Introduction

Quorum sensing (QS) is a cell-to-cell communication mechanism that mediates coordinated adaptation of bacteria by regulating the expression of numerous genes. QS has two distinct characteristics. The first relates to its complexity, which is reflected by the wide variety of signaling molecules exhibiting various functions and by the different types of communication between different QS systems. The second is the diversity of QS systems in terms of distribution, signaling molecules, and mechanisms as well as induction and transport of signaling molecules (Grandclement et al. 2016; Hawver et al. 2016; Ma et al. 2017a). QS systems are currently divided into different classes: the LuxR-LuxI systems of Gram-negative bacteria, the auto-inducing peptide (AIP) of Gram-positive bacteria, and the LuxS/AI-2 system found in both Gram-positive and Gram-negative bacteria. Many studies have shown that QS systems are responsible for group behavior, cell luminescence, antibiotic resistance, plasmid transfer, virulence factor gene expression, and biofilm formation (Asif and Acharya 2012; Kalia et al. 2015; Miller et al. 2015; Singh et al. 2016). Streptococcus suis is a zoonotic pathogen that is mainly responsible for septicemia, meningitis, pneumonia, arthritis, and even sudden death in humans and pigs worldwide (Segura et al. 2016). S. suis can be divided into 33 serotypes, of which serotype 2 (S. suis 2) is the most pathogenic and the most harmful to the swine industry, especially in East and Southeast Asian countries such as Thailand, Vietnam, and China (Hatrongjit et al. 2015; Wertheim et al. 2009; Yu et al. 2006). To date, S. suis has spread to more than 30 countries and regions around the world (Feng et al. 2014). According to reports, there were more than 1600 cases of human-infected S. suis in the world, and the vast majority of the patients were from Southeast Asia (Chatzopoulou et al. 2015). Consequently, S. suis is a constant threat to the health of humans and pigs. The LuxS/AI-2 QS system is believed to be involved in the virulence and drug resistance of S. suis.

QS mechanism

Bacteria possessing a QS system are capable of producing and releasing signaling molecules known as autoinducers (AIs), which increase from concentration as a function of cell density (Rémy et al. 2016; Tan et al. 2017). Bacteria can monitor variations in the concentrations of autoinducers to track changes in cell numbers and to collectively alter the global pattern of gene expression. QS signaling molecules include acyl-homoserine lactones (AHLs), AIP, autoinducer-2 (AI-2), diffusible signal factors (DSFs), fatty acids, and partial ester compounds. Many Gram-negative bacteria rely on the secretion of AHLs to communicate and coordinate group behavior, such as the production of extracellular enzymes and toxins (Williams 2007), biofilm formation (Alagely et al. 2011; Kim et al. 2014; Parsek and Greenberg 2005; Shih and Huang 2002), antibiotic resistance (Bainton et al. 1992), and bacterial motility (Eberl et al. 1996). The structure of AIP in Gram-positive bacteria differs according to the species (Malone et al. 2007; Novick and Muir 1999). It cannot be transported on its own across the cell wall and generally requires an ATP-binding-cassette (ABC) transport system or other membrane channel proteins for translocation outside the cell. The binary signal system of bacteria can regulate the transcriptional expression of target genes through a complex signal transduction process (Zollmann et al. 2015). The LuxS/AI-2-mediated QS system mediates interspecies and intraspecies information exchanges between Gram-positive and Gram-negative bacteria (He et al. 2015; Thompson et al. 2015).

LuxS

The luxS gene is highly conserved in bacteria. For instance, a comparative sequence analysis of luxS in Streptococcus mutans and Streptococcus pyogenes showed that the two genes share 84% identity and 92% similarity (Merritt et al. 2003). High identity and similarity of the above luxS genes were also found with those from Streptococcus pneumoniae, Lactococcus lactis, Clostridium perfringens, Neisseria meningitidis, Escherichia coli, and Haemophilus influenza (Merritt et al. 2003).

The LuxS protein is a homodimeric metallo-enzyme that contains two identical tetrahedral metal-binding sites. The core of each metal binding site contains a divalent zinc ion with two histidines, a cysteine and a water molecule, which structure is similar to active sites identified in some amidases and peptidases (Hilgers and Ludwig 2001). The LuxS system is involved in the synthesis of autoinducer AI-2, which is a furanosyl borate diester (Sperandio et al. 2003).

