Putting Humpty Dumpty Back Together Again: What Does Protein Quantification Mean in Bottom-Up Proteomics?

Bottom-up proteomics provides peptide measurements and has been invaluable for moving proteomics into large-scale analyses. Commonly, a single quantitative value is reported for each protein-coding gene by aggregating peptide quantities into protein groups following protein inference or parsimony. However, given the complexity of both RNA splicing and post-translational protein modification, it is overly simplistic to assume that all peptides that map to a singular protein-coding gene will demonstrate the same quantitative response. By assuming that all peptides from a protein-coding sequence are representative of the same protein, we may miss the discovery of important biological differences. To capture the contributions of existing proteoforms, we need to reconsider the practice of aggregating protein values to a single quantity per protein-coding gene.

High-Density Lipoprotein Carries Markers That Track With Recovery From Stroke.

RATIONALE: Prospective cohort studies question the value of HDL-C (high-density lipoprotein cholesterol) for stroke risk prediction. OBJECTIVE: Investigate the relationship between long-term functional recovery and HDL proteome and function. METHODS AND RESULTS: Changes in HDL protein composition and function (cholesterol efflux capacity) in patients after acute ischemic stroke at 2 time points (24 hours, 35 patients; 96 hours, 20 patients) and in 35 control subjects were measured. The recovery from stroke was assessed by 3 months, the National Institutes of Health Stroke Scale and modified Rankin scale scores. When compared with control subject after adjustments for sex and HDL-C levels, 12 proteins some of which participate in acute phase response and platelet activation (APMAP [adipocyte plasma membrane-associated protein], GPLD1 [phosphate inositol-glycan specific phospholipase D], APOE [apolipoprotein E], IHH [Indian hedgehog protein], ITIH4 [inter-alpha-trypsin inhibitor chain H4], SAA2 [serum amyloid A2], APOA4 [apolipoprotein A-IV], CLU [clusterin], ANTRX2 [anthrax toxin receptor 2], PON1 [serum paraoxonase/arylesterase], SERPINA1 [alpha-1-antitrypsin], and APOF [apolipoprotein F]) were significantly (adjusted P<0.05) altered in stroke HDL at 96 hours. The first 8 of these proteins were also significantly altered at 24 hours. Consistent with inflammatory remodeling, cholesterol efflux capacity was reduced by 32% (P<0.001) at both time points. Baseline stroke severity adjusted regression model showed that changes within 96-hour poststroke in APOF, APOL1, APMAP, APOC4 (apolipoprotein C4), APOM (apolipoprotein M), PCYOX1 (prenylcysteine oxidase 1), PON1, and APOE correlate with stroke recovery scores (R2=0.38-0.73, adjusted P<0.05). APOF (R2=0.73) and APOL1 (R2=0.60) continued to significantly correlate with recovery scores after accounting for tPA (tissue-type plasminogen activator) treatment. CONCLUSIONS: Changes in HDL proteins during early acute phase of stroke associate with recovery. Monitoring HDL proteins may provide clinical biomarkers that inform on stroke recuperation.

Accurately Assigning Peptides to Spectra When Only a Subset of Peptides Are Relevant.

The standard proteomics database search strategy involves searching spectra against a peptide database and estimating the false discovery rate (FDR) of the resulting set of peptide-spectrum matches. One assumption of this protocol is that all the peptides in the database are relevant to the hypothesis being investigated. However, in settings where researchers are interested in a subset of peptides, alternative search and FDR control strategies are needed. Recently, two methods were proposed to address this problem: subset-search and all-sub. We show that both methods fail to control the FDR. For subset-search, this failure is due to the presence of "neighbor" peptides, which are defined as irrelevant peptides with a similar precursor mass and fragmentation spectrum as a relevant peptide. Not considering neighbors compromises the FDR estimate because a spectrum generated by an irrelevant peptide can incorrectly match well to a relevant peptide. Therefore, we have developed a new method, "subset-neighbor search" (SNS), that accounts for neighbor peptides. We show evidence that SNS controls the FDR when neighbors are present and that SNS outperforms group-FDR, the only other method that appears to control the FDR relative to a subset of relevant peptides.

