MS-EmpiRe Utilizes Peptide-level Noise Distributions for Ultra-sensitive Detection of Differentially Expressed Proteins

Baldur: Bayesian Hierarchical Modeling for Label-Free Proteomics with Gamma Regressing Mean-Variance Trends

Label-free proteomics is a fast-growing methodology to infer abundances in mass spectrometry proteomics. Extensive research has focused on spectral quantification and peptide identification. However, research toward modeling and understanding quantitative proteomics data is scarce. Here we propose a Bayesian hierarchical decision model (Baldur) to test for differences in means between conditions for proteins, peptides, and post-translational modifications. We developed a Bayesian regression model to characterize local mean-variance trends in data, to estimate measurement uncertainty and hyperparameters for the decision model. A key contribution is the development of a new gamma regression model that describes the mean-variance dependency as a mixture of a common and a latent trend-allowing for localized trend estimates. We then evaluate the performance of Baldur, limma-trend, and t test on six benchmark datasets: five total proteomics and one post-translational modification dataset. We find that Baldur drastically improves the decision in noisier post-translational modification data over limma-trend and t test. In addition, we see significant improvements using Baldur over the other methods in the total proteomics datasets. Finally, we analyzed Baldur's performance when increasing the number of replicates and found that the method always increases precision with sample size, while showing robust control of the false positive rate. We conclude that our model vastly improves over popular data analysis methods (limma-trend and t test) in several spike-in datasets by achieving a high true positive detection rate, while greatly reducing the false-positive rate.

Low-dose Bisphenol A exposure alters the functionality and cellular environment in a human cardiomyocyte model

Early embryonic development represents a sensitive time-window during which the foetus might be vulnerable to the exposure of environmental contaminants, potentially leading to heart diseases also later in life. Bisphenol A (BPA), a synthetic chemical widely used in plastics manufacturing, has been associated with heart developmental defects, even in low concentrations. This study aims to investigate the effects of environmentally relevant doses of BPA on developing cardiomyocytes using a human induced pluripotent stem cell (hiPSC)-derived model. Firstly, a 2D in vitro differentiation system to obtain cardiomyocytes from hiPSCs (hiPSC-CMs) have been established and characterised to provide a suitable model for the early stages of cardiac development. Then, the effects of a repeated BPA exposure, starting from the undifferentiated stage throughout the differentiation process, were evaluated. The chemical significantly decreased the beat rate of hiPSC-CMs, extending the contraction and relaxation time in a dose-dependent manner. Quantitative proteomics analysis revealed a high abundance of basement membrane (BM) components (e.g., COL4A1, COL4A2, LAMC1, NID2) and a significant increase in TNNC1 and SERBP1 proteins in hiPSC-CMs treated with BPA. Network analysis of proteomics data supported altered extracellular matrix remodelling and provided a disease-gene association with well-known pathological conditions of the heart. Furthermore, upon hypoxia-reoxygenation challenge, hiPSC-CMs treated with BPA showed higher rate of apoptotic events. Taken together, our results revealed that a long-term treatment, even with low doses of BPA, interferes with hiPSC-CMs functionality and alters the surrounding cellular environment, providing new insights about diseases that might arise upon the toxin exposure. Our study contributes to the current understanding of BPA effects on developing human foetal cardiomyocytes, in correlation with human clinical observations and animal studies, and it provides a suitable model for New Approach Methodologies (NAMs) for environmental chemical hazard and risk assessment.

Proteomics and disease network associations evaluation of environmentally relevant Bisphenol A concentrations in a human 3D neural stem cell model

