Gossip in the gut: Quorum sensing, a new player in the host-microbiota interactions.

Bacteria are known to communicate with each other and regulate their activities in social networks by secreting and sensing signaling molecules called autoinducers, a process known as quorum sensing (QS). This is a growing area of research in which we are expanding our understanding of how bacteria collectively modify their behavior but are also involved in the crosstalk between the host and gut microbiome. This is particularly relevant in the case of pathologies associated with dysbiosis or disorders of the intestinal ecosystem. This review will examine the different QS systems and the evidence for their presence in the intestinal ecosystem. We will also provide clues on the role of QS molecules that may exert, directly or indirectly through their bacterial gossip, an influence on intestinal epithelial barrier function, intestinal inflammation, and intestinal carcinogenesis. This review aims to provide evidence on the role of QS molecules in gut physiology and the potential shared by this new player. Better understanding the impact of intestinal bacterial social networks and ultimately developing new therapeutic strategies to control intestinal disorders remains a challenge that needs to be addressed in the future. ©The Author(s) 2021. Published by Baishideng Publishing Group Inc. All rights reserved.

Core Tip: Host-microbiota interactions play a crucial role in the pathophysiology of many intestinal diseases. While biological components have been repeatedly described, a largely overlooked component is quorum sensing (QS), a density-dependent system able to coordinate bacterial responses and interact with host cells constantly exposed to bacteria. This review intends to describe the different QS systems to show evidence that QS is part of the intestinal ecosystem and highlight its impact on intestinal epithelial barrier function, inflammation, and intestinal carcinogenesis. From this report, we open up a new area of intestinal physiology.

INTRODUCTION

Gut microbiota mutually interacts with coevolved host epithelial and immune cells in a beneficial reciprocal relationship[1]. The advent of multi-omics sequencing in the past decade has allowed researchers to investigate the complexity of the intestinal microbiota in various human disorders[2]. Many lines of evidence support a role for alteration of gut microbiota (dysbiosis) in the development or perpetuation of inflammatory and metabolic disorders; recent data pointed out the consequences of dysbiosis on host-microbiota interactions in this setting[3,4]. Currently, gut microbiota metabolites recognized as the main drivers of the impact of gut microbiota on hosts are short-chain fatty acids (SFCAs), branched-chain amino acids, trimethylamine N-oxide, bile acids, tryptophan (Trp), and indole derivatives[3,5]. A largely overlooked component is diffusible signaling molecules, which modulate the physiological response in the three domains of life[6]. A particular class of these signaling compounds is represented by bacterial quorum sensing (QS) molecules called autoinducers (AIs). QS is a density-dependent mechanism allowing bacterial populations to coordinate gene expression and physiology by modulating, for example, metabolic pathways, secretion of virulence factors, or biofilm formation in response to AIs[7]. Drawing on its density-dependent nature, it can be hypothesized that the production of bacterial signaling molecules is abundant in the highly densely populated environment of the mammalian intestinal tract.

Moreover, since several eukaryotic systems from fungi to plants and animals are known to recognize and respond to bacterial signaling compounds, it seems likely that human intestinal cells constantly exposed to bacterial compounds might also have developed response mechanisms to AIs with consequences on intestinal physiology[8]. The purpose of this current review is to provide clues to consider bacterial QS as a new actor of host-microbiota interactions. We will start by presenting the bacterial QS systems and evidence of QS in the gut. We will then provide an overview of the impact of QS molecules on host cell functions within the gut. Finally, we will investigate how modulation of the QS could be thought of as a therapeutic option, determine the key challenges, and suggest directions for future QS research.

