Neurogranin as Cerebrospinal Fluid Biomarker for Alzheimer Disease: An Assay Comparison Study.
The measurements of CSF neurogranin concentrations in the aforementioned publications have been performed using 3 independent assays (7, 8, 15), which all target different epitopes of the protein (Fig. 1A). Yet, the discriminative power to distinguish patients from controls is strikingly comparable for all assays, even though the absolute neurogranin concentration ranges vary widely among the assays, e.g., approximately 90-250 pg/mL (15) compared with approximately 1000-4000 pg/mL (7) and approximately 100-340 pg/mL (8). These differences could be explained by differences in cohorts and by differences in assays, e.g., the use of various types of calibrators. To allow direct comparison of results across studies, it is important to understand the assay specifics and to directly compare the outcomes within one cohort of patients with different dementia diagnoses.
In this study, we aimed to link the 3 commonly used neurogranin assays developed at Washington University, St. Louis, MO (WashU), ADx NeuroSciences (ADx), and Gothenburg University, Molndal, Sweden (UGot). First, we characterized the 3 assays through an analysis of their calibrators and antibodies. Next, we compared CSF neurogranin concentrations measured in the same clinical dementia cohort. To compare the discriminative potentials of the assays, we analyzed neurogranin concentrations in the CSF of patients with dementia with Lewy bodies (DLB; n = 22), vascular dementia (VaD; n = 20), frontotemporal dementia (FTD; n = 22), or dementia due to AD (n = 22), and in the CSF of controls (n = 22). Bridging these CSF neurogranin assays will help translate neurogranin findings in multicenter comparisons and improve our understanding of the use of neurogranin as a biomarker.
Brain homogenate preparation. Two different brain homogenate samples were used: the first to characterize the 3 assays using Western blot, prepared in RIPA (radioimmunoprecipitation assay) buffer, and the second for the immunoassay measurements, prepared in MPER (Mammalian Protein Extraction Reagent) buffer (details in Methods in the Data Supplement that accompanies the online version of this article at http://www.clinchem. org/content/vol64/issue6).
Clinical samples. We collected and processed 108 CSF samples from patients of the Amsterdam Dementia Cohort (20) according to international consensus guidelines (21, 22). Further, 22 patients with a diagnosis of dementia due to AD were compared with 22 controls (individuals with subjective cognitive decline) and with patients with a diagnosis of FTD (n = 22), DLB (n = 22), or VaD (n = 20; all diagnosed according to the consensus criteria (23-26); Table 1). CSF biomarkers amyloid-[beta] 1-42 (A[beta]1-42; <550 pg/mL for AD), total Tau (t-Tau; >375 pg/mL for AD), phosphorylated Tau at P181 (p-Tau181; >52 pg/mL for AD) measured using INNOTEST[R] (Fujirebio) were used to support the clinical diagnosis as described in the diagnostic guidelines (23, 27). All patients signed informed consent, and the study was approved by the local ethical committee. For power calculation and sample distribution, see Methods in the online Data Supplement.
For the composition of the neurogranin calibrators, antibody epitopes, protein, and peptides, see Fig 1A.
Washington University in-house neurogranin Erenna[R] assay. A sandwich immunoassay was developed for CSF by means of a Single Molecule Counting[TM] Erenna[R] system (Singulex) by using the epitope-specific rabbit antibodies recognizing N-terminal epitope S10-D23 and C-terminal epitope G49-G60 (for antibody development and assay method, see (7)) The C-terminal--specific antibody (P-4793) was coupled with magnetic beads and used as the capture antibody, and the N-terminal--specific antibody (P-4794) was labeled with a fluorescent dye and used as the capping/detection antibody. Synthetic 78-mer human neurogranin used as the immunoassay standard was prepared and characterized by AAPPTec, LLC, by use of C18-reversed-phase high performance liquid chromatography and electrospray ionization--mass spectrometry and quantified using amino acid analysis (AAA Service Laboratory). The assay uses 5-[micro]L CSF per reaction, 10-fold diluted in buffer (50 mmol/L Tris, 150 mmol/L NaCl, pH 8.1, supplemented with 2-g rabbit IgG and 0.1% Triton X) in a total capture step reaction volume of 150 [micro]L containing antibody-coated magnetic beads, and assayed in triplicate. Curve fitting was done by weighted regression of the 3 signal types (detected events, event photons, and total photons) using an algorithm (SMDCurve Fit, Singulex Software SGX Link), resulting in a 5-parameter logistic equation for interpolation. Mean spike recovery for 8 CSF samples supplemented with 1000 pg/mL neurogranin was 101% (range, 97%-106%).
ADx NeuroSciences neurogranin ELISA. This sandwich ELISA combines 2 monoclonal mouse antibodies, as previously reported (15): one [ADx403 (clone ADxNGCI2)] directed against the R53-A64 sequence of neurogranin, the other [ADx451 (clone ADxNGCT1)] directed against the C-terminus of the protein, truncated at P75 specifically, i.e., G62-P75. CSF samples were incubated (15-[micro]L undiluted sample + 100-[micro]L biotinlabeled detection mAb) in a PBS-based buffer with stabilizing proteins and detergent (EuroImmun Laboratories) in duplicate, for 3 h at room temperature on a capture mAb precoated plate. Further details on the test procedure are as described by De Vos et al. (15). Final concentrations of neurogranin were interpolated [log(X); 4-parameter logistic] on the basis of a synthetic calibrator, custom-made, and characterized by Proteogenix, covering the C-terminal sequence truncated at P75. To ensure no matrix effects are present, 2 nonclinical CSF samples with different neurogranin concentrations were serially diluted in an assay diluent and measured in quadruplicate. The overall %CV on all back-calculated concentrations, including dilution factor, was 4.4% CV.