Structure and activity of LuxS in S. suis

Analysis of the three-dimensional crystal structure of S. suis LuxS revealed the presence of four LuxS monomers in each asymmetric unit (PDB ID 4XCH) (Fig. 1). The LuxS monomers consist of four antiparallel β-sheets and four antiparallel alpha helices in the order H1-S1-S2-H2-S3-S4-H3-H4 (Wang et al. 2015). Moreover, ion coupled plasma mass spectrometry has shown that Zn2+ is the main component of the active center of S. suis LuxS. This is consistent with the study by Ruzheinikov et al. (2001), who investigated the Bacillus subtilis LuxS. In a previous study, Zhu et al. (2003) determined the activity of LuxS by measuring the release of homocysteine using Ellman’s reagent. They found that the ligand environment of LuxS metalloproteinase is similar to the envelopellase of polypeptides; LuxS in the form of Fe2+ rapidly deactivates under aerobic conditions. However, the LuxS protein composed of Zn2+ or Co2+ is very stable, and the catalytic ability of Co-LuxS was the same as that of Fe-LuxS and 10 times stronger than that of Zn-LuxS. Rajan et al. (2005) also reported the presence of Fe2+ in the LuxS of Bacillus subtilis. These results suggest that different bacterial LuxS proteins may have different metal ions and that this may impact their catalytic efficiency. A bioinformatic analysis has revealed that there are possible evolutionary mutations at amino acid positions 80 and 87, which are located near the substrate binding site (Wang et al. 2015). These two amino acid mutations have a marked effect on the binding of the substrate, the catalytic activity of the enzyme, the production of AI-2 by Streptococcus spp., and the ability to form biofilms. In vitro and in vivo tests have shown that the absence of mutations in these two amino acids can inhibit the production of AI-2 and the formation of biofilm (Wang et al. 2011).

Fig. 1
figure 1

LuxS protein crystal structure of S. suis2 HA9801. Four different colors represent different LuxS monomer proteins; Zn2+ ions are represented by gray spheres

Full size image

Metabolism

LuxS is not only involved in the production of the AI-2 signaling molecule but also plays an important role in central bacterial metabolism and is part of the activated methyl cycle (Trappetti et al. 2017). LuxS is mainly responsible for the hydrolysis of S-adenosine homocysteine to S-adenosylmethionine (SAMe). SAMe is a ubiquitous biomolecule that serves primarily as a methyl donor. It is the main route by which bacteria recycle methyl groups and is key to polyamine formation and vitamin synthesis by bacteria (Bonhoure et al. 2015). A mutation or deletion of luxS results in the loss of SAMe function and the inhibition of AI-2 synthesis, indicating that QS activity will likely be affected if the luxS mutation causes a change in the relevant phenotype. The induction of luxS can also lead to changes in the extracellular concentrations of other metabolites. A recombinant LuxS assay showed that the extracellular concentration of S-ribosyl homocysteine (SAMe with a LuxS function) is significantly higher in the culture supernatants of LuxS-deficient strains than in the wild-type strain (Challan Belval et al. 2006). Based on the above results, it is possible that numerous SAMe pathway intermediates inside and outside LuxS mutants are modified.

LuxS/AI-2 system

LuxS plays a fundamental role in LuxS/AI-2-mediated QS system, while AI-2 is a by-product of bacterial methyl metabolism and balances the metabolism in activated methyl cycles (AMC) (Yang and Defoirdt 2015). In this process, the methyl group is removed from SAMe and is converted into thioglucoside homocysteine (SAH), which represents a toxic metabolite. SAH is subsequently converted by a 5′-methylthioadenosine/S-adenosyl homocysteine nucleosidase (Pfs) into adenine and thioglycoside-type homocysteine (SRH). SRH is then converted into 4,5-dihydroxy-2,3-pentanedione (DPD) and homocysteine acid (HCY) by LuxS (Malladi et al. 2011; Tavender et al. 2008). Bassler et al. studied luminescence in Vibrio harveyi and found that S-glycosylated homocystease is encoded by luxS (Bassler et al. 2010). LuxS catalyzes the cleavage of the thioether linkage of SRH to produce HCY and DPD. DPD is formed by self-cyclization, yielding AI-2 (Fig. 2) (Han and Lu 2009c; Ma et al. 2015; Miller et al. 2004; Winzer et al. 2003). As the bacterial density increases, AI-2-mediated QS is activated, LuxO phosphorylation ceases, and LuxO and LuxR act together to transcribe the luxCDABE operon, causing the bacteria to emit light. There are currently more than 55 known luxS homologs, but only a few of the molecular structures of AI-2 have been clearly determined. Han et al. and Cao et al. analyzed the S. suis genome and showed that luxS can produce AI-2 signaling molecules (Cao et al. 2011; Han and Lu 2009a).