Bisphenol A (BPA) exposure is associated with a plethora of neurodevelopmental abnormalities and brain disorders. Previous studies have demonstrated BPA-induced perturbations to critical neural stem cell (NSC) characteristics, such as proliferation and differentiation, although the underlying molecular mechanisms remain under debate. The present study evaluated the effects of a repeated-dose exposure of environmentally relevant BPA concentrations during the in vitro 3D neural induction of human induced pluripotent stem cells (hiPSCs), emulating a chronic exposure scenario. Firstly, we demonstrated that our model is suitable for NSC differentiation during the early stages of embryonic brain development. Our morphological image analysis showed that BPA exposure at 0.01, 0.1 and 1 µM decreased the average spheroid size by day 21 (D21) of the neural induction, while no effect on cell viability was detected. No alteration to the rate of the neural induction was observed based on the expression of key neural lineage and neuroectodermal transcripts. Quantitative proteomics at D21 revealed several differentially abundant proteins across all BPA-treated groups with important functions in NSC proliferation and maintenance (e.g., FABP7, GPC4, GAP43, Wnt-8B, TPPP3). Additionally, a network analysis demonstrated alterations to the glycolytic pathway, potentially implicating BPA-induced changes to glycolytic signalling in NSC proliferation impairments, as well as the pathophysiology of brain disorders including intellectual disability, autism spectrum disorders, and amyotrophic lateral sclerosis (ALS). This study enhances the current understanding of BPA-related NSC aberrations based mostly on acute, often high dose exposures of rodent in vivo and in vitro models and human GWAS data in a novel human 3D cell-based model with real-life scenario relevant prolonged and low-level exposures, offering further mechanistic insights into the ramifications of BPA exposure on the developing human brain and consequently, later life neurological disorders.

Accurate label-free quantification by directLFQ to compare unlimited numbers of proteomes

Recent advances in mass spectrometry (MS)-based proteomics enable the acquisition of increasingly large datasets within relatively short times, which exposes bottlenecks in the bioinformatics pipeline. Whereas peptide identification is already scalable, most label-free quantification (LFQ) algorithms scale quadratic or cubic with the sample numbers, which may even preclude the analysis of large-scale data. Here we introduce directLFQ, a ratio-based approach for sample normalization and the calculation of protein intensities. It estimates quantities via aligning samples and ion traces by shifting them on top of each other in logarithmic space. Importantly, directLFQ scales linearly with the number of samples, allowing analyses of large studies to finish in minutes instead of days or months. We quantify 10,000 proteomes in 10 minutes and 100,000 proteomes in less than two hours - a thousand-fold faster than some implementations of the popular LFQ algorithm MaxLFQ. In-depth characterization of directLFQ reveals excellent normalization properties and benchmark results, comparing favorably to MaxLFQ for both data-dependent acquisition (DDA) and data-independent acquisition (DIA). Additionally, directFQ provides normalized peptide intensity estimates for peptide-level comparisons. It is an important part of an overall quantitative proteomic pipeline that also needs to include high sensitive statistical analysis leading to proteoform resolution. Available as an open-source Python package and a GUI with a one-click installer, it can be used in the AlphaPept ecosystem as well as downstream of most common computational proteomics pipelines.

Proteomics- and Metabolomics-Based Analysis of Metabolic Changes in a Swine Model of Pulmonary Hypertension

Pulmonary vein stenosis (PVS) causes a rare type of pulmonary hypertension (PH) by impacting the flow and pressure within the pulmonary vasculature, resulting in endothelial dysfunction and metabolic changes. A prudent line of treatment in this type of PH would be targeted therapy to relieve the pressure and reverse the flow-related changes. We used a swine model in order to mimic PH after PVS using pulmonary vein banding (PVB) of the lower lobes for 12 weeks to mimic the hemodynamic profile associated with PH and investigated the molecular alterations that provide an impetus for the development of PH. Our current study aimed to employ unbiased proteomic and metabolomic analyses on both the upper and lower lobes of the swine lung to identify regions with metabolic alterations. We detected changes in the upper lobes for the PVB animals mainly pertaining to fatty acid metabolism, reactive oxygen species (ROS) signaling and extracellular matrix (ECM) remodeling and small, albeit, significant changes in the lower lobes for purine metabolism.

MS-DAP Platform for Downstream Data Analysis of Label-Free Proteomics Uncovers Optimal Workflows in Benchmark Data Sets and Increased Sensitivity in Analysis of Alzheimer's Biomarker Data

In the rapidly moving proteomics field, a diverse patchwork of data analysis pipelines and algorithms for data normalization and differential expression analysis is used by the community. We generated a mass spectrometry downstream analysis pipeline (MS-DAP) that integrates both popular and recently developed algorithms for normalization and statistical analyses. Additional algorithms can be easily added in the future as plugins. MS-DAP is open-source and facilitates transparent and reproducible proteome science by generating extensive data visualizations and quality reporting, provided as standardized PDF reports. Second, we performed a systematic evaluation of methods for normalization and statistical analysis on a large variety of data sets, including additional data generated in this study, which revealed key differences. Commonly used approaches for differential testing based on moderated t-statistics were consistently outperformed by more recent statistical models, all integrated in MS-DAP. Third, we introduced a novel normalization algorithm that rescues deficiencies observed in commonly used normalization methods. Finally, we used the MS-DAP platform to reanalyze a recently published large-scale proteomics data set of CSF from AD patients. This revealed increased sensitivity, resulting in additional significant target proteins which improved overlap with results reported in related studies and includes a large set of new potential AD biomarkers in addition to previously reported.