Genetic control of the mouse HDL proteome defines HDL traits, function, and heterogeneity.

HDLs are nanoparticles with more than 80 associated proteins, phospholipids, cholesterol, and cholesteryl esters. The potential inverse relation of HDL to coronary artery disease (CAD) and the effects of HDL on myriad other inflammatory conditions warrant a better understanding of the genetic basis of the HDL proteome. We conducted a comprehensive genetic analysis of the regulation of the proteome of HDL isolated from a panel of 100 diverse inbred strains of mice (the hybrid mouse diversity panel) and examined protein composition and efflux capacity to identify novel factors that affect the HDL proteome. Genetic analysis revealed widely varied HDL protein levels across the strains. Some of this variation was explained by local cis-acting regulation, termed cis-protein quantitative trait loci (QTLs). Variations in apoA-II and apoC-3 affected the abundance of multiple HDL proteins, indicating a coordinated regulation. We identified modules of covarying proteins and defined a protein-protein interaction network that describes the protein composition of the naturally occurring subspecies of HDL in mice. Sterol efflux capacity varied up to 3-fold across the strains, and HDL proteins displayed distinct correlation patterns with macrophage and ABCA1-specific cholesterol efflux capacity and cholesterol exchange, suggesting that subspecies of HDL participate in discrete functions. The baseline and stimulated sterol efflux capacity phenotypes were associated with distinct QTLs with smaller effect size, suggesting a multigenetic regulation. Our results highlight the complexity of HDL particles by revealing the high degree of heterogeneity and intercorrelation, some of which is associated with functional variation, and support the concept that HDL-cholesterol alone is not an accurate measure of HDL's properties, such as protection against CAD.

Extended Multiplexing of Tandem Mass Tags (TMT) Labeling Reveals Age and High Fat Diet Specific Proteome Changes in Mouse Epididymal Adipose Tissue.

The lack of high-throughput methods to analyze the adipose tissue protein composition limits our understanding of the protein networks responsible for age and diet related metabolic response. We have developed an approach using multiple-dimension liquid chromatography tandem mass spectrometry and extended multiplexing (24 biological samples) with tandem mass tags (TMT) labeling to analyze proteomes of epididymal adipose tissues isolated from mice fed either low or high fat diet for a short or a long-term, and from mice that aged on low versus high fat diets. The peripheral metabolic health (as measured by body weight, adiposity, plasma fasting glucose, insulin, triglycerides, total cholesterol levels, and glucose and insulin tolerance tests) deteriorated with diet and advancing age, with long-term high fat diet exposure being the worst. In response to short-term high fat diet, 43 proteins representing lipid metabolism (e.g. AACS, ACOX1, ACLY) and red-ox pathways (e.g. CPD2, CYP2E, SOD3) were significantly altered (FDR < 10%). Long-term high fat diet significantly altered 55 proteins associated with immune response (e.g. IGTB2, IFIT3, LGALS1) and rennin angiotensin system (e.g. ENPEP, CMA1, CPA3, ANPEP). Age-related changes on low fat diet significantly altered only 18 proteins representing mainly urea cycle (e.g. OTC, ARG1, CPS1), and amino acid biosynthesis (e.g. GMT, AKR1C6). Surprisingly, high fat diet driven age-related changes culminated with alterations in 155 proteins involving primarily the urea cycle (e.g. ARG1, CPS1), immune response/complement activation (e.g. C3, C4b, C8, C9, CFB, CFH, FGA), extracellular remodeling (e.g. EFEMP1, FBN1, FBN2, LTBP4, FERMT2, ECM1, EMILIN2, ITIH3) and apoptosis (e.g. YAP1, HIP1, NDRG1, PRKCD, MUL1) pathways. Using our adipose tissue tailored approach we have identified both age-related and high fat diet specific proteomic signatures highlighting a pronounced involvement of arginine metabolism in response to advancing age, and branched chain amino acid metabolism in early response to high fat feeding. Data are available via ProteomeXchange with identifier PXD005953.

