FlpS, the FNR-Like Protein of Streptococcus suis Is an Essential, Oxygen-Sensing Activator of the Arginine Deiminase System

Streptococcus (S.) suis is a zoonotic pathogen causing septicemia and meningitis in pigs and humans. During infection S. suis must metabolically adapt to extremely diverse environments of the host. CcpA and the FNR family of bacterial transcriptional regulators are important for metabolic gene regulation in various bacteria. The role of CcpA in S. suis is well defined, but the function of the FNR-like protein of S. suis, FlpS, is yet unknown. Transcriptome analyses of wild-type S. suis and a flpS mutant strain suggested that FlpS is involved in the regulation of the central carbon, arginine degradation and nucleotide metabolism. However, isotopologue profiling revealed no substantial changes in the core carbon and amino acid de novo biosynthesis. FlpS was essential for the induction of the arcABC operon of the arginine degrading pathway under aerobic and anaerobic conditions. The arcABC-inducing activity of FlpS could be associated with the level of free oxygen in the culture medium. FlpS was necessary for arcABC-dependent intracellular bacterial survival but redundant in a mice infection model. Based on these results, we propose that the core function of S. suis FlpS is the oxygen-dependent activation of the arginine deiminase system.


Introduction
Streptococcus suis is one of the most important porcine pathogens causing meningitis, septicemia, and polyarthritis. S. suis infections lead to high financial losses in the swine-producing industry worldwide [1]. In addition, S. suis is considered a neglected zoonotic pathogen that can cause pathologies such as meningitis and septicemia, as well as a streptococcal toxic shock-like syndrome in humans [2][3][4][5].
is well defined, but the function of the FNR-like regulator FlpS is yet unknown. The predicted gene for the FNR-like protein of S. suis, FlpS, is located directly upstream of the arcABC operon [23], suggesting a role of FlpS in ADS regulation. In the present study, using growth experiments, transcriptome and gene expression analyses, isotopologue profiling, and mice infection models, we investigated the role of FlpS in a serotype 2 strain of S. suis in more detail.