The hibernation-derived compound SUL-138 shifts the mitochondrial proteome towards fatty acid metabolism and prevents cognitive decline and amyloid plaque formation in an Alzheimer's disease mouse model

BACKGROUND

Alzheimer's disease (AD) is the most prevalent neurodegenerative disease worldwide and remains without effective cure. Increasing evidence is supporting the mitochondrial cascade hypothesis, proposing that loss of mitochondrial fitness and subsequent ROS and ATP imbalance are important contributors to AD pathophysiology.

METHODS

Here, we tested the effects of SUL-138, a small hibernation-derived molecule that supports mitochondrial bioenergetics via complex I/IV activation, on molecular, physiological, behavioral, and pathological outcomes in APP/PS1 and wildtype mice.

RESULTS

SUL-138 treatment rescued long-term potentiation and hippocampal memory impairments and decreased beta-amyloid plaque load in APP/PS1 mice. This was paralleled by a partial rescue of dysregulated protein expression in APP/PS1 mice as assessed by mass spectrometry-based proteomics. In-depth analysis of protein expression revealed a prominent effect of SUL-138 in APP/PS1 mice on mitochondrial protein expression. SUL-138 increased the levels of proteins involved in fatty acid metabolism in both wildtype and APP/PS1 mice. Additionally, in APP/PS1 mice only, SUL-138 increased the levels of proteins involved in glycolysis and amino acid metabolism pathways, indicating that SUL-138 rescues mitochondrial impairments that are typically observed in AD.

CONCLUSION

Our study demonstrates a SUL-138-induced shift in metabolic input towards the electron transport chain in synaptic mitochondria, coinciding with increased synaptic plasticity and memory. In conclusion, targeting mitochondrial bioenergetics might provide a promising new way to treat cognitive impairments in AD and reduce disease progression.

Analysis of proteome adaptation reveals a key role of the bacterial envelope in starvation survival

Bacteria reorganize their physiology upon entry to stationary phase. What part of this reorganization improves starvation survival is a difficult question because the change in physiology includes a global reorganization of the proteome, envelope, and metabolism of the cell. In this work, we used several trade-offs between fast growth and long survival to statistically score over 2,000 Escherichia coli proteins for their global correlation with death rate. The combined ranking allowed us to narrow down the set of proteins that positively correlate with survival and validate the causal role of a subset of proteins. Remarkably, we found that important survival genes are related to the cell envelope, i.e., periplasm and outer membrane, because the maintenance of envelope integrity of E. coli plays a crucial role during starvation. Our results uncover a new protective feature of the outer membrane that adds to the growing evidence that the outer membrane is not only a barrier that prevents abiotic substances from reaching the cytoplasm but also essential for bacterial proliferation and survival.

The Amino Acid Homoarginine Inhibits Atherogenesis by Modulating T-Cell Function

BACKGROUND

Amino acid metabolism is crucial for inflammatory processes during atherogenesis. The endogenous amino acid homoarginine is a robust biomarker for cardiovascular outcome and mortality with high levels being protective. However, the underlying mechanisms remain elusive. We investigated the effect of homoarginine supplementation on atherosclerotic plaque development with a particular focus on inflammation.

METHODS

Female ApoE-deficient mice were supplemented with homoarginine (14 mg/L) in drinking water starting 2 weeks before and continuing throughout a 6-week period of Western-type diet feeding. Control mice received normal drinking water. Immunohistochemistry and flow cytometry were used for plaque- and immunological phenotyping. T cells were characterized using mass spectrometry-based proteomics, by functional in vitro approaches, for example, proliferation and migration/chemotaxis assays as well as by super-resolution microscopy.