Gothenburg University in-house neurogranin ELISA. Ninety-six well microtiter plates were coated with an inhouse monoclonal antibody (Ng7, epitope G52-G65, 1 [micro]g/mL, 100 [micro]L/well) in carbonate buffer (50 mmol/L NaHCO3, pH 9,6) overnight at +4[degrees]C, whereas subsequent steps were performed at room temperature. After washing 4 times with 350 [micro]L/well of PBS containing 0.05% Tween-20 (PBST), the plates were blocked for 1 h with a blocking reagent (1% BSA in PBST, 250 [micro]L/ well). The calibrator, synthetic full-length neurogranin (custom-made by ProteoGenix), was prepared in the diluent (1% BSA in PBST) at an initial concentration of 8000 pg/mL and subsequent 2-fold dilutions were made. CSF samples and calibrators (50[micro]L/well) were simultaneously incubated with an equal volume of a polyclonal antibody diluted 1:10 000 (cat# 07-425, end-specific to the D78 terminus, Merck Millipore) (6) for 1 h on a plate shaker (700 rpm) followed by overnight incubation. After washing, 100 [micro]L/well of horseradish peroxidase--conjugated donkey antirabbit antibody (Pierce 31458, Thermo Scientific) diluted 1:20 000 was added. After incubation for 5 h, the plates were washed and developed using 100 [micro]L/well of TMB (TMB Peroxidase EIA substrate kit Bio-Rad, cat# 172-1067). After 30 min at room temperature, the reaction was stopped with 1 mol/L sulfuric acid (100 [micro]L/well) and the plates were read in a Vmax microplate reader (Molecular Devices) at 450 nm (650 nm as the reference wavelength). Concentrations were calculated using a 4-parameter curve fitted to the calibrators (SoftMax v4.0, Molecular Devices). Dilution linearity and parallelism parameter results were <15% CV.
Analytical performance. The lower limit of quantification (LLOQ) had been previously determined in each assay by the mean concentration of 16 blanks plus 10 times SD. Clinical samples with undetermined values lower than LLOQ were assigned the LLOQ value for inclusion in the statistical analyses. The intra-assay CV was defined as the mean of the replicate CVs of all patient CSF samples used in this study. Inter-assay CV was defined as the mean of 4-6 in-house-prepared quality control CSF pools with high and low neurogranin concentrations.
CHARACTERIZATION OF ASSAYS
Gel electrophoresis. Calibrators, 2 control CSF samples from the BIODEM laboratory (Institute Born-Bunge, University of Antwerp), and brain homogenate were separated on the basis of their molecular weights by use of a 12% Bis-Tris gel (Bolt[TM], Thermo Fisher Scientific) with 1x MES SDS running buffer (Thermo Fisher). Samples were prepared in 1x loading buffer with 0.05-mol/L dithiothreitol, and the same amounts of protein were loaded. The gel was run at 200 V.
Silver stain. Protein quantities in calibrator solutions were estimated using the Pierce[R] Silver Stain Kit (Thermo Fisher) according to manufacturer's protocol. In short, the gel was fixed in 30% ethanol and 10% acetic acid, sensitized, and stained overnight, and the color was developed for 2-3 min and stopped with 5% acetic acid. Bands were quantified in ImageJ (28).
Western blot. Proteins were transferred to a 0.2-[micro]m nitrocellulose membrane using a dry blotting system (iBlot[R] 2 Transfer Device; Thermo Fisher) and visualized using immunofluorescent secondary antibodies (LICOR[R] Biotechnology; see Methods in the online Data Supplement for details).
Passing--Bablok regression analyses were performed to compare the WashU, ADx, and UGot assays on proportional and systematic differences based on the 108 clinical samples.
For clinical performance validation, rank-transformed neurogranin concentrations were used because assumptions for normality of the model residuals were not met. To compare neurogranin concentrations among diagnostic groups, an analysis of covariance (ANCOVA) corrected for age and sex was performed per assay, followed by Bonferroni adjusted post hoc comparisons. The effect size of every assay was defined as the partial [[eta].sub.2] of these ANCOVA models. Spearman correlations were used to correlate neurogranin concentrations with Mini-Mental State Examination (MMSE) scores and AD biomarker values for each assay. Analyses were done in R version 3.4.0 (29).
CALIBRATOR QUANTIFICATION USING SILVER STAIN AFTER SDS-PAGE GEL ELECTROPHORESIS
All calibrators were separated by gel electrophoresis and visualized by silver stain (Fig. 1B). The silver stain showed a single sharp band at 14 kDa for the WashU calibrator only for the highest calibrator concentration (100 ng). For the ADx calibrator, a clean band at 6 kDa was observed when 100 or 20 ng was loaded. The UGot calibrator showed a large smear at higher molecular weights and lower clean bands at 14 kDa and at 6 kDa, although the 6-kDa band was only observed in the highest calibrator concentration (100 ng). The UGot calibrator contains by far the largest amount of protein, around 7-fold more than the WashU and the ADx calibrators. The WashU and ADx calibrators both had similar intensities of staining, indicating that the same relative amounts of calibrator were present in the samples.