QS: GOSSIP IN A BACTERIAL WORLD

In the conventional view of prokaryotic existence, bacteria live as unicellular organisms, with responses to external stimuli limited to detecting chemical and physical signals of environmental origin. This view of bacteriology is now recognized as overly simplistic because bacteria communicate through small ‘hormone-like’ organic compounds. QS is a bacterial cell-cell communication process that involves the production, detection, and response to extracellular signaling molecules called AIs. AIs enable bacteria to perceive and respond to temporal and contiguous environments and coordinate the behavior of colonies by altering gene expression. QS controls genes that direct beneficial activities when performed by groups of bacteria acting in synchrony. Processes controlled by QS include bioluminescence, sporulation, competence, antibiotic production, biofilm formation, and virulence factor secretion (Figure 1A). Similar to languages between humans, these signals vary between species. Some bacterial species can interpret many different signals, while others respond to a few. The first such system was described in 1979 in Vibrio fischeri[9], a symbiotic species that provides marine eukaryotic hosts with light. Light emission depends on transcription of the luciferase operon, which occurs when the cell population density is sufficient to produce a threshold accumulation of a secreted AI, a specific N-acyl Homoserine Lactone (AHL). It was only in 1994 that the term QS was first used to introduce the idea of a minimal population (quorum) that was needed to trigger a group behavior thanks to a signal[10]. The science of studying group behavior in microorganisms has even been named "sociomicrobiology"[11].

Figure 1

Main known mechanisms of quorum sensing activation in bacteria. A: Quorum sensing (QS) signaling depends on the release of autoinducers (AIs) in the environment. Above a threshold concentration depending on bacterial density, QS is activated and triggers gene expression. QS can be classified into three categories: self-talk (i.e., one species “talking” to itself), crosstalk (i.e., different species communicating using common AI), and eavesdropping, which refers to “listening” by species unable to produce AI by itself; B: The acyl-homoserine lactone (AHL) used by Gram-negative bacteria is produced by the synthase complex, and AHL can freely diffuse through the membrane. AHL is recognized by its intracellular receptor, and the complex binds to target gene regulatory elements; C: The AI-2 system is used by both Gram-negative and Gram-positive bacteria. AI-2 needs a transporter protein to exit and enter the cell. For both AHLs and AI-2, there is a positive feedback loop, allowing the expression of the synthase complex and receptor of AIs. AHL: Acyl-homoserine lactone; AI: Autoinducer.

Different ways of talking

Bacterial QS is highly complex and mediates communications thanks to the diversity of its different systems. QS systems can be divided into systems specific to species and mediating communication between Gram-positive bacteria, Gram-negative bacteria, and interspecies systems (Table 1 and Figure 1).