PCSK9 Association With Lipoprotein(a).

RATIONALE: Lipoprotein(a) [Lp(a)] is a highly atherogenic low-density lipoprotein-like particle characterized by the presence of apoprotein(a) [apo(a)] bound to apolipoprotein B. Proprotein convertase subtilisin/kexin type 9 (PCSK9) selectively binds low-density lipoprotein; we hypothesized that it can also be associated with Lp(a) in plasma. OBJECTIVE: Characterize the association of PCSK9 and Lp(a) in 39 subjects with high Lp(a) levels (range 39-320 mg/dL) and in transgenic mice expressing either human apo(a) only or human Lp(a) (via coexpression of human apo(a) and human apolipoprotein B). METHODS AND RESULTS: We show that PCSK9 is physically associated with Lp(a) in vivo using 3 different approaches: (1) analysis of Lp(a) fractions isolated by ultracentrifugation; (2) immunoprecipitation of plasma using antibodies to PCSK9 and immunodetection of apo(a); (3) ELISA quantification of Lp(a)-associated PCSK9. Plasma PCSK9 levels correlated with Lp(a) levels, but not with the number of kringle IV-2 repeats. PCSK9 did not bind to apo(a) only, and the association of PCSK9 with Lp(a) was not affected by the loss of the apo(a) region responsible for binding oxidized phospholipids. Preferential association of PCSK9 with Lp(a) versus low-density lipoprotein (1.7-fold increase) was seen in subjects with high Lp(a) and normal low-density lipoprotein. Finally, Lp(a)-associated PCSK9 levels directly correlated with plasma Lp(a) levels but not with total plasma PCSK9 levels. CONCLUSIONS: Our results show, for the first time, that plasma PCSK9 is found in association with Lp(a) particles in humans with high Lp(a) levels and in mice carrying human Lp(a). Lp(a)-bound PCSK9 may be pursued as a biomarker for cardiovascular risk.

Data Independent Acquisition to Inform the Development of Targeted Proteomics Assays Using a Triple Quadrupole Mass Spectrometer.

Mass spectrometry based targeted proteomics methods provide sensitive and high-throughput analysis of selected proteins. To develop a targeted bottom-up proteomics assay, peptides must be evaluated as proxies for the measurement of a protein or proteoform in a biological matrix. Candidate peptide selection typically relies on predetermined biochemical properties, data from semi-stochastic sampling, or by empirical measurements. These strategies require extensive testing and method refinement due to the difficulties associated with prediction of peptide response in the biological matrix of interest. Gas-phase fractionated (GPF) narrow window data-independent acquisition (DIA) aids in the development of reproducible selected reaction monitoring (SRM) assays by providing matrix-specific information on peptide detectability and quantification by mass spectrometry. To demonstrate the suitability of DIA data for selecting peptide targets, we reimplement a portion of an existing assay to measure 98 Alzheimer's disease proteins in cerebrospinal fluid (CSF). Peptides were selected from GPF-DIA based on signal intensity and reproducibility. The resulting SRM assay exhibits similar quantitative precision to published data, despite the inclusion of different peptides between the assays. This workflow enables development of new assays without additional up-front data acquisition, demonstrated here through generation of a separate assay for an unrelated set of proteins in CSF from the same dataset.

Development of highly multiplex targeted proteomics assays in biofluids using the Stellar mass spectrometer.

