freeze/thaw cycles, thawing conditions, precision of replicate aSyn quantification in independent experiments, impaired quantification of aSyn in CSF and showed low variation. Adsorption of lipophilic proteins to surfaces of storage vials may result in reduced detectability of these respective proteins (Lewczuk et al., 2006). According to recent recommendations (del Campo et al., 2012) we stored CSF samples in polypropylene tube to minimize protein adsorption. Our experiments using different aliquot storage volumes clearly showed that neither adsorption nor different storage volumes had an influence on aSyn quantification. According to the defined criteria of the BIOMARKAPD consortium recovery of spiked analyte should be in the range of 80–120% (Andreasson et al., in press). We found slightly lower recoveries with 70% on average and therefore this is the only variable for which the aSyn ELISA did not perform as recommended. Nevertheless as the recovery rates were comparable over the entire range of the spiked aSyn concentrations we feel that these results will allow for reliable and comparable quantification of aSyn in CSF (Lee et al., 2006; Food and Drug Administration, 2001). The only cause of variations identified is usage of different ELISA kit lots. We noticed that the second lot used in this study gave significantly lower concentration assignments than the first lot. The reason for this
Fig. 10. Standard curves from two different ELISA kit production lots perform differently.
N. Kruse, B. Mollenhauer / Journal of Immunological Methods 426 (2015) 70–75
observation is that with the second lot of the ELISA higher signal reading were obtained than with the first lot. At similar signal readings for individual samples this shift of the standard curve therefore lead to lower concentration assignments. This observation indicates that internal quality control samples of known aSyn concentration should be run in parallel with samples to be analyzed. These can either be home-made or better should be included in the kit. Similar lot-to-lot inconsistencies have been found in several previous studies using assays for other biomarkers (Hoefner and Yeo, 2002; Teunissen et al., 2010; Vos et al., 2014). Erythrocytes express high level of aSyn (Barbour et al., 2008) which may be released from the cells upon destruction due to freezing. Since the CSF samples for our experiments were devoid of elevated levels of red and white blood cells as well as hemoglobin it is not unexpected that centrifugation did not show any effect on aSyn quantitative results. Nevertheless we recommend centrifugation of CSF samples to remove minor contamination of erythrocytes from CSF. In a recent study performed with the same aSyn ELISA Heegaard et al. (Heegaard et al., 2014) induced some analytical modifications leading to reduced background without changing the detection limit. This observation indicates that further optimization of the assay procedure might lead to improved performance of the aSyn ELISA. Inclusion of internal quality control samples will help to verify accurate performance of test runs. Further round robin test and automation of assay procedure will certainly leads to widespread acceptance of this new aSyn ELISA. Acknowledgments The authors want to thank Birgit Otte for the excellent technical assistance. This is an EU Joint Programme — Neurodegenerative Disease Research (JPND) project. The project is supported through the following funding organizations under the aegis of JPND — www.jpnd.eu. References Andreasson, U., Perret-Liaudet, A., van Waalwijk van Doorn, L., Blennow, K., Chiasserini, D., Engelborghs, S., Fladby, T., Genc, S., Kruse, N., Kuiperij, B., Kulic, L., Lewczuk, P., Mollenhauer, B., Mroczko, B., Parnetti, L., Vanmechelen, E., Verbeek, M., Winblad, B., Zetterberg, H., Koel-Simmelink, M., Teunissen, C.E., 2015. A practical guide to immunoassay method validation. Frontiers in Neurology (in press). Barbour, R., Kling, K., Anderson, J.P., Banducci, K., Cole, T., Diep, L., Fox, M., Goldstein, J.M., Soriano, F., Seubert, P., Chilcote, T.J., 2008. Red blood cells are the major source of alpha-synuclein in blood. Neurodegener. Dis. 5 (2), 55. http://dx.doi.org/10.1159/ 000112832. Del Campo, M., Mollenhauer, B., Bertolotto, A., Engelborghs, S., Hampel, H., Simonsen, A.H., Kapaki, E., Kruse, N., Le Bastard, N., Lehmann, S., Molinuevo, J.L., Parnetti, L., Perret-Liaudet, A., Saez-Valero, J., Saka, E., Urbani, A., Vanmechelen, E., Verbeek, M., Visser, P.J., Teunissen, C., 2012. Recommendations to standardize preanalytical confounding factors in Alzheimer's and Parkinson's disease cerebrospinal fluid biomarkers: an update. Biomark. Med 6 (4), 419. Food and Drug Administration, 2001. Guidance for Industry: Bioanalytical Method Validation. U.S. Department of Health and Human Services (http://www.fda.gov/ downloads/Drugs/…/Guidances/ucm070107.pdf). Heegaard, N.H.H., Tanassi, J.T., Bech, S., Salvesen, L., Jensen, P.L., Montezinho, L.C.P., Winge, K., 2014. CSF-α-synuclein in the differential diagnosis of parkinsonian syndromes. Future Neurol. 9 (5), 525. http://dx.doi.org/10.221/FNL.14.51.