FlpS is Involved in the Regulation of the arcABC Operon and Metabolic Genes
In order to dissect the role of FlpS in S. suis, we first determined the FlpS regulon. For this, we generated a flpS knock-out strain (10∆flpS) by insertion mutagenesis. The functional knock-out of the flpS gene was controlled by Northern blot analysis ( Figure S1B). Following this, we compared the transcriptome of strain 10∆flpS and its parental wild-type strain 10 grown to the exponential (exp) or to the early stationary (stat) growth phase ( Figure S1A) by cDNA microarray analyses. Microarray data were additionally validated by RT-qPCR experiments for a selected subset of genes in the wild-type strain 10 and strain 10∆flpS (Table S2). Overall, 292 and 318 genes were found to be differentially expressed in strain 10∆flpS in the exp and stat growth phase, respectively ( Figure 1B). In the exp growth phase, higher expression of 208 genes and lower expression of 84 genes suggested both a repressing and activating regulatory role for FlpS. In the stat growth phase this distribution changed to 172 higher-expressed and 142 lower-expressed genes. Classification of clusters of orthologous groups (COG) indicated that, apart from genes that could not be assigned to any specific function, the majority of differentially expressed genes in both growth phases of strain 10∆flpS were associated with carbohydrate transport/metabolism, transcriptional regulation, translation and nucleotide transport/metabolism ( Figure 1A, Tables S3-S5).
The flpS knock-out markedly affected the expression of the arginine deiminase operon (arcABC). Thus, it was approximately 150-fold lower-expressed in the stat growth phase when compared to the wild-type strain, indicating that FlpS is an essential activator of arcABC expression. The influence of FlpS on the expression of the other differentially expressed genes was considerably lower, which suggests a less dominant role of FlpS in their regulation.
We have previously shown that the regulator ArgR is an essential, system-specific regulator of the arcABC operon [33]. Thus, the argR knock-out specifically affected arcABC operon expression without influencing the expression of any other gene. This finding suggested a functional relevance of the weak but differentially regulated genes in strain 10∆flpS. Thus, we analyzed these gene subsets in more detail. Among those, several genes encoded for PTS systems for the uptake of alternative carbohydrate sources. In the exp growth phase, PTS for maltose (SSU0395) and lactose (SSU0892-0893) were higher expressed. Accordingly, the genes for the degradation of these carbohydrate sources to glycolytic intermediates SSU0890-0899 for lactose (tagatose pathway) and SSU0353-0354 for maltose degradation were upregulated. In the exp growth phase the predicted PTS for fructose (SSU0768), trehalose (SSU0217) and glucose/mannose (SSU1583), and in the stat growth phase the PTS for predicted transport of mannose (SSU0199-201), galactose (SSU0403-0406), and N-acetylgalactosamines (SSU1055-1057) were significantly lower-expressed in strain 10∆flpS. In both growth phases, no substantial changes in gene expression were observed for the majority of the genes of the central carbon metabolic routes such as glycolysis, the pentose phosphate pathway (PPP), the fragmentary tricarboxylic acid cycle (TCA), and genes needed for homolactic and mixed-acid fermentation of S. suis. Only gapA and gpmA (both glycolysis) in the exp growth phase, and pflBD and adhE (both mixed acid fermentation) in both growth phases were differentially expressed. The genes related to amino acid biosynthesis were not differentially expressed in strain 10∆flpS. The attenuation of arginine degradation by the affected arcABC expression results in losses of ATP and carbamoyl-phosphate which is generally produced by this pathway. Carbamoyl-phosphate is needed for pyrimidine biosynthesis but can also be synthesized from glutamine by carAB, and in line with the attenuated arcABC expression, these genes and several genes of pyrimidine biosynthesis (pyrABEFGPR, deoA, cdd) were more highly expressed in strain 10∆flpS. Since S. suis strain 10 is auxotrophic for glutamine, the higher demand of glutamine for driving the reaction catalyzed by the carAB gene product may be satisfied by the upregulation of different glutamine transporters. Accordingly, the glutamine transporter genes glnQ1, glnQ4, glnQ5 were more highly expressed in the mutant.
Pathogens 2016, 5, 51 4 of 17 auxotrophic for glutamine, the higher demand of glutamine for driving the reaction catalyzed by the carAB gene product may be satisfied by the upregulation of different glutamine transporters. Accordingly, the glutamine transporter genes glnQ1, glnQ4, glnQ5 were more highly expressed in the mutant. Summary of significantly differentially expressed genes during exp and stat growth of S. suis strain 10ΔflpS and classification of clusters of orthologous groups (COG). C, energy production and conversion; D, cell cycle control, cell division; E, amino acid transport and metabolism; F, nucleotide transport and metabolism; G, carbohydrate transport and metabolism; H, coenzyme transport and metabolism; I, lipid transport and metabolism; J, translation; K, transcription; L, replication, recombination and repair; M, cell wall/membrane biogenesis; O, post-translational modification, protein turnover, chaperones; P, inorganic ion transport and metabolism; R, general function prediction only; S, function unknown; T, signal transduction mechanisms; V, defense mechanisms; [−], no prediction; (B) Venn diagram illustration of the number of significant differentially expressed genes during exp and stat growth of S. suis strain 10ΔflpS; (C) Venn diagram illustration of the number of significant differentially expressed genes during exp and stat growth of S. suis strains 10ΔflpS and 10ΔccpA [34].
We assumed that a FNR/Flp regulon may overlap with a regulon of related global regulators Crp/CcpA. Therefore, we compared the FlpS regulon with a previously published CcpA regulon which comprises not only several genes of the central carbon metabolism but also virulenceassociated genes [34]. We observed in the exp growth phase of strain 10ΔflpS and 10ΔccpA a group of genes whose expression was affected in both mutants ( Figure 1C). These were the genes for PTS systems (SSU0215-SSU0217, SSU1308-1310), a phosphoglucomutase (pgmA), unknown membrane and classification of clusters of orthologous groups (COG). C, energy production and conversion; D, cell cycle control, cell division; E, amino acid transport and metabolism; F, nucleotide transport and metabolism; G, carbohydrate transport and metabolism; H, coenzyme transport and metabolism; I, lipid transport and metabolism; J, translation; K, transcription; L, replication, recombination and repair; M, cell wall/membrane biogenesis; O, post-translational modification, protein turnover, chaperones; P, inorganic ion transport and metabolism; R, general function prediction only; S, function unknown; T, signal transduction mechanisms; V, defense mechanisms; [´], no prediction; (B) Venn diagram illustration of the number of significant differentially expressed genes during exp and stat growth of S. suis strain 10∆flpS; (C) Venn diagram illustration of the number of significant differentially expressed genes during exp and stat growth of S. suis strains 10∆flpS and 10∆ccpA [34].
We assumed that a FNR/Flp regulon may overlap with a regulon of related global regulators Crp/CcpA. Therefore, we compared the FlpS regulon with a previously published CcpA regulon which comprises not only several genes of the central carbon metabolism but also virulence-associated genes [34]. We observed in the exp growth phase of strain 10∆flpS and 10∆ccpA a group of genes whose expression was affected in both mutants ( Figure 1C). These were the genes for PTS systems (SSU0215-SSU0217, SSU1308-1310), a phosphoglucomutase (pgmA), unknown membrane protein complexes (SSU1070-SSU1073, SSU1924-1928), and stress-related genes (hrcA, grpE, dnaJK). However, most of them displayed opposite regulation. In the stat growth phase there was a high number of similarly regulated genes (n = 70). Among them were genes encoding for ribosomal proteins, PTS systems (SSU0198-SSU0199, SSU0402-0407), sugar conversions (SSU0999-1001, SSU1915) and ABC transporters (SSU1194, SSU1233, SSU1372, SSU1574-1575, SSU1851-1853). Genes whose expression was exclusively affected in the FlpS mutant could be assigned to nucleotide metabolism (SSU0311, SSU0734, SSU1823) and ABC transporters (SSU0114-0115, SSU1498-1499, SSU1824-1825), whereas those in the CcpA mutant were assigned to carbohydrate transport (SSU0165-0167, SSU1234), carbohydrate conversions (SSU0676, glgABP, SSU1171, SSU1230, SSU1214) and metabolism (ldh, eno). The majority of genes were regulated in only one growth phase for each mutant strain, indicating that each of the knock-outs predominantly exhibit individual gene regulation. Taken together, FlpS is strongly involved in the activation of the arcABC operon and seems to be involved in the regulation of a selected group of metabolic genes.

