Comparative Phenotypic, Proteomic, and Phosphoproteomic Analysis Reveals Different Roles of Serine/Threonine Phosphatase and Kinase in the Growth, Cell Division, and Pathogenicity of Streptococcus suis

Eukaryote-like serine/threonine kinases (STKs) and cognate phosphatases (STPs) comprise an important regulatory system in many bacterial pathogens. The complexity of this regulatory system has not been fully understood due to the presence of multiple STKs/STPs in many bacteria and their multiple substrates involved in many different physiological and pathogenetic processes. Streptococci are the best materials for the study due to a single copy of the gene encoding STK and its cognate STP. Although several studies have been done to investigate the roles of STK and STP in zoonotic Streptococcus suis, respectively, few studies were performed on the coordinated regulatory roles of this system. In this study, we carried out a systemic study on STK/STP in S. suis by using a comparative phenotypic, proteomic, and phosphoproteomic analysis. Mouse infection assays revealed that STK played a much more important role in S. suis pathogenesis than STP. The ∆stk and ∆stp∆stk strains, but not ∆stp, showed severe growth retardation. Moreover, both ∆stp and ∆stk strains displayed defects in cell division, but they were abnormal in different ways. The comparative proteomics and phosphoproteomics revealed that deletion of stk or stp had a significant influence on protein expression. Interestingly, more virulence factors were found to be downregulated in ∆stk than ∆stp. In ∆stk strain, a substantial number of the proteins with a reduced phosphorylation level were involved in cell division, energy metabolism, and protein translation. However, only a few proteins showed increased phosphorylation in ∆stp, which also included some proteins related to cell division. Collectively, our results show that both STP and STK are critical regulatory proteins for S. suis and that STK seems to play more important roles in growth, cell division, and pathogenesis.


Introduction
Reversible protein phosphorylation is a key mechanism that regulates many cellular processes in both prokaryotes and eukaryotes [1]. In bacteria, protein phosphorylation, which is mainly mediated by histidine kinases (HKs), eukaryote-like serine/threonine defined minimal medium (CDM) supplemented with 1% glucose was prepared as described previously [44].

Construction of Plasmids, Mutants, and Complemented Strains
All primers used in this study are listed in Table S2. Because stp overlaps with stk, to avoid affecting stk when deleting stp, the coding sequence of the erythromycin resistance gene was used to replace the CDS of stp [36]. Briefly, upstream and downstream 1000 bp fragments flanking stp and/or stk were amplified from the S. suis genome using primers Pup_F/R/Pdown_F/R, Kup_F/R/Kdown_F/R, and pKup_F/R/pKdown_F/R, and the erm cassette was amplified from plasmid pAT18 [45] using primers Erm_F/R. The fragments were cloned into plasmid pSET4s [46] by seamless cloning using the ClonExpress MultiS One Step Cloning Kit (Vazyme, Nanjing, China), generating the plasmids of pSET4s-P/K/PK. S. suis competent cells were transformed with pSET4s derived plasmids, and the single exchanged and mutant strains were screened as previously described [47]. Considering the variation in plasmid copy numbers, the complement strains were constructed by integrating the target gene into the chromosome [48]. The stp, stk, and stp/stk flanked with the respective upstream and downstream fragments were cloned into pSET4s, and the complement strains were constructed as described above. The expression of STP or STK in each strain was detected by real time quantitative PCR (RT-qPCR) and Western blotting.

RT-qPCR
To test the expression of stp or stk, total RNA was extracted from WT, ∆stp, and ∆stk strains. Briefly, RNA was isolated using the SV Total RNA Isolation System (Promega, Madison, WI, USA) according to the manufacturer's instructions. HiScript Q Select RT SuperMix (+gDNA wiper) (Vazyme, Nanjing, China) was used to remove residual genomic DNA and synthesize cDNA according to the manufacturer's instructions. cDNA was used as the template for RT-qPCR using the TB Green™ Premix Ex Taq™ II kit (Takara) according to the manufacturer's instructions with an ABI 7500 HT Sequence Detection System (Applied Biosystems, Waltham, MA, USA). The expression level of the tested gene was calculated using the 2 −∆∆Ct method and normalized to the housekeeping gene gapdh of S. suis [49].

Western Blotting
To further confirm the mutant and complemented strains, we performed Western blot analysis to detect the expression of STP and STK in each strain as previously described [50]. Cells at the mid-log phase of each strain were collected, washed with PBS, and lysed by sonication. The amount of total protein loaded was normalized using a Micro BCA protein assay kit (Cwbiotech, Beijing, China), followed by Western blot analysis using anti-STP, anti-STK, or anti-GidA serum, respectively, which were produced in mice as previously described [51]. Antibody-tagged protein bands were detected by using the Western ECL Substrate Kit (Cat# 1705060, Bio-Rad, Hercules, CA, USA).