QUANTITATIVE AND QUALITATIVE COMPARISONS OF NEUROGRANIN EPITOPE RECOGNITION BY CAPTURE AND DETECTION ANTIBODIES OF IMMUNOASSAYS
After separation by gel electrophoresis, the calibrators of the 3 assays, as well as CSF and brain homogenate samples, were immunoblotted with the antibodies of all 3 assays (Fig. 1C-H). The experiments shown for the WashU calibrator (Fig. 1G) were inadvertently performed with calibrator stored for 2 days at 4[degrees]C and not freshly thawed material. The experiment was repeated with freshly thawed material (see Fig. 1 in the online Data Supplement) and gave similar results. All antibodies showed the strongest and cleanest bands when exposed to their respective assay calibrator. The N-terminal WashU antibody (S10-D23) recognized both its own calibrator and the UGot calibrator at 14 kDa, although the latter gave the strongest signal. The ADx calibrator was not recognized (Fig. 1C). The G49-G60 WashU antibody recognized its own and the UGot calibrator at 14 kDa, and again the UGot calibrator gave the strongest signal. The ADx calibrator was detected at 6 kDa (Fig. 1D). The R53-V64 ADx antibody recognized its own calibrator at 6 kDa and the UGot calibrator at 14 kDa. The WashU calibrator showed a very weak band at 14kDa(Fig. 1E).The G62-P75 antibody, on the contrary, only recognized its own calibrator at 6 kDa (Fig. 1F). The G52-G65 UGot antibody recognized its own calibrator and the WashU calibrator at 14 kDa, and it weakly stained the ADx calibrator at 6 kDa (Fig. 1G and see Fig. 1 in the online Data Supplement). The V66-D78 UGot antibody detected its own and the WashU calibrator at 14 and 6 kDa (Fig. 1H and see Fig. 6 in the online Data Supplement), whereas the ADx calibrator gave no signal (Fig. 1H). Three of the bands at higher molecular weights in the UGot calibrator were recognized by all neurogranin antibodies, except for the ADx451 antibody that recognized only the neoepitope at the neurogranin truncated at P75.
Next, we assessed the affinity of the neurogranin antibodies to CSF and brain lysate samples (Fig. 1C-H). Neurogranin was recognized in CSF by the WashU and UGot antibodies, i.e., bands were shown, unexpectedly, around 60 or 70 kDa, but not by the ADx antibodies. Neurogranin in brain lysate was recognized as a single band at 14 kDa by most antibodies, whereas the WashU antibodies P-4794 and, to a far lesser extent, P-4793 recognized 2 or 3 bands at higher molecular weights, and the ADx451 antibody recognized very weakly a band around 27 kDa.
NEUROGRANIN EPITOPE RECOGNITION BY THE 3 IMMUNOASSAYS
To better understand what forms of neurogranin are recognized by the 3 different immunoassays, the calibrators of each assay were measured in the other 2 assays (Table 2). The WashU neurogranin assay did not detect the P75-truncated ADx calibrator. The WashU assay fully recognized the UGot calibrator, and the neurogranin values obtained by the WashU assay were approximately 30% higher than the concentrations obtained by the UGot assay itself. The ADx assay, designed to detect neurogranin truncated at P75, did not detect the calibrator of WashU or the calibrator ofUGot, although, for the latter, lower concentrations at the border of the LLOQ were detected for all calibrator dilutions. The UGot assay detected smaller amounts of the P75-truncated ADx calibrator, whereas WashU calibrator was detected above the upper limit of detection in all dilutions (12-399 pg/mL), presumably owing to the use of added rabbit IgG as a carrier protein in the WashU calibrator.
All assays recognized neurogranin in the dilutions of the control brain lysate sample (Fig. 2). The WashU and UGot assay quantified neurogranin in the brain lysate at similar concentrations, whereas the ADx assay detected approximately 70-fold lower neurogranin concentrations compared with the WashU and UGot assays.
ANALYTICAL PERFORMANCE OF THE ASSAYS
Neurogranin concentrations could be determined in all samples by using the WashU and ADx assays, whereas 9 out of 108 samples were below the LLOQ by using the UGot assay. All replicate neurogranin measurements had CVs < 20% in the WashU and ADx assays, whereas 20% of the samples measured with the Gothenburg assay had CVs > 20%, mainly in samples with concentrations near the LLOQ. Intra-assay/inter-assay CVs were 5%/6% (WashU), 7%/7% (ADx), and 6%/8% (UGot). Neurogranin values ranged from [median (range)] 1881 (330-8320) pg/mL for WashU, 372 (71-1191) pg/mL for ADx, and 416 (115-1481) pg/mL for UGot. The strongest correlation for neurogranin in clinical samples was found between the ADx and WashU assays, with a Spearman [rho] of 0.95, whereas for UGot vs WashU and UGot vs ADx, Spearman p was 0.87 and 0.81, respectively. Passing-Bablok regression analysis demonstrated proportional differences among all 3 assays (Fig. 3), whereas the WashU assay showed about 5 times higher neurogranin concentrations than the ADx and UGot assays. A systematic difference was observed only between the ADx vs WashU assays.