Metabolic Characterization of the FlpS Knock-Out Strain 10∆flpS
In contrast to the arcABC-specific regulator ArgR, FlpS seems to be additionally involved in the regulation of metabolic genes. Therefore, we analyzed if the differential expression of metabolic genes in strain 10∆flpS has any relevance for metabolic activity. Since we observed that several PTS and ABC transporters were differentially expressed in strain 10∆flpS, we further analyzed its ability to use different carbohydrate sources for growth in a chemically defined medium ( Figure 2A). The experiments revealed reduced maximal growth of strain 10∆flpS on glucose, cellobiose, and trehalose (α-α 1 -diglucoside). These phenotypes corresponded to the affected expression of the respective genes found by our microarray analysis. SSU0357 and SSU1583 presumably encode for glucose-specific PTS whereas SSU0217 and SSU1309 encode for trehalose-and β-glucoside PTS, respectively. protein complexes (SSU1070-SSU1073, SSU1924-1928), and stress-related genes (hrcA, grpE, dnaJK). However, most of them displayed opposite regulation. In the stat growth phase there was a high number of similarly regulated genes (n = 70). Among them were genes encoding for ribosomal proteins, PTS systems (SSU0198-SSU0199, SSU0402-0407), sugar conversions (SSU0999-1001, SSU1915) and ABC transporters (SSU1194, SSU1233, SSU1372, SSU1574-1575, SSU1851-1853). Genes whose expression was exclusively affected in the FlpS mutant could be assigned to nucleotide metabolism (SSU0311, SSU0734, SSU1823) and ABC transporters (SSU0114-0115, SSU1498-1499, SSU1824-1825), whereas those in the CcpA mutant were assigned to carbohydrate transport (SSU0165-0167, SSU1234), carbohydrate conversions (SSU0676, glgABP, SSU1171, SSU1230, SSU1214) and metabolism (ldh, eno). The majority of genes were regulated in only one growth phase for each mutant strain, indicating that each of the knock-outs predominantly exhibit individual gene regulation. Taken together, FlpS is strongly involved in the activation of the arcABC operon and seems to be involved in the regulation of a selected group of metabolic genes.

Metabolic Characterization of the FlpS Knock-Out Strain 10ΔflpS
In contrast to the arcABC-specific regulator ArgR, FlpS seems to be additionally involved in the regulation of metabolic genes. Therefore, we analyzed if the differential expression of metabolic genes in strain 10ΔflpS has any relevance for metabolic activity. Since we observed that several PTS and ABC transporters were differentially expressed in strain 10ΔflpS, we further analyzed its ability to use different carbohydrate sources for growth in a chemically defined medium ( Figure 2A). The experiments revealed reduced maximal growth of strain 10ΔflpS on glucose, cellobiose, and trehalose (α-α′-diglucoside). These phenotypes corresponded to the affected expression of the respective genes found by our microarray analysis. SSU0357 and SSU1583 presumably encode for glucose-specific PTS whereas SSU0217 and SSU1309 encode for trehalose-and β-glucoside PTS, respectively.  Glucose and glucose-containing di-and oligosaccharides are the major energy sources of S. suis. Since expression of gapA and gpmA was slightly affected in strain 10ΔflpS, we were interested to know if the reduced growth of strain 10ΔflpS on glucose might result from an altered central carbon metabolism due to changes in carbon fluxes. For this, we performed 13 C-isotopologue profiling experiments in which 13 C-labeled [U-13 C6]glucose was supplemented for bacterial growth. In S. suis, metabolization of [U-13 C6]glucose led to the highest molar 13 C rates in alanine, aspartate, serine, and threonine [36]. After growth on [U-13 C6]glucose we did not observe any significant differences in the overall molar 13 C excess ( Figure 2B) or the isotopologue pattern (Table S6) of these four amino acids between the wild-type strain, strain 10ΔccpA, strain 10ΔflpS and a 10ΔflpSΔccpA double mutant, indicating no substantial changes in the activity of the main carbon metabolism.