Animal Infection Experiments
All animal experiments were approved by the Laboratory Animal Monitoring Committee of Huazhong Agricultural University (HZAUMO-2019-074) and performed according to the recommendations in the Guide for the Care and Use of Laboratory Animals of Hubei Province, China. Six-to seven-week-old C57BL/6 mice with similar body weights were used for all animal infection experiments. Mice were randomly divided into different groups and injected intraperitoneally. The survival rate of the mice was recorded every 24 h post-infection for seven days. For the bacterial load assays, mice were euthanized at each indicated time point, and the organs were collected, weighed, homogenized in sterile saline, and plated onto TSA for cell counting. A competitive infection assay [52], an accurate and sensitive approach to determine relative virulence, was used to compare the pathogenicity of the three mutants, in which mice were infected intraperitoneally Microorganisms 2021, 9,2442 4 of 21 with a 1:1 mixture of the ∆stp and ∆stk (2.85 × 10 7 CFU in total), ∆stp and ∆stp∆stk (2.8 × 10 7 CFU in total), ∆stk and ∆stp∆stk (3.05 × 10 7 CFU in total), followed by bacterial load enumeration as described above. The competitive index (CI) was calculated by output (log 10 CFU1/log 10 CFU2)/Input (log 10 CFU1/log 10 CFU2). ∆stp, which was inframe substituted with the Erm r coding sequence, has erythromycin resistance, which can be differentiated by plating on an erythromycin plate. The numbers of the ∆stk and ∆stp∆stk strains in the recovered colonies were determined by using PCR.

Growth Assay
The growth assay was measured in a Bioscreen C system (Lab Systems Helsinki, Vantaa, Finland). To detect bacterial growth in TSB and CDM, overnight grown cell culture was diluted in TSB and CDM medium in a Bioscreen plate to an initial OD 600 nm of 0.1. The OD 600 nm was automatically monitored every 30 min at 37 • C with shaking. The assay was performed in triplicate.

Morphological Analysis
Light microscopy, scanning electron microscopy (SEM), and dual-color structured illumination microscopy (SIM) were used to analyze bacterial morphology. Bacteria were cultured in TSB containing 10% inactivated NBS overnight at 37 • C, diluted 1: 100 into fresh medium, grown to the mid-log phase, collected, and washed three times with PBS, stained with Gram staining reagents according to established procedures, and observed under a light microscope. Scanning electron microscopy (SEM) was performed as previously described [53], with some modifications. Bacteria at the mid-log phase were washed three times with PBS, spotted onto glass coverslips, fixed with 2.5% glutaraldehyde overnight at 4 • C, dehydrated with increasing concentrations of ethanol of 30% (15 min), 50% (15 min), 70% (15 min), 90% (15 min), and 100% (15 min, twice), air-dried, covered with a 10 nm gold/platinum layer (JSM-6390LV, JEOL, Japan), and observed with SEM (JFC-1600, JEOL, Tokyo, Japan). FDL, which is a kind of fluorescent D-amino acid dye (FDAA) [54], was used to label the newly synthesized peptidoglycan. Alexa Fluor™ 647 NHS Ester (AF-647) (Thermo Scientific, Waltham, MA, USA) was used to label the outline of S. suis. Bacteria were washed three times with PBS, mixed with FDL at a final concentration of 200 µM, incubated at 37 • C for 10 min, and washed three times with PBS. The resuspension was mixed with AF-647 at a final concentration of 20 nM, incubated at 28 • C for 30 min, washed three times with PBS, and imaged with dual-color structured illumination microscopy (SIM) (Nikon Instruments, Inc., Tokyo, Japan) (FDL, excitation at 490 nm and emission at 525 nm; AF-647, excitation at 651 nm and emission at 672 nm).

Trypsin Digestion
Triplicate cultures were harvested at the mid-log phase, resuspended in lysis buffer, and ultrasonicated (PTM Bio, Zhejiang, China). The protein was precipitated with precooled 20% trichloroacetic acid (TCA) for 2 h at −20 • C and redissolved in 8 M urea. The protein was digested overnight with trypsin (1: 50 trypsin: protein mass ratio).

TMT Labeling
The peptides were desalted using a Strata X C18 SPE column (Phenomenex), vacuumdried, and reconstituted with 0.5 M TEAB according to the manufacturer's protocol (Thermo Fisher Scientific, Waltham, MA, USA). Each sample was labeled with different tags, incubated for 2 h at room temperature, pooled, desalted, and dried by vacuum centrifugation.

HPLC Fractionation and Enrichment of Phosphorylated Peptides
Peptide fractions were acquired on a Thermo Betasil C18 column (5 µm particles, 10 mm ID, 250 mm length) through high pH reversed-phase HPLC. For enrichment of phosphomodified peptides, tryptic peptide mixtures were mixed with IMAC microspheres Microorganisms 2021, 9, 2442 5 of 21 in loading buffer with gentle vibration. After centrifugation, the supernatant was removed, and IMAC microspheres with bound phosphopeptides were acquired. These IMAC microspheres were washed with 50% acetonitrile/6% trifluoroacetic acid and 30% acetonitrile/0.1% trifluoroacetic acid continuously to remove nonspecifically adsorbed peptides. To elute the enriched phosphopeptides from the IMAC microspheres, elution buffer containing 10% NH 4 OH was added, and the enriched phosphopeptides were eluted with vibration. Finally, the peptides were lyophilized for LC-MS/MS analysis.

LC-MS/MS Analysis
After the peptides were dissolved using 0.1% formic acid and separated with an EASY-nLC 1000 ultrahigh-performance liquid phase system, they were subjected to the NSI source followed by Orbitrap Fusion TM system analysis. The NCE (set to 28) and Orbitrap (resolution = 17,500) were used to select peptides for MS/MS and detect the fragments.