Fig. 2
figure 2

The synthetic pathway of AI-2 molecules in bacteria. S-adenosylmethionine (SAM), the major methyl donor in metabolic processes, is responsible for the production of AI-2. SAM transfers methyl groups to methyl acceptors with the aid of methyltransferases to produce S-adenosylhomocysteine (SAH). SAH is hydrolyzed to S-ribose homocysteine (SRH) with the involvement of the nucleoside Pfs with a concomitant release of adenine. The thioether bond of SRH is cleaved under the action of the LuxS protein and produces homocysteine and 4,5-dihydroxy-2,3-pentanedione (DPD). It is reported that AI-2 is the result of DPD self-cycling reaction, involving least two reactions. SAM can also form the main triamine spermidine by decarboxylation with diamine and putrescine. The reaction is also accompanied by the release of the toxic by-product 5′-thiomethyladenosine (MTA), which is then modified by Pfs, producing adenine and 5′-thiomethylribose (MTR). The exact export mechanism of AI-2 is unknown

Full size image

Gene expression levels of LuxS/AI-2 system in S. suis

The absence of luxS leads to changes in the gene expression and phenotype of bacteria, and the LuxS protein directly regulates the gene expression of bacteria and the functions of various proteins, including enzymes involved in biofilm formation, bacteriocin synthesis, cell competence, and acid resistance. Hasona et al. (2007) used gene chips to analyze the decreased expression of LuxS protein in Streptococcus mutans, and found that the expression of 81 genes related to cell competence, including chaperones, proteases, cell envelope synthetases, DNA repair, and replication enzymes, was upregulated. On the other hand, the downregulation of 35 genes involved in ribosomal proteins, and biosynthetic enzymes of amino acids and proteins was also observed. Sztajer et al. (2008), through a global transcriptome analysis of a luxS null mutant of S. mutans, found that the expressions of 585 genes were still affected after adding chemically pure DPD to the medium. It suggested that the LuxS enzyme plays an important role in the methyl transfer metabolism.

The Pfs enzyme converts S-adenosylhomocysteine into S-ribosylhomocysteine and adenine and plays an important role in bacterial metabolism and synthesis of AI-2 (Cao et al. 2011; Han and Lu 2009a). AI-2 activity and the transcription of pfs reach their highest levels in the late logarithmic growth phase of S. suis 2 while the transcription of luxS is higher during the stationary phase. On the other hand, the transcription of pfs and the production of AI-2 remain the same while the transcription of luxS and the production of AI-2 do not (Han and Lu 2009b). The differences in transcription between the S. suis wild-type and the luxS deletion mutant (ΔluxS) strains were determined using an Agilent microarray. The results showed that 312 genes are differentially expressed in wild-type cells, of which 144 were upregulated and 168 were downregulated. By introducing DPD into the ΔluxS strain, 79 genes were differentially expressed. Of these, 29 were related to the production of S. suis virulence factors and the uptake of iron were regulated by the LuxS/AI-2-mediated QS system (Cao et al. 2011). We also constructed a S. suis luxS+ overexpressing strain and showed by real-time PCR that the expression of luxS increases throughout all the growth phases, while the level of pfs expression remains unchanged. Overexpressing luxS did not increase the level of pfs expression or AI-2 production. The overexpressing strain formed more biofilms, which increased with the incubation time. However, the bacteria grew more than the wild-type strain, indicating that the production of AI-2 is not correlated with the transcription of luxS. Although the expression of luxS is constitutive, the transcription of pfs is possibly correlated with AI-2 production in S. suis (Wang et al. 2013). These results suggest that the production of AI-2 is controlled by the interaction between luxS and pfs and increases only when LuxS does not increase the production of AI-2.