75
Hoefner, D.M., Yeo, K.-T.J., 2002. Lot-to-lot inconsistency of anticardiolipin reagents. Clin. Chem. 48 (9), 1625. Kang, J.H., Irwin, D.J., Chen-Plotkin, A.S., Siderowf, A., Caspell, C., Coffey, C.S., Waligorska, T., Taylor, P., Pan, S., Frasier, M., Marek, K., Kieburtz, K., Jennings, D., Simuni, T., Tanner, C.M., Singleton, A., Toga, A.W., Chowdhury, S., Mollenhauer, B., Trojanowski, J.Q., Shaw, L.M., 2013. Association of cerebrospinal fluid beta-amyloid 1–42, T-tau, P-tau181, and alpha-synuclein levels with clinical features of drug-naive patients with early Parkinson disease. JAMA Neurol. 70 (10), 1277. http://dx.doi.org/10. 1001/jamaneurol.2013.3861. Kruse, N., Persson, S., Alcolea, D., Bahl, J.M.C., Baldeiras, I., Capello, E., Chiasserini, D., Bocchio Chiavetto, L., Emersic, A., Engelborghs, S., Eren, E., Fladby, T., Frisoni, G., García-Ayllón, M.-S., Genc, S., Gkatzima, O., Heegaard, N.H.H., Janeiro, A.M., Kováčech, B., Kuiperij, H.B., Leitão, M.J., Lleó, A., Martins, M., Matos, M., Mollergard, H.M., Nobili, F., Öhrfelt, A., Parnetti, L., Resende de Oliveira, C., Rot, U., Sáez-Valero, J., Struyfs, H., Tanassi, J.T., Taylor, P., Tsolaki, M., Vanmechelen, E., Verbeek, M.M., Zilka, N., Blennow, K., Zetterberg, H., Mollenhauer, B., 2015. Validation of a quantitative cerebrospinal fluid alpha-synuclein assay in a European-wide interlaboratory study. Neurobiol. Aging 36 (9), 2587. http://dx.doi.org/10.1016/j. neurobiolaging.2015.05.003. Lee, J.W., Devanarayan, V., Barrett, Y.C., Weiner, R., Allinson, J., Fountain, S., Keller, S., Weinryb, I., Green, M., Duan, L., Rogers, J.A., Millham, R., O'Brien, P.J., Sailstad, J., Khan, M., Ray, C., Wagner, J.A., 2006. Fit-for-purpose method development and validation for successful biomarker measurement. Pharm. Res. 23 (2), 312. http://dx.doi.org/10.1007/s11095-005-9045-3. Lewczuk, P., Beck, G., Esselmann, H., Bruckmoser, R., Zimmermann, R., Fiszer, M., Bibl, M., Maler, J.M., Kornhuber, J., Wiltfang, J., 2006. Effect of sample collection tubes on cerebrospinal fluid concentrations of tau proteins and amyloid beta peptides. Clin. Chem. 52 (2), 332. Mattsson, N., Andreasson, U., Persson, S., Carrillo, M.C., Collins, S., Chalbot, S., Cutler, N., Dufour-Rainfray, D., Fagan, A.M., Heegaard, N.H., Robin Hsiung, G.Y., Hyman, B., Iqbal, K., Lachno, D.R., Lleo, A., Lewczuk, P., Molinuevo, J.L., Parchi, P., Regeniter, A., Rissman, R., Rosenmann, H., Sancesario, G., Schroder, J., Shaw, L.M., Teunissen, C.E., Trojanowski, J.Q., Vanderstichele, H., Vandijck, M., Verbeek, M.M., Zetterberg, H., Blennow, K., Kaser, S.A., 2013. CSF biomarker variability in the Alzheimer's Association quality control program. Alzheimers Dement. 9 (3), 251. http://dx.doi. org/10.1016/j.jalz.2013.01.010. Mollenhauer, B., Cullen, V., Kahn, I., Krastins, B., Outeiro, T.F., Pepivani, I., Ng, J., Schulz-Schaeffer, W., Kretzschmar, H.A., McLean, P.J., Trenkwalder, C., Sarracino, D.A., Vonsattel, J.P., Locascio, J.J., El-Agnaf, O.M., Schlossmacher, M.G., 2008. Direct quantification of CSF alpha-synuclein by ELISA and first cross-sectional study in patients with neurodegeneration. Exp. Neurol. 213 (2), 315. http://dx. doi.org/10.1016/j.expneurol.2008.06.004. Mollenhauer, B., El-Agnaf, O.M., Marcus, K., Trenkwalder, C., Schlossmacher, M.G., 2010. Quantification of alpha-synuclein in cerebrospinal fluid as a biomarker candidate: review of the literature and considerations for future studies. Biomark. Med 4 (5), 683. http://dx.doi.org/10.2217/bmm.10.90. Mollenhauer, B., Locascio, J.J., Schulz-Schaeffer, W., Sixel-Doring, F., Trenkwalder, C., Schlossmacher, M.G., 2011. alpha-Synuclein and tau concentrations in cerebrospinal fluid of patients presenting with parkinsonism: a cohort study. Lancet Neurol. 10 (3), 230. http://dx.doi.org/10.1016/S1474-4422(11)70014-X. Mollenhauer, B., Trautmann, E., Taylor, P., Manninger, P., Sixel-Doring, F., Ebentheuer, J., et al., 2013. Total CSF alpha-synuclein is lower in de novo Parkinson patients than in healthy subjects. Neurosci. Lett. 532, 44. http://dx.doi.org/10.1016/j. neulet.2012.11.004. Sako, W., Murakami, N., Izumi, Y., Kaji, R., 2014. Reduced alpha-synuclein in cerebrospinal fluid in synucleinopathies: evidence from a meta-analysis. Mov. Disord. 29 (13), 1599. http://dx.doi.org/10.1002/mds.26036. Teunissen, C.E., Verwey, N.A., Kester, M.I., van Uffelen, K., Blankenstein, M.A., 2010. Standardization of assay procedures for analysis of the CSF biomarkers amyloid β((1–42)), tau, and phosphorylated tau in Alzheimer's disease: report of an International Workshop. Int. J. Alzheimers Dis. 2010. http://dx.doi.org/10.4061/ 2010/635053 (Sep 27, pii: 635053). Vos, S.J., Visser, P.J., Verhey, F., Aalten, P., Knol, D., Ramakers, I., Scheltens, P., Rikkert, M.G., Verbeek, M.M., Teunissen, C.E., 2014. Variability of CSF Alzheimer's disease biomarkers: implications for clinical practice. PLoS One 9 (6), e100784. http://dx. doi.org/10.1371/journal.pone.0100784 (Jun 24, eCollection 2014).
