key: cord-0257390-3bcwtj98 authors: Ping, Lingyan; Kundinger, Sean R.; Duong, Duc M.; Yin, Luming; Gearing, Marla; Lah, James J.; Levey, Allan I.; Seyfried, Nicholas T. title: Global quantitative analysis of the human brain proteome and phosphoproteome in Alzheimer’s disease date: 2020-05-21 journal: bioRxiv DOI: 10.1101/2020.05.19.105197 sha: a94753edfe3e0178ec9178f64769b55bb005022f doc_id: 257390 cord_uid: 3bcwtj98 Alzheimer’s disease (AD) is characterized by an early, asymptomatic phase (AsymAD) in which individuals exhibit amyloid-beta (Aβ) plaque accumulation in the absence of clinically detectable cognitive decline. Here we report an unbiased multiplex quantitative proteomic and phosphoproteomic analysis using tandem mass tag (TMT) isobaric labeling of human post-mortem cortex (n=27) across pathology-free controls, AsymAD and symptomatic AD individuals. With off-line high-pH fractionation and liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS) on an Orbitrap Lumos mass spectrometer, we identified 11,378 protein groups across three TMT 11-plex batches. Immobilized metal affinity chromatography (IMAC) was used to enrich for phosphopeptides from the same TMT-labeled cases and 51,736 phosphopeptides were identified. Of these, 48,992 were quantified by TMT reporter ions representing 33,652 unique phosphosites. Two reference standards in each TMT 11-plex were included to assess intra- and inter-batch variance at the protein and peptide level. This comprehensive human brain proteome and phosphoproteome dataset will serve as a valuable resource for the identification of biochemical, cellular and signaling pathways altered during AD progression. and volumes were normalized with additional lysis buffer. Samples were reduced with 1 mM dithiothreitol (DTT) for 30 min, followed by 5 mM iodoacetamide (IAA) alkylation in the dark for another 30 min. Samples were diluted 4-fold with 50 mM triethylammonium bicarbonate (TEAB) before incubating with Lysyl endopeptidase (Wako) at 1:100 (w/w) for 12 hr. Trypsin (Promega) was then added at a 1:50 (w/w) ratio and digestion was carried out for another 12 hr after the urea concentration was diluted to 1 M with 50 mM TEAB. The peptide solutions were desalted with a C18 Sep-Pak column (Waters). Briefly, the Sep-Pak columns were activated with 3 × 1.5 mL of methanol, then equilibrated with 6 × 1.5 mL 0.1% triflouroacetic acid (TFA). The samples were loaded after acidification to a final concentration of 1% formic acid (FA) and 0.1% TFA. Each column was washed with 6 × 1.5 mL 0.1% TFA. Elution was performed with 2 × 1.5 mL 50% acetonitrile. A 600 µL aliquot from each sample was pooled and the mixture was divided into 6 global internal standard (GIS) samples with a total volume of 2400 µL each, consistent with our previous work 22 , and peptide solutions were dried by vacuum (Labconco). All 27 samples from 3 groups and 6 GIS were divided and labeled using two sets of 5 mg 11 plex TMT reagents (Thermo Scientific A34808, Lot No for TMT 10plex: SI258088, and 131C channel: SJ258847). The batch arrangement is provided in Supplementary Table 2 . Briefly, each of the TMT reagents were dissolved in 256 μL anhydrous acetonitrile and the same channel was combined together from two 5 mg reagents. The samples were reconstituted in 400 μL of 100 mM TEAB buffer and mixed with 3.2 mg (164 μL) of the corresponding labeling reagent channel. The reaction was incubated for 1 hr and subsequently quenched with 32 μL of 5% hydroxylamine (Pierce). For each TMT plex, labeled peptides from all 11 channels were mixed and desalted with a 500 mg Sep-Pak column (Waters). The labeled peptide mixture was eluted in 4.5 mL of 50% acetonitrile and dried by vacuum. The off-line high pH fractionation method was adapted from the CPTAC protocol 23 . Briefly, dried samples were re-suspended in high pH loading buffer (1 mM ammonium formate, 2% (v/v) acetonitrile) and loaded onto a ZORBAX Extend 300 C18 columns (4.6 mm x 250 mm with 5 µm beads) from Agilent. An Agilent 1100 HPLC system was used to carry out the fractionation. Ammonium