Table 1

Examples of bacterial quorum sensing autoinducer and corresponding systems

	AI	Example of producing bacteria	QS system	Bacterial QS-regulated processes	Ref.
Gram +	AI peptide	Staphylococcus aureus	agr	Virulence	Novick et al[155]
		Listeria monocytogenes	agr	Virulence	Autret et al[156]
		Clostridium perfringens	agr	Virulence	Ohtani et al[157]
		Enterococcus faecalis	FsR	Virulence	Sifri et al[158]
		Bacillus subtilis	com	Competence	Magnuson et al[159]
	γ-butyrolactone	Streptomyces genus	scb	Antibiotics	Takano et al[13]
			scg	Metabolism	Du et al[14]
Gram -	AI-1 (acyl-homoserine lactones)	Vibrio fischeri	LuxI/LuxR	Luminescence	Engebrecht et al[16]
		Vibrio harveyi	LuxLM/LuxN	Luminescence	Mok et al[160]
				Virulence	Waters and Bassler[161]
		Pseudomonas aeruginosa	LasI/LasR	Virulence and biofilm	Gambello and Iglewski[162], Gambello et al[163], Winson et al[164], and Chapon-Hervé et al[165]
			RhlI/RhlR
	PQS	Pseudomonas aeruginosa	PqsABCD/PqsR	QS regulation	Pesci et al[166]
				Pyocyanin	Gallagher et al[167]
				Iron homeostasis	Bredenbruch et al[168] and Diggle et al[169]
				Virulence	Gallagher et al[167] and Cao et al[170]
				Biofilm	Diggle et al[171]
	IQS	Pseudomonas aeruginosa	AmbBCDE/IqsR	Response to stress	Lee et al[172]
	CAI	Vibrio (cholerae)	CqsA/CqsS	Virulence	Ng et al[173]
	AI-3	EHEC O157:H7	Qse/QseBC	Attachment-effacement	Sperandio et al[19], Walters et al[21], and Kim et al[22]
		EPEC O26:H11	Qse/unknown	Unknown	Kim et al[22], and Kaper and Sperandio[40]
		AIEC LF82	Qse/unknown	Unknown	Kim et al[22]
		Escherichia coli MG1655	Unknown	Unknown	Kim et al[22]
		Escherichia coli BW25113	Unknown	Unknown	Kim et al[22]
		Salmonella enterica	Qse/unknown	Unknown	Kim et al[22], Kaper and Sperandio[40], and Walters and Sperandio[174]
		Shigella flexneri	Qse/unknown	Unknown	Kim et al[22], Kaper and Sperandio[40], and Walters and Sperandio[174]
		Yersinia sp.	Qse/unknown	Unknown	Kim et al[22], Kaper and Sperandio[40], and Walters and Sperandio[174]
Gram + and -	AI-2	Vibrio harveyi	LuxS/LuxPQ	Bioluminescence, TSS, protease	Surette et al[24], Mok et al[160], and Schauder et al[175]
		Vibrio cholerae	LuxS/LuxPQ	Virulence and Biofilm	Schauder et al[175], Zhu et al[176], and Hammer and Bassler[177]
		Enterococcus faecalis	LuxS/LuxPQ	Unknown	Surette et al[24], and Schauder et al[175]
		EHEC	LuxS/LsrB (?)	Attachment-effacement	Schauder et al[175], and Bansal et al[178]
		Salmonella enterica	LuxS/LsrB	Pathogenicity and invasion	Miller et al[26], Schauder et al[175], and Choi et al[179]

AI: Autoinducer; AIEC: Adherent-invasive Escherichia coli; AIP: AutoInducer peptides; CAI: Cholera autoinducer-1; EHEC: Enterohemorrhagic Escherichia coli; EPEC: Enteropathogenic Escherichia coli; IQS: Integrated quorum sensing; PQS: Pseudomonas quinolone signal; QS: Quorum sensing.

QS in Gram-positive species is driven in most cases by 5-17 amino acid oligopeptides (AIPs for AutoInducer Peptides), which are detected by membrane receptors belonging to the histidine kinase family and are involved in virulence or competence[12]. In addition, γ-butyrolactones are produced and integrated by Streptomyces sp. as signals controlling antibiotic production[13] or metabolism[14].

QS in Gram-negative bacteria relies on a high diversity of different systems, with some bacteria, such as Pseudomonas aeruginosa (P. aeruginosa), possessing several QS systems (Table 1). The expression of over 300 genes is regulated by QS. The most common system is driven by AI-1 molecules belonging to the AHL family, which are constituted by a homoserine lactone ring carrying a 4-18 carbon acyl chain. The lengths and modifications of the acyl chain give each AHL its species specificity[15]. The first described model is the Vibrio fischeri system, in which N-3-oxohexanoyl-homoserine lactone (3-oxo-C6) is synthesized by LuxI synthase, passively diffuses out of the cell and enters another bacterium in which it binds its receptor LuxR[16] (Figure 1B). Above a threshold, the AHL-receptor complex binds a consensus DNA sequence, thus triggering luciferase expression[17]. This model applies to all AHL systems (Table 1 and Figure 1B). The system involves a positive feedback loop, thus promoting QS activation at the population scale (Figure 1B). To date, numerous homologous systems (i.e., genes coding synthases and receptors) have been described in many Gram-negative bacteria, including over 70 Proteobacteria species[18] (Table 1).

