Virulent B. abortus 2308 was obtained from Tecon Biological Co., Ltd. (Urumqi, China) and cultured in tryptone soy agar (TSA). The colonies were cultured in TSA medium at 37°C in a shaking incubator. Before performing analysis of PTMs, we extracted the whole proteins of Brucella in exponential stage and stationary growth stage in three separate preliminary experiments, and found that the modification level of proteins in exponential stage is higher than that in stationary growth stage through detecting by western blot. Besides, Brucella has better growth activity (Yang et al., 2021) and invasiveness (Rossetti et al., 2009) in exponential stage, so we finally chose Brucella in exponential stage as research object in this study (Detailed data are listed in supplemental Supplementary Figure S1). The bacteria solution was sonicated three times on ice using a high intensity ultrasonic processor in lysis buffer (8 M urea, 1% protease Inhibitor Cocktail) (protease Inhibitor Cocktail, Beyotime, China, P1005). The remaining cell debris was removed by centrifugation at 12,000 g at 4°C for 10 min (Wang et al., 2019). The supernatant was collected, and the protein concentration was determined using a BCA kit (Beyotime, China, P0012) according to the manufacturer’s instructions.

For digestion, the protein solution was reduced with 5 mM dithiothreitol for 30 min at 56 °C and alkylated with 11 mM iodoacetamide for 15 min at 25°C in the dark. The protein sample was then diluted by adding 100 mM TEAB (tetraethyl-ammonium bromide) to a urea concentration of less than 2 M. Finally, trypsin was added at a 1:50 trypsin-to-protein mass ratio for the first digestion overnight and 1:100 trypsin-to-protein mass ratio for a second 4 h digestion.

To enrich lysine 2-hydroxyisobutyrylation, lysine succinylation, lysine crotonylation, lysine acetylation, and lysine malonylation modified peptides, tryptic peptides dissolved in NETN buffer (100 mM NaCl, 1 mM EDTA, 50 mM Tris-HCl, 0.5% NP-40, pH 8.0) were incubated with pre-washed antibody beads (PTM-804 for 2-hydroxyisobutyrylation, PTM-409 for succinylation, PTM-503 for crotonylation, PTM-104 for acetylation, and PTM-901 for malonylation, PTM Bio) at 4°C overnight with gentle shaking. Then the beads were washed four times with NETN buffer and twice with H₂O. The bound peptides were eluted from the beads with 0.1% trifluoroacetic acid. Finally, the eluted fractions were combined and vacuum-dried. For LC-MS/MS analysis, the resulting peptides were desalted with C18 ZipTips (Millipore) according to the manufacturer’s instructions.

The tryptic peptides were dissolved in 0.1% formic acid (solvent A) and directly loaded onto a homemade reversed-phase analytical column (15-cm length, 75-μm internal diameter). The gradient comprised an increase from 6 to 23% solvent B (0.1% formic acid in 98% acetonitrile) over 26 min, then 23–35% over 8 min, and 35–80% over 3 min. Elution was completed by holding at 80% for the last 3 min. The procedure was performed at a constant flow rate of 400 nL/min using an EASY-nLC 1000 ultra-performance liquid chromatography (UPLC) system.

The peptides were subjected to NSI source followed by MS/MS in Q ExactiveTM Plus (Thermo Fisher Scientific, Waltham, United States) coupled online to the UPLC. The applied electrospray voltage was 2.0 kV. The m/z scan range was 350 to 1,800 for a full scan, and intact peptides were detected in the Orbitrap at a resolution of 70,000. Peptides were then selected for MS/MS using a NCE setting of 28, and the fragments were detected in the Orbitrap at a resolution of 17,500. We followed a data-dependent procedure that alternated between one MS scan and 20 MS/MS scans with a 15.0 s dynamic exclusion. Automatic gain control was set at 5E4. The fixed first mass was set to 100 m/z.