Characteristics of LuxS/AI-2 system in S. suis

The LuxS/AI-2 system impacts on various biological properties of bacteria. It is considered as one of the most important global regulatory networks in bacteria; the regulation of the expression of the corresponding genes is modulated as the bacterial concentrations change. In general, regulating gene expression often causes a variety of phenotypic changes. Many physiological functions of bacteria have been found to be regulated by the LuxS/AI-2 system, including bacterial luminescence, antibiotic susceptibility, plasmid transfer, virulence, gene expression, and biofilm formation (Ma et al. 2015; Wang et al. 2017). In S. suis, current research is mainly focused on the impacts of the LuxS/AI-2-mediated QS system on the regulation of the biological characteristics described below.

Bacterial growth

By observing the growth curves and the morphological changes of bacterial mutants and wild-type cells, physiological changes displayed by the mutants can be more intuitively understood. The growth rate of a ΔluxS S. suis mutant was found to be lower than that of the wild-type strain. The logarithmic growth phase of the ΔluxS S. suis mutant lagged behind that of the wild-type strain, and the growth rate of the mutant was lower than that of the wild-type strain. The wild-type strain reached the quiescent period 90 min ahead of the ΔluxS strain. Light microscopy showed that cells of the ΔluxS strain aggregate and that the length of the bacterial chains was shorter than that of the wild-type strain. Electron microscopy revealed that the thickness of the capsule of the ΔluxS mutant was thinner than that of the wild-type strain (Cao et al. 2011).

Cell adhesion

The first steps in the infection of a host cell by a pathogen are usually adhesion and then colonization. By using site-specific integration, isogenic mutations were generated in luxS, and the resulting Lactobacillus acidophilus mutants resulted in a 58% reduction in adhesion to Caco-2 cells (Buck et al. 2009). The same phenomena also appear in Actinobacillus pleuropneumoniae (Li et al. 2011) and Lactobacillus plantarum (Jia et al. 2018).

Comparative analysis of adhesion of the S. suis wild-type strain and the ΔluxS mutant may provide insight regarding regulation of adherence by the LuxS/AI-2 system. The ability of the luxS mutant to adhere to human epithelial laryngeal carcinoma cells and human umbilical vein endothelial cells was found to be lower than that of the wild-type strain. Moreover, low concentrations of exogenous AI-2 added to the S. suis wild-type strain and luxS mutant increased adhesion to the host cells, while high concentrations of AI-2 decreased adhesion, with 4 and 6 μM AI-2 being the most efficient at improving adhesion (Yang et al. 2014).

Biofilm formation

Biofilms are produced by bacteria and embed the entire bacterial population. Biofilms have a significant effect on bacterial virulence and drug resistance. In recent years, studies have found that QS systems and two-component regulatory systems play an important regulatory role in the formation and development of bacterial biofilms (Christiaen et al. 2014; O'Loughlin et al. 2013). Among them, the LuxS/AI-2-mediated QS system has attracted more and more attention due to its widespread existence in Gram-positive and Gram-negative bacteria.

The ability of S. suis to form biofilm was significantly increased when low amount of AI-2 was added to the growth medium, while the addition of high concentrations of AI-2 suppressed the ability to form biofilm. The addition of 2 μM AI-2 significantly increased biofilm formation at 24 h but had no effect at 48 h (Yang et al. 2014). These results showed that the ability of S. suis to form biofilm is enhanced by the overexpression of AI-2 and the incubation time (Wang et al. 2013).

Hemolytic activity

Hemolysis can cause anemia, sepsis, and other symptoms. As such, bacterial hemolytic activity is an important indicator of pathogenicity. Pecharki et al. (2008) found that the hemolytic activity of a Streptococcus intermedius luxS mutant was five times lower than that of the wild-type strain. Supplementing the culture medium with the AI-2 precursor 4,5-dihydroxy-2,3-pentanedione restored the hemolytic activity of the mutant. In our study, it was found that the maximum dilution ratio of culture supernatant required to lyse 50% of erythrocytes is 1:16, 1:2, and 1:16 for the wild-type, ΔluxS, and luxS complement strains of S. suis 2 HA9801, respectively. This suggests that luxS can modulate the hemolytic ability of bacteria (Wang et al. 2011).