formate (pH 10), diluted to a concentration of 4.5 mM in 2% (v/v) acetonitrile, was used as basic-pH reverse phase solvent A, whereas 4.5 mM ammonium formate (pH 10) in 90% (v/v) acetonitrile was used as paired solvent B. The peptides were eluted in a 96-min gradient with 0%-16% mobile phase B from 7-13 min, 16%-40% B from 13-73 min, 40%-44% B from 73-77 min, 44%-60% B from 77-82 min and kept at 60% B until the end with a flow rate of 0.8 mL/min. A total of 77 individual fractions were collected across the gradient from 14 to 91 minutes, at a rate of 1 fraction per minute. The 77 individual fractions were then pooled into 24 fractions by combining 3 off-line fractions into one with alternating combinations, except for the 24th fraction, which also contained the last 5 remaining individual fractions (Supplementary Figure 1) . Fractions were acidified to a final concentration of 1% FA, and 5% of sample volumes were dried and reserved for proteome analysis. For phosphorylated peptide enrichment, 95% of the 24 high-pH fractions were further combined into 12 fractions in an alternating manner (1 and 13, 2 and 14, etc.). Peptide amounts were assumed to be equally distributed in all fractions. The IMAC enrichment method was performed according to CPTAC protocol with some minor modifications 23 . Briefly, 1200 μL of slurry, in which the beads/solvent ratio is 1:1 (v/v), was utilized for one batch of TMT fractions. Beads were stripped of nickel with 8 mL of 100 mM EDTA and then equilibrated with 8 mL of 50 mM FeCl 3 both by end-to-end rotation for 30 min. To remove excess Fe 3+ ion, beads were washed with 3 × 8 mL of water and resuspended in 2.4 mL of 1:1:1 (v/v/v) ratio of acetonitrile/methanol/0.01%(v/v) acetic acid. The beads were re-rinsed with 2.4 mL of 100% acetonitrile/0.1% TFA and divided into 12 tubes. The supernatant was removed before the peptide mixture was added. All 12 dried fractions were reconstituted in 0.4 mL of 50% acetonitrile/0.1% TFA and then diluted 1:1 with 100% acetonitrile/0.1% TFA to obtain a final 75% acetonitrile/0.1% TFA peptide solution at a concentration of 0.5 μg/μl. The peptide mixture was incubated with treated beads for 30 min with endto-end rotation. Enriched IMAC beads were resuspended in 100 μL of 80% acetonitrile/0.1% TFA before the stage tips were conditioned. Stage tips were equilibrated with 2 × 100 μL methanol washes, 2 × 100 μL 100% acetonitrile/0.1% trifluoroacetic acid washes, followed by 2 × 100 μL of 1% FA washes. The IMAC bead slurry was loaded onto the stage tips and washed with 3 × 100 μL of 80% acetonitrile/0.1% TFA, then 3 × 100 μL of 1% FA. The phosphorylated peptides were released from IMAC beads by 3 × 100 μL 500 mM dibasic sodium phosphate (Na 2 HPO 4, Sigma, S9763), pH 7.0, and washed by 3 × 100 μL 1% FA. The phosphorylated peptides were eluted from stage tips by 3 × 100 μL 50% acetonitrile/0.1% FA. The phosphorylated peptide solutions were dried with vacuum. Both proteome and phosphoproteome samples were run on a Fusion Lumos equipped with a NanoFlex nano-electrospray source (ThermoFisher). The same volume of loading buffer (19 μL of 0.1% TFA) was added to each of the fractions assuming equal distribution of peptide concentration across all 24 proteomic subfractions. Therefore, an equal 2 μL (1 μg equivalent) of each fraction was loaded for proteomic analysis. It was assumed that phosphorylated peptides were ~ 1% (w/w) of all peptides. The same volume of loading buffer (7 μL, 0.1% TFA) was added to all IMAC elution samples, and of this, 2 μL (1 μg equivalent) was analyzed by mass spectrometry. All proteome and phosphoproteome samples were separated on 25 cm long 75 μm ID fused silica columns (New Objective, Woburn, MA) packed in-house with 1.9 μm Reprosil-Pur C18-AQ resin (Dr Maisch). All fractions were eluted over a 140 minute gradient using an Easy nLC 1200 (Thermofisher). The gradient started with 1% buffer B (A: water with 0.1% formic acid and B: 80% acetonitrile in water with 0.1% formic acid) and went to 7% in 3 minutes, then increased from 7% to 30% in 137 