Loss of FlpS, but not CcpA, Is Dispensable for Virulence Properties of S. suis in Mice Infection Models
The data above suggest that FlpS and CcpA are transcriptional regulators involved in several mostly different metabolic processes. In order to prove if a knock-out of FlpS and/or CcpA has an influence on bacterial fitness and virulence properties of S. suis in vivo, we conducted two different types of mice infection experiments. Firstly, we used an established intranasal infection model to study S. suis colonization of the murine nasopharynx [37]. As expected, none of the mice showed any clinical signs after intranasal infection, since intranasal infection leads to subclinical infection. As a read-out parameter for colonization of the upper respiratory tract, the bacterial load was determined after tracheonasal lavage (TNL). In contrast to the infection with the capsular polysaccharidedeficient strain 10ΔcpsEF which was not able to colonize the nasopharyngeal cavity, we observed effective colonization of the murine nasopharynx by wild-type strain 10 (in nine of 10 mice) and strain Glucose and glucose-containing di-and oligosaccharides are the major energy sources of S. suis. Since expression of gapA and gpmA was slightly affected in strain 10∆flpS, we were interested to know if the reduced growth of strain 10∆flpS on glucose might result from an altered central carbon metabolism due to changes in carbon fluxes. For this, we performed 13 C-isotopologue profiling experiments in which 13 C-labeled [U-13 C 6 ]glucose was supplemented for bacterial growth. In S. suis, metabolization of [U- 13 C 6 ]glucose led to the highest molar 13 C rates in alanine, aspartate, serine, and threonine [36]. After growth on [U-13 C 6 ]glucose we did not observe any significant differences in the overall molar 13 C excess ( Figure 2B) or the isotopologue pattern (Table S6) of these four amino acids between the wild-type strain, strain 10∆ccpA, strain 10∆flpS and a 10∆flpS∆ccpA double mutant, indicating no substantial changes in the activity of the main carbon metabolism.

Loss of FlpS, but not CcpA, Is Dispensable for Virulence Properties of S. suis in Mice Infection Models
The data above suggest that FlpS and CcpA are transcriptional regulators involved in several mostly different metabolic processes. In order to prove if a knock-out of FlpS and/or CcpA has an influence on bacterial fitness and virulence properties of S. suis in vivo, we conducted two different types of mice infection experiments. Firstly, we used an established intranasal infection model to study S. suis colonization of the murine nasopharynx [37]. As expected, none of the mice showed any clinical signs after intranasal infection, since intranasal infection leads to subclinical infection. As a read-out parameter for colonization of the upper respiratory tract, the bacterial load was determined after tracheonasal lavage (TNL). In contrast to the infection with the capsular polysaccharide-deficient strain 10∆cpsEF which was not able to colonize the nasopharyngeal cavity, we observed effective colonization of the murine nasopharynx by wild-type strain 10 (in nine of 10 mice) and strain 10∆flpS (in all mice) ( Figure 3A). No significant differences were observed between bacterial loads in the TNL after infection with the wild-type strain 10, strain 10∆flpS, strain 10∆ccpA∆flpS and 10∆ccpA. 10ΔflpS (in all mice) ( Figure 3A). No significant differences were observed between bacterial loads in the TNL after infection with the wild-type strain 10, strain 10ΔflpS, strain 10ΔccpAΔflpS and 10ΔccpA.  Secondly, we performed intravenous infection of mice. As shown by the Kaplan-Meier diagrams for mortality ( Figure 3B), strain 10∆flpS was as virulent as the wild-type strain 10, whereas strains 10∆ccpA and 10∆cpsEF were strongly attenuated. Furthermore, the significant difference between strain 10∆flpS and strain 10∆ccpA∆flpS indicates that the phenotype of the latter is dominated by the knock-out of ccpA. In addition, we observed no significant differences between the bacterial loads in the inner organs after the infection with wild-type strain 10 and strain 10∆flpS. However, bacterial loads in systemic organs were, in most cases, significantly higher in mice infected with strain 10∆flpS in comparison to strains 10∆ccpA and 10∆cpsEF. Taken together, results from both the intranasal as well as the intravenous infection model indicated that the loss of CcpA but not FlpS affects the virulence properties of S. suis in mice.