Contents lists available at ScienceDirect
Journal of Immunological Methods journal homepage: www.elsevier.com/locate/jim
Research paper
Validation of a commercially available enzyme-linked immunoabsorbent assay for the quantification of human α-Synuclein in cerebrospinal fluid Niels Kruse a,⁎, Brit Mollenhauer a,b a b
Department of Neuropathology, University Medical Center, Robert-Koch-Str. 40, 37075 Göttingen, Germany Paracelsus-Elena-Klinik, Klinikstr. 16, 34128 Kassel, Germany
a r t i c l e
i n f o
Article history: Received 15 May 2015 Received in revised form 10 July 2015 Accepted 4 August 2015 Available online 10 August 2015 Keywords: alpha-Synuclein ELISA Validation
a b s t r a c t The quantification of α-Synuclein in cerebrospinal fluid (CSF) as a biomarker has gained tremendous interest in the last years. Several commercially available immunoassays are emerging. We here describe the full validation of one commercially available ELISA assay for the quantification of α-Synuclein in human CSF (Covance alpha-Synuclein ELISA kit). The study was conducted within the BIOMARKAPD project in the European initiative Joint Program for Neurodegenerative Diseases (JPND). We investigated the effect of several pre-analytical and analytical confounders: i.e. (1) need for centrifugation of freshly drawn CSF, (2) sample stability, (3) delay of freezing, (4) volume of storage aliquots, (5) freeze/thaw cycles, (6) thawing conditions, (7) dilution linearity, (8) parallelism, (9) spike recovery, and (10) precision. None of these confounders influenced the levels of α-Synuclein in CSF significantly. We found a very high intra-assay precision. The inter-assay precision was lower than expected due to different performances of kit lots used. Overall the validated immunoassay is useful for the quantification of α-Synuclein in human CSF. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Levels of α-Synuclein (aSyn) in cerebrospinal fluid (CSF) in Parkinson's disease and related synucleinopathies have been analyzed in several studies (for review see Sako et al., 2014). However, not all of these studies reached the same conclusions. The reason for variable results relies largely on the use of home-made ELISA assays using different antibody combinations (Mollenhauer et al., 2010). In a recent ring trial we investigated the performance of a newly released commercially available human specific ELISA (Covance alpha-Synuclein ELISA kit) to quantify aSyn in cerebrospinal fluid in 18 different laboratories (Kruse et al., 2015). In this study pre-analytical confounders were largely excluded through the centrally provided samples. In the current manuscript we extended this initial validation study at one center and analyzed the effect of various preanalytical and analytical manipulations on the aSyn quantification with this assay. This study was conducted in accordance to the standard operating procedure developed within the BIOMARKAPD consortium (Andreasson et al., in press). In particular we analyzed (1) the need for centrifugation of freshly drawn CSF, (2) sample stability under various conditions, (3) delay of freezing, (4) volume of storage aliquots, (5) freeze/thaw cycles, (6) thawing conditions, (7) dilution linearity, (8) parallelism, (9) spike recovery, and (10) precision of replicate ⁎ Corresponding author. E-mail addresses: [email protected] (N. Kruse), [email protected] (B. Mollenhauer).
http://dx.doi.org/10.1016/j.jim.2015.08.003 0022-1759/© 2015 Elsevier B.V. All rights reserved.