RESULTS

Homoarginine supplementation led to a 2-fold increase in circulating homoarginine concentrations. Homoarginine-treated mice exhibited reduced atherosclerosis in the aortic root and brachiocephalic trunk. A substantial decrease in CD3⁺ T cells in the atherosclerotic lesions suggested a T-cell-related effect of homoarginine supplementation, which was mainly attributed to CD4⁺ T cells. Macrophages, dendritic cells, and B cells were not affected. CD4⁺ T-cell proteomics and subsequent pathway analysis together with in vitro studies demonstrated that homoarginine profoundly modulated the spatial organization of the T-cell actin cytoskeleton and increased filopodia formation via inhibition of Myh9 (myosin heavy chain 9). Further mechanistic studies revealed an inhibition of T-cell proliferation as well as a striking impairment of the migratory capacities of T cells in response to relevant chemokines by homoarginine, all of which likely contribute to its atheroprotective effects.

CONCLUSIONS

Our study unravels a novel mechanism by which the amino acid homoarginine reduces atherosclerosis, establishing that homoarginine modulates the T-cell cytoskeleton and thereby mitigates T-cell functions important during atherogenesis. These findings provide a molecular explanation for the beneficial effects of homoarginine in atherosclerotic cardiovascular disease.

Ultra‐high sensitivity mass spectrometry quantifies single‐cell proteome changes upon perturbation

Single‐cell technologies are revolutionizing biology but are today mainly limited to imaging and deep sequencing. However, proteins are the main drivers of cellular function and in‐depth characterization of individual cells by mass spectrometry (MS)‐based proteomics would thus be highly valuable and complementary. Here, we develop a robust workflow combining miniaturized sample preparation, very low flow‐rate chromatography, and a novel trapped ion mobility mass spectrometer, resulting in a more than 10‐fold improved sensitivity. We precisely and robustly quantify proteomes and their changes in single, FACS‐isolated cells. Arresting cells at defined stages of the cell cycle by drug treatment retrieves expected key regulators. Furthermore, it highlights potential novel ones and allows cell phase prediction. Comparing the variability in more than 430 single‐cell proteomes to transcriptome data revealed a stable‐core proteome despite perturbation, while the transcriptome appears stochastic. Our technology can readily be applied to ultra‐high sensitivity analyses of tissue material, posttranslational modifications, and small molecule studies from small cell counts to gain unprecedented insights into cellular heterogeneity in health and disease.

Differential Effects of Insulin-Deficient Diabetes Mellitus on Visceral vs. Subcutaneous Adipose Tissue-Multi-omics Insights From the Munich MIDY Pig Model

Adipose tissue (AT) is no longer considered to be responsible for energy storage only but is now recognized as a major endocrine organ that is distributed across different parts of the body and is actively involved in regulatory processes controlling energy homeostasis. Moreover, AT plays a crucial role in the development of metabolic disease such as diabetes. Recent evidence has shown that adipokines have the ability to regulate blood glucose levels and improve metabolic homeostasis. While AT has been studied extensively in the context of type 2 diabetes, less is known about how different AT types are affected by absolute insulin deficiency in type 1 or permanent neonatal diabetes mellitus. Here, we analyzed visceral and subcutaneous AT in a diabetic, insulin-deficient pig model (MIDY) and wild-type (WT) littermate controls by RNA sequencing and quantitative proteomics. Multi-omics analysis indicates a depot-specific dysregulation of crucial metabolic pathways in MIDY AT samples. We identified key proteins involved in glucose uptake and downstream signaling, lipogenesis, lipolysis and β-oxidation to be differentially regulated between visceral and subcutaneous AT in response to insulin deficiency. Proteins related to glycogenolysis, pyruvate metabolism, TCA cycle and lipogenesis were increased in subcutaneous AT, whereas β-oxidation-related proteins were increased in visceral AT from MIDY pigs, pointing at a regionally different metabolic adaptation to master energy stress arising from diminished glucose utilization in MIDY AT. Chronic, absolute insulin deficiency and hyperglycemia revealed fat depot-specific signatures using multi-omics analysis. The generated datasets are a valuable resource for further comparative and translational studies in clinical diabetes research.