NEUROGRANIN RESULTS DIFFER BETWEEN DIAGNOSTIC GROUPS
ANCOVA showed a specific increase in neurogranin concentrations in AD patients than in all other clinical groups in each assay (P < 0.05; Fig. 4). Effect size was the highest in the UGot assay: [[eta].sup.2] for the WashU, ADx, and UGot assays was 0.14, 0.11, and 0.21, respectively. Post hoc Bonferroni tests showed a specific increase in neurogranin concentrations in the AD group compared than in the DLB group (P < 0.05 for WashU and ADx, P < 0.01 for UGot) and VaD group (P < 0.01 for WashU, P < 0.05 for ADx, P < 0.001 for UGot), whereas the differences between AD patients and controls were only significant for the WashU (P < 0.05) and UGot assays (P < 0.01). Sex and age did not significantly influence neurogranin values (see Materials 2-3 in the online Data Supplement). Apolipoprotein E (APOE) [epsilon]4 carriers had slightly higher neurogranin concentrations than APOE [epsilon]4 noncarriers, but this difference reached significance in only the UGot assay (see Material 4 in the online Data Supplement).
RELATION OF NEUROGRANIN WITH MMSE SCORES AND OTHER AD BIOMARKERS IN THE 3 ASSAYS
A weak negative correlation between neurogranin and MMSE score was found in the UGot assay ([rho] = -0.23, P < 0.05) butnotin the WashU or ADx assays (see Material 5 in the online Data Supplement). None of the neurogranin assays showed a correlation with A[beta]1-42, but all had positive correlations with t-Tau and p-Tau181, respectively: Spearman [rho] = 0.78 and 0.81 for WashU, 0.76 and 0.80 for ADx, and 0.73 and 0.71 for UGot (all P < 0.0001; see Material 6 in the online Data Supplement).
We compared 3 commonly used neurogranin assays with regard to their calibrators, their epitope reactivity, and their clinical performances. All assays detected different neurogranin epitopes: the WashU assay targets the N-terminal and C-terminal part of the protein, whereas the UGot ELISA specifically targets the C-terminal end. Yet, the WashU and UGot assays could mutually recognize their calibrators, on Western blot and in the immunoassays. On the other hand, the ADx assay detects a specific form of C-terminally truncated neurogranin, the calibrator of which was not detected by the other assays and the ADx assay could not detect the other calibrators. Results from our clinical cohort showed that CSF neurogranin concentrations measured by the different assays correlated well among each other, but showed large differences in absolute values. Neurogranin values were increased in AD patients than in controls (although not significantly in the ADx assay), DLB patients, and VaD patients, but not in FTD patients.
We characterized and quantified the calibrators of the 3 assays using silver stain to better understand the large differences in absolute values observed in previous studies. The WashU and ADx calibrators showed clean bands at 14 and 6 kDa, respectively, and the UGot calibrator showed many protein bands at different molecular weights on silver stain, likely due to BSA that was added to the calibrator solution. Three of these higher-molecular-weight bands were recognized by all neurogranin antibodies on Western blot, except for the ADx451 antibody that was directed toward neurogranin truncated at P75. These higher-molecular-weight bands might represent fusion proteins comprising neurogranin, calmodulin, and other calmodulin-binding proteins (30). The 14-kDa band is typically seen for neurogranin on SDS-PAGE gel (31), and we hypothesize that this is a dimer formed by disulfide bonds between cysteine residues at positions 3, 4, or 9 of neurogranin. The ADx calibrator lacks the N-terminal part of neurogranin, including these cysteines, and stains a monomeric 6-kDa band on SDS-PAGE gel. The differences in absolute neurogranin concentrations were, however, not explained by the amount of protein found in the calibrators because the UGot calibrator contained the most protein, whereas the WashU assays produced the highest absolute values. A potential explanation could be the differences in value assignment of the calibrators. Another cause could be the different antibody affinities among the antibodies in the assays to their calibrator.
The neurogranin values in CSF and brain homogenate detected by the immunoassays were not reflected in high staining intensities in the Western blot experiments. To illustrate, CSF samples on Western blot did not show the typical 14-kDa neurogranin band, except for a weak signal detected by the N-terminal WashU antibody, whereas neurogranin in CSF is abundantly detected by all immunoassays. This is probably owing to altered conformational states of the protein in the different experiments, as samples were denatured and reduced for the SDS-PAGE gel and Western blot compared with native conditions used in the immunoassays. Alternatively, referring to the concentrations of neurogranin detected in CSF with ELISAs, i.e., in the picogram per milliliter range, it is possible that Western blot technology is not sensitive enough to demonstrate neurogranin in CSF.
Our results confirmed that CSF neurogranin concentrations were specifically increased in AD patients than in controls and in patients with other types of dementia, although the neurogranin concentrations in patients with FTD were slightly increased as well, and we observed an overlap among the different disease groups. These findings are consistent with findings from previous studies (13, 16-19), indicating that neurogranin may have limited value for the differential diagnosis of dementia.
Importantly, neurogranin concentrations among all 3 assays correlated well in all clinical dementia groups (overall, [rho] = 0.95 between ADx and WashU, [rho] = 0.87 between UGot and WashU, and [rho] = 0.81 between UGot and ADx). Nevertheless, we do not support the use of a fixed conversion factor between the neurogranin results of the different assays, as they have affinity for different neurogranin epitopes. That was not reflected in the clinical neurogranin concentrations, although it could explain our finding that neurogranin concentrations measured by UGot correlated with MMSE score and related to APOE[epsilon]4 carriership, whereas the concentrations measured by WashU and ADx did not. Clinical performance of the immunoassays compared through their effect sizes of ANCOVA was the highest in the UGot assay, followed by the WashU and ADx assays. A limitation of the UGot assay was that 8% of the samples were at or below the LLOQ value, which artificially reduced the variance in statistical comparisons.