minutes, then to 95% within 5 minutes and finally staying at 95% for 25 minutes. (GAIIGLMVGGVV), Aβ42 (GAIIGLMVGGVVIA), and Aβ43 (GAIIGLMVGGVVIAT). Sequences that map to tau microtubule-binding repeat (MTBR) domains were also set as an additional entry encompassing residues 224-370 (tau 2N4R Isoform, 1-441) in the Uniprot sequence, while all tau isoform sequences were modified by removing MTBR peptides and replicated as new "deltaMTBR" entries 29, 30 . Separation and quantification of these peptide sequences facilitated the investigation of APOE allele, Aβ peptide and MTBR peptide-specific regulation of biology in AD datasets 29 [Data Citation 3]. The respective FASTA database used in this study was deposited on Synapse (syn20820455). The SEQUEST HT search engine was used and parameters were identical for both total and IMAC proteomes and specified as the following: fully-tryptic specificity; maximum of two missed cleavages; minimum peptide length of 6; fixed modifications for TMT tags on lysine residues and peptide N-termini (+ 229.162932 Da) and carbamidomethylation of cysteine residues (+ 57.02146 Da); variable modifications for oxidation of methionine residues (+ 15.99492 Da); deamidation of asparagine and glutamine (+ 0.984 Da); phosphorylation of serine, threonine and tyrosine (+ 79.9663 Da); precursor mass tolerance of 20 ppm; fragment mass tolerance of 0.05 daltons. Percolator was used to filter peptide spectral matches (PSM) and peptides to a false discovery rate (FDR) of less than 1% using target-decoy strategy. The phosphosite localization site threshold was set to 0.75, ensuring <5% false-localization rate (FLR) of PTM assignments as described 31 . Following spectral assignment, peptides were assembled into proteins and were further filtered based on the combined probabilities of their constituent peptides to a final FDR of 1%. In cases of redundancy, shared peptides were assigned to the protein sequence in adherence with the principles of parsimony. Reporter ions were quantified from MS2 scans using an integration tolerance of 20 ppm with the most confident centroid setting. The search results and TMT quantification are included [Data Citation 4]. All files have been deposited on Synapse (syn20820053).These include sample trait (syn20820456, Data Citation 1), all mass spectrometry raw files (n=108) from both total proteome and phosphoproteome We utilized a modified version of the CPTAC protocol to identify the total proteome and phosphoproteome from the same cases across different stages of AD. Control, AsymAD and AD tissues were randomized across the 3 batches (each containing 11 samples) with 9 individual cases per batch (Figure 1a and Supplementary tables 1 and 2). Two TMT channels in each batch were dedicated to global reference internal standards (GIS), representing an equivalent amount of pooled peptides from all cases, which allows assessment of the intra-and inter-batch variance 22 . To reduce sample complexity and increase proteome depth prior to LC-MS/MS, we employed off-line high-pH reversed-phase fractionation essentially as described in the CPTAC protocol 23 . A total of 77 individual fractions were collected and combined into 24 fractions for total proteome analysis (Supplementary Figure 1) for each batch. A step-wise concatenation strategy was used for pooling the fractions. A total of 5% of the material by volume was used for total proteome analysis, and the remaining 95% of the sample was used for phosphopeptide enrichment by immobilized affinity chromatography (IMAC) with Fe 3+ -loaded nitrilotriacetic acid (NTA) beads. Each of the 24 fractions were pooled into 12 subfractions prior to IMAC. Both the total proteome (n=72 fractions across 3 batches) and phosphoproteome (n=36 fractions across 3 batches) were analyzed by LC-MS/MS with high-resolution precursor and MS/MS scans on an Orbitrap Fusion Lumos mass spectrometer. For the total proteome runs, a total of 164,034 unique peptides were identified that mapped to 11,378 protein groups at a 1% FDR on the peptide spectrum match (PSM) level across