FlpS is an Oxygen-Sensing Regulator and Essential for arcABC Operon Expression In Vitro
The activity of FNR and FNR-like regulators is tightly controlled by the redox state in bacteria [14]. In S. suis, the activity of ArcABC is subjected to CCR and strongly activated at the stat growth phase in which free glucose is consumed [38]. We have previously shown that the activity of the arginine deiminase in S. suis is induced under anaerobic conditions [23]. Our microarray experiments revealed FlpS as a regulator of the arcABC operon since its expression was attenuated in the stat growth phase of strain 10∆flpS. Therefore, we selected the arcABC operon as a read out to dissect FlpS activation in more detail. To show the relevance of FlpS for anaerobic expression of the ADS, we performed Western blot analysis with cell lysates of the wild-type strain 10, strain 10∆flpS and a complemented mutant strain (10∆flpScomp) grown anaerobically to stat growth phase. As shown in Figure 4A, ArcB expression was not detectable in strain 10∆flpS whereas the wild-type expression level was restored in strain 10∆flpScomp in which the flpS gene was provided in trans. The data indicated that FlpS is essential for ArcABC expression under anaerobic growth conditions. Following, we performed qRT-PCR analyses of arcABC operon mRNA from S. suis grown aerobically or anaerobically. This data confirmed that FlpS is essential for arcABC operon expression under aerobic and anaerobic condition ( Figure 4B,C). The expression of the arcABC operon of the wild-type strain was induced at standard batch and anaerobic conditions as soon as the repressing effect of CCR ceased ( Figure 4B). Remarkably, higher levels of expression were observed in anaerobically grown cultures when compared to batch cultures, suggesting that more active FlpS is present under anaerobic conditions. To test the influence of oxygen on FlpS activity, we increased the levels of oxygen in the aerobic growth experiments by intensive shaking of the S. suis cultures. The data shown in Figure 4C demonstrate that induction of arcABC expression was dependent on FlpS, even if the culture was shaken. However, compared to the time kinetic of the batch cultures, arcABC operon induction was delayed in shaking cultures, indicating that higher oxygen tension reduces FlpS activity. To further elucidate the regulatory function of FlpS in arcABC operon expression, we performed GFP reporter studies. For this, the wild-type strain 10 and strain 10∆flpS were transformed with a plasmid carrying a gfp gene under transcriptional control of the arcABC promoter. Bacteria were grown in a minimal medium to the stationary growth phase and gfp activity was then determined by fluorescence measurement ( Figure 4D). In contrast to the wild-type construct 10::arcA-gfp which is able to induce gfp expression, almost no gfp activity could be detected in strain 10∆flpS::arcA-gfp, indicating that FlpS is essential for reporter gene activity. Taken together, these data confirm that FlpS is essential for arcABC operon expression and they provide evidence that FlpS activity is regulated by the environmental oxygen concentration.  We have previously shown that the arcABC operon contributes to the survival of S. suis in epithelial cells. For neutralization of endosomal acidification, ADS activity is important because it generates ammonia and ATP which can be used by the bacterial F 1 F o -ATPase to actively extrude protons. Since FlpS is essential for arcABC operon expression, we performed infection experiments with HEp-2 epithelial cells. We performed these experiments with a flpS-deficient strain in a non-encapsulated background (strain 10∆cpsEF∆flpS), as the polysaccharide capsule inhibits S. suis uptake by epithelial cells. As shown in Figure 4E, strain 10∆cpsEF was able to survive and multiply intracellularly, whereas significantly lower survival rates were recorded for strain 10∆cpsEF∆flpS. To analyze whether the reduced survival of the flpS-deficient strain correlated with its inability to prevent acidification, due to defective arcABC operon expression, HEp-2 cells were treated with bafilomycin to inhibit endosomal acidification before infection. Compared with the infection of untreated cells, the pretreatment of the cells with bafilomycin significantly increased the survival rate of strain 10∆cpsEF∆flpS. These data suggest that despite having minor relevance in S. suis mice colonization and infection, FlpS contributes to S. suis resistance against endosomal acidification.