aSyn quantification in independent experiments. The most important outcome is that aSyn seems to be very stable under all conditions applied. 2. Material and methods 2.1. Collection of CSF samples For validation experiments randomly selected CSF samples were used. Details on sample processing and diagnostic criteria were published recently (Mollenhauer et al., 2011). Our study was done in accordance with the Declaration of Helsinki and with informed written consent provided by all patients or by their next of kin in the case of cognitive impairment (mini-mental status examination score 23 of 25 or below). Our study was approved by the ethics committees at the University of Göttingen and the Landesärztekammer Hessen (Germany). CSF samples had no elevation in red or white blood cell counts and semiquantitative hemoglobin levels. 2.2. Preparation of CSF samples Preparation of CSF samples has been described recently (Mollenhauer et al., 2011). Depending on the CSF volume available after performing routine diagnostics CSF aliquots (except for Aliquot volume samples (see Section 2.2.4 below)) were prepared at 375 μl or 60 μl in 0.5 ml polypropylene reaction tubes (Sarstedt, Nümbrecht, Germany; Ref.# 72.699). For most of the variables to be analyzed
N. Kruse, B. Mollenhauer / Journal of Immunological Methods 426 (2015) 70–75
71
Fig. 3. Delay of freezing has no effect on quantification of CSF aSyn.
Fig. 1. No effect of different centrifugation conditions on CSF aSyn quantification results.
(centrifugation, precision, stability, freeze/thaw cycles, delay of freezing, and thawing conditions), multiple 60 μl volumes of CSF were sufficient, while larger amounts were necessary of other experiments (aliquot volumes, dilution linearity, and parallelism). After preparation all CSF aliquots were stored at − 80 °C until analysis. 2.2.1. Centrifugation samples Three different aliquots were prepared from CSF obtained from six donors. (1) One aliquot was left after lumbar puncture (LP) without centrifugation. (2) One aliquot was centrifuged at 1600 ×g and room temperature (RT) for 10 min to obtain the RT samples and (3) the remaining CSF was centrifuged at 1600 ×g at 4 °C.
2.2.4. Aliquot volume samples CSF samples (n = 7 donors) were prepared at 125 μl, 250 μl, and 375 μl in 500 μl tubes (corresponding to 25%, 50%, and 75% of the storage tube volumes) and left for one year at −80 °C before analysis. 2.2.5. Freeze/thaw cycle samples Three aliquots were thawed at RT for 60 min and refrozen at −80 °C. After at least 12 h at − 80 °C this procedure was repeated twice on two aliquots and another two times on one aliquot, respectively. CSF samples from eight donors were used for this purpose and generated samples being thawed up to five times before final thawing for analysis. 2.2.6. Thawing conditions Aliquots from four different CSF samples were thawed under different conditions: either rapidly by keeping them in the hand or thawing them for up to 3 h on ice, at 5–8 °C in a refrigerator or at RT before performing the assay.
2.2.2. Stability samples Nine aliquots were prepared to test CSF stability under different conditions. CSF from seven donors was used for this purpose. Aliquot 1 was left at − 80 °C until analysis. Aliquots 2 were left at 5–8 °C for 24 h before being frozen at − 80 °C. Aliquots 3 were left at 5–8 °C for one week before being frozen at −80 °C. Aliquots 4 were left at RT for 24 h before being frozen at −80 °C. Aliquots 5 were left at −20 °C for one month before being frozen at −80 °C.
2.2.7. Dilution linearity samples CSF samples (n = 6 donors) were spiked with approximately 100-times the endogenous concentration of aSyn and then serially diluted two-fold before analysis.
2.2.3. Delay of freezing samples After lumbar puncture CSF aliquots (n = 5 donors) were left for 2 h, 24 h, 48 h, and five days at RT or 4 °C before being frozen at −80 °C.
2.2.9. Spike recovery samples Aliquots of CSF samples (n = 5 donors) were spiked with aSyn standards at final concentrations of 1000 pg/ml, 500 pg/ml, and
Fig. 2. High reproducibility of CSF aSyn quantification under various storage conditions.
Fig. 4. No effect of storage aliquot volume on CSF aSyn quantification.
2.2.8. Parallelism samples CSF samples (n = 5 donors) were serially diluted two-fold for analysis starting with non-diluted CSF as the highest concentration.
72
N. Kruse, B. Mollenhauer / Journal of Immunological Methods 426 (2015) 70–75
version (Mollenhauer et al., 2008). In brief, CSF samples were diluted 1:10 with 1 × Reagent Diluent for analysis. Plates were washed four times with 300 μl wash buffer. aSyn standards and CSF samples were applied at 200 μl volumes to reaction wells. The reaction plate was incubated overnight at 4 °C without shaking. The next day the content of the wells was discarded and plates were washed as described above. 50 μl biotinylated primary antibody was added at 1000-fold dilution of the stock solution in 1 × Reagent Diluent. Plates were then incubated for 2 h at RT with shaking at 300 rpm. After washing again, 200 μl streptavidin-coupled horse-radish peroxidase, diluted 15,000-fold in 1 × Reagent Diluent, was added and incubated for 1 h at RT with shaking at 300 rpm. Plates were washed again, 100 μl chemiluminescence substrate was added and plates were read in a Tecan Infinite 200Pro electrochemiluminescence reader. Experiments were performed using kit lots #D13IC02685 and D14IK01932. Fig. 5. No effect of multiple freeze/thaw cycles on CSF aSyn quantification.