Other Gram-negative QS systems involve the AI-3 molecule, initially identified in enterohemorrhagic Escherichia coli (E. coli) (EHEC) serotype O157:H7[19]. AI-3 regulates flagellar genes and pathogenicity[20,21] and is thought to be present in other enteropathogens (Table 1). A recent study[22] uncovered the structure of AI-3 and its natural analogs, including the prominent analog in mouse feces in vivo, which belongs to the pyrazinone family. The authors showed that various gram-negative and Gram-positive bacteria produce AI-3 analogs, thus redefining the specificity of AI-3 molecules.

Last, the third type of QS system has been identified in Gram-negative bacteria such as EHEC or Vibrio species and Gram-positive bacteria such as Salmonella enterica[23,24]. It relies on AI-2 molecules such as S-THMF-borate [for (2S,4S)-2-methyl-2,3,3’,4-tetrahydroxy-tetrahydrofurane-borate][25] and R-THMF [for (2R,4S)-2-methyl-2,3,3’,4-tetrahydroxy-tetrahydro furane][26]. AI-2 has now been found in various bacterial species in which it regulates many processes[27] and is proposed to mediate poly-species communication (Figure 1C).

In addition, indole is produced from Trp by Gram-negative and Gram-positive commensal and pathogenic bacteria displaying tryptophanase activity[28-31]. As the source of Trp is supplied by the diet and cannot be synthesized endogenously, either by bacteria or by the host, indole is a bacterial byproduct of Trp metabolism. However, in recent years, some authors have considered indole to be a QS molecule, as it is produced in a density-dependent manner and regulates several bacterial physiological processes, such as the formation of spores or biofilms, virulence traits, bacterial motility, and drug resistance[29,32,33].

The versatility of QS systems and their AI molecules highlights the complexity of communication and thus emphasizes the key role QS could play in a diverse ecosystem: the intestinal microbiota.

QS in the gut

The study of QS in the gut is still a relatively recent matter of interest, as QS is generally addressed from the pathogenic bacterium P. aeruginosa point of view in the lung ecosystem. However, many arguments suggest that QS is a new player in the gut ecosystem.

AHLs in the gut: As part of the eavesdropping mechanism, some bacteria from the human gut can sense AHLs from other species (Figure 1A). Gram-negative bacilli such as E. coli, Enterobacter, or Klebsiella express the receptor SdiA, which can sense AHL, without producing such a signal[34]. The opportunistic pathogen P. aeruginosa, which targets the digestive tract in severely immune-compromised patients[35,36], and the more common enteropathogen Yersinia enterolitica are known to produce AHLs[37]. Analyses from sequencing databases have shown the presence of LuxI/LuxR homologs in a few commensals: Hafnia alvei, Edwardsiella tarda, and Ralstonia sp. strain 5_7_47FAA[37]. However, the latter article did not demonstrate the presence of AHLs but only homologs of the genes encoding the synthase complex and the receptor. A cohort low sample size pediatric study (n = 4) demonstrated, thanks to bacterial reporter systems, the presence of AHLs in the feces of patients without identifying them[38]. Our team investigated the question of AHLs in the human gut in the context of inflammatory bowel disease (IBD). With high-resolution mass spectrometry, we identified approximately ten AHLs in the feces of healthy patients, IBD patients in remission, and flare[39]. We also found a never-described AHL, 3-oxo-C12:2-HSL, that was less represented in IBD patients, especially in flares, than in healthy subjects[39].

AI-2 in the gut: AI-2 presence in the gut has been reported by several articles[19,40,41] but is mainly linked to pathogenic bacteria such as enterohemorrhagic E. coli[19]. As AI-2 is considered a “universal language”, it is not surprising to find this AI in an ecosystem as diverse as the gut microbiota.

