INTRODUCTION
Proteins in organisms from all domains of life can be functionally altered by posttranslational modifications (PTMs) that are often species dependent (1). PTMs in bacteria generally occur at much lower levels than in eukaryotes (2). PTMs serve to regulate activity in response to environmental conditions, as they generally occur faster than transcription-translation (2); they can also be used to tune phenotypic diversity (3). Among the high variety of PTMs (1), protein phosphorylation is one of the best-studied examples, partially attributable to major developments in phosphoproteomics.
To date, phosphorylation has been identified on the side chains of different amino acids (aa): serine (Ser), threonine (Thr), tyrosine (Tyr), histidine (His), arginine (Arg), lysine (Lys), aspartate (Asp), and cysteine (Cys) (1). The chemistries involved in the phosphorylation of different amino acids are different. Phosphorylation on Ser/Thr/Tyr results in more stable phosphoester bonds, whereas His/Lys/Arg phosphorylation results in thermodynamically unstable phosphoamidates (1). Phosphorylation on Asp leads to a mixed phosphoacylanhydride, and modification of cysteine, finally, leads to a phosphothiolate (1). Ser/Thr/Tyr phosphorylation appears to be most common in bacteria (4), but this may relate to their thermodynamic stability.
In canonical protein phosphorylation, a phosphate group is transferred to a protein through the action of a kinase and can be removed by a phosphatase; sometimes the kinase and phosphatase functions are encoded by the same bifunctional enzyme (1). Often, the response of bacteria to environmental stimuli is dependent on so-called two-component systems that consist of a membrane-associated two-component sensor histidine kinase (HK) and a cytosolic response regulator (RR). When triggered, autophosphorylation of the HK on a histidine residue results in the transfer of the phosphate group to an aspartate residue on the response regulator. Subsequently, the phosphorylated RR can bind to DNA to regulate the transcription of target genes (1). Other His-phosphorylated proteins include the phosphoenolpyruvate-dependent sugar phosphotransferase system (PTS) proteins, with the exception of the EIIB component, which can be phosphorylated on Cys residues (5). As the name suggests, the donor for the phosphorylation of PTSs is not a kinase but phosphoenol pyruvate (PEP). The PTS phosphorylation cascade includes an unusual bifunctional kinase/phosphatase, called HPr, that in its
Phosphorylation on Ser/Thr is mediated by eukaryote-type Ser/Thr kinases (eSTKs), also known as Hanks-type kinases (6). Interestingly, there is some evidence that eSTKs may outnumber two-component systems (7). Phosphorylation on Tyr in bacteria is mediated by a unique family of proteins known as BY kinases (2, 8, 9).
The exploration of bacterial phosphorylation has greatly been stimulated by the development of mass spectrometry (MS)-based phosphoproteomics techniques that provide a snapshot of the site-resolved phosphorylation state of proteins under given conditions (10, 11). These techniques rely on strategies to enrich for phosphorylated peptides through, for instance, anti-phospho antibodies, strong cation exchange chromatography, chemical modifications, or immobilized metal affinity chromatography (IMAC) (2, 4). Overall, the picture that emerges from these experiments is that both the numbers of phosphorylation sites and the relative abundances of different types of phosphorylation differ between organisms (2, 4, 11, 12). Additionally, it was found that one protein can be dynamically phosphorylated on multiple different sites (4, 13).
In bacteria, phosphorylation has been implicated in diverse processes, such as cell cycle regulation, cellular differentiation, morphogenesis, metabolism, persistence, and virulence (1). Whereas phosphoproteomics initially focused on model organisms, such as the Gram-negative bacterium
Despite the wealth of phosphoproteomic studies (16, 17), phosphorylation in clostridia is largely unexplored. A single study has been performed with
Here, we report the first phosphoproteomic analysis of
RESULTS
A one-step IMAC procedure allows the identification of phosphopeptides from
The number of phosphorylated proteins in bacteria is generally much lower than in eukaryotic cells (1). To benchmark our IMAC workflow, we therefore first analyzed a human cell line (JY cells) where we were expecting a relatively high number of phosphopeptides compared to the number in
Next, we analyzed the phosphoproteome of
FIG 1
Global overview of phosphoproteins in
To compare the different samples, we selected unique peptide sequences that were found to be phosphorylated, without considering phospho-isomers (i.e., phosphorylation at different sites in the same peptide). In general, we observed a progressive increase in protein phosphorylation during growth. We identified most phosphopeptides at later growth phases, with the most prominent increase between the samples taken at the onset of stationary growth phase and the samples from the 24-h cultures (Fig. 1B).