250 pg/ml. Spiking was performed by adding one volume of standards to nine volumes of CSF. Endogenous protein concentrations were determined by spiking one volume Reagent Diluent to nine volumes CSF to account for the dilution factor. Furthermore, nine volumes of Reagent Diluent were spiked with one volume of protein standards to confirm spiked protein concentrations. 2.2.10. Precision samples 24 samples were prepared from CSF obtained from seven donors for analysis of intra-assay and inter-assay precision. 12 aliquots from individual CSF samples were measured in one experiment followed by quantification of aSyn in three additional aliquots in four independent experiments. Intra-assay precision refers to the comparability of results performed on the same day and inter-assay variability refers to the comparability of results from independent experiments performed on different days.
2.4. Data analysis Standards and samples were analyzed in duplicates. Mean and standard deviations were calculated for each duplicate. Background values were subtracted from standard and sample readings. Standard curves were plotted using Microsoft Excel 2003. Resulting curve parameters of linear regression curves were used to calculate aSyn concentrations in individual samples, as is considered acceptable according to the kit insert. Concentrations were adjusted for dilution factors. Results from duplicates with b15% SD were considered acceptable. Reliable results were defined as those with CVs for replicate determinations of analyte concentration of b20%. In some analysis aSyn concentrations were normalized for clarity. Normalization was either done for the dilution factor (Figs. 7 and 8) or to samples thawed quickly (Fig. 6). 2.5. Calculations All calculations were done using Microsoft Excel software.
2.3. Quantification of aSyn
3. Results
A general outline of the validation strategy including documentation of results will be presented elsewhere (Andreasson et al., in press). Determination of aSyn concentrations in CSF was done with one commercially available alpha-Synuclein ELISA kit according to the manufacturer's instruction (Covance, Dedham, MA, USA; Cat.-# SIG38974) (Mollenhauer et al., 2013). This assay is based on a former
In this study we performed a full validation of the Covance Alpha-Synuclein ELISA kit. According to the kit instructions aSyn concentrations down to 6.1 pg/ml can be quantified. In a previous European-wide study (Kruse et al., 2015) reliable quantification over the entire detection range of the standard curve was confirmed. Our data agreed to those reported by Kang et al. (2013) and were
Fig. 6. Different thawing conditions do not influence quantification results of aSyn in CSF.
N. Kruse, B. Mollenhauer / Journal of Immunological Methods 426 (2015) 70–75
73
reproduced in this study. All aSyn concentrations measured in this study were significantly higher than the lowest concentration of the standard curve and we therefore conclude that these were reliable. 3.1. Centrifugation Results of our different centrifugation conditions are shown in Fig. 1 and demonstrate that quantification under all conditions gave similar results. When concentrations were normalized to the concentration of the reference treatment (centrifugation at 4 °C) none of the samples deviated by more than 20% (range: 1.7–13.5%). When mean concentrations were calculated for all three conditions CVs varied between 1.23% and 6.41%. We therefore conclude that centrifugation has no effect on the aSyn concentration determined in these samples without artificial blood contamination. Fig. 8. Parallelism of serially diluted CSF samples and standards.
3.2. Stability Stability of aSyn in CSF was determined under various conditions as described in the Materials and methods section. These analyses clearly demonstrate that nearly all samples were quantified with similar results for individual CSF samples. With the exception of two aliquots (sample 1, 24 h room temperature and one month − 20 °C) all quantification results were in the expected range (Fig. 2). These results indicate that CSF may be stored for extended time periods at temperatures other than −80 °C, what is usually recommended. 3.3. Delay of freezing After having demonstrated that CSF samples may be stored for extended time period at room temperature or 5–8 °C we looked at further time points for storage at either room temperature and 5–8 °C before they were frozen at − 80 °C. Under all conditions we found highly consistent aSyn concentrations indicating that aSyn in CSF is very stable. CVs were 10% or less for all CSF samples (Fig. 3). 3.4. Aliquot volume Due to a potential increased concentration of analyte by evaporation of solvent we investigated if this occurs with aSyn in CSF by preparing three different aliquot volumes (25%, 50%, and 75% of the maximal vial volume) of CSF. This procedure also allowed us to investigate if absorption is an issue when variable storage volumes are used. For 7 different CSF samples analyzed after storage for approximately one year at − 80 °C we did not observe significant changes in aSyn
Fig. 7. Dilution linearity of serially diluted CSF samples after spiking with high concentration of aSyn.
concentration (Fig. 4). Mean values calculated from all three volumes from each CSF samples exhibited CV values below 20% (range: 1.6–14.3%). 3.5. Freeze/thaw cycles The effect of multiple freeze/thaw cycles was negligible: with up to five freeze/thaw cycles eight CSF samples showed highly comparable aSyn concentrations among themselves (Fig. 5). CVs from comparison of three aliquots per CSF sample after increasing numbers of freeze/thaw cycles were 1.2–11.0%. 3.6. Thawing conditions Different thawing procedures did not impair quantification of aSyn as compared to quickly frozen aliquots used for normalization. Even after leaving CSF samples for up to 3 h at room temperature all concentrations were determined to be within 20% of the reference aliquots (Fig. 6). 3.7. Dilution linearity Fig. 7 shows the results from experiments performed to determine the dilution linearity. aSyn concentrations were normalized for the results of the aliquots diluted 1:8 as this dilution closely matched the 1:10 dilution recommended in the assay instructions. We found that aliquots diluted four-fold to 32-fold gave comparable results. Most concentrations were within 20% of that of the reference aliquot. Only
Fig. 9. Reproducibility of CSF aSyn quantification in independent experiments.