The development of targeted assays that monitor biomedically relevant proteins is an important step in bridging discovery experiments to large scale clinical studies. Targeted assays are currently unable to scale to hundreds or thousands of targets. We demonstrate the generation of large-scale assays using a novel hybrid nominal mass instrument. The scale of these assays is achievable with the Stellar™ mass spectrometer through the accommodation of shifting retention times by real-time alignment, while being sensitive and fast enough to handle many concurrent targets. Assays were constructed using precursor information from gas-phase fractionated (GPF) data-independent acquisition (DIA). We demonstrate the ability to schedule methods from an orbitrap and linear ion trap acquired GPF DIA library and compare the quantification of a matrix-matched calibration curve from orbitrap DIA and linear ion trap parallel reaction monitoring (PRM). Two applications of these proposed workflows are shown with a cerebrospinal fluid (CSF) neurodegenerative disease protein PRM assay and with a Mag-Net enriched plasma extracellular vesicle (EV) protein survey PRM assay.

Mag-Net: Rapid enrichment of membrane-bound particles enables high coverage quantitative analysis of the plasma proteome.

Membrane-bound particles in plasma are composed of exosomes, microvesicles, and apoptotic bodies and represent ~1-2% of the total protein composition. Proteomic interrogation of this subset of plasma proteins augments the representation of tissue-specific proteins, representing a "liquid biopsy," while enabling the detection of proteins that would otherwise be beyond the dynamic range of liquid chromatography-tandem mass spectrometry of unfractionated plasma. We have developed an enrichment strategy (Mag-Net) using hyper-porous strong-anion exchange magnetic microparticles to sieve membrane-bound particles from plasma. The Mag-Net method is robust, reproducible, inexpensive, and requires <100 µL plasma input. Coupled to a quantitative data-independent mass spectrometry analytical strategy, we demonstrate that we can collect results for >37,000 peptides from >4,000 plasma proteins with high precision. Using this analytical pipeline on a small cohort of patients with neurodegenerative disease and healthy age-matched controls, we discovered 204 proteins that differentiate (q-value < 0.05) patients with Alzheimer's disease dementia (ADD) from those without ADD. Our method also discovered 310 proteins that were different between Parkinson's disease and those with either ADD or healthy cognitively normal individuals. Using machine learning we were able to distinguish between ADD and not ADD with a mean ROC AUC = 0.98 ± 0.06.

A framework for quality control in quantitative proteomics.

A thorough evaluation of the quality, reproducibility, and variability of bottom-up proteomics data is necessary at every stage of a workflow from planning to analysis. We share real-world case studies applying adaptable quality control (QC) measures to assess sample preparation, system function, and quantitative analysis. System suitability samples are repeatedly measured longitudinally with targeted methods, and we share examples where they are used on three instrument platforms to identify severe system failures and track function over months to years. Internal QCs incorporated at protein and peptide-level allow our team to assess sample preparation issues and to differentiate system failures from sample-specific issues. External QC samples prepared alongside our experimental samples are used to verify the consistency and quantitative potential of our results during batch correction and normalization before assessing biological phenotypes. We combine these controls with rapid analysis using Skyline, longitudinal QC metrics using AutoQC, and server-based data deposition using PanoramaWeb. We propose that this integrated approach to QC be used as a starting point for groups to facilitate rapid quality control assessment to ensure that valuable instrument time is used to collect the best quality data possible.

Brain proteomic analysis implicates actin filament processes and injury response in resilience to Alzheimer's disease.

Resilience to Alzheimer's disease is an uncommon combination of high disease burden without dementia that offers valuable insights into limiting clinical impact. Here we assessed 43 research participants meeting stringent criteria, 11 healthy controls, 12 resilience to Alzheimer's disease and 20 Alzheimer's disease with dementia and analyzed matched isocortical regions, hippocampus, and caudate nucleus by mass spectrometry-based proteomics. Of 7115 differentially expressed soluble proteins, lower isocortical and hippocampal soluble Aß levels is a significant feature of resilience when compared to healthy control and Alzheimer's disease dementia groups. Protein co-expression analysis reveals 181 densely-interacting proteins significantly associated with resilience that were enriched for actin filament-based processes, cellular detoxification, and wound healing in isocortex and hippocampus, further supported by four validation cohorts. Our results suggest that lowering soluble Aß concentration may suppress severe cognitive impairment along the Alzheimer's disease continuum. The molecular basis of resilience likely holds important therapeutic insights.