The major strength of this study is the comparison of the calibrators of the immunoassays directly on silver stain and Western blot and that the assays were directly compared using a similar set of clinical CSF samples.
A limitation of this study was the large difference in assay calibrator composition and quantities for comparison on Western blot. High-abundance proteins required other transfer conditions than did low-abundance proteins, as did the high-molecular-weight complexes compared with the small 14-kDa neurogranin form. This hampered semiquantification of the bands observed in Western blot, limiting the quantification to the calibrator bands that were stained in the gel by silver staining.
The finding that the different neurogranin values detected by the 3 assays are not differentially expressed between dementia subtypes suggests alternative hypotheses regarding neurogranin's role in dementia pathology. Potentially, synaptic loss, which is mainly described as the hallmark of AD (32, 33), also plays a prominent role in other types of dementia (34, 35). Studies using ratios of biomarkers, for example, neurogranin together with the presynaptic protein BACE1 (15), the amyloid pathology marker A[beta]1-42 (13), or the neurodegeneration marker t-Tau (12), could yield better discriminatory power among differential diagnoses of dementia. Moreover, the different forms of neurogranin were not differentially expressed in this cross-sectional design, but it could have increased value in longitudinal designs focused on disease progression.
In conclusion, we have shown that different epitopes of neurogranin in CSF are measured using the 3 assays described here. WashU targets full-length neurogranin, ADx targets P75-truncated neurogranin, and UGot targets neurogranin at the C-terminus. By directly comparing these 3 commonly used assays, we are one step closer to implementation of neurogranin as an additional CSF biomarker for dementia as all the assays can predict dementia although their absolute values differ. A possible next step is to perform quantitative SRM (selected reaction monitoring) analysis based on trypsin-digested CSF peptides, to compare the repeatedly and independently observed trypsin-specific neurogranin peptide (G55R68, www.srmatlas.org; (36,37)) to the neurogranin concentrations from the different neurogranin assays. Also, insight regarding the relative performance of the assays compared with each other provides new opportunities for studying neurogranin in the progression of dementia, because the assays measure relatively similar neurogranin differences among clinical dementia groups but use different neurogranin epitopes.
Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contribution to the conception and design, acquisition of data, or analysis and interpretation o fdata; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article.
Authors' Disclosures or Potential Conflicts of Interest: Upon manuscript submission, all authors completed the author disclosure form. Disclosures and/or potential conflicts of interest:
Employment or Leadership: E. Vanmechelen, ADx NeuroSciences. Consultant or Advisory Role: None declared.
Stock Ownership: A. De Vos, ADx NeuroSciences; E. Vanmechelen, ADx NeuroSciences.
Honoraria: None declared.
Research Funding: E.A.J. Willemse, BIODEM laboratory, Institute Born-Bunge, University of Antwerp; A. De Vos, institutional funding from Agency for Innovation by Science and Technology. ZonMW. This study was part of the BIOMARKAPD project in the JPND programme (www.jpnd.eu) and financially supported by EMIF-AD, and the Agency for Innovation by Science and Technology (IWT O&O 140105). The work performed at the Institute Born-Bunge in Antwerp (Silver stain and Western blot experiments) was financially supported by a travel grant from Alzheimer Nederland. The brain banks of VU University Medical Center Amsterdam (The Netherlands) and the Institute Born-Bunge, University of Antwerp (Belgium) donated the brain homogenate sample.
Expert Testimony: None declared.
Patents: J. Ladenson, biomarkers of brain injury (Patent 7,985,555), Alzheimer's diagnosis (Patent 8,481,277).
Role of Sponsor: The funding organizations played no role in the design of study, choice of enrolled patients, review and interpretation of data, and final approval of manuscript.
Acknowledgments: The authors thank both Marta del Campo-Milan from VU University Medical Center Amsterdam (the Netherlands) and Joery Goossens from the Institute Born-Bunge, University of Antwerp (Belgium) and the brain banks of both institutes for their kind donation of the brain homogenate sample.
(1.) Gerendasy DD, Sutcliffe JG. RC3/neurogranin, a postsynaptic calpacitinforsetting the response threshold to calcium influxes. Mol Neurobiol 1997;15:131-63.
(2.) Kaleka KS, Gerges NZ. Neurogranin restores amyloid [beta]-mediated synaptic transmission and long-term potentiation deficits. Exp Neurol 2016;277:115-23.
(3.) Davidsson P, Blennow K. Neurochemical dissection of synaptic pathology in Alzheimer's disease. Int Psychogeriatr 1998;10:11-23.
(4.) Reddy PH, Mani G, Park BS, Jacques J, Murdoch G, Whetsell W, et al. Differential loss of synaptic proteinsin Alzheimer's disease: implications for synaptic dysfunction. J Alzheimers Dis 2005;7:103-17-80.
(5.) Thorsell A, Bjerke M, Gobom J, Brunhage E, Vanmechelen E, Andreasen N, et al. Neurogranin in cerebrospinal fluid as a marker of synaptic degeneration in Alzheimer's disease. Brain Res 2010;1362:13-22.
(6.) De Vos A, Jacobs D, Struyfs H, Fransen E, Andersson K, Portelius E, et al. C-terminal neurogranin isi ncreased in cerebrospinal fluid but unchanged in plasma in Alzheimer's disease. Alzheimers Dement 2015;11:1461-9.
(7.) Kester MI, Teunissen CE, Crimmins DL, Herries EM, Ladenson JH, Scheltens P, et al. Neurogranin as a cerebrospinal fluid biomarker for synaptic loss insymptomatic Alzheimerdisease. JAMA Neurol 2015;72:1275-80.
