Background

To combat the increasing threat of drug-resistant and life-threatening pathogens, new antimicrobial strategies must be developed. One important approach to achieving this is to understand bacterial adaptation to the host immune response. Recent studies that build on prior work (Shatalin et al., 2011) suggest that bacterial H₂S biogenesis may well be a clinically important adaptive response during infection (Mironov et al., 2017; Shukla et al., 2017; Luhachack et al., 2019; Toliver-Kinsky et al., 2019; Saini et al., 2020). The beneficial traits attributed to H₂S are likely due to highly oxidized sulfur species termed reactive sulfur species (RSS), e.g., organic persulfides (RSSH), which have been shown to be potent antioxidants in mammalian cells (Ida et al., 2014). Recently, RSS has been quantitated in a number of different bacterial pathogens, suggesting that the antioxidant and cytoprotective findings in mammalian cells may well extend to bacterial cells (Peng et al., 2017a and 2017b; Shen et al., 2018; Walsh et al., 2020).

The mechanism by which H₂S/RSS signals in the cell is not completely understood but almost certainly involves the protein post-translational modification (PTM) termed S-sulfuration (persulfidation) (Filipovic et al., 2018; Walsh and Giedroc, 2020). This PTM may provide protection against over-oxidation of protein thiols or perform a regulatory role by altering the chemistry of active-site thiols, thiols in regulatory domains, or those found in transcriptional regulators. Investigating the regulatory nature of protein persulfidation is complicated by the fact that protein thiols can form a number of other redox modifications, e.g., S-sulfenylation (RSOH) or S-nitrosation (RSNO), rendering it difficult to distinguish those proteins for which persulfidation is regulatory versus a consequence of a highly reactive or surface-exposed thiol. Most published methods for the detection of protein persulfidation incorporate: i) streptavidin-based alkylation probes to enrich for thiols and persulfide-harboring peptides followed by selective reduction of the mixed disulfide originating from alkylated persulfides (Gao et al., 2015; Peng et al., 2017b); and/or ii) a tag-switch approach that further exploits the unique reactivity of this mixed disulfide toward a specific nucleophile (Zhang et al., 2014; Park et al., 2015; Zivanovic et al., 2019). Recently, low-pH quantitative thiol reactivity profiling (QTRP) was introduced to directly detect protein persulfides and thiols in the same sample, an important step toward defining the ratio of persulfide to thiol at all cysteines (Fu et al., 2020).

The proteomics workflow presented here was recently applied to the major human pathogen Acinetobacter baumannii, in which several hundred proteins were identified as sites of persulfidation in both untreated and Na₂S-treated cells (Walsh et al., 2020). This method relies on the alkylation of both thiol and persulfide residues using a biotinylated iodoacetamide probe. The proteins are then digested into peptides and enriched using streptavidin bead capture, after which peptides containing a mixed disulfide that is derived only from persulfidated residues are selectively reduced, alkylated, and identified by LC-MS/MS (Figure 1A). Although false positives can result from the reaction of these electrophilic probes with sulfenylated cysteines (Reisz et al., 2013), lower concentrations of reagent and shorter alkylation times can minimize this. In addition, it is important to point out that H₂S reacts directly with a sulfenylated cysteine to form a persulfide (Cuevasanta et al., 2015 and 2019); therefore, the extent to which sulfenylated cysteines would be present in a proteome following incubation of bacterial cell cultures with excess H₂S is unknown, but may well be relatively low.

In a parallel experiment, protein abundance is determined for the entire bacterial lysate without enrichment using a widely used label-free quantitation method (Zhu et al., 2010). The two data sets are then used to calculate or simply compare the change in persulfidation status normalized to the change in protein abundance (Figure 1B), thus pinpointing cysteine thiols for which persulfidation may well be regulatory in nature. This workflow leads to the identification of over 40 proteins in A. baumannii characterized by a ≥2-fold change in the normalized persulfidation status, a significant number of which are transcriptional regulators, including two implicated in biofilm formation (Walsh et al., 2020).

Figure 1. Proteome persulfidation profiling. A. General workflow for the enrichment and identification of proteome persulfides (Peng et al., 2017b). Both untreated (WT) and treated (WT+Na₂S) samples are processed using the same workflow and analyzed independently of one another. B. Resulting data are plotted as a change in persulfidation status for a cysteine-containing peptide in a protein versus a change in the normalized abundance of that protein in WT vs. Na₂S-treated cells (Walsh et al., 2020). Abbreviations: BfmR, biofilm response regulator; Crp-like, cAMP-activated regulator; FabD, ACP-malonyl transferase; FbaA, fructose-bisphosphate aldolase; MnmA, tRNA 2-thiouridine synthase; PDO1, persulfide dioxygenase-1; PrpC, 2-methylcitrate synthase; SQR, sulfide-quinone oxidoreductase.