Data Analysis
In this study, both proteomics and phosphoproteomics assays were conducted on the same cohort with the same batch of samples. The relative quantitative value of the phosphopeptide is divided by the relative quantitative value of corresponding proteins to remove the influence from protein expression. The MaxQuant search engine (v.1.5.2.8) was used to analyze the MS/MS data. Parameter setting: UniProt_Streptococcus_suis_strain_SC84 (2104 sequences). Image Pro Plus 6.0 was used for chain-length measurement. ImageJ was used to measure cell length and width.

Statistical Analysis
GraphPad Prism (version 8) software was used for statistical analysis. The Student's t-test was used to analyze the differences between the two groups. Survival rates between different groups in the animal infection assay were analyzed by the log-rank (Mantel-Cox) test [55].

Construction of S. suis ∆stp, ∆stk, and ∆stp∆stk Mutants
To investigate the functions of protein phosphorylation and dephosphorylation mediated by STK and STP in S. suis, we constructed three mutant strains, ∆stp, ∆stk, and ∆stp∆stk of S. suis, as well as their complement strains C∆stp, C∆stk, and C∆stp∆stk. As previously reported that stp and stk were cotranscribed [20], it was critical to ensure that deletion of stp did not affect the expression of stk. Therefore, the coding sequence (CDS) of the erythromycin resistance gene was used to replace the CDS of stp to avoid downstream polar effects ( Figure 1A), and the expression of stp and stk in each strain was analyzed by RT-qPCR and Western blot. Figure 1B,C show that the expression level of STK was comparable in the ∆stp and WT strains at both the mRNA and protein levels. The complemented strains also showed restored expression of STP and STK, respectively. These results suggested that the mutant strains and their complement strains were constructed successfully.

The Role of STP, STK, and STP/STK in Bacterial Pathogenicity in Mice
STP and STK are believed to be cognate pairs responsible for reversible protein phosphorylation, and they have both been shown to be involved in bacterial pathogenesis [34,35,56]. However, the exact role of each protein in S. suis pathogenesis remains to be further investigated. Therefore, we used a mouse infection model to compare the virulence of ∆stp, ∆stk, and ∆stp∆stk strains with that of the WT strain of S. suis. In the mouse survival assay, only 7% (1/15) of the mice intraperitoneally injected with 8 × 10 8 CFU of the WT strain survived. In comparison, the survival rates for the ∆stp, ∆stk, and ∆stp∆stk strains were 27% (4/15), 73% (11/15), and 67% (10/15), respectively, indicating that stk played a more important role than stp in regulating the virulence of S. suis (Figure 2A).
We further performed an in vivo colonization assay in mice with the WT strain and three mutants. The results showed that from 12 h post-infection (hpi) to 24 hpi, the bacterial load of the WT strain in each organ slowly decreased, indicating a strong ability to colonize the host. In contrast, the bacterial loads of the mutant strains showed significantly faster clearance in each organ. In particular, at 24 hpi, the ∆stk strain was almost completely cleared in the lung, spleen, and brain ( Figure 2B), which was consistent with the results of the mouse survival assay. In contrast, the ∆stp and ∆stp∆stk strains showed a significantly higher ability to colonize mice than the ∆stk strain. To more accurately compare the virulence of the mutants, we performed a series of competitive infection assays. As shown in Table 1 and Figure 2C-E, ∆stp significantly outcompeted ∆stk (CI 6h = 1.16746, p < 0.05; CI 9h = 1.24600, p < 0.05) and ∆stp∆stk (CI 6h = 1.54531, p < 0.05; CI 9h = 1.39062, p < 0.05). ∆stp∆stk showed a significantly higher bacterial load than ∆stk at 6 hpi (CI 6h = 1.26691, p < 0.05), but no significant difference in bacterial load was observed at 9 hpi (CI 9h = 1.24696, p > 0.05). Collectively, our data demonstrated that the deletion of stk significantly attenuated the virulence of S. suis, which was consistent with the observations in our previous reports [17]. In comparison, stp deletion resulted in reduced impact compared to the stk deletion on impairing the virulence of S. suis. These results suggested that the mutant strains and their complement strains were constructed successfully. Total RNA was extracted from the WT, ∆stp, and ∆stk strains, respectively, and reverse-transcribed to cDNA, which was used as the template for RT-qPCR. The gapdh gene was used as the internal reference for normalization. Data are representative of three independent experiments with the mean ± standard deviation, and fold changes were calculated relative to wild-type expression levels. Fold change ≥ 1.5 and p value ≤ 0.05 were considered to represent a differential expression. (C) Detection of STP and STK expression by Western blot. Cells of each indicated strain at the mid-log phase were collected, washed with PBS, and lysed by sonication, followed by immunoblot analysis using STP, STK, and GidA polyclonal antibodies. Protein samples were normalized before loading. The GidA antibody was used as a control. C∆stp, C∆stk, and C∆stp∆stk indicate the corresponding complemented strains.

