系统生物学

Memory is a fundamental process, regulated by protein–protein interactions within neuronal proteome networks. Learning-dependent changes in specific brain regions important for memory have been detected by mass spectrometry, by comparing proteins from animals trained to learn with mock-trained controls. Detection through this method relies on relative protein abundance; brain dissection is readily available for macroscopic animals to spatially control protein identification by mass spectrometry. In the nematode C. elegans, however, such spatial control is limited due to its microscopic size, hindering its utilization in proteomics. A protocol to address this limitation would strengthen an already excellent model to study memory, given that many proteins for learning are evolutionarily conserved in the worm and single-cell expression is uniquely defined across all 302 neurons. We modified existing protocols to enable (i) proximity labeling detection of neuronal proteins in C. elegans and (ii) high-throughput enrichment of these proteins from >3,000 whole worm bodies simultaneously, to assess trained vs. mock-trained proteomes. This involved the biotin ligase enzyme TurboID, which promiscuously labels nearby proteins with its substrate biotin. Enzyme expression was transgenically restricted to the nervous system, and biotin supplementation was limited to the training (or mock training) period in a classical (gustatory) conditioning paradigm. Labeled proteins were enriched by pull-down using streptavidin, which has a high binding affinity to biotin, and then processed for mass spectrometry runs and qualitative data analysis. This protocol is uniquely advantageous in that it minimizes proteins present before a temporal window of interest (training/mock training), improving the detection of lowly abundant proteins from a specific tissue in the worm (neurons). We have demonstrated that the protocol can sufficiently detect novel learning regulators, thus providing a useful framework to interrogate proteomes in microscopic brains.

Chromatin modifications regulate genome function by recruiting proteins that control transcription, genome organization, and DNA repair. Identifying the proteins associated with specific chromatin modifications is therefore essential for understanding how these regulatory processes operate. Traditional approaches, including chromatin immunoprecipitation and affinity purification coupled to mass spectrometry, have uncovered many chromatin-associated proteins. However, they often rely on crosslinking and chromatin fragmentation, which can disrupt native chromatin architecture and limit the detection of transient interactions. Here, we describe a proximity-labeling protocol for identifying the chromatin-dependent protein interactome associated with specific chromatin marks, termed ChromID. ChromID uses engineered chromatin readers (eCRs) fused to a promiscuous biotin ligase, which labels proteins in the immediate vicinity of the targeted chromatin mark. The protocol includes in vivo biotin labeling, nuclear extract preparation, streptavidin-based enrichment, and tryptic digestion for downstream LC-MS/MS analysis. The protocol has been validated across multiple cell types and chromatin contexts and can be extended to other chromatin-associated proteins, providing a versatile approach to profile chromatin-associated proteomes within their native cellular environment.

Bottom-up proteomics workflows encompass several key stages, including sample preparation, data acquisition, and data analysis. Of these, sample preparation is the initial and critical stage, as it significantly influences the depth, reproducibility, and reliability of subsequent mass spectrometry–based analyses. While several main digestion strategies exist, including in-gel, in-solution, and filter-aided methods, each presents distinct trade-offs in terms of throughput, contamination removal, and applicability to complex biological matrices. The Suspension Trapping (S-Trap) method offers a compelling alternative by efficiently capturing and digesting proteins while removing interferents like sodium dodecyl sulfate (SDS), which can compromise downstream LC–MS/MS performance. This protocol details a S-Trap workflow optimized for biofluid proteomics, specifically plasma, serum, and cerebrospinal fluid (CSF). We describe two complementary formats: a manual tube-based procedure for individual or small-batch samples and a 96-well-plate-based system enabling high-throughput processing. The protocol integrates optional high-abundance protein depletion to enhance coverage of low-abundance analytes and includes steps for reduction, alkylation, digestion, and peptide elution for low total protein content samples, such as plasma, serum, and cerebrospinal fluid. By providing a detailed protocol, this work aims to improve the consistency and accessibility of S-Trap-based sample preparation, facilitating robust and reproducible discoveries in bottom-up proteomics.

Extracellular vesicles (EVs) are nanoscale particles secreted by all cells and present in all biological fluids, where they carry molecular cargo reflective of health and disease states. Their diagnostic potential is often obscured by the high abundance of non-EV proteins and lipoproteins (e.g., albumin, apolipoproteins) that complicate proteomic analysis of primary biofluids, such as ascites fluid. Conventional isolation strategies face a persistent trade-off between EV purity and yield. To overcome this, a magnetic bead-based protocol (Mag-Net) to enrich EVs according to electrochemical surface charge using strong anion-exchange chemistry (SAX) was adapted for proteomics. Our workflow is specifically adapted to ascites fluid from human or murine sources. This approach effectively separates EVs from high-abundance proteins and lipoproteins, enabling proteomic profiling from as little as 2 μL of ascites fluid. Demonstrated in both murine and human ovarian cancer models, Mag-Net offers a reproducible, scalable, and automation-ready solution for EV isolation from various biofluids.

Spatial proteomics enables the mapping of protein distribution within tissues, which is crucial for understanding cellular functions in their native context. While spatial transcriptomics has seen rapid advancement, spatial proteomics faces challenges due to protein non-amplifiability and mass spectrometry sensitivity limitations. This protocol describes a sparse sampling strategy for spatial proteomics (S4P) that combines multi-angle tissue strip microdissection with deep learning–based image reconstruction. The method achieves whole-tissue slice coverage with significantly reduced sampling requirements, enabling mapping of over 9,000 proteins in mouse brain tissue at 525 μm resolution within 200 h of mass spectrometry time. Key advantages include reduced sample processing time, deep proteome coverage, and applicability to centimeter-sized tissue samples.