all batches, which represented 10,373 coding gene products. The total numbers of identified peptides, proteins and PSMs for all batches in the total proteome are listed in Table 1 . For each batch, there were approximately 10,000 protein groups identified (Figure 1b) , which was comparable to the depth achieved in the CPTAC protocol using different tissue sources 32, 33 . The confidence of identification for peptide and protein is highly related with the number of PSMs and unique peptides. In the total proteome dataset, more than 77% of the proteins were identified with 2 or more unique peptides (Figure 2a) , while each unique peptide averaged approximately 6 PSMs (Table 1) . Approximately 93% of all proteins were identified with at least 2 PSMs (Figure 2b) . To obtain deep coverage of the phosphoproteome, enrichment strategies are usually applied due to the relatively low abundance of phosphorylation. To assess the quality of our IMAC phosphopeptide enrichment method, we calculated the percent phosphopeptide content (peptide level) in both the total proteome and IMAC phosphoproteome datasets (Tables 1 and 2, and Figure 1d ). The total proteome identified a total of 164,034 unique peptides, and approximately 2% were phosphopeptides (Figure 1d ). Although the IMAC dataset identified less peptides overall (n=72,138), approximately 71% of the IMAC proteome were phosphopeptides ( Table 2 , and Figure 1d ). The IMAC enrichment method therefore led to an 18-fold increase in phosphopeptide identification using half of the instrument time. We set the threshold of phosphosite identification was set to 0.75 by SEQUEST, estimating less than 5% false localization rate (FLR) of each assigned site. After filtering, approximately 83% of all phosphosites identified had localization scores greater than or equal to 0.99. There were 51,736 phosphorylated peptides with 33,652 unique individual phosphosites in total mapping to 8,415 proteins. A total of 34,379 of the phosphorylated peptides were identified in at least two of the three IMAC TMT batches (Figure 1c) . These figures are similar to the depth reported using the same protocol from breast cancer tissue 34 . The numbers of identified peptides in each IMAC batch are listed in Table 2 with calculation of phosphorylation enrichment at both the total peptide and PSM level. Of note, the phosphorylated peptides showed slightly higher level of enrichment at the PSM level (83.55% as average) than peptide level (71.72% as average), which indicates that phosphopeptides were more intense and thus more frequently sequenced by LC-MS/MS than an average unmodified peptide in the phosphoproteome. Indeed, each non-phosphopeptide was identified by an average of 3.5 PSMs, whereas each phosphopeptide was identified by an average of 7 PSMs ( Table 2 ). In total, 74% of all phosphopeptides identified by IMAC enrichment had two or more PSMs, which is consistent with the frequency of PSMs for peptides identified from the total proteome (Figure 2c and 2d) . Thus, although phosphopeptides were highly-enriched in the IMAC proteome, they were sampled at a rate generally consistent with non-phosphopeptides from the total proteome, allowing greater sequencing depth of the phosphoproteome. A major advantage of TMT approaches is the ability to quantify multiple samples in a single run, thereby critically reducing overall MS instrument time. This becomes especially important when the total number increases to dozens or even hundreds of samples 30, 35, 36 . Typically, one or more TMT channels are dedicated for global internal standard(s) (GIS) and included in all batches, which can be used to normalize the measurement for protein or peptide signal from all samples across all batches 22 . In this study, we included two pooled reference standards in each 11-plex TMT batch (channels 126 and 131C), which allows normalization within and across TMT batches (Figure 1a , Supplementary Tables 1 and 2, data citation 1). The two reference standards essentially serve as technical replicates (i.e., a null-experiment), which can