We assessed the reproducibility of the identifications between the different samples at the various time points. At the mid-exponential growth phase, 232 phosphopeptides were identified in all three samples, corresponding to 41% of the total number of peptides identified at this time point (Fig. 1C). For the second time point, 191 phosphopeptides were found in all three samples (20% of the total number identified at this time point). This relatively low percentage was due mainly to one of the samples, in which we found a lower number of phosphopeptides (Fig. 1B, culture 3), as the overlap between phosphopeptides was much higher between the other two cultures at this time point (63% of the total number). Of note, this was not due to technical variation, as analysis of a second, independently processed, sample from the same culture at the same time point showed similar low numbers of phosphopeptides. After 24 h, 1,422 phosphopeptides were common to all three samples (66% of the total number identified at this time point). Overall, this shows significant congruence between our phosphopeptide identifications, despite individual processing of all three biological replicates and time points.
Within one culture, most phosphopeptides observed at the earlier time points were also observed at later stages (as exemplified for culture 2 in Fig. 1D), suggesting a gradual increase in the repertoire of phosphoproteins rather than a large-scale reprogramming. This is also clear from the fact that 155 phosphopeptides were found in all samples at all time points. Nevertheless, we also observed a limited set of proteins that appear to be phosphorylated predominantly, if not exclusively, in the exponential growth phase (Table S1) and are absent from late-stationary-growth-phase samples. One such example is discussed further below.
We performed protein/gene set enrichment analysis of the phosphorylated proteins identified and determined the molecular functions and biological processes associated with these proteins. Figure 2 represents the enrichment maps for each of the three time points, providing an overview by aggregating the enriched terms according to the shared protein set. Comparing the enrichment maps side by side revealed that during the first two stages, there were concentrated, interconnected clusters of functions/processes, i.e., performed by the same set of proteins, a trend that was lost in the late stationary growth phase. At the mid-exponential growth phase, these clusters were largely associated with metabolic and biosynthetic processes, including a few related to phosphorylation, as well as translation-associated functions and processes. At the beginning of the stationary growth phase, additional functions/processes start to be visible, including those related to ion/cation channel activities and ATP-related processes, as well as cell motility and carbohydrate metabolic processes. At the late stationary growth phase, multiple parallel, unconnected functions and processes, like DNA repair, transferases, and catalytic activities, were enriched, as were nucleotide biosynthetic processes that were already visible at the earlier time points. Of note, only the top 30 terms are shown in these maps; i.e., the functions and processes that are visible on top in the early stages likely are still active at the late stage, but they do not appear in the top 30.
FIG 2
Enrichment maps of gene ontology terms obtained from
Phosphorylation related to annotated kinases.
We used the UniProt annotation of the
SpoIIAB is an anti-sigma factor which in other bacteria is known to phosphorylate the anti-sigma factor F antagonist SpoIIAA during the process of sporulation (26, 32). In line with a role during sporulation, SpoIIAA phosphopeptides were exclusively identified in the 24-h sample, but not in samples taken at the mid-exponential or early stationary growth phase. The phosphorylation of SpoIIAA was found on Ser56 (LP = 0.83), Ser57 (LP = 0.55), Ser77 (LP = 1.00), Ser82 (LP = 0.84), and Ser83 (LP = 0.97) (Table S1). Of these, Ser56 corresponds to the equivalent serine of
FIG 3
Spectra of conserved phosphorylation in regulatory proteins of
We also found phosphorylation of the kinase SpoIIAB itself (Table S1) at Ser13 (LP = 0.98), Ser17 (LP = 1.0), and Ser106 (LP = 1.00). Lower-confidence sites in SpoIIAB include Ser123 (LP = 0.50), Ser124 (LP = 0.49), and Thr111 (LP = 0.62). Of all these sites, only Ser106 was found consistently in all three 24-h cultures, and the assignment of the site was confirmed by manual inspection of the MS/MS data (Fig. S3A).