74
N. Kruse, B. Mollenhauer / Journal of Immunological Methods 426 (2015) 70–75
4. Discussion
Table 1 Intra-assay and inter-assay precision. CSF sample
Intra-assay precision
Inter-assay precision
CSF 2 CSF 3 CSF 4a CSF 5a CSF 9 CSF 10a CSF 11a
4.2% 5.3% 4.2% 5.0% 4.5% 2.6% 3.7%
13.0% 10.8% 24.0% 29.3% 20.3% 64.9% 29.8%
a
Experiments on these CSF samples were performed with different kit lots.
one CSF sample had slightly larger variability (CSF 5 with 78% at four-fold dilution and 123% at 32-fold dilution). 3.8. Parallelism Parallelism of serially diluted CSF samples is demonstrated in Fig. 8. Normalization of aSyn concentrations was again relative to those obtained for the 1:8 dilutions. We found that CSF diluted 1:4 to 1:16 in most cases gave reliable results. 3.9. Spike recovery aSyn concentrations were measured in CSF aliquots spiked with three different concentrations of aSyn standards (i.e. 1000 pg/ml, 500 pg/ml and 250 pg/ml). The concentration of the spike solutions and the endogenous concentration of the CSF samples were measured in parallel. After subtraction of the endogenous aSyn concentration mean recovery rates were calculated to be 70.1 ± 4.1% (range: 66.6–78.7%) (data not shown). 3.10. Precision Fig. 9 and Table 1 show information for intra-assay precision and inter-assay precision from 24 aliquots of seven different samples in five independent experiments. Intra-assay precision was very robust with CVs between 2.6 and 5.3%. Inter-assay precision was more inconsistent with variation between 10.8 and 20.3% for one lot and 24.0–64.9% for two different lots (Table 1). The reason for this observation is that these analyses were performed on ELISA kits from two different lots (Fig. 10). Compared to ELISA kit lot 1 much higher signal values were detected for individual standard concentrations. At similar signal readings for individual samples this shift of the standard curve therefore lead to lower concentration assignments.
Knowledge on pre-analytical and analytical confounders on quantification of CSF biomarkers is of importance as large multicenter biomarker studies need to compare measurements across different centers with expected variation and to minimize further avoidable source of variability. In this study we performed a comprehensive validation of one ELISA assay for the quantification of human aSyn in CSF. For other biomarkers, mainly in the field of Alzheimer's disease (AD) similar studies have been enormous helpful to design biomarker studies and to compare studies form different centers (Mattsson et al., 2013). The biomarker field in PD is lagging behind. Some characteristics of CSF aSyn are similar to beta amyloid (such as the lipophilic nature and in consequence the propensity to stick to certain tube and pipette tip material) but some characteristics are particular in CSF aSyn from other analytes, e.g. the high abundance on erythrocytes. To our knowledge this is the first publication analyzing various pre-analytical and analytical confounders related to aSyn quantification. We found that none of the confounders analyzed in this study did negatively influence the quantitative results. This indicates that aSyn is a very stable protein and robust for multicenter assessments. None of the potential sources of variation, i.e. the need for centrifugation of freshly drawn CSF, sample stability under various conditions, delay of freezing at − 80 °C, volume of storage aliquots, repeated freeze/thaw cycles, thawing conditions, precision of replicate aSyn quantification in independent experiments, impaired quantification of aSyn in CSF and showed low variation. Adsorption of lipophilic proteins to surfaces of storage vials may result in reduced detectability of these respective proteins (Lewczuk et al., 2006). According to recent recommendations (del Campo et al., 2012) we stored CSF samples in polypropylene tube to minimize protein adsorption. Our experiments using different aliquot storage volumes clearly showed that neither adsorption nor different storage volumes had an influence on aSyn quantification. According to the defined criteria of the BIOMARKAPD consortium recovery of spiked analyte should be in the range of 80–120% (Andreasson et al., in press). We found slightly lower recoveries with 70% on average and therefore this is the only variable for which the aSyn ELISA did not perform as recommended. Nevertheless as the recovery rates were comparable over the entire range of the spiked aSyn concentrations we feel that these results will allow for reliable and comparable quantification of aSyn in CSF (Lee et al., 2006; Food and Drug Administration, 2001). The only cause of variations identified is usage of different ELISA kit lots. We noticed that the second lot used in this study gave significantly lower concentration assignments than the first lot. The reason for this
Fig. 10. Standard curves from two different ELISA kit production lots perform differently.