(8.) Kvartsberg H, Duits FH, Ingelsson M, Andreasen N, Ohrfelt A, Andersson K, et al. Cerebrospinal fluid levels of the synaptic protein neurogranin correlates with cognitive decline in prodromal Alzheimer's disease. Alzheimers Dement 2015;11:1180-90.
(9.) Kvartsberg H, Portelius E, Andreasson U, Brinkmalm G, Hellwig K, Lelental N, et al. Characterization of the postsynaptic protein neurogranin in paired cerebrospinal fluid and plasma samples from Alzheimer's disease patients and healthy controls. Alzheimers Res Ther 2015; 7:40.
(10.) Portelius E, Zetterberg H, Skillback T, Tornqvist U, Andreasson U, Trojanowski JQ, et al. Cerebrospinal fluid neurogranin: relation to cognition and neurodegeneration in Alzheimer's disease. Brain 2015;138: 3373-85.
(11.) Sanfilippo C, Forlenza O, Zetterberg H, Blennow K. Increased neurogranin concentrations in cerebrospinal fluid of Alzheimer's disease and in mild cognitive impairment due to AD. J Neural Transm 2016;123: 1443-7.
(12.) Mattsson N, Insel PS, Palmqvist S, Portelius E, Zetterberg H, Weiner M, et al. Cerebrospinal fluid tau, neurogranin, and neurofilament light in Alzheimer's disease. EMBO Mol Med 2016;8:1184-96.
(13.) Tarawneh R, D'Angelo G, Crimmins D, Herries E, Griest T, Fagan AM, et al. Diagnostic and prognostic utility of the synaptic marker neurogranin in Alzheimer disease. JAMA Neurol 2016;73:561-71.
(14.) Remnestal J, Just D, Mitsios N, Fredolini C, Mulder J, Schwenk JM, et al. CSF profiling of the human brain-enriched proteome reveals associations of neuromodulin and neurogranin to Alzheimer's disease. Proteomics Clin Appl 2016;10:1242-53.
(15.) De Vos A, Struyfs H, Jacobs D, Fransen E, Klewansky T, De Roeck E, et al. The cerebrospinal fluid neurogranin/ BACE1 ratio is a potential correlate of cognitive decline in Alzheimer's disease. J Alzheimers Dis 2016;53: 1523-38.
(16.) Wellington H, Paterson RW, Portelius E, Tornqvist U, Magdalinou N, Fox NC, et al. Increased CSF neurogranin concentration is specific to Alzheimer disease. Neurology 2016;86:829-35.
(17.) Hellwig K, Kvartsberg H, Portelius E, Andreasson U, Oberstein TJ, Lewczuk P, et al. Neurogranin and YKL-40: independent markers of synaptic degeneration and neuroinflammation in Alzheimer's disease. Alzheimers Res Ther 2015;7:74.
(18.) Janelidze S, Hertze J, Zetterberg H, Landqvist Waldo M, Santillo A, Blennow K, Hansson O. Cerebrospinal fluid neurogranin and YKL-40 as biomarkers of Alzheimer's disease. Ann Clin Transl Neurol 2015;3:12-20.
(19.) Lista S, Toschi N, Baldacci F, Zetterberg H, Blennow K, Kilimann I, et al. Cerebrospinal fluid neurogranin as a biomarker of neurodegenerative diseases: a cross-sectional study. J Alzheimer's Dis 2017;59:1327-34.
(20.) van der Flier WM, Pijnenburg YAL, Prins N, Lemstra AW, Bouwman FH, Teunissen CE, et al. Optimizing patient care and research: the Amsterdam Dementia Cohort. J Alzheimers Dis 2014;41:313-27.
(21.) Teunissen CE, Petzold A, Bennett JL, Berven FS, Brundin L, Comabella M, et al. A consensus protocol for the standardization of cerebrospinal fluid collection and biobanking. Neurology 2009;73:1914 -22.
(22.) Teunissen C, Menge T, Altintas A, Alvarez-Cermeno JC, Bertolotto A, Berven FS, et al. Consensus definitions and application guidelines for control groups in cerebrospinal fluid biomarker studies in multiple sclerosis. Mult Scler 2013;19:1802-9.
(23.) McKhann GM, Knopman DS, Chertkow H, Hyman BT, Jack CR, Kawas CH, et al. The diagnosis of dementia due to Alzheimer's disease: recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimer's Dement 2011;7:263-9.
(24.) Neary D, Snowden JS, Gustafson L, Passant U, Stuss D, Black S, et al. Frontotemporal lobar degeneration: a consensus on clinical diagnostic criteria. Neurology 1998;51:1546-54.
(25.) McKeith IG, Dickson DW, Lowe J, Emre M, O'Brien JT, Feldman H, et al. Diagnosis and management of dementia with Lewy bodies: third report of the DLB Consortium. Neurology 2005;65:1863-72.
(26.) Roman GC, Tatemichi TK, Erkinjuntti T, Cummings JL, Masdeu JC, Garcia JH, et al. Vascular dementia: diagnostic criteria for research studies. Report of the NINDS-AIREN International Workshop. Neurology 1993;43: 250-60.
(27.) Mulder C, Verwey NA, van der Flier WM, Bouwman FH, Kok A, van Elk EJ, et al. Amyloid-beta(1- 42), total tau, and phosphorylated tau as cerebrospinal fluid biomarkers for the diagnosis of Alzheimer disease. Clin Chem 2010;56:248 -53.