Discussion
In Gram-positive bacteria, catabolite repression can be linked to the activity of the CcpA regulator, which was characterized previously in S. suis by us and others [34,39]. FNR-like proteins (FLP) are found in Gram-positive bacteria and have been shown to be involved in oxygen-regulated gene expression [14,40]. However, knowledge about Flp proteins in Gram-positive bacteria and especially in S. suis is very limited.
By an in silico homology search, we located the gene encoding for an FNR-like protein of S. suis (FlpS) closely upstream of the arcABC operon [23]. In the present study we generated a flpS knock-out mutant strain and analyzed its gene expression in exp and stat growth phases in a complex medium. These analyses showed that flpS expression was essential for arcABC operon expression. Additionally, depletion of FlpS considerably influenced the expression level of multiple other genes in both growth phases. The higher and lower expression of genes suggested a repressing and activating regulatory role for FlpS. However, many other transcription factor genes were influenced (exp, n = 27; stat, n = 24) in the flpS mutant, so their contributions to these expression profiles have to be considered. On the other hand, the strongly increasing number of lower-expressed genes from the exp to stat growth phase might suggest that FlpS is an activator of gene expression and that FlpS itself is activated by the changing environment during bacterial growth. The majority of differentially expressed genes in both growth phases of strain 10∆flpS were associated with carbohydrate transport/metabolism and nucleotide transport/metabolism. In Gram-negative bacteria (e.g., E. coli), FNR and CRP can regulate the expression of overlapping regulons of target genes in response to environmental (oxygen/redox) or metabolic stimuli (carbon source), respectively [14]. Therefore, we assumed that FNR/Flp regulation may overlap with the regulation of the related regulator CcpA. Hence, we compared the FlpS regulon with a previously published CcpA regulon [34]. These data revealed a group of genes whose expression was affected in both mutants, especially in the exp growth phase. However, most of them displayed opposite regulation in strains 10∆flpS and 10∆ccpA. Nevertheless, this comparison enabled us to dissect the metabolic role of FlpS more precisely. The genes whose expression was exclusively affected in the CcpA mutant assigned to carbohydrate transport, carbohydrate conversions and metabolism reflect the sovereign role of CcpA in CCR. In contrast, the genes exclusively affected in the FlpS mutant assigned to nucleotide metabolism and ABC transporters indicate a role of FlpS in the regulation of these processes. An influence on glycolysis by FlpS might have been deduced from the differential expression of gapA and gpmA in the mutant but this could not be confirmed by the isotopologue pattern of alanine, an amino acid which is directly synthesized from the end product of glycolysis. Furthermore, in correlation with the isotopologue data, we did not observe any significant changes in amino acid preferences of strain 10∆flpS in growth experiments in which we subsequently omitted each of the main 20 amino acids in CDM media (data not shown). Overall, these experiments indicate that under our experimental conditions, the impact of FlpS in S. suis metabolism is rather low. This is further underlined by the results of our mice infection experiments. In contrast to the CcpA mutant whose attenuation is presumably driven by alteration virulence-associated factor expression [38], the flpS mutant was as virulent as the wild-type strain.
The FlpS gene is located closely upstream of the arcABC operon and our data indicate that FlpS is essential for arcABC expression. We have previously shown that arcABC operon expression in S. suis depends essentially on the regulator ArgR [33] and is subjected to CCR, however, not as a direct target of CcpA [34]. The CCR of the arcABC operon explains its exclusive expression at the late stage of bacterial growth when glucose is almost consumed. FlpS does not influence argR expression; therefore, the loss of arcABC operon expression in the absence of ArgR [33] or FlpS indicates that both regulators cannot compensate for each other, and that they may synergistically induce the arcABC operon, as also proposed for other bacteria [20,25,26,41]. Similar to ArgR, the regulatory activity of FlpS can be restricted to the promoter region of the arcABC operon; nevertheless, future studies will have to show whether FlpS directly regulates arcABC operon expression by binding to the promoter. FlpS shows homologies to the FNR-like regulators FLP of S. gordonii, ArcR of Bacillus licheniformis, and ArcR of Staphylococcus aureus [23], all of which have been shown to regulate the arginine deiminase genes in the respective species under anaerobic conditions [13,20,26]. Accordingly, we found that strain 10∆flpS was not able to induce arcABC expression when grown under anaerobic conditions. The fact that FlpS is also essential for arcABC expression in S. suis grown under aerobic conditions denotes FlpS as a regulator whose activity is dependent on the redox state of the cell which might be influenced by free oxygen or by an as-yet-unknown mechanism.
Our data provide evidence for an involvement of FlpS in oxygen-and/or redox-related responses of S. suis. Thus, the FNR regulator in E. coli is known to regulate the expression of genes under anaerobic conditions, including fumarate reductase, nitrate and nitrite reductase, and cytochrome d oxidase genes, all of which are enzymes involved in processes of the metabolic transition between aerobic and anaerobic growth [42][43][44]. The FlpS regulon of S. suis comprises genes whose expression was influenced in S. intermedius and S. mutans when grown under anaerobic conditions [45,46]. Similar to S. intermedius, we found upregulation of glycolytic genes, pyruvate-formate lyase, alcohol dehydrogenase, the arginine deaminase pathway and the nucleotide synthesis/salvage pathways [45]. In addition, similar to the response of S. intermedius and S. mutans [45,46], we found altered expression of PTS and ABC transporters for utilization of C5 and C6 sugars (e.g., glucose, trehalose) in strain 10∆flpS. Furthermore, among the oxidative stress-related genes, the gene of the Dps-like peroxide resistance protein (Dpr) was differentially expressed in strain 10∆flpS. Lastly, even the enhanced arcABC expression when S. suis was grown under anaerobic conditions and the delay in arcABC expression when the oxygen tension in the culture medium was increased indicate the oxygen-dependent activity of FlpS.
FlpS of S. suis carries two cysteine residues at the N-terminus which might contribute to iron-sulfur cluster binding. After the exchange of one cysteine residue by a serine at position 129 (C129S), we did not observe any differences in arcABC expression in comparison to the complemented mutant strain carrying the wild-type flpS allele (data not shown). This observation, however, does not exclude oxygen-dependent regulation of FlpS since FlpS of S. suis, similar to the Flp protein of Lactobacillus casei, might be active in the absence of iron-sulfur cluster formation [14]. On the other hand, FlpS activity might require binding of an oxygen-labile iron-sulfur cluster by N-terminal non-cysteinyl residues, which has been described for the FNR-like protein FlpA of Lactococcus lactis [15]. Further studies are needed to functionally dissect the mechanism of redox regulation of S. suis FlpS.
A particular role of FlpS for virulence has not been shown yet. Like other pathogenic streptococci, S. suis is able to infect and survive in epithelial cells. This feature is largely associated with a functional ADS [23,33,47]. Our data show that FlpS contributes to intracellular survival of S. suis. The reduced survival of the flpS mutant in HEp-2 cells might be explainable by its absolute necessity for arcABC expression. In contrast to the cell culture infections, the flpS knock-out seemed not to affect S. suis colonization and survival in two experimental mouse models. Rather, results from intravenously infected mice suggest better survival of strain 10∆flpS than the wild-type strain 10 in the inner organs. However, these data were not statistically significant and in the CcpA/FlpS double mutant, the knock-out of flpS could not compensate the loss of CcpA. Overall, these data suggest that the phenotype of strain 10∆flpS is not advantageous when S. suis has to counteract a complex immune response or has to metabolically adapt to the in vivo environment in mice. The discrepancy between in vitro and in vivo data emphasizes that the role of FlpS in infection might be limited to a particular niche in its primary host, the pig, where the bacterium is exposed to specific redox conditions. This hypothesis, however, awaits further elucidation.