be used to assess the variance in measurements. Thus, the degree of the signal variation between two internal GIS channels can be used as a threshold to further filter out the poor quantitation data. Indeed, the signal of proteins and peptides from 126 and 131C channels were very consistent and showed very good linear correlation across all three batches (Figure 3a) . We also consistently observed a strong correlation at both the peptide level from the total proteome (Figure 3b ) and phosphoproteome (Figure 3c) . Notably, some peptides exhibited large variation in signal between the two pooled standard channels, especially those peptides with lower total signal abundance as previously described 22 . According to the central limit theorem, the log 2 ratio for the two GIS channels (log 2 TMT channels 126/131C) should fit a standard Gaussian distribution with the mean at or near zero (Supplemental Figure 2a) , which can be used to assess the technical variation of measurements 37, 38 . This allows end-users of the datasets to impose a filtering criterion that can be used to remove peptides or proteins that do not meet variance metrics (>2 standard deviations (SD) from the mean). Following this filtering criteria, a total of 1,123 peptides were filtered out of the analyses due to large variance, equivalent to ~4% of all quantitated peptides in IMAC Batch 1. In batch 2 and batch 3, there were 2,465 and 2,289 peptides filtered out by the >2SD standard, representing 5% of all peptides identified respectively (Supplementary Figure 2b) . It is also worth noting that batch effects may be significant when the sample number is large and the variation due to sample preparation cannot be ignored. In this case, post-hoc data normalization strategies should be employed to remove these batch effects 39, 40 . In this project, however, this step was not necessary given the relatively modest sample size (n=27) and since all samples were digested at the same time. Aβ plaque and hyperphosphorylated tau neurofibrillary tangle (NFT) accumulation in the brain are the core pathological hallmarks of AD 41, 42 . Thus, as a quality control of our measurements, we assessed the levels of Aβ and tau in our dataset. To confirm increased Aβ levels in diseased cases, the ion intensities from first two tryptic peptides of Aβ were used as a surrogate for amyloid levels in the brain 5 , corresponding to residues 6-16 (Peptide 1) and 17-28 (Peptide 2) of the Aβ sequence, since the C-terminal non-tryptic peptides were not stably detected in all batches. Indeed, both of these two peptides showed significant increase in AsymAD and AD groups comparing with control samples (Figure 4a, Data Citation 5) . Additionally, measurements of Peptide 1 and Peptide 2 were highly correlated (Figure 4b) . Given this, the sum intensity of the two peptides was used to represent Aβ levels in each sample, which showed significant increase in both AsymAD and AD samples when compared to Control (Figure 4c) . Another hallmark of AD is hyperphosphorylated tau 43 , which is the core component of neurofibrillary tangles (NFTs) in diseased neurons. Remarkably, 21 phosphorylated tau peptides were detected in total proteome even without IMAC enrichment, which highlights the robust phosphorylation of this protein in AD brain (Figure 5a) . After IMAC enrichment, a total of 112 tau phosphopeptides were identified. Of note, there were 47 peptides containing two or more phosphosites, which was close to 42% of all phosphopeptides mapped to tau (Figure 5a ). Since the MTBR domains form the core of neurofibrillary tangles and is required to seed tau aggregation 44, 45 , it was set as an additional protein entry as MAPT-MTBR within the database, while all other tau isoforms were replaced as new "deltaMTBR" entries after the MTBR sequence was removed from the original sequences. As shown in Figure 5b and 5c, both MAPTdeltaMTBR (MAPTΔMTBR) and MAPT-MTBR show differences between AD and control groups. However, as expected, the effect size (log 2 fold change) for the tau MTBR is larger than the tau