Thompson et al[42] showed that most Firmicutes contain LuxS protein orthologs, an important enzyme that allows AI-2 production. In contrast, its presence is less represented in Bacteroidetes. Mutant E. coli engineered to regulate AI-2 levels in the mouse gut counteract antibiotic-induced dysbiosis[42]. AI-2 produced by E. coli benefits Firmicutes while restraining Bacteroidetes representation[42], suggesting an important role of AI-2 in the gut ecosystem.

Other QS signals: Concerning other QS signals, there is less evidence of their implications in gut microbiota. A recent study showed a correlation between indole and Clostridioides difficile (C. difficile) infection (CDI) with higher indole concentrations for patients with CDI than for CDI-negative patients with diarrhea[43]. C. difficile induces indole production through overexpression of the tryptophanase gene tnaA in enterotoxigenic E. coli and other indole-producing anaerobes. This increased indole level has been shown to be detrimental to some of the beneficial bacteria and favors C. difficile colonization.

These mechanisms collectively suggest a complex bacteria-bacteria QS network in the gut ecosystem. The key issue is now to decode every language to fully understand its potential in host-microbiota interactions.

Gossip in the gut: Direct dialog with the host and indirect effects through bacterial behavior

In an ecosystem as networked and complex as that in the gut, bacterial communication has to be seen from a large perspective, with multiple bacterial populations crosstalking to each other through eavesdropping or crosstalking between species (Figure 1A). Therefore, the question of how QS affects the host can be addressed in two ways (Figure 2).

Figure 2

Interkingdom dialog between bacteria and the host through quorum sensing molecules. When reaching a threshold concentration within a bacterial community, quorum sensing (QS) autoinducers synchronize group behaviors such as virulence and attachment-effacement strategies as in enterohemorrhagic Escherichia coli, thus indirectly affecting the host (dotted line arrow, middle). QS molecules can impact the host through direct contacts (full arrow, left) with host cells, such as epithelial or immune cells, as has been extensively shown for the Pseudomonas aeruginosa QS molecule 3-oxo-C12-HSL, which freely enters mammalian cells. In addition, QS molecules can indirectly modify the host (dotted line arrow, right) through effects on other bacterial populations with different metabolic properties. QS: Quorum sensing.

QS modulates microorganism metabolism, which in turn can affect the host’s physiology; one could consider this to be an indirect effect of QS molecules on the host (Figure 2). Bacterial metabolism modifies beneficial byproducts such as SFCAs and bile acids[42]. By modulating intestinal microbiota composition, QS can indirectly influence gut physiology by promoting deleterious or beneficial bacteria (Figure 2). Thompson et al[42] demonstrated that AI-2 modulates dysbiotic microbiota composition by enhancing Firmicutes growth. Several reports in vitro and in vivo describe how enteropathogens can signal through QS to commensals and trigger the expression of toxins, virulence factors, and biofilm formation[44-46]. In addition, AI-3 controls the genes that enable enterohemorrhagic Escherichia coli to cause lesions by the attachment and effacing process[19].

The question addressed in this review is how quorum-sensing molecules can directly affect the host, independent of the producing bacteria (Figure 2).

HOST: “YOU’VE GOT MAIL”

As discussed above, it remains largely unknown how the microbial communities hosted in the gut lumen use QS communication systems. However, there is evidence that at least several bacterial species commonly found in the gastrointestinal tract have the capacity to synthesize QS molecules[6,8,39].

Studies on the impact of QS molecules on the biology of intestinal host cells have focused on key actors of barrier function and the immune response. Intestinal barrier function includes the ability of epithelial cells to form: (1) A selective barrier whose permeability is controlled by cell-cell junctions[47], (2) Synthesize a protective mucus layer and antimicrobial peptides[48,49], and (3) Secrete cytokines and chemokines allowing appropriate crosstalk with the underlying immune compartment. The intestinal immune system is involved in tolerogenic or inflammatory responses to the commensal microbiota or pathobionts/pathogens[50,51], and it represents the largest immune organ in the body. Intestinal epithelial cells, intraepithelial lymphocytes, and immune cells located in the lamina propria are involved in the modulation of immunity and inflammation by microbiota[52].