N. Kruse, B. Mollenhauer / Journal of Immunological Methods 426 (2015) 70–75
observation is that with the second lot of the ELISA higher signal reading were obtained than with the first lot. At similar signal readings for individual samples this shift of the standard curve therefore lead to lower concentration assignments. This observation indicates that internal quality control samples of known aSyn concentration should be run in parallel with samples to be analyzed. These can either be home-made or better should be included in the kit. Similar lot-to-lot inconsistencies have been found in several previous studies using assays for other biomarkers (Hoefner and Yeo, 2002; Teunissen et al., 2010; Vos et al., 2014). Erythrocytes express high level of aSyn (Barbour et al., 2008) which may be released from the cells upon destruction due to freezing. Since the CSF samples for our experiments were devoid of elevated levels of red and white blood cells as well as hemoglobin it is not unexpected that centrifugation did not show any effect on aSyn quantitative results. Nevertheless we recommend centrifugation of CSF samples to remove minor contamination of erythrocytes from CSF. In a recent study performed with the same aSyn ELISA Heegaard et al. (Heegaard et al., 2014) induced some analytical modifications leading to reduced background without changing the detection limit. This observation indicates that further optimization of the assay procedure might lead to improved performance of the aSyn ELISA. Inclusion of internal quality control samples will help to verify accurate performance of test runs. Further round robin test and automation of assay procedure will certainly leads to widespread acceptance of this new aSyn ELISA. Acknowledgments The authors want to thank Birgit Otte for the excellent technical assistance. This is an EU Joint Programme — Neurodegenerative Disease Research (JPND) project. The project is supported through the following funding organizations under the aegis of JPND — www.jpnd.eu. References Andreasson, U., Perret-Liaudet, A., van Waalwijk van Doorn, L., Blennow, K., Chiasserini, D., Engelborghs, S., Fladby, T., Genc, S., Kruse, N., Kuiperij, B., Kulic, L., Lewczuk, P., Mollenhauer, B., Mroczko, B., Parnetti, L., Vanmechelen, E., Verbeek, M., Winblad, B., Zetterberg, H., Koel-Simmelink, M., Teunissen, C.E., 2015. A practical guide to immunoassay method validation. Frontiers in Neurology (in press). Barbour, R., Kling, K., Anderson, J.P., Banducci, K., Cole, T., Diep, L., Fox, M., Goldstein, J.M., Soriano, F., Seubert, P., Chilcote, T.J., 2008. Red blood cells are the major source of alpha-synuclein in blood. Neurodegener. Dis. 5 (2), 55. http://dx.doi.org/10.1159/ 000112832. Del Campo, M., Mollenhauer, B., Bertolotto, A., Engelborghs, S., Hampel, H., Simonsen, A.H., Kapaki, E., Kruse, N., Le Bastard, N., Lehmann, S., Molinuevo, J.L., Parnetti, L., Perret-Liaudet, A., Saez-Valero, J., Saka, E., Urbani, A., Vanmechelen, E., Verbeek, M., Visser, P.J., Teunissen, C., 2012. Recommendations to standardize preanalytical confounding factors in Alzheimer's and Parkinson's disease cerebrospinal fluid biomarkers: an update. Biomark. Med 6 (4), 419. Food and Drug Administration, 2001. Guidance for Industry: Bioanalytical Method Validation. U.S. Department of Health and Human Services (http://www.fda.gov/ downloads/Drugs/…/Guidances/ucm070107.pdf). Heegaard, N.H.H., Tanassi, J.T., Bech, S., Salvesen, L., Jensen, P.L., Montezinho, L.C.P., Winge, K., 2014. CSF-α-synuclein in the differential diagnosis of parkinsonian syndromes. Future Neurol. 9 (5), 525. http://dx.doi.org/10.221/FNL.14.51.
75
Hoefner, D.M., Yeo, K.-T.J., 2002. Lot-to-lot inconsistency of anticardiolipin reagents. Clin. Chem. 48 (9), 1625. Kang, J.H., Irwin, D.J., Chen-Plotkin, A.S., Siderowf, A., Caspell, C., Coffey, C.S., Waligorska, T., Taylor, P., Pan, S., Frasier, M., Marek, K., Kieburtz, K., Jennings, D., Simuni, T., Tanner, C.M., Singleton, A., Toga, A.W., Chowdhury, S., Mollenhauer, B., Trojanowski, J.Q., Shaw, L.M., 2013. Association of cerebrospinal fluid beta-amyloid 1–42, T-tau, P-tau181, and alpha-synuclein levels with clinical features of drug-naive patients with early Parkinson disease. JAMA Neurol. 70 (10), 1277. http://dx.doi.org/10. 