The Role of STP, STK, and STP/STK in Bacterial Pathogenicity in Mice
STP and STK are believed to be cognate pairs responsible for reversible protein phosphorylation, and they have both been shown to be involved in bacterial pathogenesis [34,35,56]. However, the exact role of each protein in S. suis pathogenesis remains to be further investigated. Therefore, we used a mouse infection model to compare the viru- Transcription levels of the stp and stk genes in the WT, ∆stp, and ∆stk strains. Total RNA was extracted from the WT, ∆stp, and ∆stk strains, respectively, and reverse-transcribed to cDNA, which was used as the template for RT-qPCR. The gapdh gene was used as the internal reference for normalization. Data are representative of three independent experiments with the mean ± standard deviation, and fold changes were calculated relative to wild-type expression levels. Fold change ≥ 1.5 and p value ≤ 0.05 were considered to represent a differential expression. (C) Detection of STP and STK expression by Western blot. Cells of each indicated strain at the mid-log phase were collected, washed with PBS, and lysed by sonication, followed by immunoblot analysis using STP, STK, and GidA polyclonal antibodies. Protein samples were normalized before loading. The GidA antibody was used as a control. C∆stp, C∆stk, and C∆stp∆stk indicate the corresponding complemented strains. p < 0.05; CI9h = 1.24600, p < 0.05) and ΔstpΔstk (CI6h = 1.54531, p < 0.05; CI9h = 1.39062, p < 0.05). ΔstpΔstk showed a significantly higher bacterial load than Δstk at 6 hpi (CI6h = 1.26691, p < 0.05), but no significant difference in bacterial load was observed at 9 hpi (CI9h = 1.24696, p > 0.05). Collectively, our data demonstrated that the deletion of stk significantly attenuated the virulence of S. suis, which was consistent with the observations in our previous reports [17]. In comparison, stp deletion resulted in reduced impact compared to the stk deletion on impairing the virulence of S. suis.  Mice were intraperitoneally injected with 8 × 10 8 CFU of WT, ∆stp, ∆stk, or ∆stp∆stk, respectively, and PBS was used as a negative control. Significant differences in survival rates between different groups were analyzed by the log-rank (Mantel-Cox) test. Fifteen mice were used in each group. (B) Bacterial colonization assay. Mice were intraperitoneally injected with 9 × 10 7 CFU of WT, ∆stp, ∆stk, or ∆stp∆stk, respectively. Mice were euthanized at each indicated time point. The spleen, lung, kidney, and brain were collected, resuspended in PBS, homogenized, and plated on TSA plates for colony enumeration. Five mice were used in each group. Data are presented as the mean ± standard deviations. Statistical significance is determined by two-tailed, unpaired Student's t-tests (ns, p value > 0.05; *, p value < 0.05; **, p value < 0.01; ***, p value < 0.001). (C-E) Competitive infection assay. Mice were infected with a 1:1 mixture of ∆stp and ∆stk (C), ∆stp and ∆stp∆stk (D), or ∆stp∆stk and ∆stk (E). Spleens were collected, resuspended in PBS, homogenized, and plated on TSA plates for colony enumeration. Five mice were used in each group. Statistical significance was determined by two-tailed, unpaired Student's t-tests (ns, p value > 0.05; *, p value < 0.05). Note: n is the number of animals in each group. CI = Output (log 10 CFU1/log 10 CFU2)/Input (log 10 CFU1/log 10 CFU2). * indicates p value < 0.05; NS indicates no statistical significance.

The Role of STP and STK in Growth and Cell Division
We further tested other important phenotypes, including growth and cell division of the ∆stp, ∆stk, and ∆stp∆stk strains. The growth of each mutant and WT strain in TSB medium or chemically defined minimal medium (CDM) was tested. As shown in Figure 3A, the growth of WT, ∆stp, and ∆stp∆stk was comparable in TSB, whereas ∆stk presented retarded growth and reached a lower OD 600 nm in the stationary phase ( Figure 3A). Surprisingly, the ∆stk strain could barely grow in CDM ( Figure 3B). However, the growth of ∆stp was similar to that of the WT strain, and the growth of ∆stp∆stk was much slower than that of WT but better than that of ∆stk ( Figure 3B). To determine the cause of such growth defects, CDM supplemented with different carbon sources, casamino acids (CAAs), vitamins, purines, or complex nutrients (tryptone or peptone) was used to test the growth of the ∆stk strain ( Figure 3C and Figure S1). As shown in Figure 3C, in addition to the observation that the growth of ∆stk was greatly increased in CDM supplemented with tryptone or peptone, it was interesting to see that its growth in CDM supplemented with multivitamins was also significantly recovered. These results suggested that STK was more important than STP for the growth of S. suis.
STK has been reported to play important roles in cell division in Streptococcus [23,29], while limited information about the role of STP in cell division has been reported. Therefore, we subsequently analyzed the morphology, cell division, and nascent peptidoglycan synthesis of WT and mutant strains using light microscopy (LM), scanning electron microscopy (SEM), and dual-color structured illumination microscopy (SIM). One of the most significant morphological changes among the mutants was the increase in the length of bacterial chains. All three mutants formed long chains, with each chain comprising over ten cells compared with the WT strain, which showed a normal chain length with each chain harboring three to five cells ( Figure 4A-D). The morphology of all the complemented strains was restored ( Figure S2). These results revealed the presence of defects in cell division or daughter cell separation. Further analysis with SIM revealed abnormal cell division in both ∆stp and ∆stk strains. By staining with Alexa Fluor 647 dye (AF-647), which specifically labels the cytoplasmic membrane, and fluorescein-D-lysine (FDL), which labels nascent peptidoglycan (PG), followed by visualization using SIM, the cell morphology and division septum could be clearly observed ( Figure 4C). Compared with the WT cells that displayed a standard ellipsoidal shape, 37.3% (142/381) of ∆stp cells showed a rounded morphology due to a decreased length-width ratio ( Figure 4E-G). Moreover, in 9.4% (36/381) of ∆stp cells, the Z-ring was not localized to the mid-cell, and the angles of the Z-rings with respect to the cell long axis were incorrect, leading to the formation of twisted cell chains ( Figure 4C, marked by yellow arrows). The ∆stk cells also showed abnormal cell division; however, they displayed a distinctly aberrant morphology. Approximately 6.8% (26/385) of the ∆stk cells underwent disordered division, leading to the formation of asymmetric daughter cells, and 10.4% (40/385) formed elongated cells containing multiple, unconstricted cell division rings ( Figure 4C, marked with yellow arrows). In contrast, the ∆stp∆stk strain showed relatively normal cell division. These results demonstrated that STP and STK both played an essential role in cell division. STK has been reported to play important roles in cell division in Streptococcus [23,29], while limited information about the role of STP in cell division has been reported. Therefore, we subsequently analyzed the morphology, cell division, and nascent peptidoglycan synthesis of WT and mutant strains using light microscopy (LM), scanning electron microscopy (SEM), and dual-color structured illumination microscopy (SIM). One of the most significant morphological changes among the mutants was the increase in the length of bacterial chains. All three mutants formed long chains, with each chain comprising over ten cells compared with the WT strain, which showed a normal chain length with each chain harboring three to five cells ( Figure 4A-D). The morphology of all the complemented strains was restored ( Figure S2). These results revealed the presence of defects in cell division or daughter cell separation. Further analysis with SIM revealed abnormal cell division in both Δstp and Δstk strains. By staining with Alexa Fluor 647 dye (AF-647), which specifically labels the cytoplasmic membrane, and fluorescein-D-lysine (FDL), which labels nascent peptidoglycan (PG), followed by visualization using SIM, the cell morphology and division septum could be clearly observed ( Figure 4C). Compared with the WT cells that displayed a standard ellipsoidal shape, 37.3% (142/381) of Δstp cells showed a rounded morphology due to a decreased length-width ratio ( Figure 4E-G). Moreover, in 9.4% (36/381) of Δstp cells, the Z-ring was not localized to the mid-cell, and the angles of the Z-rings with respect to the cell long axis were incorrect, leading to the formation of twisted cell chains ( Figure 4C, marked by yellow arrows). The Δstk cells also showed abnormal cell division; however, they displayed a distinctly aberrant morphology. Approximately 6.8% (26/385) of the Δstk cells underwent disordered division, leading