(28.) Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ: 25 years of image analysis. Nat Methods 2012;9:671-5.
(29.) R Core Team. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing, 2017. https://www.r-project.org/ (Accessed October 2017).
(30.) Hoffman L, Chandrasekar A, Wang X, Putkey JA, Waxham MN. Neurogranin alters the structure and calcium binding properties of calmodulin. J Biol Chem 2014; 289:14644 -55.
(31.) Diez-Guerra FJ. Neurogranin, a link between calcium/ calmodulin and protein kinase C signaling in synaptic plasticity. IUBMB Life 2010;62:597-606.
(32.) Terry RD, Masliah E, Salmon DP, Butters N, DeTeresa R, Hill R, et al. Physical basis of cognitive alterations in Alzheimer's disease: synapse loss is the major correlate of cognitive impairment. Ann Neurol 1991;30: 572-80.
(33.) Blennow K, Bogdanovic N, Alafuzoff I, Ekman R, Davidsson P. Synaptic pathology in Alzheimer's disease: relation to severity of dementia, but not to senile plaques, neurofibrillary tangles, or the ApoE4 allele. J Neural Transm 1996;103:603-18.
(34.) Clare R, King VG, Wirenfeldt M, Vinters H V. Synapse loss in dementias. J Neurosci Res 2010;88:2083-90.
(35.) Herms J, Dorostkar MM. Dendritic spine pathology in neurodegenerative diseases. Annu Rev Pathol 2016; 11:221-50.
(36.) Hansson KT, Skillback T, Pernevik E, Kern S, Portelius E, Hoglund K, et al. Expanding the cerebrospinal fluid endopeptidome. Proteomics 2017;17:1600384.
(37.) Kusebauch U, Campbell DS, Deutsch EW, Chu CS, Spicer DA, Brusniak MY, et al. Human SRMAtlas: a resource of targeted assays to quantify the complete human proteome. Cell 2016;166:766-78.
Eline A.J. Willemse, [1,2,3] * Ann De Vos,  Elizabeth M. Herries,  Ulf Andreasson,  Sebastiaan Engelborghs,  Wiesje M. van der Flier, [2,7] Philip Scheltens,  Dan Crimmins,  [dagger] Jack H. Ladenson,  Eugeen Vanmechelen,  Henrik Zetterberg, [6,8,9] Anne M. Fagan,  Kaj Blennow,  Maria Bjerke,  and Charlotte E. Teunissen 
 Neurochemistry laboratory, Department of Clinical Chemistry, Amsterdam Neuroscience, VU University Medical Center Amsterdam, the Netherlands;  Alzheimer Center, Department of Neurology, Amsterdam Neuroscience, VU University Medical Center Amsterdam, the Netherlands;  Reference Center for Biological Markers of Dementia (BIODEM), Department of Biomedical Sciences, Institute Born-Bunge, University of Antwerp, Antwerp, Belgium;  ADx NeuroSciences, Ghent, Belgium;  Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, MO;  Clinical Neurochemistry Laboratory, Institute of Neuroscience and Physiology, the Sahlgrenska Academy at the University of Gothenburg, Molndal, Sweden;  VU University Medical Center, Epidemiology & Biostatistics, Amsterdam, the Netherlands;  UCL Institute of Neurology, Department of Molecular Neuroscience, Queen Square, London, United Kingdom;  UK Dementia Research Institute, London, United Kingdom;  Department of Neurology, Knight Alzheimer's Disease Research Center, Hope Center for Neurodegenerative Disorders, Washington University School of Medicine, St. Louis, MO.
* Address correspondence to this author at: Room PK2Br129, Laboratory of Neurochemistry, Department of Clinical Chemistry and Alzheimer center, VU University Medical Center, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands. E-mail firstname.lastname@example.org.
Received October 24, 2017; accepted February 6, 2018.
Previously published online at DOI: 10.1373/clinchem.2017.283028
[C] 2018 American Association for Clinical Chemistry
 Nonstandard abbreviations: AD, Alzheimer disease; CSF, cerebrospinal fluid; WashU, Washington University; ADx, ADx Neuro Sciences; UGot, University of Gothenburg; DLB, dementia with Lewy bodies; FTD, frontotemporal dementia; VaD, Vascular dementia; PBST, PBS containing 0.05%Tween-20; LLOQ, lower limit of quantification; ANCOVA, analysis of covariance; MMSE, Mini-Mental State Examination.
Caption: Fig. 1. Composition of the neurogranin assay calibrator and the epitopes of the neurogranin antibodies, schematically, on silver stain, and on Western blot.
Amino acid sequence of full-length human neurogranin ("parent protein") and its fragments found in CSF and brain tissue (A). The 3 rows above ("antibodies") indicate what epitopes are tagged by the capture and detection antibodies used in the WashU assay (upper row: S10-D23 and G49-G60), ADx assay (middle row: R53-A64 and G62-P75), and UGot assay (bottom row: G52-G65 and V66 -D78). The upper 3 rows depict the amino acid sequences of the calibrators of the assays, of which the numbering is based on the parent protein. Adapted from (9). Silver staining of the WashU, ADx, and UGot calibrators (B). Western blot analysis of neurogranin calibrators, CSF, and brain homogenate stained by the antibodies of the 3 immunoassays(C-H).
Caption: Fig. 2. Neurogranin concentrations in dilutions of 1 brain lysate sample measured by the 3 immunoassays.
y-axis on log2 scale.