The impact of QS molecules on barrier function and the immune response has been mainly studied in the context of host-pathogen interactions, probably because most of the data rely on AHLs produced by P. aeruginosa. However, evidence of the presence of QS molecules in the healthy intestinal lumen has led to further study on their effects on the host compartment, including barrier function, inflammatory process, and carcinogenesis.

Effects of quorum-sensing molecules on intestinal epithelial barrier function

AHLs: 3-oxo-C12-HSL produced by P. aeruginosa is probably the QS molecule whose effects on the barrier function of epithelial cells have been the most studied during the last two decades[53]. P. aeruginosa synthesizes various virulence factors, which act synergistically with QS molecules to destabilize cell-cell junctions and promote bacterial transmigration across epithelial and endothelial barriers[54].

3-oxo-C12-HSL induces an increase in paracellular permeability to ions and macromolecules[55-60] (Table 2). This deleterious effect of 3-oxo-C12-HSL on barrier function is accompanied by an alteration of tight junctions (TJs) (Figure 3). In the Caco-2 intestinal epithelial cell line, 3-oxo-C12-HSL induced a decrease in the expression, as well as mislocalization, of the TJ proteins occludin, tricellulin, ZO- 1, ZO-3, JAM-A, and of the adherent junction proteins E-cadherin and β-catenin[55,58-61] (Table 2). Loss of occludin/ZO-1 and tricellulin/ZO-1 interaction at the plasma membrane suggested the dismantling of TJ protein complexes[61]. In addition, hyperphosphorylation of occludin, ZO-1, ZO-3, JAM-A, E-cadherin, and β-catenin on tyrosine residues (as well as serine and threonine for E-cadherin and ZO-1) was reported in the presence of 3-oxo-C12-HSL, whereas the serine and threonine residues of occludin, JAM-A and β-catenin were less phosphorylated[58,59].

Table 2

Effects of quorum sensing molecules on different parameters of the intestinal epithelial barrier function

QS molecule	Effects	Ref.
Effects on the intestinal epithelial migration
3-oxo-C12-HSL	Increased migration at low concentrations (1.5-12 μmol/L) vs inhibition at 200 μmol/L	Karlsson et al[72]
	Interaction with IQGAP1 and increase in Rac1/Cdc42 (1.5-200 μmol/L)	Karlsson et al[72]
Effects on the intestinal epithelial permeability and intercellular junctions
3-oxo-C12-HSL	Increased permeability to ions and macromolecules (100-400 μmol/L)	Eum et al[55], Vikström et al[58-60], and Aguanno et al[61]
	Activation of p38 and p42/44 and calcium signaling (100-200 μmol/L)	Vikström et al[58-60]
	Decreased expression levels of tight junction genes (100-400 μmol/L); Disassembly of tight and adherens junctions (modification of their phosphorylation status and involvement of MMP-2 and -3)	Eum et al[55], Vikström et al[58-60], and Aguanno et al[61]
	Decreased levels of tight junction proteins occludin and tricellulin (100-400 μmol/L)	Eum et al[55]
	Decreased protein levels of extracellular matrix and tight junction proteins (400 μmol/L)	Tao et al[62]
3-oxo-C12:2-HSL	No deleterious effects on permeabilityProtection of tight junction integrity and maintenance of junctional complexes at the plasma membrane under pro-inflammatory conditions	Landman et al[39] and Aguanno et al[61]
3-oxo-C14-HSL	Decreased protein levels of extracellular matrix and tight junction proteins (400 μmol/L)	Tao et al[62]
Indole and indole derivatives	Decreased permeability to ions and increased expression of genes coding tight junction and cytoskeleton proteins	Bansal et al[76] and Shimada et al[77]
	Decreased permeability to macromolecules	Venkatesh et al[79]
	Increased transcripts levels of genes coding tight junction proteins	Shin et al[78]
Effects on the mucus layer components
3-oxo-C12-HSL	Decreased MUC3 mRNA levels (30 μmol/L)	Taguchi et al[70]
	Decrease in Muc2 production in goblet cell-like cell line (100 μmol/L) vs increase in colonic cell line (400 μmol/L)	Tao et al[67]
Indole	Increased expression of genes involved in the production of mucins	Bansal et al[76]
Effects on intestinal epithelial cell viability
3-oxo-C12-HSL	Mitochondrial dysfunction and induction of apoptosis in goblet cell-like cell line (100 μmol/L) and in colonic cell line (30-100 μmol/L)	Tao et al[67-69], and Taguchi et al[70]
	Induction of apoptosis, mitochondrial dysfunction, oxidative stress and blocking of cell cycle (400 μmol/L)	Tao et al[62]
3-oxo-C14-HSL	Induction of apoptosis, mitochondrial dysfunction, oxidative stress and blocking of cell cycle (400 μmol/L)	Eum et al[55], Vikström et al[58-60], Aguanno et al[61], and Tao et al[62]
CSF	Reduction of oxidative stress-induced cell death and loss of the epithelial barrier (involving HSP27 and p38/MAPK pathway)	Fujiya et al[74]