with the ΔMTBR in AD. The tau MTBR sequenced from the phosphoproteome, which contained stronger phosphopeptides signal, yielded even better separation between AD and control groups compared with MTBR sequenced from the total proteome. A one-way ANOVA of peptide levels across three groups (CTL, AsymAD and AD) was also performed (Data Citation 5), and peptide volcano plots were calculated, showing log 2 fold changes and log 10transformed p-values of peptides from both the total proteome and phosphoproteome datasets (Figure 5d ). In total proteome peptide data, we observed both Peptide 1 and Peptide 2 from Aβ to be significantly increased in AD when compared to controls. Additionally, the tau phosphopeptides were among the most changed peptides between AD and control, with significantly increased peptide abundances (Figure 5d ). In agreement with this, tau phosphopeptides were among the most significantly-changed peptides in the IMAC proteome as well. To illustrate this, all tau phosphopeptides quantified in more than two batches were colored according to the degree of fold change between AD and control from IMAC proteome (Figure 5e ). Importantly, the IMAC enrichment allowed deep sequencing and quantification of phosphopeptides mapping to the Proline-rich (Pro-rich) domain (residues 103-244) and MTBR (residues 244-368) domain 46 . Both these regions showed the most consistently increasing in abundance in AD compared with other regions of the tau protein. Ultimately, these deep human brain proteomic and phosphoproteomic datasets serve as a valuable resource for a variety of research endeavors including, but not limited to, the following applications: This dataset provides a reference for relative protein abundance in brain, especially if an investigator wants to determine whether their protein of interest is abundantly expressed in human brain. [Data Citation 4]. There were three separate clinical and pathological groups of human post-mortem tissues representing three stages of AD. One can compare the expression differences between different stages at the protein, peptide or phosphopeptide level. The volcano plots shown in Figure 5d displays the substantive changes in peptide levels between AD and control groups. The same analysis between AsymAD and control can also be applied. This analysis also includes the quantification of peptides with and without phosphorylation sites in the same peptide within the same sample, which can greatly benefit the investigators working to fully describe the phosphorylation stoichiometry of certain proteins [Data Citation 4]. In this dataset, there were more than 10,000 proteins quantified which is more than enough to conduct systems-level analysis. WeiGhted Co-expression Network Analysis (WGCNA) and related algorithms can be utilized for systems-based network analyses, which generate modules of proteins clustered by correlated expression patterns 5, 30, 36 . The protein clusters can then be correlated to molecular functions and pathways. These programs can also be used to correlate expression clusters to various biological traits. Furthermore, the cell-type specificity of individual proteins may be investigated according to the module membership of a protein and the brain cell-type enrichment data for that particular module. Pathway analysis is routine with software 47 or web services 48, 49 to analyze different high-throughput omics data, like genomics, transcriptomics, proteomics, lipidomics and metabolomics. Pathway analyses help to organize a list of proteins into a cohesive list of pathway maps to interpret proteomics results. These analyses have proved to be a very powerful interpretation tool in biological research, facilitating novel insights in disparate fields including development 50 , apoptosis 51 , cancer 52, 53 , and other diseases 54, 55 . Several biological pathways have been linked to AD using similar methods 30, [56] [57] [58] . Given the excellent coverage of the AD proteome and phosphoproteme from the same samples described here, this dataset may therefore serve as a useful