Bacterial Strains and Growth Conditions
The highly virulent S. suis serotype 2 strain 10 (Smith et al., 1999) and respective mutant strains used in this study are listed in Table S1. S. suis was cultured at 37˝C on Columbia blood agar base (Oxoid) containing 6% (v/v) sheep blood or horse blood supplemented with spectinomycin (100 µg/mL), and/or erythromycin (2 µg/mL) if necessary. If not stated otherwise, broth cultures of S. suis were obtained by overnight culturing in Todd-Hewitt Broth (THB, Becton Dickinson Diagnostics). On the next day overnight culture were diluted in prewarmed THB medium to an OD 600 = 0.02 for growth experiments. Growth experiments (supplementation of different carbohydrate substrates; omitting single amino acids) in chemically defined medium (CDM) were performed in a 96-well microplate reader as previously described [36].

Mutagenesis of S. suis
The inactivation of the flpS gene (SSU0579) was done by insertion mutagenesis in the parental S. suis strain 10, the capsular polysaccharide deficient strain 10∆cpsEF::spc R (10∆cpsEF) or the ccpA deficient strain 10∆ccpA::Em R (10∆ccpA). For this, two mutagenesis plasmids each carrying a different antibiotic resistance cassette were constructed. In both cases a 1509-nucleotide PCR fragment comprising the flpS locus was amplified by PCR from chromosomal DNA of S. suis strain 10 with oligonucleotide primers AdR1 and AdR2. The PCR product was subcloned in the E. coli cloning vector pGEM-T Easy (Promega) to create pGEMAdR. Then, plasmid pGEMAdR was linearized by SnaBI and ligated to a EcoRV-SmaI fragment derived from vector pICspc containing the spectinomycin resistance cassette [48], resulting in pGEMFlpSspc. By the insertion of the antibiotic resistance cassette the flpS gene becomes disrupted. In order to construct a erythromycin resistance conferring mutagenesis plasmid, linearized pGEMAdR was ligated to the PvuII fragment derived containing the erythromycin resistance cassette from vector pICerm [49], resulting in pGEMFlpSermR. Electrotransformation of the above listed S. suis strains with pGEMFlpSspc or pGEMFlpSermR (see also Table S1) was performed as previously described [23]. Successful mutation was verified by PCR (data not shown) and Northern Blot analysis [38].

Episomal Complementation of S. suis 10∆flpS
For complementation experiments, flpS was amplified from S. suis DNA with the primer pair flpSagg-SalI and flpSend-PstI and subsequently cloned into the S. suis-E. coli shuttle vector pORI23 [50]. Thereby, the flpS gene was set under control of a constitutive P23 lactococcal promoter and the original ribosome binding-site (RBS) 'ggagt' in this promoter was changed to the optimal RBS of gram-positive bacteria 'ggagg', finally yielding plasmid pORI23-flpS. For the site directed exchange of a cysteine to a serine in the amino acid sequence of FlpS at position 129 we used pORI23-flpS as a template and applied primer pair FlpS_C129S_for/FlpS_C129S_rev according to the manual instructions of the QuikChange ® Site-Directed Mutagenesis Kit (Stratagene). The resulting plasmid pORI23-flpS(C129S) and also pORI23-flpS were then introduced into electro-competent S. suis strain 10∆flpS::spc R . Transformants were confirmed by PCR and DNA sequencing.

Microarray Analysis
Comparison of gene expression between the wild-type and flpS-deficient strain was done by microarray analyses. For this, S. suis strain 10 and strain 10∆flpS were grown in THB medium to the exponential growth and early stationary growth phase, harvested and RNA extracted from bacteria as described [33]. cDNA microarray experiments and subsequent data analysis was performed as previously specified [33]. All microarray data have been submitted MIAME complied with ArrayExpress under Accession No. E-MTAB-3997. Venn diagrams were constructed with VENNY 2.1.