Unveiling Resilience to Alzheimer's Disease: Insights From Brain Regional Proteomic Markers.

Studying proteomics data of the human brain could offer numerous insights into unraveling the signature of resilience to Alzheimer's disease. In our previous study with rigorous cohort selection criteria that excluded 4 common comorbidities, we harnessed multiple brain regions from 43 research participants with 12 of them displaying cognitive resilience to Alzheimer's disease. Based on the previous findings, this work focuses on 6 proteins out of the 33 differentially expressed proteins associated with resilience to Alzheimer's disease. These proteins are used to construct a decision tree classifier, enabling the differentiation of 3 groups: (i) healthy control, (ii) resilience to Alzheimer's disease, and (iii) Alzheimer's disease with dementia. Our analysis unveiled 2 important regional proteomic markers: Aß peptides in the hippocampus and PA1B3 in the inferior parietal lobule. These findings underscore the potential of using distinct regional proteomic markers as signatures in characterizing the resilience to Alzheimer's disease.

A peptide-centric quantitative proteomics dataset for the phenotypic assessment of Alzheimer's disease.

Alzheimer's disease (AD) is a looming public health disaster with limited interventions. Alzheimer's is a complex disease that can present with or without causative mutations and can be accompanied by a range of age-related comorbidities. This diverse presentation makes it difficult to study molecular changes specific to AD. To better understand the molecular signatures of disease we constructed a unique human brain sample cohort inclusive of autosomal dominant AD dementia (ADD), sporadic ADD, and those without dementia but with high AD histopathologic burden, and cognitively normal individuals with no/minimal AD histopathologic burden. All samples are clinically well characterized, and brain tissue was preserved postmortem by rapid autopsy. Samples from four brain regions were processed and analyzed by data-independent acquisition LC-MS/MS. Here we present a high-quality quantitative dataset at the peptide and protein level for each brain region. Multiple internal and external control strategies were included in this experiment to ensure data quality. All data are deposited in the ProteomeXchange repositories and available from each step of our processing.

Elevated Lipoprotein(a) Levels Lower ABCA1 Cholesterol Efflux Capacity.

CONTEXT: Elevated serum lipoprotein(a) [Lp(a)] levels are associated with increased cardiovascular disease risk. ABCA1-mediated cholesterol efflux from macrophages may be an antiatherogenic process. Plasminogen (PLG) is a driver of ABCA1-mediated cholesterol efflux, and its action is inhibited by purified human Lp(a). OBJECTIVE: To determine the effects of Lp(a) in human serum on ABCA1 cholesterol efflux. METHODS: Cholesterol efflux capacity (CEC) was measured with two different cell-culture models using serum from 76 patients with either low (<50 mg/dL) or high (>50 mg/dL) Lp(a) levels. RESULTS: Using cAMP-stimulated J774 macrophages or baby hamster kidney fibroblasts overexpressing human ABCA1, we show that CEC was lower in patients with high Lp(a) levels compared with patients with low levels (-30.6%, P = 0.002 vs -24.1%, P < 0.001, respectively). Total-serum CEC negatively correlated with Lp(a) levels (r = -0.433, P = 0.0007 vs r = -0.505, P = 0.0011, respectively). These negative associations persisted after adjusting for serum cholesterol, age, sex, and statin use in a multiple linear regression model (adjusted R2 = 0.413 or 0.405, respectively) and were strengthened when further adjusting for the interaction between Lp(a) and PLG levels (adjusted R2 = 0.465 and 0.409, respectively). Total-serum and isolated Lp(a) from patients with high Lp(a) inhibited PLG-mediated ABCA1 cholesterol efflux. CONCLUSION: Total-serum CEC is reduced in patients with high Lp(a) levels. This is in part due to the inhibition of PLG-mediated ABCA1 cholesterol efflux by Lp(a). Our findings suggest an atherogenic role for Lp(a) through its ability to inhibit CEC.