Comparative Proteomic Analysis Revealing the Influence of stp and stk on Protein Expression
To investigate the influence of STP and STK on protein expression in S. suis, we performed a comparative proteomics analysis with the WT, ∆stp, and ∆stk strains using liquid chromatography mass spectrometry/mass spectrometry (LC-MS/MS), as illustrated in Figure S3A. A total of 1427 proteins were identified, constituting 72% of the proteins encoded by the S. suis SC19 genome. Among these identified proteins, the abundance of 1379 proteins was quantitatively compared between the WT and each mutant strain ( Figure  S3B). As shown in Figure 5A,B, compared with the WT strain, the ∆stk strain possessed the most dramatic change in protein abundance, in which 139 genes were differentially expressed, with 37 proteins upregulated and 102 proteins downregulated. For the ∆stp strain, 82 genes were differentially expressed compared with the WT strain, including 47 upregulated and 35 downregulated genes (fold change > 1.5, p < 0.05). to the formation of asymmetric daughter cells, and 10.4% (40/385) formed elongated cells containing multiple, unconstricted cell division rings ( Figure 4C, marked with yellow arrows). In contrast, the ΔstpΔstk strain showed relatively normal cell division. These results demonstrated that STP and STK both played an essential role in cell division.

Comparative Proteomic Analysis Revealing the Influence of stp and stk on Protein Expression
To investigate the influence of STP and STK on protein expression in S. suis, we performed a comparative proteomics analysis with the WT, Δstp, and Δstk strains using liquid chromatography mass spectrometry/mass spectrometry (LC-MS/MS), as illustrated in Figure S3A. A total of 1427 proteins were identified, constituting 72% of the proteins encoded by the S. suis SC19 genome. Among these identified proteins, the abundance of 1379 proteins was quantitatively compared between the WT and each mutant strain ( Figure  S3B). As shown in Figure 5A,B, compared with the WT strain, the Δstk strain possessed the most dramatic change in protein abundance, in which 139 genes were differentially The EggNOG v5.0 database was used to analyze the functions of the differentially expressed proteins [57]. As shown in Figure 6, apart from the proteins with unknown functions, the differentially expressed proteins between the WT strain and each mutant were found to be distributed in almost all functional groups. ∆stk harbored the most downregulated proteins, in which the differentially expressed proteins most frequently fell in categories including K (Transcription), P (Inorganic ion transport and metabolism), J (Translation, ribosomal structure, and biogenesis), and G (Carbohydrate transport and metabolism). In contrast, ∆stp presented fewer downregulated proteins but had more proteins with increased abundance in which the differentially expressed proteins most frequently fell in categories including E (Amino acid transport and metabolism), P (Inorganic ion transport and metabolism), K (Transcription), and C (Energy production and conversion) ( Figure 6). The data suggested that both STP and STK were global regulators of protein expression. As ∆stk showed retarded growth in CDM but the growth was restored in CDM supplemented with multivitamins, by analyzing the differentially expressed genes in the mutants, we found that a coA-binding protein (NZ_CP020863.1_prot_WP_012027318.1_1268) was significantly downregulated in ∆stk but was not differentially expressed in ∆stp (Table S3).
(B9H01_06780), fHBP (B9H01_01430), and Permease (B9H01_06785), and only one, ArcC (B9H01_03140), was upregulated in Δstk. In contrast, in Δstp, only three VFs were significantly downregulated, including MRP (B9H01_03725), ATP-binding cassette (B9H01_06780), and Permease (B9H01_06785), and five VFs were upregulated, including FeoB (B9H01_06805), PurD (B9H01_00260), GroEL (B9H01_00900), SLY (B9H01_06770), and AdhE (B9H01_01465), compared with those in the WT strain. These results corresponded with the virulence phenotype associated with the disruption of STK with significantly attenuated virulence of S. suis, while deletion of STP showed a gently reduced virulence.      As there was a difference in virulence between ∆stp and ∆stk strains, the abundance of several known virulence factors (VFs) in the mutant strains was compared with that in the WT strain [58]. As shown in Figure