RsbW is the anti-sigma factor for σB, an important regulator of the stress response in Gram-positive bacteria (27, 31). The interaction of RsbW with σB is inhibited through the binding of the anti-σ factor antagonist RsbV to RsbW (27, 31). The interaction between RsbV and RsbW is negatively regulated by the RsbW-dependent phosphorylation of RsbV (27). We found two high-LP phosphopeptides for RsbV, indicative of phosphorylation of Ser57 (LP = 0.93) and Ser84 (LP = 1.00), and low-LP phosphorylation of Thr58 (LP = 0.56). Work with
HPr kinase (HprK/ScoC) can phosphorylate the phosphocarrier protein HPr (PtsH), a protein involved in the import of monosaccharides as part of the phosphotransferase system (PTS) (5). In
Whereas phosphorylation of Hanks-type serine-threonine kinases in other organisms has been described, this is not the case for the
A single substrate for PrkC has been characterized: CwlA (CD630_11350) (34). For this protein, phosphorylation was reported on Ser136 and Thr405, with the latter being specific for PrkC. In our analyses, we found multiple Ser-phosphorylated peptides, including Ser136 (LP = 1.00) and Thr405 (LP = 1.00) (Table S1). Additionally, it appears that PrkC is involved in cell wall homeostasis and antimicrobial resistance based on phenotypes of a
Dynamic phosphorylation within a protein.
PrkC may also regulate cell division, such as DivIVA (33, 43–46). We identified high-confidence phosphopeptides derived from
We were interested to see if more proteins demonstrate dynamic phosphorylation. Manual inspection of the list of phosphopeptides identified two hypothetical proteins that show different patterns of phosphorylation in exponential phase from those in stationary growth phase, particularly with CD630_27170 and CD630_28470 (Table S1). Phosphorylation of CD630_27170 occurs exclusively on Ser573 (LP = 1.00) in exponential growth phase, at Ser573 and Ser286 (LP = 1.00) at the onset of stationary growth phase, and at multiple other sites at 24 h. CD630_28470 is phosphorylated at Ser186 (LP = 1.00) in exponential growth phase and at the onset of stationary phase but not in late stationary phase (Table S1). Together, these examples indicate that posttranslational modification within a particular protein in
Indirect detection of cysteine phosphorylation.
In addition to using protein kinase-dependent phosphorylation, in which the phosphate group is donated by ATP, bacteria can use the phosphate group from phosphoenolpyruvate (PEP) in a process catalyzed by a PTS. This is of particular relevance for the import of monosaccharides into the cell, during which the sugars are converted to phospho-sugars (5). The initial enzymes involved in this cascade of events (enzyme I [PtsI] and Hpr [PtsH]) are phosphorylated at conserved histidine residues and do not have specificity for the different monosaccharides. The subunits of the enzyme II complex (EIIA, EIIB, and EIIC) provide the necessary specificity for different monosaccharides. In our phosphoproteomic data set, we found that many of these subunits of the different EII complexes are phosphorylated at serine and threonine residues (Table S1). Activation of the monosaccharide during uptake, however, requires the transfer of phosphate from a conserved cysteine residue of the EIIA and EIIB subunits, respectively. We accidently put carbamidomethylation of Cys as a variable, instead of as a fixed, modification during one of our database searches and serendipitously observed cysteine-containing peptides from EII complex subunits that did not appear to be phosphorylated (as expected based on our enrichment strategy) or carbamidomethylated (as expected from our proteomics workflow). Of note, when we searched data from a HeLa tryptic digest, which we use as a standard for our liquid chromatography (LC)-MS/MS setup, with the same parameters, we found very few free cysteines (18 out of 2,753 peptide spectrum matches of peptides containing a cysteine).
As an example of a tryptic peptide with a free cysteine in our
In conclusion, it appears that a one-step Fe-IMAC enrichment procedure can indirectly detect at least a subset of cysteine-phosphorylated proteins.