resource for pathway analysis. A protein domain or motif is a part of a given protein sequence that serves as a substrate for kinases or other enzymes to recognize and chemically modify, and is replicated in other sequences in the proteome, playing conserved roles in protein function 59, 60 . Recent advances in genomics and proteomics sequencing following the development of bioinformatics 61,62 making possible large-scale domain or motif analyses. As kinases reliably phosphorylate motif sequences specific to that enzyme, the altered phosphorylation of certain motifs may reflect impaired kinase dynamics in AD. Given the enhanced coverage of the AD proteome and phosphoproteome, this dataset can be an excellent tool for AD-related domain or motif analysis. Due to the multiplexing nature of the TMT method, proteomic sample processing has become increasingly high throughput and a more popular mode of research. As innovative technical advances in instrumentation, computing and processing have steadily improved, TMT-labeled peptide analyses have begun to be applied to targeted proteomic methods, such as TOMAHAQ 63 . In this dataset, we have identified 164,034 peptides and 51,736 phosphopeptides through TMT isobaric labeling. Importantly, this dataset includes peptidespecific characteristics such as intensity, charge and modification state, which can serve as a resource to reference for targeted proteome analyses in the future. (b) The adjusted correlation of normalized abundance from 126 and 131 channels after log 2 ratio of (normalized abundance of 126/131) data were filtered with 2 standard deviations. Those peptides that were >2 standard deviation (grey points) were excluded from further analyses, while peptides with ≤2 SD were included (black dots). After filtering out variable peptides from the IMAC proteome, the GIS channels were better correlated. Log 2 (ratio of two GIS channels) Log 2 (ratio of two GIS channels) Alzheimer's Disease Facts and Figures Report Alzheimer's Disease International (ADI), ADI website Toward defining the preclinical stages of Alzheimer's disease: recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease The impact of amyloid-beta and tau on prospective cognitive decline in older individuals A Multi-network Approach Identifies Protein-Specific Co-expression in Asymptomatic and Symptomatic Alzheimer's Disease Alternative splicing and the evolution of phenotypic novelty The emerging era of genomic data integration for analyzing splice isoform function Quantitative, high-resolution proteomics for data-driven systems biology Targeted proteome investigation via selected reaction monitoring mass spectrometry Protein Phosphorylation is a Key Mechanism in Alzheimer's Disease Multiplexed Phosphoproteomic Study of Brain in Patients with Alzheimer's Disease and Age-Matched Cognitively Healthy Controls Pharmacologic inhibition of ROCK2 suppresses amyloid-beta production in an Alzheimer's disease mouse model Glycogen synthase kinase-3 induces Alzheimer's disease-like phosphorylation of tau: Generation of paired helical filament epitopes and neuronal localisation of the kinase Cdk5 Is a Key Factor in Tau Aggregation and Tangle Formation In Vivo Decreased levels of protein kinase C in Alzheimer brain Microtubule-Affinity Regulating Kinase (MARK) Is Tightly Associated with Neurofibrillary Tangles in Alzheimer Brain: A Fluorescence Resonance Energy Transfer Study Grundke-Iqbal, I. & Iqbal, K. Phosphoprotein phosphatase activities in Alzheimer disease brain More than 100,000 detectable peptide species elute in single shotgun proteomics runs but the majority is inaccessible to data-dependent LC-MS/MS Proteome-Wide Evaluation of Two Common Protein Quantification Methods Identification of 2D-gel proteins: a comparison of MALDI/TOF peptide mass mapping to mu LC-ESI tandem mass spectrometry Evaluation of twodimensional gel electrophoresis-based proteome analysis technology Global quantitative analysis of the human brain proteome in Alzheimer's and Parkinson's Disease. Scientific data 5 Reproducible workflow for