Spatial discordances between mRNAs and proteins in the intestinal epithelium

The use of transcriptomes as reliable proxies for cellular proteomes is controversial. In the small intestine, enterocytes operate for 4 days as they migrate along villi, which are highly graded microenvironments. Spatial transcriptomics have demonstrated profound zonation in enterocyte gene expression, but how this variability translates to protein content is unclear. Here we show that enterocyte proteins and messenger RNAs along the villus axis are zonated, yet often spatially discordant. Using spatial sorting with zonated surface markers, together with a Bayesian approach to infer protein translation and degradation rates from the combined spatial profiles, we find that, while many genes exhibit proteins zonated toward the villus tip, mRNA is zonated toward the villus bottom. Finally, we demonstrate that space-independent protein synthesis delays can explain many of the mRNA-protein discordances. Our work provides a proteomic spatial blueprint of the intestinal epithelium, highlighting the importance of protein measurements for inferring cell states in tissues that operate outside of steady state.

A scalable, clinically severe pig model for Duchenne muscular dystrophy

Large animal models for Duchenne muscular dystrophy (DMD) are crucial for evaluation of diagnostic procedures and treatment strategies. Pigs cloned from male cells lacking DMD exon 52 (DMDΔ52) resemble molecular, clinical and pathological hallmarks of DMD, but die before sexual maturity and cannot be propagated by breeding. Therefore, we generated female DMD+/- carriers. A single founder animal had 11 litters with 29 DMDY/-, 34 DMD+/- as well as 36 male and 29 female wild-type offspring. Breeding with F1 and F2 DMD+/- carriers resulted in additional 114 DMDY/- piglets. With intensive neonatal management, the majority survived for 3-4 months, providing statistically relevant cohorts for experimental studies. Pathological investigations and proteome studies of skeletal muscles and myocardium confirmed the resemblance of human disease mechanisms. Importantly, DMDY/- pigs reveal progressive myocardial fibrosis and increased expression of connexin-43, associated with significantly reduced left ventricular ejection fraction already at age 3 months. Furthermore, behavioral tests provided evidence for impaired cognitive ability. Our breeding cohort of DMDΔ52 pigs and standardized tissue repositories provide important resources for studying DMD disease mechanisms and for testing novel treatment strategies.

Torpor enhances synaptic strength and restores memory performance in a mouse model of Alzheimer's disease

Hibernation induces neurodegeneration-like changes in the brain, which are completely reversed upon arousal. Hibernation-induced plasticity may therefore be of great relevance for the treatment of neurodegenerative diseases, but remains largely unexplored. Here we show that a single torpor and arousal sequence in mice does not induce dendrite retraction and synapse loss as observed in seasonal hibernators. Instead, it increases hippocampal long-term potentiation and contextual fear memory. This is accompanied by increased levels of key postsynaptic proteins and mitochondrial complex I and IV proteins, indicating mitochondrial reactivation and enhanced synaptic plasticity upon arousal. Interestingly, a single torpor and arousal sequence was also sufficient to restore contextual fear memory in an APP/PS1 mouse model of Alzheimer’s disease. Our study demonstrates that torpor in mice evokes an exceptional state of hippocampal plasticity and that naturally occurring plasticity mechanisms during torpor provide an opportunity to identify unique druggable targets for the treatment of cognitive impairment.

Brain Acetyl-CoA Production and Phosphorylation of Cytoskeletal Proteins Are Targets of CYP46A1 Activity Modulation and Altered Sterol Flux

Cholesterol and 24-hydroxycholesterol are the most abundant brain sterols and represent the substrate and product, respectively, of cytochrome P450 46A1 (CYP46A1), a CNS-specific enzyme. CYP46A1 controls cholesterol elimination and turnover in the brain, the two processes that determine the rate of brain sterol flux through the plasma membranes and thereby the properties of these membranes. Brain sterol flux is decreased in Cyp46a1-/- mice compared to wild-type mice and increased in 5XFAD mice (a model of Alzheimer's disease) when they are treated with a small dose of efavirenz, a CYP46A1 activator. Herein, we first assessed the brain proteome (synaptosomal fractions) and phospho-proteome (synaptosomal fractions and brain homogenates) of efavirenz-treated and control 5XFAD mice. Then, based on the pattern of protein abundance change, we conducted acetyl-CoA measurements (brain homogenates and mitochondria) and metabolic profiling (brain homogenates). The phospho-proteomics datasets were used for comparative analyses with the datasets obtained by us previously on mice with the same changes (efavirenz-treated and control 5XFAD mice from a different treatment paradigm) or with changes in the opposite direction (Cyp46a1-/- vs wild-type mice) in brain sterol flux. We found that CYP46A1 activity or the rate of brain sterol flux affects acetyl-CoA-related metabolic pathways as well as phosphorylation of cytoskeletal and other proteins. Knowledge of the key roles of acetyl-CoA and cytoskeletal phosphorylation in cell biology expands our understanding of the significance of CYP46A1-mediated cholesterol 24-hydroxylation in the brain and provides an additional explanation for why CYP46A1 activity modulations are beneficial in mouse models of different brain diseases.