DISCUSSION
In this study, we characterized the phosphoproteome of a laboratory strain of
Many different methods have been used to analyze bacterial phosphoproteomes (1, 2, 11). A crucial step in these analyses is the method used to enrich for phosphopeptides (2, 4). In our study, we employed Fe-IMAC, as this has recently been shown to be a very effective way for analyzing bacterial phosphoproteomes (35, 36). The number of phosphoproteins identified here is in line with these studies and indicates that this is a suitable method for the analysis of phosphoproteins in
Our work provides a starting point for experimental validation of the identified phosphoproteins, as well as a dissection of the contribution of individual protein kinases and phosphatases. Generally, the actions of protein kinases are better understood than those of phosphatases (16). We identified site-specific phosphorylation in the protein kinase PrkC (Table S1). A knockout of the gene encoding this kinase demonstrates pleiotropic effects (33), and comparing wild-type and
The identified phosphopeptides also offer a possibility of characterizing the effect of phosphorylation on substrate proteins, by mutating residues in these proteins to nonphosphorylatable analogs (alanine for serine/threonine and phenylalanine for tyrosine) or phosphomimetic negatively charged amino acids (aspartate and glutamate) (2, 10). Such an approach may be preferable to mutating the kinases and phosphatases, as these can have overlapping specificities (57). However, mutating specific phosphorylation sites can lead to phosphorylation on neighboring sites, due to which the phenotypic consequences of the mutation may be only partially penetrant.
In previous phosphoproteomic analyses, a large divergence was often observed in the lists of phosphorylated proteins even within a single genus or species (2, 58, 59). In our experiments, we observed apparently conserved phosphorylation events on several proteins, including SpoIIA, SpoIIAB, RsbV, and others (Fig. 3; Table S1), which may appear in contrast to these observations. Part of the reason for the limited overlap in previous experiments may have been the diversity in enrichment methods and experimental protocols, limited sensitivities of the workflows, and variability in sampling methods and time points. Our workflow allows the reliable identification of a large set of phosphoproteins in comparison with those found by other studies (11, 12, 16, 17). Moreover, we applied rigorous quality control of phosphosite assignment. It is likely that developments in mass spectrometry instrumentation, data acquisition, and analysis pipelines will reveal that many more processes in fact may be regulated by conserved phosphorylation events. For example, a recent phosphoproteomic analysis of
Our work indicates that levels of protein phosphorylation in
We focused our in-depth analyses on those proteins for which the phosphosite assignment could be done with high probability; where necessary, we manually verified the automatic identifications. Nevertheless, we also observed peptides for which it was not possible to assign the modification to a particular residue, similar to what others did (59). For those proteins, alternative fragmentation techniques (65) or processing of the sample—using, for instance, proteases other than trypsin—might yield better results.
The phosphoproteome described here complements existing omics approaches for
MATERIALS AND METHODS
Chemicals.
Unless noted otherwise, chemicals were obtained from Sigma-Aldrich Chemie.
Cell culture.
JY cells were grown at 37°C in Iscove's modified Dulbecco's medium supplemented with 10% heat-inactivated fetal bovine serum and
Sample preparation.
Cell pellets were resuspended in 3 mL lysis buffer [8 M urea–50 mM Tris-HCl, pH 7.4, 1 mM orthovanadate, 5 mM Tris(2-carboxyethyl)phosphine (TCEP), 30 mM chloroacetamide (CAA), PhosSTOP phosphatase inhibitor (Thermo Fischer Scientific), cOmplete mini EDTA-free protease inhibitor, 1 mM MgCl2] and incubated for 20 min at room temperature. Cells were lysed by sonication (Soniprep 150 ultrasonic disintegrator [MSE]; 5 × 30 s; amplitude, 12 μm). In between sonication steps, samples were cooled on ice for 30 s. Next, samples were centrifuged for 15 min at 7,200 ×
Proteins were precipitated by first adding 16 mL of methanol (Actu-All Chemicals) and mixing, followed by the addition of 4 mL of chloroform (Merck Millipore) and mixing. After the addition of 12 mL of Milli-Q water (obtained from an Elga Pure Lab Chorus 1 machine), the samples were mixed by vortexing. Samples were then centrifuged for 15 min at 11,000 ×
Trypsin was added at a ratio of 1:25 (wt/wt), and overnight digestions were performed at 37°C. The next day, samples were centrifuged for 10 min at 11,000 ×
Phosphopeptide enrichment.