The BspF gene was amplifified by PCR and cloned into pET-GST (Bioon, Shenzhen, China, zt208) to produce GST-tagged fusion proteins with a PreScission protease cleavage site between GST and the target proteins. The proteins were expressed in E. coli strain Rosetta and induced by 0.2 mM isopropyl-b-D-thiogalactopyranoside (IPTG) when the cell density reached an OD_600nm of 0.8. After growth at 16°C for 12 h, the cells were harvested, re-suspended in lysis buffer (1×PBS, 2 mM DTT and 1 mM PMSF) and lysed by sonication. The cell lysate was centrifuged at 20,000 g for 45 min at 4°C to remove cell debris. The supernatant was applied onto a self-packaged GST-affifinity column (2 ml glutathione Sepharose 4B; GE Healthcare, United States) and contaminant proteins were removed with wash buffer (lysis buffer plus 200 mM NaCl). The fusion protein was then digested with PreScission protease at 4°C overnight. The protein was eluted with lysis buffer. The eluant was concentrated and further purifified using a Superdex-200 (GE Healthcare, United States) column equilibrated with a buffer containing 10 mM Tris–HCl pH 7.8, 500 mM NaCl, and 5 mM DTT. Refer to Supplementary Figure S2.

HEK-293T and HeLa cells were cultured in Dulbecco’s minimal essential medium containing 10% fetal bovine serum (FBS, Gemin, United States, A73D00E) at 37°C and 5% CO₂. Plated cells were cultured in 10 ml of 10% growth medium for 1 day before transfection, until they reached 70–90% confluency. Lipofectamine 2000 reagent and the plasmids were diluted in Opti-MEM medium (Gibco, Thermo Fisher Scientific, United States, 2085119) and incubated for 10 min. The diluted plasmids and Lipofectamine 2000 were gently mixed, incubated for 30 min at room temperature, and then added to the cells. Six hours after transfection, the growth medium was removed from the cells and replaced with 10 ml of 2% FBS maintenance medium. Test plates were transfected with 16 μg of HA-RicA plasmid (constructed by our lab), and the control plates were transfected with 16 μg of pCMV-HA (Bioon, Shenzhe, China, zt296) plasmid. The cells were collected 30 h after transfection. Protein expression level was assessed using western blotting.

Cell lysates were harvested at 30 h, and the protein concentration was determined. Equivalent quantities of cell lysates were denatured in 5X loading buffer and boiled at 100°C for 10 min. Equal amounts of proteins were separated by 10% SDS-PAGE (EpiZyme, China, PG112) and transferred to polyvinylidene fluoride (PVDF) membranes (Merck Millipore, United States, ISEQ00010). The membranes were then washed in Tris-buffered saline with Tween 20 (TBST) and blocked in TBST containing 5% skimmed milk for 2 h at 25°C. Membranes were incubated overnight at 4 °C with antibodies for detecting Cr (PTM Bio, China, PTM-545RM), HA (Beyotime, China, AF0039) and β-actin (Beyotime, China, AF5001), followed by incubation with HRP-conjugated secondary antibody (Beyotime, China, A0192) at 25°C for 2 h. Signals were detected with Clarity ECL (Enhanced Chemiluminescence) reagents (Beyotime, China, P0018FS).

The MS/MS data were processed using the MaxQuant search engine (v.1.5.2.8). Tandem mass spectra were searched against the SwissProt database concatenated with the reverse decoy database. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD030621. Trypsin was specified as a cleavage enzyme, allowing up to two missing cleavages. The mass tolerance for precursor ions was set as 20 ppm in the first search and 5 ppm in the main search, and the mass tolerance for fragment ions was set as 0.02 Da. Carbamidomethyl on cysteine was specified as a fixed modification, and oxidation on methionine was specified as a variable modification. FDR (false discovery rate) was adjusted to <1%, and the minimum score for peptides was set at > 40.

The GO annotation proteome was derived from the UniProt-GOA database (http://www.ebi.ac.uk/GOA/). First, the protein identity (ID) was converted to UniProt ID and then mapped to GO IDs by protein ID. If some identified proteins were not annotated by the UniProtGOA database, the InterProScan software was used to annotate the GO function of the protein based on the protein sequence alignment method. Subsequently, proteins were classified by Gene Ontology annotation based on three categories: biological process, cellular component, and molecular function. The KEGG (Kyoto Encyclopedia of Genes and Genomes) database was used to annotate protein pathways. First, the KEGG online service tool KAAS (KEGG Automatic Annotation Server) was used to annotate the proteins’ KEGG database description. The annotation results were mapped to the pathway database using the online service tool KEGG mapper.