The cellular compartments of eukaryotic cells are defined by their specific protein compositions. Different strategies are used for the identification of the subcellular proteomes, such as fractionation by differential centrifugation of cellular extracts. The localization of mitochondrial proteins is particularly challenging, as mitochondria consist of two membranes of different protein composition and two aqueous subcompartments, the intermembrane space (IMS) and the matrix. Previous studies identified subcompartment-specific proteomes by using combinations of hypotonic swelling and protease digestion followed by mass spectrometry. Here, we present an alternative, more unbiased method to identify the proteomes of mitochondrial subcompartments by use of an improved ascorbate peroxidase (APEX2) that is targeted to the IMS and the matrix. This method allows the subcompartment-specific labeling of proteins in mitochondria isolated from cells of the baker’s yeast Saccharomyces cerevisiae, followed by their purification on streptavidin beads. With this method, the proteins located in the different mitochondrial subcompartments of yeast cells can be efficiently and comprehensively identified.

The cellular secretome is a rich source of biomarkers and extracellular signaling molecules, but proteomic profiling remains challenging, especially when processing culture volumes greater than 5 mL. Low protein abundance, high serum contamination, and sample loss during preparation limit reproducibility and sensitivity in mass spectrometry–based workflows. Here, we present an optimized and scalable protocol that integrates (i) 50 kDa molecular weight cutoff ultrafiltration, (ii) spin column depletion of abundant serum proteins, and (iii) acetone/TCA precipitation for protein recovery. This workflow enables balanced recovery of both low- and high-molecular-weight proteins while reducing background from serum albumin, thereby improving sensitivity, reproducibility, and dynamic range for LC–MS/MS analysis. Validated in human mesenchymal stromal cell cultures, the protocol is broadly applicable across diverse cell types and experimental designs, making it well-suited for biomarker discovery and extracellular proteomics.

Protein S-nitrosylation is a critical post-translational modification that regulates diverse cellular functions and signaling pathways. Although various biochemical methods have been developed to detect S-nitrosylated proteins, many suffer from limited specificity and sensitivity. Here, we describe a robust protocol that combines a modified biotin-switch technique (BST) with streptavidin-based affinity enrichment and quantitative mass spectrometry to detect and profile nitrosylated proteins in cultured cells. The method involves blocking free thiols, selective reduction of nitrosothiols, biotin labeling, enrichment of biotinylated proteins, and identification by tandem mass tag (TMT)-based quantitative mass spectrometry. Additionally, site-directed mutagenesis is employed to generate “non-nitrosylable” mutants for functional validation of specific nitrosylation sites. This protocol provides high specificity, quantitative capability, and versatility for both targeted and global analysis of protein nitrosylation.

In plants, the apoplast contains a diverse set of proteins that underpin mechanisms for maintaining cell homeostasis, cell wall remodeling, cell signaling, and pathogen defense. Apoplast protein composition is highly regulated, primarily through the control of secretory traffic in response to endogenous and environmental factors. Dynamic changes in apoplast proteome facilitate plant survival in a changing climate. Even so, the apoplast proteome profiles in plants remain poorly characterized due to technological limitations. Recent progress in quantitative proteomics has significantly advanced the resolution of proteomic profiling in mammalian systems and has the potential for application in plant systems. In this protocol, we provide a detailed and efficient protocol for tandem mass tag (TMT)-based quantitative analysis of Arabidopsis thaliana secretory proteome to resolve dynamic changes in leaf apoplast proteome profiles. The protocol employs apoplast flush collection followed by protein cleaning using filter-aided sample preparation (FASP), protein digestion, TMT-labeling of peptides, and mass spectrometry (MS) analysis. Subsequent data analysis for peptide detection and quantification uses Proteome Discoverer software (PD) 3.0. Additionally, we have incorporated in silico–generated spectral libraries using PD 3.0, which enables rapid and efficient analysis of proteomic data. Our optimized protocol offers a robust framework for quantitative secretory proteomic analysis in plants, with potential applications in functional proteomics and the study of trafficking systems that impact plant growth, survival, and health.

Recent advances in single-cell technologies have provided limited insight into the role of protein phosphorylation in T-cell fate and function. Dysregulated protein phosphorylation is associated with adverse clinical outcomes, emphasizing the need for reliable methods to unravel the complexities of T-cell signal transduction and disease-related alterations. While flow cytometry is widely used, it is constrained by spectral overlap, limiting the number of protein targets for simultaneous analysis. To overcome this, we present a robust protocol for whole blood T-cell stimulation and subsequent analysis using mass cytometry by time-of-flight (CyTOF). CyTOF minimizes spillover into adjacent channels by employing highly pure, stable, heavy metal–conjugated antibodies for protein detection. This protocol offers a high-dimensional approach for phenotypic and phospho-protein characterization of key signaling pathways, including JAK/STAT, MAPK, PI3K/mTOR, PKC, and NF-κB. A key feature is the T-cell stimulation reagent, which mimics endogenous activation by engaging the T-cell receptor (TCR)/CD3 complex and providing co-stimulation via an anti-CD28 antibody. Further, we enhance reproducibility and enable batch processing through the implementation of the Prot1/Thaw-Lyse system for immediate cryopreservation of stimulated blood samples. By employing CyTOF, this method permits the simultaneous analysis of 31 protein targets with single-cell resolution, minimizing spillover and providing superior specificity, sensitivity, and resolution over flow cytometric methods. This approach facilitates the robust assessment of TCR activation and its effect on bystander populations, which has been challenging with spectral flow cytometry due to the limited availability of methanol-resistant fluorophores. This protocol is a precise and reproducible method for elucidating the downstream effects of T-cell stimulation and immune status, with significant potential for clinical applications, including the assessment of T-cell-targeted therapies.