A one-step phosphopeptide IMAC enrichment procedure was performed using a 4- by 50-mm ProPac IMAC-10 analytical column (Thermo Fisher Scientific). For the JY cells, an equivalent of 5 mg of protein was used for one IMAC purification, while for the
Tryptic peptides (resuspended in 260 μL solvent A, of which 250 μL was injected) were loaded, and nonphosphopeptides were removed by washing the column for 20 min with solvent A. Peptide separation was performed using a linear gradient from 0% to 45% of solvent B (0.5% [vol/vol] NH4OH). The complete peak fraction of phosphopeptides between 46 and 49 min was manually collected and lyophilized prior to mass spectrometric analysis.
LC-MS/MS analysis.
Lyophilized peptides were reconstituted in 100 μL water-formic acid (100/0.1 [vol/vol]) and analyzed by on-line C18 nano-high-performance liquid chromatography (nano-HPLC) MS/MS with a system consisting of an Easy nLC 1200 gradient high-performance liquid chromatography (HPLC) system (Thermo Fisher Scientific, Bremen, Germany) and an Orbitrap Fusion Lumos Tribrid mass spectrometer (Thermo Scientific). Samples (10 μL, in duplicate) were injected onto a homemade precolumn (100 μm by 15 mm; Reprosil-Pur C18-AQ, 3 μm; A. Maisch, Ammerbuch, Germany) and eluted on a homemade analytical nano-HPLC column (30 cm by 75 μm; Reprosil-Pur C18-AQ, 3 μm). The gradient was run from 2% to 36% solvent B (water-acetonitrile-formic acid [20/80/0.1, vol/vol/vol]) in 120 min. The nano-HPLC column was drawn to a tip of ∼5 μm, which acted as the electrospray needle of the MS source.
The Lumos mass spectrometer was operated in data-dependent MS/MS mode for a cycle time of 3 s, with the higher-energy C-trap dissociation (HCD) collision energy at 32 V and recording of the master scan 2 (MS2) spectrum in the Orbitrap. In MS1, the resolution was 120,000 and the scan range
Data analysis.
MaxQuant software (version 1.5.1.2) was used to process the raw data files, which were searched against the
The MaxQuant output table “phospho (STY)Sites.txt” (Table S1) was used for further analysis of the phosphopeptides. For the stringent phosphorylation site assignment, only peptides with a localization probability score of >0.95 were selected. Because we used the match-between-runs option, we only used unique peptides without considering the phosphorylation site (i.e., not considering phospho-isomers) for comparison of different samples.
Gene ontology protein/gene set enrichment analysis was performed using a sorted list of all identified phosphoproteins at each analysis time point, i.e., mid-exponential, beginning stationary, and late stationary growth phases. The sorting was based on the abundance of the phosphoproteins identified and determined by the ion intensity, as reported by the mass spectrometer in the output of MaxQuant (cutoff at a score of 40 and false-discovery rate at 1%). We considered the phosphoproteins that were reported at least once in any of the three replicates. Each protein identification was considered only once in the sorted list, with a rank determined by the most intense results that we have obtained in any of the three replicates by any protein-associated phosphopeptide. The enrichment was performed using the weighted Kolmogorov-Smirnov-like statistic as implemented in the common gene set enrichment analysis. We limited our analysis to the top 30 enriched gene ontology terms. We generated enrichment maps that represent the relationship between the enriched terms as a network. The edges of an enrichment map correspond to the number of shared proteins between the associated terms. These maps give a higher-level overview of the enrichment analysis and allow identifying functional modules with ontology terms that are related to each other by the underlying protein set used. For the functional analysis, we used all gene ontology annotations of
Data availability.
All data are contained within the manuscript or the associated supplemental material or are available from the authors upon request. The mass spectrometry proteomics data have been deposited into the ProteomeXchange Consortium via the PRIDE partner repository (68) with the data set identifier PXD029475.
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
Copyright © 2022 Smits et al. This work is published under https://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Abstract
ABSTRACT
Phosphorylation is a posttranslational modification that can affect both housekeeping functions and virulence characteristics in bacterial pathogens. In the Gram-positive enteropathogen
IMPORTANCE In this paper, we present a comprehensive analysis of protein phosphorylation in the Gram-positive enteropathogen
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer