Figure 1.Proteomic tools used in the study of pathogenic infection
(A) Overview of the replication cycle of intracellular pathogens. General steps are shown for cytosolic‐replicating non‐enveloped viruses (a), nuclear‐replicating enveloped viruses (b), and cytosolic bacteria (c). Viruses enter the cell through endocytosis (1a) or fusion with the host cell membrane (1b). The viral genome is extruded from the capsid to the cytosol (2a) or nucleoplasm (2b). Viral genes are then expressed (3) to produce viral proteins (4). Viral proteins facilitate immune evasion, viral genome replication (5a and 5b), viral genome encapsidation (6a and 6b), and envelopment (7). Fully assembled infectious viruses are secreted through lysis (8a) or through exocytosis (8b). Bacteria become endocytosed by the host cell (1c), followed by release from endocytic vesicles mediated by bacteria‐secreted proteins (2c). Then, the bacteria acquire nutrients directly from the host cytosol (3c). Bacteria replicate in the cytosol (4c) and exit the cell to the extracellular space or directly to neighboring cells (5c). (B) Overview of proteomic tools that have been utilized in the of study host–pathogen interactions.
Figure 2.Proteomic tools to study protein–protein interactions in pathogenic infections
(A) Detection of PPIs via shotgun IP‐MS. Quantitation can be done via label‐free (top), isobaric tagging (middle), or SILAC (bottom) strategies. (B) Detection of PPIs via top‐down MS. This method differentiates multiple intact different complexes containing the same protein of interest and can provide stoichiometry. (C) Detection of direct PPIs via Y2H. This method allows detection of direct interactions without the complexity of a cross‐linker, but at the expense of a high false‐positive rate and non‐infection contexts. (D) Characterization of PPIs by hydrogen/deuterium exchange. This method identifies the regions of each protein that interact in vitro and can be used to derive kinetic information about the interaction. (E) Detection of direct PPIs via cross‐linker. This method also identifies the regions of interaction on each protein and can be used in cells or in vitro.
Figure 3.Proteomic tools to study whole‐cell or subcellular proteome alterations during infection
Proteomic methods can be used to study alterations throughout infection in protein abundance or in protein subcellular localization. (A) SILAC‐based workflow to define proteome alterations upon infection. Differentially labeled uninfected and infected cells are mixed and processed directly for MS analysis (whole‐cell temporal proteomics) or preceded by a subcellular fractionation step (organelle temporal proteomics). (B) An alternative method using label‐free quantification or isobaric tags (e.g., TMT) to define proteome alterations at multiple time points of infection. Cells or subcellular organelles are harvested at different infection times, and following digestion, peptides from each fraction are analyzed by MS (label‐free quantification) or labeled with isobaric tags and mixed for multiplexed MS analysis and quantification. (C) Defining protein localization during infection. Discrete subcellular fractions (e.g., differential centrifugation or density gradient fractionation) are collected from infected and uninfected cells, and analyzed by MS. (D) Proteomic approach to assign proteins to specific organelles and determine alterations in protein subcellular localization. Multiple organelles from infected or uninfected cells are partially separated using a density gradient. Fractions are analyzed by quantitative MS, resulting in spatial profiles of proteins across the gradient. Well‐known organelle residents are used as organelle markers for the spatial profiles. The remaining proteins are assigned to organelles using classification algorithms (e.g., machine learning). The libraries of predicted protein localizations from infected and uninfected cells are compared to determine candidate proteins undergoing infection‐induced translocations between organelles.
Figure 4.Post‐translational modifications (PTMs) involved in the context of infection
Infections induce a series of dynamic PTMs on host and pathogen proteins that act in pro‐ and anti‐pathogen responses. (1) Virulence factors promote ubiquitination and SUMOylation of host proteins, resulting in degradation by the proteasome. (2) Acetylation is used by pathogens to block immune response. (3) Pathogen invasion causes phosphorylation of transcription factors, altering host and pathogen gene expression. (4) Viral genome can evade immune response by disrupting nuclear bodies via SUMOylation of PML and other nuclear body proteins. (5) Histones, finely‐tuned by PTMs, bind to host and viral genomes, modulating host and viral gene expression. (6) Viral glycoproteins on envelope are modified by glycosylation, facilitating viral entry and dampening immune response. (7) Pathogen infection activates PPR receptors, which relay the signal to transcription factors via adaptors to induce immune responses.
Figure 5.Integrative omic approaches for the study of host–pathogen interactions
(A) Proteomic tools have been integrated with other omic approaches to refine models of pathogen genomes. Proteomic data can be acquired from purified virions, bacteria culture under host physiological conditions, and from infected host cells. Transcriptomic data are acquired from bacteria cultures or virus‐infected host cells. Genomic data are acquired from purified viral particles and bacteria cultures. In genomic‐informed proteomic experiments, MS data are searched through a genome database containing translated sequences from known or predicted open reading frames (ORFs) to provide evidence of protein expression. In transcriptomic‐informed proteomic experiments, the search database is built from RNA‐seq and ribosome profiling to identify uncommon transcripts difficult to predict from genomic sequences. Proteogenomic approaches use both genomic and transcriptomic data to build a customized database search useful in identifying peptides from 5′ untranslated regions (UTRs), 3′ UTRs, alternative junction splicing, intron sequences, short ORFs, and alternative reading frames to refine pathogen gene models. (B) Integrating proteomics with other omic tools and phenotype information to identify key virulence factors. Genomic data are used to identify strain‐specific mutations and associate them with differences in protein expression. The differences in protein expression are then related to phenotype alterations using phenotypic assays on wild‐type strains and those harboring the mutations. Proteomics, glycomics, and glycopeptidomics are used to identify glycosylation patterns associated with specific protein sequences and viral strains. These glycosylation patterns are then associated with phenotypic variations using phenotypic assays on strain having different glycosylation patterns. (C) Using multi‐omic datasets to define the host response to an infection. Quantitative multi‐omic datasets are mapped to known metabolic networks to identify pathways that are up‐ or downregulated upon infection. Alternatively, multi‐omic datasets are integrated with phenotype data to construct correlation networks. Nodes represent individual genes, metabolites or phenotype measurements, or they can also represent module eigengenes. Edges represent correlations between nodes from quantitative measurements. Networks can be analyzed to identify novel associations, including novel clusters, and key nodes, such as hubs and bridges.
Proteomic approaches in the clinic directed toward infectious diseases. Diagram showing proteomic approaches that have been applied for the discovery, diagnosis, and prognosis of pathogens and infectious diseases. (Top) Protein arrays (e.g., HD‐NAPPA) exposing different viral proteins. If an antibody targeting these viral proteins is present in the serum of the individual, it binds and is detected using fluorescent secondary antibodies. (Middle) Using shotgun MS approaches, proteins can be readily detected in the patient body fluids. Significantly enriched proteins in diseased individuals are candidates as biomarkers. (Bottom) These biomarkers are then monitored using a targeted MS approach in diseased patients for diagnosis and prognosis. Additionally, a targeted MS approach can be used to readily identify pathogens from patient samples.