1001/jamaneurol.2013.3861. Kruse, N., Persson, S., Alcolea, D., Bahl, J.M.C., Baldeiras, I., Capello, E., Chiasserini, D., Bocchio Chiavetto, L., Emersic, A., Engelborghs, S., Eren, E., Fladby, T., Frisoni, G., García-Ayllón, M.-S., Genc, S., Gkatzima, O., Heegaard, N.H.H., Janeiro, A.M., Kováčech, B., Kuiperij, H.B., Leitão, M.J., Lleó, A., Martins, M., Matos, M., Mollergard, H.M., Nobili, F., Öhrfelt, A., Parnetti, L., Resende de Oliveira, C., Rot, U., Sáez-Valero, J., Struyfs, H., Tanassi, J.T., Taylor, P., Tsolaki, M., Vanmechelen, E., Verbeek, M.M., Zilka, N., Blennow, K., Zetterberg, H., Mollenhauer, B., 2015. Validation of a quantitative cerebrospinal fluid alpha-synuclein assay in a European-wide interlaboratory study. Neurobiol. Aging 36 (9), 2587. http://dx.doi.org/10.1016/j. neurobiolaging.2015.05.003. Lee, J.W., Devanarayan, V., Barrett, Y.C., Weiner, R., Allinson, J., Fountain, S., Keller, S., Weinryb, I., Green, M., Duan, L., Rogers, J.A., Millham, R., O'Brien, P.J., Sailstad, J., Khan, M., Ray, C., Wagner, J.A., 2006. Fit-for-purpose method development and validation for successful biomarker measurement. Pharm. Res. 23 (2), 312. http://dx.doi.org/10.1007/s11095-005-9045-3. Lewczuk, P., Beck, G., Esselmann, H., Bruckmoser, R., Zimmermann, R., Fiszer, M., Bibl, M., Maler, J.M., Kornhuber, J., Wiltfang, J., 2006. Effect of sample collection tubes on cerebrospinal fluid concentrations of tau proteins and amyloid beta peptides. Clin. Chem. 52 (2), 332. Mattsson, N., Andreasson, U., Persson, S., Carrillo, M.C., Collins, S., Chalbot, S., Cutler, N., Dufour-Rainfray, D., Fagan, A.M., Heegaard, N.H., Robin Hsiung, G.Y., Hyman, B., Iqbal, K., Lachno, D.R., Lleo, A., Lewczuk, P., Molinuevo, J.L., Parchi, P., Regeniter, A., Rissman, R., Rosenmann, H., Sancesario, G., Schroder, J., Shaw, L.M., Teunissen, C.E., Trojanowski, J.Q., Vanderstichele, H., Vandijck, M., Verbeek, M.M., Zetterberg, H., Blennow, K., Kaser, S.A., 2013. CSF biomarker variability in the Alzheimer's Association quality control program. Alzheimers Dement. 9 (3), 251. http://dx.doi. org/10.1016/j.jalz.2013.01.010. Mollenhauer, B., Cullen, V., Kahn, I., Krastins, B., Outeiro, T.F., Pepivani, I., Ng, J., Schulz-Schaeffer, W., Kretzschmar, H.A., McLean, P.J., Trenkwalder, C., Sarracino, D.A., Vonsattel, J.P., Locascio, J.J., El-Agnaf, O.M., Schlossmacher, M.G., 2008. Direct quantification of CSF alpha-synuclein by ELISA and first cross-sectional study in patients with neurodegeneration. Exp. Neurol. 213 (2), 315. http://dx. doi.org/10.1016/j.expneurol.2008.06.004. Mollenhauer, B., El-Agnaf, O.M., Marcus, K., Trenkwalder, C., Schlossmacher, M.G., 2010. Quantification of alpha-synuclein in cerebrospinal fluid as a biomarker candidate: review of the literature and considerations for future studies. Biomark. Med 4 (5), 683. http://dx.doi.org/10.2217/bmm.10.90. Mollenhauer, B., Locascio, J.J., Schulz-Schaeffer, W., Sixel-Doring, F., Trenkwalder, C., Schlossmacher, M.G., 2011. alpha-Synuclein and tau concentrations in cerebrospinal fluid of patients presenting with parkinsonism: a cohort study. Lancet Neurol. 10 (3), 230. http://dx.doi.org/10.1016/S1474-4422(11)70014-X. Mollenhauer, B., Trautmann, E., Taylor, P., Manninger, P., Sixel-Doring, F., Ebentheuer, J., et al., 2013. Total CSF alpha-synuclein is lower in de novo Parkinson patients than in healthy subjects. Neurosci. Lett. 532, 44. http://dx.doi.org/10.1016/j. neulet.2012.11.004. Sako, W., Murakami, N., Izumi, Y., Kaji, R., 2014. Reduced alpha-synuclein in cerebrospinal fluid in synucleinopathies: evidence from a meta-analysis. Mov. Disord. 29 (13), 1599. http://dx.doi.org/10.1002/mds.26036. Teunissen, C.E., Verwey, N.A., Kester, M.I., van Uffelen, K., Blankenstein, M.A., 2010. Standardization of assay procedures for analysis of the CSF biomarkers amyloid β((1–42)), tau, and phosphorylated tau in Alzheimer's disease: report of an International Workshop. Int. J. Alzheimers Dis. 2010. http://dx.doi.org/10.4061/ 2010/635053 (Sep 27, pii: 635053). Vos, S.J., Visser, P.J., Verhey, F., Aalten, P., Knol, D., Ramakers, I., Scheltens, P., Rikkert, M.G., Verbeek, M.M., Teunissen, C.E., 2014. Variability of CSF Alzheimer's disease biomarkers: implications for clinical practice. PLoS One 9 (6), e100784. http://dx. doi.org/10.1371/journal.pone.0100784 (Jun 24, eCollection 2014).