Comparative Phosphoproteomic Analysis
STK and STP are serine/threonine kinases and phosphatases, respectively. Therefore, we next analyzed the changes in the abundance of serine/threonine phosphorylated proteins in Δstp and Δstk compared with the WT strain. In Δstp, a total of 73 unique phosphosites involving 50 proteins showed a significantly altered abundance (fold change > 1.5, p < 0.05), among which 14 proteins were less abundant and 38 proteins were more abundant compared with those in the WT strain. Between the Δstk and WT strains, 87 unique phosphosites involving 67 proteins were identified with a changed abundance (fold change > 1.5, p < 0.05), among which 29 proteins were less abundant and 39 proteins were more abundant ( Figure 8A and Table S4). A Venn diagram analysis showed that a total of 37 proteins were identified with changed abundance in both the Δstp and Δstk strains, suggesting potential overlapping functions of STP and STK ( Figure 8B).
By COG analysis, it was shown that the phosphoproteins with changed abundance in both Δstp and Δstk strains were mainly related to "Translation, ribosomal structure and

Comparative Phosphoproteomic Analysis
STK and STP are serine/threonine kinases and phosphatases, respectively. Therefore, we next analyzed the changes in the abundance of serine/threonine phosphorylated proteins in ∆stp and ∆stk compared with the WT strain. In ∆stp, a total of 73 unique phosphosites involving 50 proteins showed a significantly altered abundance (fold change > 1.5, p < 0.05), among which 14 proteins were less abundant and 38 proteins were more abundant compared with those in the WT strain. Between the ∆stk and WT strains, 87 unique phosphosites involving 67 proteins were identified with a changed abundance (fold change > 1.5, p < 0.05), among which 29 proteins were less abundant and 39 proteins were more abundant ( Figure 8A and Table S4). A Venn diagram analysis showed that a total of 37 proteins were identified with changed abundance in both the ∆stp and ∆stk strains, suggesting potential overlapping functions of STP and STK ( Figure 8B). cell division-related proteins, including MapZ, GpsB, MapZ, FtsZ, DivIVA, SepF, FtsW, MltG, Jag, and GlmS. The remaining proteins were mainly involved in metabolism, such as Hpr, NeuB, PfKA, ADK, GT, PhoH, and LmrC, and protein translation, such as InfB, EF-P, EF-G, and BipA (Table 2). In contrast, among the fourteen proteins that showed increased abundance after Ser/Thr phosphomodified peptide enrichment in the Δstp strain, only four proteins, including DivIVA, STK, MltG, and GlmM, were involved in cell division, and the others were involved in metabolism, such as NADP, PrfA, GAPDH, OppF, Hprk, ArgS, SecA, and IDH (Table 2). These results suggested that STP and STK might affect the cell division, energy metabolism, and protein translation of S. suis in a manner dependent on Ser/Thr protein phosphorylation.  By COG analysis, it was shown that the phosphoproteins with changed abundance in both ∆stp and ∆stk strains were mainly related to "Translation, ribosomal structure and biogenesis", "Carbohydrate transport and metabolism" and "Cell cycle control, cell division, chromosome partitioning". In most of the COG categories, ∆stk presented more phosphoproteins with altered abundance than ∆stp ( Figure 8C and Table S3), which coincided with the above observations that deletion of stk caused more severe phenotypic changes than stp deletion, including the impact on growth, metabolism, and virulence of S. suis.
Specifically, as shown in Table S4, several important proteins involved in central energy metabolism were identified to contain phosphosites with altered abundance in both ∆stp and ∆stk strains. HprK is a metabolic regulator that mediates the phosphorylation of HPr to control the metabolic rate [59]. The phosphorylation at S/T residues of Hpr was downregulated in ∆stp and ∆stk, while that of an S residue of HprK was upregulated only in ∆stp. The phosphorylation levels of GAPDH and GlmM were increased in both mutant strains compared with the WT strain, while the phosphorylation level of PfkA was decreased in both mutant strains compared with the WT strain. The phosphosite abundances of PtsEI and Pyk were increased in ∆stk and did not change in ∆stp. However, Eno exhibited an opposite situation (Figure 9).