multiplexed deep-scale proteome and phosphoproteome analysis of tumor tissues by liquid chromatography-mass spectrometry Use of high-pH (basic/alkaline) mobile phases for LC-MS or LC-MS/MS bioanalysis Multidimensional chromatography coupled to mass spectrometry in analysing complex proteomics samples Deep Profiling of Proteome and Phosphoproteome by Isobaric Labeling, Extensive Liquid Chromatography, and Mass Spectrometry The Consortium to Establish a Registry for Alzheimer's Disease (CERAD). Part II. Standardization of the neuropathologic assessment of Alzheimer's disease Neuropathological stageing of Alzheimer-related changes Effects of APOE Genotype on Brain Proteomic Network and Cell Type Changes in Alzheimer's Disease. Frontiers in molecular neuroscience 11 Large-scale proteomic analysis of Alzheimer's disease brain and cerebrospinal fluid reveals early changes in energy metabolism associated with microglia and astrocyte activation Comparison of alternative MS/MS and bioinformatics approaches for confident phosphorylation site localization Evaluation of NCI-7 Cell Line Panel as a Reference Material for Clinical Proteomics Proteogenomics connects somatic mutations to signalling in breast cancer Large-scale proteomic analysis of human brain identifies proteins associated with cognitive trajectory in advanced age Integrated Proteomics Reveals Brain-Based Cerebrospinal Fluid Biomarkers in Asymptomatic and Symptomatic Alzheimer's Disease. bioRxiv Multiplex SILAC analysis of a cellular TDP-43 proteinopathy model reveals protein inclusions associated with SUMOylation and diverse polyubiquitin chains Quantitative phosphoproteomics of Alzheimer's disease reveals cross-talk between kinases and small heat shock proteins Multibatch TMT Reveals False Positives, Batch Effects and Missing Values Large-scale proteomic analysis of Alzheimer's disease brain and cerebrospinal fluid reveals early changes in energy metabolism associated with microglia and astrocyte activation Alzheimer's disease: initial report of the purification and characterization of a novel cerebrovascular amyloid protein Abnormal phosphorylation of the microtubule-associated protein tau (tau) in Alzheimer cytoskeletal pathology Core candidate neurochemical and imaging biomarkers of Alzheimer's disease Structure and pathology of tau protein in Alzheimer disease Conformation determines the seeding potencies of native and recombinant Tau aggregates Cryo-EM structures of tau filaments from Alzheimer's disease EasyGO: Gene Ontology-based annotation and functional enrichment analysis tool for agronomical species Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists STRING v9.1: protein-protein interaction networks, with increased coverage and integration Subcellular transcriptomes and proteomes of developing axon projections in the cerebral cortex Functional interplay between caspase cleavage and phosphorylation sculpts the apoptotic proteome Functional proteomics to dissect tyrosine kinase signalling pathways in cancer Proteomic contributions to personalized cancer care Temporal analysis of phosphotyrosinedependent signaling networks by quantitative proteomics A proteomic approach to obesity and type 2 diabetes The role of Wnt signaling in neuronal dysfunction in Alzheimer's Disease Deep Multilayer Brain Proteomics Identifies Molecular Networks in Alzheimer's Disease Progression The role of mitogen-activated protein kinase pathways in Alzheimer's disease A million peptide motifs for the molecular biologist Comparative analysis of 1196 orthologous mouse and human full-length mRNA and protein sequences PSSMSearch: a server for modeling, visualization, proteome-wide discovery and annotation of protein motif specificity determinants Leitmotif: protein motif scanning 2.0 TomahaqCompanion: A Tool for the Creation and Analysis of Isobaric Label Based Multiplexed Targeted Assays 1 2 3 6 5 4 9 8 7 11 12 13 16 15 14 19 18 17 10 21 22 23 24 20 1 2 3 6 5 4 9 8 7 11 12 13 16 15 14 19 18 17 10 21 22 23 24 20 1 2 3 6 5 4 9 8 7 11 12 13 16 15 14 19 18 17 10 21 22 23 24 20 24 24 24 24 24