Protein quality control and regulated proteolysis in the genome-reduced organism Mycoplasma pneumoniae

Protein degradation is a crucial cellular process in all-living systems. Here, using Mycoplasma pneumoniae as a model organism, we defined the minimal protein degradation machinery required to maintain proteome homeostasis. Then, we conditionally depleted the two essential ATP-dependent proteases. Whereas depletion of Lon results in increased protein aggregation and decreased heat tolerance, FtsH depletion induces cell membrane damage, suggesting a role in quality control of membrane proteins. An integrative comparative study combining shotgun proteomics and RNA-seq revealed 62 and 34 candidate substrates, respectively. Cellular localization of substrates and epistasis studies supports separate functions for Lon and FtsH. Protein half-life measurements also suggest a role for Lon-modulated protein decay. Lon plays a key role in protein quality control, degrading misfolded proteins and those not assembled into functional complexes. We propose that regulating complex assembly and degradation of isolated proteins is a mechanism that coordinates important cellular processes like cell division. Finally, by considering the entire set of proteases and chaperones, we provide a fully integrated view of how a minimal cell regulates protein folding and degradation.

Block Design with Common Reference Samples Enables Robust Large-Scale Label-Free Quantitative Proteome Profiling

Label-free quantitative proteomics has become an increasingly popular tool for profiling global protein abundances. However, one major limitation is the potential performance drift of the LC-MS platform over time, which, in turn, limits its utility for analyzing large-scale sample sets. To address this, we introduce an experimental and data analysis scheme based on a block design with common references within each block for enabling large-scale label-free quantification. In this scheme, a large number of samples (e.g., >100 samples) are analyzed in smaller and more manageable blocks, minimizing instrument drift and variability within individual blocks. Each designated block also contains common reference samples (e.g., controls) for normalization across all blocks. We demonstrated the robustness of this approach by profiling the proteome response of human macrophage THP-1 cells to 11 engineered nanomaterials at two different doses. A total of 116 samples were analyzed in six blocks, yielding an average coverage of 4500 proteins per sample. Following a common reference-based correction, 2537 proteins were quantified with high reproducibility without any imputation of missing values from 116 data sets. The data revealed the consistent quantification of proteins across all six blocks, as illustrated by the highly consistent abundances of house-keeping proteins in all samples and the high levels of correlation among samples from different blocks. The data also demonstrated that label-free quantification is robust and accurate enough to quantify even very subtle abundance changes as well as large fold-changes. Our streamlined workflow is easy to implement and can be readily adapted to other large cohort studies for reproducible label-free proteome quantification.

From command-line bioinformatics to bioGUI

Bioinformatics is a highly interdisciplinary field providing (bioinformatics) applications for scientists from many disciplines. Installing and starting applications on the command-line (CL) is inconvenient and/or inefficient for many scientists. Nonetheless, most methods are implemented with a command-line interface only. Providing a graphical user interface (GUI) for bioinformatics applications is one step toward routinely making CL-only applications available to more scientists and, thus, toward a more effective interdisciplinary work. With our bioGUI framework we address two main problems of using CL bioinformatics applications: First, many tools work on UNIX-systems only, while many scientists use Microsoft Windows. Second, scientists refrain from using CL tools which, however, could well support them in their research. With bioGUI install modules and templates, installing and using CL tools is made possible for most scientists-even on Windows, due to bioGUI's support for Windows Subsystem for Linux. In addition, bioGUI templates can easily be created, making the bioGUI framework highly rewarding for developers. From the bioGUI repository it is possible to download, install and use bioinformatics tools with just a few clicks.