Bacterial virulence

Virulence factor gene expression by bacterial pathogens is required to cause infections. Virulence gene expression must be precisely regulated if pathogens are to infect a host and cause a disease. LuxS has been reported to play a critical role in regulating both bacterial virulence and interspecies QS in a broad variety of bacteria (Jones et al. 2010; Li et al. 2011; Ma et al. 2017b). The expression of the virulence genes gdh, cps, mrp, gapdh, sly, fbps, and ef was found to be 0.66-, 0.61-, 0.45-, 0.48-, 0.29-, 0.57-, and 0.38-fold lower, respectively, in the ΔluxS mutant strain than in the wild-type strain of S. suis. Zebrafish infection experiments have shown that the virulence of ΔluxS strain is 10 times lower than that of the wild-type strain and that complemented strains showed a partially restored level of virulence (Wang et al. 2011). Experiments in a pig model have also shown that the number of ΔluxS bacteria infecting the pig lung, brain, and joints is significantly lower than that of the wild-type strain (Cao et al. 2011).

Stress responses

Bacterial stress responses, referring to changes in the physiology and biochemistry of bacteria, are caused by variations in their normal living environment (Pham et al. 2017); it may also cause morphologic changes of the bacteria (Oh et al. 2015). The stress response is a spontaneous adaptation of the bacteria to the external environment. A previous study showed that inactivation of the luxS gene results in a wide range of phenotypic changes, including an increase in tolerance to H2O2 (Cao et al. 2011). Similar results were also observed in our laboratory (unpublished data). Under oxidative stress conditions, the wild-type strain of S. suis treated with H2O2 exhibited sensitivity to H2O2, whereas the ΔluxS was more resistant to H2O2. On the other hand, the acid resistance assay showed that the inactivation of the luxS gene results in a rapid decrease in the number of S. suis cells. Similarly, in the presence of an iron chelator, the growth of ΔluxS was significantly decreased (unpublished data).

Future prospects

Given the ongoing increase in the antibiotic resistance of bacterial pathogens and the important role played by the LuxS/AI-2 system in intercellular communication, metabolism, and virulence, a novel antibacterial strategy based on interference with the LuxS/AI-2-mediated QS system can be proposed. A better knowledge of the mechanisms of bacterial regulation and the regulation of AI-2 uptake is essential in order to be able to control bacteria through inhibition of pathogen signaling. More specifically, in-depth studies are required to identify the genes involved in the uptake of AI-2 by S. suis, the genes associated with AI-2, the target proteins of S. suis that are regulated by AI-2, the downstream target proteins regulated following AI-2 binding to the receptor protein, and the regulation of the QS signaling pathway. These issues need to be urgently addressed. Gene chip technology is a widely used DNA analytical approach for detecting gene transcription levels through nucleic acid hybridization. Many studies have used this approach to identify a large number of genes affected by LuxS and AI-2 in S. suis. Investigations by our group to identify the S. suis AI-2 receptor are ongoing. It is planned to construct a random S. suis mutant library using the transposon Tn917 and find the gene-mediated downstream regulatory network of AI-2 and its receptor through gene chip analysis and proteomics. With the deepening research on QS systems, researchers have developed a variety of quorum sensing inhibitors (QSIs), including both natural and synthetic agents. These QSIs can be divided into three categories: (1) non-peptide small-molecule substances such as garlic extract, thiolactones, and AI-2 analogues; (2) peptides such as leader peptide for oligopeptide signal molecules (AgrD); and (3) proteins such as QS degrading enzymes. These inhibitors can block QS by inhibiting signal production, blocking signal receptors or disrupting QS signals, and also providing an alternative way of controlling microbial pathogenesis (Pan and Ren 2009). In addition, the advantage of using QSI is not only to prevent bacterial growth but also to avoid the production of bacterial resistant strains (Saurav et al. 2017). Consequently, QSIs inhibitors appear to have good prospects for the prevention and control of pathogens.