The subcellular localization prediction software wolfpsort was used to predict subcellular localization. Wolfpsort is an updated version of PSORT/PSORT II for the prediction of eukaryotic sequences. For prokaryotic species, the subcellular localization prediction software CELLO was used.

Proteins were classified by GO annotation into three categories: biological process, cellular compartment, and molecular function. For each category, a two-tailed Fisher’s exact test was employed to test the significance of enrichment of the identified proteins against all proteins in the species database. The KEGG enriched pathways were similarly identified. Pathways and GO-terms with a corrected p-value < 0.05 were considered significant. Significant pathways were classified into hierarchical categories according to the KEGG website.

All identified protein database accessions or sequences were searched against the STRING database version 10.5 for protein-protein interactions. Only interactions between the proteins belonging to the searched data set were selected, thereby excluding external candidates. STRING defines a metric called “confidence score” to define interaction confidence; we identified all interactions that had a high confidence score (>0.7). The interaction network from STRING was visualized in the R package networkD3.

All data are analyzed with GraphPad Prism version 8 (GraphPad Software; San Diego, CA) and expressed as means ± SEM. Statistical significance was calculated using two-tailed Student’s t test, unless stated otherwise. p-values of <0.05 (typically ≤0.05) were regarded as statistically significant. NS stands for not statistically significant.

Combined the growth activity, invasiveness and modification level of Brucella in two stages, the Brucella in exponential stage was chosen in our study (Supplementary Figure S1). To obtain an overall view of lysine modification levels and patterns, we performed a global proteomic analysis of Brucella using tryptic digestion, antibody affinity enrichment, and high-resolution liquid chromatography-tandem mass spectroscopy (LC-MS/MS). lysine 2-hydroxyisobutyrylation (Hib), lysine succinylation (Sc), lysine crotonylation (Cr), lysine acetylation (Ac), and lysine malonylation (Ma) were assessed (Figure 1A). The circos plot shows the abundance configurations of the proteome and five lysine PTM-omics for proteins isolated from B. abortus 2308. All five modification-types have similar location distributions in chromosome I (2,121,359 bp), whereas in chromosome II (1,156,948 bp), lysine acetylation and lysine succinylation were more prominent (Figure 1B).

Across the five lysine PTM-omics, we identified 1,904 modified proteins in the B. abortus proteome. The most frequent type was Hib, while Ma was least frequent. In the identified proteins the distribution of modification sites across proteins were as follows: Hib 6,953 sites and 1,336 proteins; Sc,5,825 sites and 1,293 proteins; Cr, 5,709 sites and 1,290 proteins; Ac, 4,317 sites and 1,191 proteins; and Ma, 1,693 sites and 612 proteins (Figure 1C). The number of malonylated proteins was considerably lower than that of the other four PTM-omics. These five PTM-omics shared 548 proteins, and most proteins (≥96%) underwent more than one modification. Approximately 4% of proteins had only one type of PTMs. Specifically, 69, 62, 44, 55, and 4 proteins underwent only Sc, Hib, Ac, Cr, and Ma, respectively (Figure 1D). These proteins were involved in energy metabolism and protein synthesis. By statistical analysis of amino acid sequence before and after all Hib, Sc, Cr, Ac and Ma in samples, we calculated the trend of amino acid sequence in the region of Hib, Sc, Cr, Ac and Ma sites. Such analysis can reveal the sequence characteristics of the modified site and thus infer or identify the enzyme associated with the modification (Supplementary Figure S3).