Caption: Fig. 3. Three Passing-Bablok regression analysis of the neurogranin assays.
The ADx vs the WashU assays exhibit a proportional difference [slope (95% CI) of regression line: 5.78 (5.36; 6.09)], as well as a systematic difference [intercept (95% CI): -169.72 (-266.41; -56.46)] (A). The UGot vs the WashU assays exhibit a proportional difference [slope (95% CI) of regression line: 5.17 (4.38; 5.83)] but not a systematic difference [intercept (95% CI): -177.48 (-448.19; 45.18)] (B). The UGot vs the ADx assays exhibit a proportional difference [slope (95% CI) of regression line: (0.82 (0.71; 0.94)] but not a systematic difference [intercept (95% CI): 10.36 (-29.37; 57.93)] (C). Orange dots indicate individual CSF samples (n = 108), the dotted lines represent the equation x = y (identity line), and the blue areas show the 95% CIs of the regression lines.
Table 1. Patient sample characteristics. (a) Controls Dementia due to AD N 22 22 Sex (male) 13 (59%) 13 (59%) Age 64 (6) 65 (8) APOE [epsilon]4 carriers 4 (18) 15(68) MMSE 28 [27,29] 18 [16,23] A[beta]42, pg/mL 891 (268) 511 (160) t-Tau, pg/mL 311 (155) 800 (378) p-Tau181, pg/mL 49 (21) 98 (45) DLB FTD VaD N 22 22 20 Sex (male) 21 (96%) 15 (68%) 13 (65%) Age 68 (6) 63 (5) 68 (6) APOE [epsilon]4 carriers 12 (55) 8(36) 13(65) MMSE 25 [21,27] 24 [18,26] 24 [22,26] A[beta]42, pg/mL 730 (299) 914 (245) 586(241) t-Tau, pg/mL 366 (256) 434 (249) 404(188) p-Tau181, pg/mL 54 (33) 55 (26) 53 (22) P value N Sex (male) 0.026 Age 0.023 APOE [epsilon]4 carriers 0.004 MMSE <0.001 A[beta]42, pg/mL <0.001 t-Tau, pg/mL <0.001 p-Tau181, pg/mL <0.001 (a) Data are represented as n (%), mean (SD), or median [interquartile range]. Groups were compared using Fisher exact test for sex and APOE [epsilon]4 carriership, Kruskal-Wallis rank sum test for MMSE, and ANOVA for age, A[beta]42, t-Tau, and p-Tau181. Table 2. Recognition of neurogranin assay calibrators by the WashU assay, ADx assay, and UGotassay. (a) Recognize calibrator of: WashU assay ADx assay WashU (synthetic, full-length) - - Calibrator concentrations: 399 pg/mL <LLOQ 61 pg/mL 12 pg/mL ADx (synthetic, truncated at P75) - Calibrator concentrations: 3.3 pg/mL 500 pg/mL 1.8 pg/mL 300 pg/mL 0.8 pg/mL 150 pg/mL <LLOQ (75,30,15,5,0 pg/mL) UGot (recombinant, full-length) + + - Calibrator concentrations: 5703 pg/mL 18 pg/mL 1772 pg/mL 16 pg/mL 506 pg/mL 22 pg/mL Recognize calibrator of: UGot assay WashU (synthetic, full-length) + + Calibrator concentrations: >ULOQ (b) ADx (synthetic, truncated at P75) -- Calibrator concentrations: <LLOQ UGot (recombinant, full-length) Calibrator concentrations: 4458 pg/mL 1114 pg/mL 279 pg/mL (a) Degree of detection of a calibrator by another assay is scored from + + (very good) to - (very poor). (b) Data of WashU calibrator detection by the UGot assay are not available [>upper limit of quantification (ULOQ)] owing to interference by rabbit IgG in the buffer used to prepare the standards. Fig. 4. Boxplots of CSF neurogranin concentrations (pg/mL) in dementia differential diagnosis groups measured with the WashU assay (A), ADx assay (B), and UGot assay (C). Absolute neurogranin ranges vary among the 3 assays, but the neurogranin concentrations within the clinical groups show the same pattern among the 3 assays. Asterisks indicate significance of * P <0.05, ** P < 0.01, and *** P <0.001 in Bonferroni posthoc comparisons following ANCOVA on the ranked neurogranin concentrations corrected for age and sex. CON AD DLB FTD VaD Median Ng 1711 2659 1594 1817 1742 (pg/ml) 1st Q-3rd Q 1141- 1007- 1135- 1561- 1107- (pg/ml) 2660 6561 2214 2538 2132 CON AD DLB FTD VaD Median Ng 330 462 314 364 325 (pg/ml) 1st Q-3rd Q 236- 367- 199- 279- 193- (pg/ml) 496 736 395 476 434 CON AD DLB FTD VaD Median Ng 371 626 385 462 364 (pg/ml) 1st Q-3rd Q 214- 490- 238- 364- 204- (pg/ml) 487 910 457 632 422
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Proteomics and Protein Markers|
|Author:||Willemse, Eline A.J.; de Vos, Ann; Herries, Elizabeth M.; Andreasson, Ulf; Engelborghs, Sebastiaan;|
|Date:||Jun 1, 2018|
|Previous Article:||Activity-Based Detection of Cannabinoids in Serum and Plasma Samples.|
|Next Article:||Ultrasensitive Detection of Chimerism by Single-Molecule Molecular Inversion Probe Capture and High-Throughput Sequencing of Copy Number Deletion...|