Figure 9.
Proteins involved in bacterial glycolysis and carbohydrate transport with altered phosphorylation levels in Δstp and Δstk. The names of the proteins with altered phosphorylation levels are shown in dark red. In the brackets, the green arrow on the left indicates the change in phosphorylation level in the ∆stp strain compared with WT (p < 0.05), and the blue arrow on the right indicates the change in the ∆stk strain compared with WT (p < 0.05). The hyphen indicates no significant change (p > 0.05). Pgi, phosphoglucose isomerase; Pfk, phosphofructokinase; Fba, fructose-1,6bisphosphate aldolase; Tpi, triosephosphate isomerase; Gap, glyceraldehyde-3-phosphate dehydrogenase; Pgk, phosphoglycerate kinase; Pgm, phosphoglycerate mutase; Eno, enolase; Pyk, pyruvate kinase; EI, PTS-enzyme I; Hpr, phosphocarrier protein; HprK, HPr kinase/phosphorylase. Table 2. Proteins with down-and up-regulated abundance after phospho-peptide enrichment in Δstk and Δstp compared with WT.

No. Protein Name Protein Description Amino Acid Position Regulated Type Ratio p Value Function
Δstk/WT   Next, we analyzed the proteins with a reduced phosphorylation level in the ∆stk strain and increased phosphorylation level in the ∆stp strain, which might be their potential direct substrate proteins. Among the 29 proteins that showed a reduced abundance after enrichment of Ser/Thr phosphomodified peptides in ∆stk, a large proportion were cell division-related proteins, including MapZ, GpsB, MapZ, FtsZ, DivIVA, SepF, FtsW, MltG, Jag, and GlmS. The remaining proteins were mainly involved in metabolism, such as Hpr, NeuB, PfKA, ADK, GT, PhoH, and LmrC, and protein translation, such as InfB, EF-P, EF-G, and BipA (Table 2). In contrast, among the fourteen proteins that showed increased abundance after Ser/Thr phosphomodified peptide enrichment in the ∆stp strain, only four proteins, including DivIVA, STK, MltG, and GlmM, were involved in cell division, and the others were involved in metabolism, such as NADP, PrfA, GAPDH, OppF, Hprk, ArgS, SecA, and IDH (Table 2). These results suggested that STP and STK might affect the cell division, energy metabolism, and protein translation of S. suis in a manner dependent on Ser/Thr protein phosphorylation. Note: The assay was performed with triplicate and the ratio was presented as the average protein abundance. The p value indicates the statistical difference of each protein between the mutant strain and the WT strain by using the Student's t-test.

Discussion
The eukaryote-like Ser/Thr protein kinases (STKs) and phosphatases (STPs) in bacteria have recently attracted extensive research interest due to their important roles in cell division, metabolism, and pathogenesis in many important bacterial pathogens [22,35,56,[60][61][62]. STKs and STPs comprise a complex regulatory system, of which the mechanism of regulation has not been fully understood. Many studies have been performed to reveal the functions of STKs and STPs, however, most of them focus on STK or STP individually. To provide a holistic understanding of this complex regulatory system, systemic investigations are needed. Therefore, we performed a comparative study in S. suis, an important zoonotic bacterial pathogen where only one copy of gene encoding STK and its cognate STP is present, with ∆stk, ∆stp, ∆stk∆stp, and the wild-type strains. Comprehensive phenotypic analysis combined with comparative proteomics and phosphoproteomics analysis were carried out to reveal the STK/STP mediated regulation network.
In the animal infection assay, we showed that STK was essential for the virulence of S. suis, which is consistent with previous studies [17,20]. Some previous studies also showed that deletion of stp led to decreased virulence [42,63]. Similarly, we presented detailed data showing that stp deletion also compromised the virulence of S. suis, however, the degree in virulence attenuation was much lower than that of stk deletion. The comparative proteomic results also revealed that more VFs showed downregulated expression in the ∆stk strain than in the ∆stp strain, which could potentially explain the difference in virulence between the two strains.
In the growth assay, it is interesting to see that the growth of the ∆stk strain was almost abolished in CDM, and the addition of multivitamins largely restored this growth defect. Among the differentially expressed proteins in ∆stk, we found that coA-binding protein was significantly downregulated (Table S3) and was not differentially expressed in ∆stp, which might explain the partial recovering effect of multivitamins on ∆stk.
It was noticed that in the results of both virulence evaluation and growth assays found that the double deletion mutant did not display a more severe attenuation than the single deletion mutant ∆stk. One possible explanation could be that potential phosphorylation systems in addition to the Ser/Thr kinase-phosphatase system may exist that have common targets. Indeed, crosstalk between different phosphorylation systems has been reported [64,65]. Under this circumstance, when STK and STP are both deleted, the substrate proteins may be phosphorylated by other phosphorylation systems, therefore giving phenotypes that are less severe than each of the single deletion mutants.
Bacterial cell division is a vital process requiring sophisticated spatial and temporal regulation to maintain accurate proliferation and proper morphology [66,67]. Our results showed that both ∆stp and ∆stk displayed a chained morphology, which was consistent with previous studies of that impaired in cell division [22,62] and peptidoglycan synthesis [23,33,68]. Comparative phosphoproteomic data in this study revealed that 31% of the proteins with a decreased phosphorylation level in the ∆stk strain were cell division related proteins, including MapZ, GpsB, FtsZ, DivIVA, SepF, FtsW, MltG [69], Jag [70], and GlmS [71]. Among these proteins, MapZ [26], GpsB [72], FtsZ [28], DivIVA [24], and Jag [73] have been demonstrated to be direct substrates of STK, while whether SepF, FtsW, MltG, and GlmS are STK substrates needs further confirmation. In the ∆stp strain, Di-vIVA, STK, MltG, and GlmM also showed increased phosphorylation levels, suggesting that these proteins may be common targets of STP and STK. The high prevalence of cell division-related proteins among the proteins with changed phosphorylation states in ∆stp and ∆stk further suggests the important roles of STP and STK in S. suis cell division.
Other than cell division-related proteins, it was observed that in ∆stk, several translationrelated proteins also showed a downregulated phosphorylation level, which is consistent with previous studies [27,37]. Although other phosphatases also exist in S. suis, our phosphoproteomic data showed that the phosphorylation level of a variety of proteins was significantly changed in ∆stp, indicating an important role of STP. We also performed a phosphatase activity assay using the method previously described [74] with the WT, ∆stp, and C∆stp strains, which showed that the total phosphatase activity of the ∆stp stain was indeed significantly lower than that of the WT strain ( Figure S4). It is worth noting that Table 2 shows the proteins with downregulated phosphorylation in the ∆stk strain and upregulated phosphorylation in the ∆stp strain, which are more likely to be direct target proteins. As shown in Table S4, other proteins had upregulated phosphorylation levels in the ∆stk strain and downregulated phosphorylation levels in the ∆stp strain. The influence of STP and STK on the phosphorylation of these proteins could be an indirect effect. When comparing the proteins with altered phosphorylation in the ∆stk strain and ∆stp strain, there were some overlapping proteins, such as DivIVA and MltG. However, many more of the proteins were found specifically in each individual mutant strain. These data suggest that STP and STK functionally overlap but may also have specific functions, which has also been reported in previous studies. For example, in S. pneumonia, protein including Pgm, ComB, and RsMF are found to have changed phosphorylation status in ∆stp but not in ∆stk by the phosphoproteomics analysis, indicating the presence of potential STP-specific substrates [8]. In S. pyogenes, STP is shown to be secreted to the host cells to potentially activate proapoptotic signaling cascades; however, its cognate STK is unlikely to interfere with the host cell signaling [39] In this study, comparative proteomics and phosphoproteomics analysis were used to reveal how STK and STP affect the growth, cell division, and virulence of S. suis. However, the omics data can just provide some clues. More experiments are needed to further reveal the underlying mechanisms. For example, questions, such as how STK affects the cell division of S. suis and which substrate proteins are responsible for the formation of the aberrant morphology and mislocalized Z-ing, and why STK deletion causes different growth defects in different growth media need further elucidation. Moreover, in the phosphoproteomics analysis, a variety of proteins showed downregulated phosphorylation in ∆stk and upregulated phosphorylation in ∆stp. These could be their potential substrate proteins. However, more studies are needed to verify them.