To validate the mass spectrometry data and to ensure that sample preparation reached standard conditions, the mass error and peptide length of the identified peptides were examined. As shown in Supplementary Figure S4, the mass error is less than 10 ppm, and the precision satisfies the bioinformatics analysis. For peptide length (Supplementary Figure S5), the majority of the distribution was between seven and twenty amino acids, which is in accordance with the characteristic length of tryptic peptides. MS/MS information related to these PTM peptides was deposited in the iProX database with the accession number.

To determine the reliability of the LC-MS/MS data, we conducted a verification experiment using an antibody against the crotonyl group. We selected two effector proteins, RicA and BspF, for validation experiments in vivo and in vitro, respectively. According to the LC-MS/MS data, RicA (BAB1-1279) underwent four types of PTMs (Cr, Ma, Sc, and Ac), and BspF (BAB1-1948) underwent two types of PTMs (Hib and Cr). Both were crotonylated, but RicA had five Cr sites, whereas BspF had only one (Figure 2A). Therefore, we investigated the Cr of these two proteins. As expected, RicA and BspF were detected with the Cr antibody in vivo and in vitro, respectively (Figures 2B,C). These results were consistent with the mass spectroscopy data.

Subcellular localization of modified proteins in B. abortus 2308 provides clues about the possible functions of the modified proteins. The results (Figure 3A) showed that the largest proportion of modified proteins was assigned to cytoplasmic (Hib, 74%; Sc, 78%; Cr, 74%; Ac, 78%; and Ma, 85%), followed by periplasmic (Hib, 15%; Sc, 12%; Cr, 15%; Ac, 13%; and Ma, 10%) and membrane (Hib, 10%; Sc, 7%; Cr, 10%; Ac, 8%; and Ma, 3%) compartments. Overview of the subcellular localization of modified proteins is shown in Figure 3B. All five types of modification had similar subcellular location distributions. Interestingly, neither Sc nor Ma affected the proteins identified in the outer membrane, which indicates that these two modifications do not occur in outer membrane proteins. For example, lptD is located in the extracellular membrane and is involved in the synthesis of Brucella Lipopolysaccharide (LPS), but neither Sc nor Ma was detected.

Gene ontology (GO) analysis and Kyoto encyclopedia of genes and genomes (KEGG) pathway enrichment analysis were used to understand the biological significance of the five types of modifications. Statistical distribution charts of modified proteins under GO categories are shown in Figure 4. Cr, Hib, and Sc were similarly distributed across biological processes, while distribution of Ac and Ma were similar. All five modifications occurred on proteins mainly located in the ribose phosphate metabolic process, ribose phosphate biosynthetic process, purine-containing compound biosynthetic process, peptide metabolic process, peptide biosynthetic process, oxidoreduction coenzyme metabolic process, nucleoside phosphate metabolic process, and amide biosynthetic process. Among the five modifications, Ma was more widely distributed, and Ac was more concentrated. The flavin-containing compound biosynthetic process, glucose-6-phosphate metabolic process, and riboflavin metabolic process were associated with Cr, Hib, and Sc. Ac and Ma were mainly involved in carboxylic acid metabolic processes, oxoacid metabolic processes, and organic acid biosynthetic processes. As for cellular components, all five modifications were mainly enriched in intracellular part, intracellular, and cytoplasmic part. Cr was the most widely distributed in the cellular components, and only Cr participated in the outer membrane, outer cell membrane, envelope, and cell envelope; Hib did not participate in non-membrane-bound organelles, and Ma was absent in the cytosolic part. Concerning molecular functions, all five modifications were closely related to catalytic activity. Cr, Hib, Sc, and Ac were associated with transport activity, but Ma was associated with structural molecular activity. In addition, all five modifications are ubiquitous in energy generation and exchange as well as in replication, recombination, and repair processes. In particular, malonylated proteins was more prominent in the processes of translation and the structure of the ribosome, but their enrichment in processes of cell cycle regulation, cell division, and chromosome segmentation were markedly lower compared to that of the other four modifications. Consequently, these five modifiers appear to collectively perform their respective functions to regulate the intracellular survival of Brucella.

