Proteomics experiments commonly aim to estimate and detect differential abundance across all expressed proteins. Within this experimental design, some of the most challenging measurements are small fold changes for lower abundance proteins. While bottom-up proteomics methods are approaching comprehensive coverage of even complex eukaryotic proteomes, failing to reliably quantify lower abundance proteins can limit the precision and reach of experiments to much less than the identified—let alone total—proteome. Here we test the ability of two common methods, a tandem mass tagging (TMT) method and a label-free quantitation method (LFQ), to achieve comprehensive quantitative coverage by benchmarking their capacity to measure 3 different levels of change (3-, 2-, and 1.5-fold) across an entire data set. Both methods achieved comparably accurate estimates for all 3-fold-changes. However, the TMT method detected changes that reached statistical significance three times more often due to higher precision and fewer missing values. These findings highlight the importance of refining proteome quantitation methods to bring the number of usefully quantified proteins into closer agreement with the number of total quantified proteins.

The budding yeast Saccharomyces cerevisiae is a model system for investigating biological processes. Cellular processes are known to be dysregulated because of shifts in carbon sources. However, the comprehensive proteomic alterations thereof have not been fully investigated. Here we examined proteomic alterations in S. cerevisiae due to the adaptation of yeast from glucose to nine different carbon sources – maltose, trehalose, fructose, sucrose, glycerol, acetate, pyruvate, lactic acid, and oleate. Isobaric tag-based mass spectrometry techniques are at the forefront of global proteomic investigations. As such, we used a TMT10-plex strategy to study multiple growth conditions in a single experiment. The SPS-MS3 method on an Orbitrap Fusion Lumos mass spectrometer enabled the quantification of over 5000 yeast proteins across ten carbon sources at a 1% protein-level FDR. On average, the proteomes of yeast cultured in fructose and sucrose deviated the least from those cultured in glucose. As expected, gene ontology classification revealed the major alteration in protein abundances occurred in metabolic pathways and mitochondrial proteins. Our protocol lays the groundwork for further investigation of carbon source-induced protein alterations. Additionally, these data offer a hypothesis-generating resource for future studies aiming to investigate both characterized and uncharacterized genes.

The integrated stress response (ISR) attenuates the rate of protein synthesis while inducing expression of stress proteins in cells. Various insults activate kinases that phosphorylate the GTPase eIF2 leading to inhibition of its exchange factor eIF2B. Vanishing White Matter (VWM) is a neurological disease caused by eIF2B mutations that, like phosphorylated eIF2, reduce its activity. We show that introduction of a human VWM mutation into mice leads to persistent ISR induction in the central nervous system. ISR activation precedes myelin loss and development of motor deficits. Remarkably, long-term treatment with a small molecule eIF2B activator, 2BAct, prevents all measures of pathology and normalizes the transcriptome and proteome of VWM mice. 2BAct stimulates the remaining activity of mutant eIF2B complex in vivo, abrogating the maladaptive stress response. Thus, 2BAct-like molecules may provide a promising therapeutic approach for VWM and provide relief from chronic ISR induction in a variety of disease contexts.

Mass spectrometry (MS) has become an accessible tool for whole proteome quantitation with the ability to characterize protein expression across thousands of proteins within a single experiment. A subset of MS quantification methods (e.g., SILAC and label-free) monitor the relative intensity of intact peptides, where thousands of measurements can be made from a single mass spectrum. An alternative approach, isobaric labeling, enables precise quantification of multiple samples simultaneously through unique and sample specific mass reporter ions. Consequently, in a single scan, the quantitative signal comes from a limited number of spectral features (≤11). The signal observed for these features is constrained by automatic gain control, forcing codependence of concurrent signals. The study of constrained outcomes primarily belongs to the field of compositional data analysis. We show experimentally that isobaric tag proteomics data are inherently compositional and highlight the implications for data analysis and interpretation. We present a new statistical model and accompanying software that improves estimation accuracy and the ability to detect changes in protein abundance. Finally, we demonstrate a unique compositional effect on proteins with infinite changes. We conclude that many infinite changes will appear small and that the magnitude of these estimates is highly dependent on experimental design.

An idealized version of a label-free discovery mass spectrometry proteomics experiment would provide absolute abundance measurements for a whole proteome, across varying conditions. Unfortunately, this ideal is not realized. Measurements are made on peptides requiring an inferential step to obtain protein level estimates. The inference is complicated by experimental factors that necessitate relative abundance estimation and result in widespread non-ignorable missing data. Relative abundance on the log scale takes the form of parameter contrasts. In a complete-case analysis, contrast estimates may be biased by missing data and a substantial amount of useful information will often go unused. To avoid problems with missing data, many analysts have turned to single imputation solutions. Unfortunately, these methods often create further difficulties by hiding inestimable contrasts, preventing the recovery of interblock information and failing to account for imputation uncertainty. To mitigate many of the problems caused by missing values, we propose the use of a Bayesian selection model. Our model is tested on simulated data, real data with simulated missing values, and on a ground truth dilution experiment where all of the true relative changes are known. The analysis suggests that our model, compared with various imputation strategies and complete-case analyses, can increase accuracy and provide substantial improvements to interval coverage.