High-Throughput Proteomics Identifies Proteins With Importance to Postantibiotic Recovery in Depolarized Persister Cells

Abstract

Bacterial populations produce phenotypic variants called persisters to survive harmful conditions. Persisters are highly tolerant to antibiotics and repopulate environments after the stress has vanished. In order to resume growth, persisters have to recover from the persistent state, but the processes behind recovery remain mostly elusive. Deciphering these processes is an essential step toward understanding the persister phenomenon in its entirety. High-throughput proteomics by mass spectrometry is a valuable tool to assess persister physiology during any stage of the persister life cycle, and is expected to considerably contribute to our understanding of the recovery process. In the present study, an Escherichia coli strain, that overproduces the membrane-depolarizing toxin TisB, was established as a model for persistence by the use of high-throughput proteomics. Labeling of TisB persisters with stable isotope-containing amino acids (pulsed-SILAC) revealed an active translational response to ampicillin, including several RpoS-dependent proteins. Subsequent investigation of the persister proteome during postantibiotic recovery by label-free quantitative proteomics identified proteins with importance to the recovery process. Among them, AhpF, a component of alkyl hydroperoxide reductase, and the outer membrane porin OmpF were found to affect the persistence time of TisB persisters. Assessing the role of AhpF and OmpF in TisB-independent persisters demonstrated that the importance of a particular protein for the recovery process strongly depends on the physiological condition of a persister cell. Our study provides important insights into persister physiology and the processes behind recovery of depolarized cells.

Growth Mode and Physiological State of Cells Prior to Biofilm Formation Affect Immune Evasion and Persistence of Staphylococcus aureus

“…CspD is induced in the stationary phase or upon carbon starvation and increases the production of persister cells [70]. Recently, this DNA-binding protein was also proposed to play a role in the postantibiotic recovery of E. coli [71]. An RNA chaperone, CspA, was also efficiently produced during both growth modes, with somewhat higher abundances on biofilm surfaces.…”

Computational Approaches in Proteomics

“…In contrast to TD and MD proteomics analysis, for BU data analysis, a deconvolution step is not required when implementing ESI, due to the rare generation of double-and triple-charged fragment ions (36). Mass spectra raw data are commonly processed by Proteome Discoverer or MaxQuant platforms using several search engines, such as Sequest, Mascot, Andromeda, X!Tandem and COMET, usually against UniProt databases (37)(38)(39). MaxQuant software can also determine protein quantitation and estimate the error of PTM false localization.…”

“…For downstream correlation and clustering analysis, the identified proteins are commonly processed in the Perseus platform (38,40,41). To reduce data complexity, principal component analysis (PCA) has been the method of choice, and also to identify the relatedness of the differentially expressed proteins within and among samples (39,41). Moreover, to interpret the potential function of the datasets obtained, the DAVID platform is commonly used to enrich them with Gene Ontology terms, KEGG pathway information and InterPro protein domains (39,42).…”

“…To reduce data complexity, principal component analysis (PCA) has been the method of choice, and also to identify the relatedness of the differentially expressed proteins within and among samples (39,41). Moreover, to interpret the potential function of the datasets obtained, the DAVID platform is commonly used to enrich them with Gene Ontology terms, KEGG pathway information and InterPro protein domains (39,42). Additionally, constructed networks are commonly visualized in Cytoscape, and in order to identify functional and physical associations among mRNA and protein data, the STRING database is used (43).…”

Proteome dynamics during antibiotic persistence and resuscitation

“…As only the living (persisting) cells are capable of incorporating the label, and the label mass shift can be easily resolved in a mass spectrometer, this approach can be used to selectively label and detect newly synthesized persister proteins in the high background of proteins originating from dead cells. Similar strategies, termed "pulsed" or "dynamic" stable isotope labeling by amino acids in cell culture (SILAC), have previously been used to analyze protein turnover in eukaryotic (Doherty et al, 2009;Schwanhäusser et al, 2011) and prokaryotic cells (Chua et al, 2016;Spanka et al, 2019).…”