Real-Time Quantitative RT-PCR (qRT-PCR)
For qRT-PCR experiments bacteria were grown in THB media under anaerobic conditions (anaerobic chamber (Don Whitely Scientific; atmosphere: 5% CO 2 , 10% H 2 , 85% N 2 ), standard batch conditions, or shaking conditions (200 rpm). Bacteria were harvested at the exponential growth phase (OD 600 = 0.3; P 0 ) and then hourly after this time point as indicated by serially increasing numbers (e.g., P 1 , one hour after P 0 ). After RNA extraction, reverse transcriptase reactions followed by quantitative PCR (qPCR), and evaluation of qPCR were done as described [34]. Relative target gene transcript levels were first normalized to gyrB transcript levels and then expressed as 2´∆ ∆CT values with regard to time point P 0 of the wild-type strain [51].

Western Blot Analysis
Bacteria were anaerobically grown overnight in THB medium, on the next day diluted 1:15 in THB and further incubated under anaerobic conditions. Eight hours later bacteria were harvested, lysed and an ArcB-specific immunoblot was performed as described [33].

Isotopologue Profiling
Bacteria were grown in THB medium supplemented with 10 mM [U-13 C 6 ]glucose to the exponential and early stationary growth phase, were then harvested and processed as specified recently [36].

Experimental Infection of Mice
The animal experiments of the present study were approved by the Committee on Animal Experiments of the Lower Saxonian State Office for Consumer Protection and Food Safety (Niedersächsisches Landesamt für Verbraucherschutz und Lebensmittelsicherheit, LAVES; approval number 33.12-42502-04-12/0935). Infection experiments were in strict accordance with the German regulations of the Society for Laboratory Animal Science (GV-SOLAS) and the European Health Law of the Federation of Laboratory Animal Science Associations (FELASA). Furthermore, all experiments were planned to fulfill '3R' rules on reduction, refinement and replacement of animal infection experiments.
For all experimental S. suis infections five-to six-week-old female C57BL/6J WT mice (Charles River WIGA, Sulzfeld, Germany) were infected with S. suis strains grown to the exponential growth phase (OD 600 = 0.2). Intranasal infection of mice was exactly done as previously described with the exception that a higher dose was applied [19]. Briefly, after acetic acid predisposition mice were intranasally infected with a dose of approximately 1ˆ10 9 CFU. Infection experiments were done in groups of five mice and repeated once (in total 10 mice per S. suis strain). After infection, mice were constantly monitored for clinical signs according to an established clinical score matrix [37] but sacrificed three days post infection (d.p.i) for comparative histological and bacteriological screenings. The determination of the bacterial load in tracheonasal lavages (TNL) and indicated inner organs was performed as described [19,37]. In case of the intravenous infection model, mice were infected with approximately 5ˆ10 8 CFU in 100 µL of PBS via the tail vein. Infected mice were sacrificed 9 d.p.i if they have not been euthanized earlier due to animal welfare reasons as also judged by the investigated clinical score matrix [37]. Bacteriological and histological screenings of mice tissues were done as described earlier [19,37]. For both infection models differences in bacterial loads in TNL, blood, or organ samples were analyzed by a Kruskal-Wallis test with a post hoc Dunn's multiple comparisons test. Statistical analysis of Kaplan-Meier diagrams was conducted with the log-rank test.
Supplementary Materials: The following are available online at: http://www.mdpi.com/2076-0817/5/3/51/s1. Figure S1: Growth of S. suis strains and verification of the flpS knock-out. (A) Growth of wild-type strain 10, strain 10∆flpS and strain 10∆flpScomp in THB under batch conditions; (B) Northern blot analysis of wild-type strain 10 and strain 10∆flpS along the growth phase. The blot was incubated with a [α-32 P]dCTP labeled flpS probe. The ethidium bromide gel showing prominent 16S and 23S rRNA bands serves as a loading control, Table S1: Bacterial strains, plasmids, and oligonucleotides used in this study, Table S2: Validation of microarray data by RT-qPCR, Table S3: Functional COG classification of differentially expressed genes deduced from comparative microarray analyses, Table S4: Overview of significantly higher-and lower-expressed genes as determined by comparative DNA microarray analysis of the S. suis wild-type strain 10 and 10∆flpS during exponential growth, Table S5: Overview of significantly higher-and lower-expressed genes as determined by comparative DNA microarray analysis of the S. suis wild-type strain 10 and 10∆flpS during stationary growth, Table S6: 13 C-excess (mol %) of amino acids in experiments with S. suis strain 10, 10∆flpS, 10∆ccpA, and 10∆ccpA∆flpS cultivated in THB media supplemented with 10 mM [U-13 C 6 ]glucose.