GM-CSF driven myeloid cells in adipose tissue link weight gain and insulin resistance via formation of 2-aminoadipate.

In a GM-CSF driven myeloid cell deficient mouse model (Csf2-/-) that has preserved insulin sensitivity despite increased adiposity, we used unbiased three-dimensional integration of proteome profiles, metabolic profiles, and gene regulatory networks to understand adipose tissue proteome-wide changes and their metabolic implications. Multi-dimensional liquid chromatography mass spectrometry and extended multiplex mass labeling was used to analyze proteomes of epididymal adipose tissues isolated from Csf2+/+ and Csf2-/- mice that were fed low fat, high fat, or high fat plus cholesterol diets for 8 weeks. The metabolic health (as measured by body weight, adiposity, plasma fasting glucose, insulin, triglycerides, phospholipids, total cholesterol levels, and glucose and insulin tolerance tests) deteriorated with diet for both genotypes, while mice lacking Csf2 were protected from insulin resistance. Regardless of diet, 30 mostly mitochondrial, branch chain amino acids (BCAA), and lysine metabolism proteins were altered between Csf2-/- and Csf2+/+ mice (FDR < 0.05). Lack of GM-CSF driven myeloid cells lead to reduced adipose tissue 2-oxoglutarate dehydrogenase complex (DHTKD1) levels and subsequent increase in plasma 2-aminoadipate (2-AA) levels, both of which are reported to correlate with insulin resistance. Tissue DHTKD1 levels were >4-fold upregulated and plasma 2-AA levels were >2 fold reduced in Csf2-/- mice (p < 0.05). GM-CSF driven myeloid cells link peripheral insulin sensitivity to adiposity via lysine metabolism involving DHTKD1/2-AA axis in a diet independent manner.

Human PCSK9 promotes hepatic lipogenesis and atherosclerosis development via apoE- and LDLR-mediated mechanisms.

AIMS: Proprotein convertase subtilisin/kexin type 9 (PCSK9) promotes the degradation of hepatic low-density lipoprotein (LDL) receptors (LDLR), thereby, decreasing hepatocyte LDL-cholesterol (LDL-C) uptake. However, it is unknown whether PCSK9 has effects on atherogenesis that are independent of lipid changes. The present study investigated the effect of human (h) PCSK9 on plasma lipids, hepatic lipogenesis, and atherosclerotic lesion size and composition in transgenic mice expressing hPCSK9 (hPCSK9tg) on wild-type (WT), LDLR⁻/⁻, or apoE⁻/⁻ background. METHODS AND RESULTS: hPCSK9 expression significantly increased plasma cholesterol (+91%), triglycerides (+18%), and apoB (+57%) levels only in WT mice. The increase in plasma lipids was a consequence of both decreased hepatic LDLR and increased hepatic lipid production, mediated transcriptionally and post-transcriptionally by PCSK9 and dependent on both LDLR and apoE. Despite the lack of changes in plasma lipids in mice expressing hPCSK9 and lacking LDLR (the main target for PCSK9) or apoE (a canonical ligand for the LDLR), hPCSK9 expression increased aortic lesion size in the absence of apoE (268 655 ± 97 972 µm² in hPCSK9tg/apoE⁻/⁻ vs. 189 423 ± 65 700 µm(2) in apoE⁻/⁻) but not in the absence of LDLR. Additionally, hPCSK9 accumulated in the atheroma and increased lesion Ly6C(hi) monocytes (by 21%) in apoE⁻/⁻ mice, but not in LDLR⁻/⁻ mice. CONCLUSIONS: PCSK9 increases hepatic lipid and lipoprotein production via apoE- and LDLR-dependent mechanisms. However, hPCSK9 also accumulate in the artery wall and directly affects atherosclerosis lesion size and composition independently of such plasma lipid and lipoprotein changes. These effects of hPCSK9 are dependent on LDLR but are independent of apoE.