KEGG is an information network that connects known molecular interactions, such as metabolic pathways, complexes, and biochemical reactions. In Brucella, these five modifications were observed for different functions, with proteins enriched in different pathways. The KEGG pathway enrichment bubble plots of proteins corresponding to the modification sites are shown in Figure 5. Sc mainly regulated the synthesis of lysine, while Hib was involved in glucose metabolism pathways, which was in agreement with previous research showing that Hib affects the glycolysis process in yeast (Huang et al., 2017). Ac mainly affected carbon and fatty acid metabolism in Brucella, whereas Cr was involved in carbon metabolism, fatty acid metabolism, and the TCA cycle, and many crotonylated proteins were related to ribosomes, which is consistent with a previous study showing that the transcription and replication of genes are closely related to lysine Cr (Wei et al., 2017a), (Wei et al., 2017b). Ma was mainly enriched in KEGG pathways related to the TCA cycle, antibiotic synthesis, and amino acid synthesis.

Through biological analysis of the PTM-omics data, we found that a large number of proteins involved in virulence were modified. To study and discuss the effect of modification on Brucella, we screened a few classes of proteins that are important during the growth, reproduction, invasion, and infection of Brucella, including virulence factors, stress particles, and essential genes, which may help Brucella to survive intracellularly in the host and cause persistent infection (Tables 1, 2). In Table 1, we list the proteins associated with immune evasion (lpxC, lpxD, fabZ, lpxA, lpxB, kdsA, htrB, kdsB, acpXL, lpxE, wboA, pgm, wbpZ, manAoAg, manCoAg, pmm, wbkA, gmd, per, wzm, wzt, wbkB, wbkC, lpsA, lpsB/lpcC, wbdA, wbpL, manBcore, manCcore, lpxK, and waaA/kdtA), the modification of these proteins may enhance the ability of Brucella to escape immune phagocytosis of host cells. About intracellular survival (CbetaG), regulation (bvrR and bvrS), and T4SS structural proteins (VirB1–11), we thus hypothesized that the modification would improve the ability of Brucella to help the bacteria survive intracellularly in the host. Notably, we observed that most of these virulence factors were associated with PTMs. In addition, 10 of 15 known effectors of T4SS (RicA, VceA, VceC, BPE043, BtpA, BspB, BspC, BspE, BspF, and SepA) underwent various modifications. VceA can suppress autophagy in the process of infection, thereby to escape the immune killing system of the host, which is conducive to the intracellular survival of Brucella. VceC can promote autophagy in the process of infection, which is unhelpful for escaping the immune killing system of the host. In Table 2, VceC is modified by all five of these modifications but VceA is not modified by Ac and Ma, we hypothesized that these two modifications may regulate the autophagy process and thus affect the intracellular survival of Brucella (Zhang et al., 2019b). Moreover, essential genes (for example, SerS, FusA, and RpsL), one of the 14 membrane proteins of Brucella (Omp31-1), and regulatory proteins (sodC), which can help bacteria escape the killing effects of phagocytes, are modified by all five of these modifications (Table 2). Consequently, it is likely that the five modifications play a crucial role in virulence and these modified sites can potentially serve as therapeutic targets for brucellosis.

Protein interaction networks are composed of individual proteins that interact with each other to participate in gene expression regulation, cell cycle regulation, biological signal transmission, energy and substance metabolism, and other processes. Among the 1904 modified proteins in Brucella, we screened 48 modified virulence-related proteins and constructed an interaction network (Figure 6). Here, we found that the LPS-associated proteins were mainly modified by Hib and Sc, with lesser amounts of Cr and Ac; proteins modified by Ma were minimal. The LPS-associated proteins, including pgm, wbdA, manBcore, wboA, wbkC, and gmd were mainly Hib and Sc, but they also exhibited Cr and Ac. However, wbkC was not modified by Ma. The modification types of the effectors and essential genes were more diverse, with prominent Hib, Sc, and Cr, and lesser amounts of Ac and Ma. Modification of RicA, BPE043, BspC, SerS, and FusA were consistent with this trend; however, the essential genes FtsZ were not modified by Ma. The physiological interaction between these modified proteins may contribute to their synergistic effect in B. abortus 2308.