Conclusions
Although the STK and cognate STP are involved in the reversible protein phosphorylation, this study suggests that STK may play a much more important role in S. suis growth and pathogenesis than STP. Both ∆stp and ∆stk were defective in cell division, where they displayed an aberrant cell morphology. The ∆stk strain presented many more differentially expressed proteins than the ∆stp strain when compared with the wild-type strain. Many more known virulence factors were downregulated in ∆stk than in ∆stp. In ∆stk a substantial number of the proteins with reduced abundance after phosphopeptide enrichment were involved in cell division, followed by energy metabolism and protein translation. However, only a few proteins showed an increased abundance after phosphopeptide enrichment, which was also related to cell division. Collectively, both STP and STK are critical regulatory proteins in S. suis, and STK may play a more important role than STP in growth, cell division, and pathogenesis.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/microorganisms9122442/s1. Figure S1: Growth analysis of ∆stk strains. (A) The growth of ∆stk strain in CDM or CDM supplemented with each indicated nutrient. Growth was monitored with an automatic plate reader with shaking at 37 • C. Data are presented as means ± SD of triplicate. Figure S2: Morphological analysis of C∆stp, C∆stk, and C∆stp∆stk strains. Gram staining (A) and scanning electron microscopy (SEM) (B) analysis of C∆stp, C∆stk, and C∆stp∆stk strains. Figure  S3: Comparative proteomic analysis between the WT and ∆stp or ∆stk strains. (A) The systematic workflow of the quantitative profiling of the global mass spectrum; (B) The basic statistics of mass spectrum data. Figure S4: Phosphatase activity assay. Cells at the mid-log phase of each indicated strain were collected, washed three times with buffer1 [20 mM Tris-HCl (pH 8.0) and 200 mM NaCl2], resuspended in lyse buffer [50 mM Tris-HCl (pH 8.0), 150 mM NaCl2, 100 U/mL mutanolysin (Sigma), and 1 mM PMSF], and lysed by sonication. The amount of total protein loaded was normalized using a Micro BCA protein assay kit (Cwbiotech, China). Samples were mixed with buffer2 [50 mM Tris-HCl (pH 8.0), 2 mM MnCl 2 , and 20 mM p-nitrophenyl phosphate (p-NPP)], and incubated at 37 • C for 10 min. The absorbance at 405 nm was then measured. Purified recombinant STP protein was used as the positive control, the reaction buffer2 used as the blank. Data are presented as means ± SD of triplicate. Statistical significance was determined by two-tailed, unpaired Student's t-tests (ns, p value > 0.05; *, p value < 0.05). Table S1: Strains and plasmids used in this study. Table S2: oligonucleotides used in this study. Table S3: The differentially expressed proteins in ∆stp and ∆stk compared with WT. Table S4: Proteins identified with changed abundance after phosphopeptide enrichment in ∆stp and ∆stk compared with WT. Data Availability Statement: The mass spectrometry proteomics data have been deposited to the Pro-teomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD029994.