CSF: Competence and sporulation factor; HSL: Homoserine lactones; HSP27: Heat shock protein 27; IQGAP1: IQ motif containing GTPase activating protein 1; MAPK: Mitogen-activated protein kinase; MMP-2/-3: Matrix metalloproteinase-2/-3; MUC: Mucin; QS: Quorum sensing; Rac1/Cdc42: Ras-related C3 botulinum toxin substrate 1/cell division control protein 42 homolog.

Figure 3

Effects of quorum sensing molecules on intestinal barrier function (see Table 2) and on the immune response (see Table 3). The Pseudomonas aeruginosa quorum sensing (QS) molecule 3-oxo-C12-HSL induces apoptosis in various cell types, including epithelial cells, promoting a breach in the intestinal barrier. In addition, 3-oxo-C12-HSL disrupts tight junctions, thus leading to increased paracellular permeability, and affects mucin production. Conversely, intestinal acyl-homoserine lactone 3-oxo-C12:2-HSL and the tryptophan metabolite indole protect tight junctions. Bacillus subtilis CSF, which binds to OCTN2, also promotes intestinal barrier integrity by reducing cell death through activation of HSP27 signaling. While 3-oxo-C12-HSL stimulates chemoattraction and phagocytosis in neutrophils and induces cell death, its pro- or anti-inflammatory effects on immune cells are more complex (see Table 3). Autoinducers (AI)-2 and AI-3 both exert proinflammatory effects on macrophages by inducing the expression of the immune mediators TNSF9 and interleukin (IL)-8, respectively, whereas 3-oxo-C12:2-HSL reduces IL-8 production by epithelial cells. It remains to be clarified how all these QS molecules could cross the intestinal barrier and/or reach immune cells in vivo in a physiological context, as illustrated by dotted lines. Last, just as QS molecules can impact eukaryotic cells, the host can interfere with QS: the hormones epinephrine/norepinephrine bind to the AI-3 receptor in EHEC; intestinal epithelial cells secrete an AI-2 mimic in addition to paraoxonase (PON) enzymes degrading homoserine lactones. CSF: Competence and sporulation factor; AI: Autoinducer; PON: Paraoxonase; AhR: Aryl hydrocarbon receptor; IL: Interleukin.

Several signaling mechanisms, including p38 and p42/44 MAP kinases[59,60], Ca2+ release[57,58], matrix metalloproteinases MMP-2 and MMP-3 via protease-activated receptor (PAR) signaling[55] and oxidative stress[62], have been implicated in 3-oxo-C12-HSL effects on junctional proteins and on the concomitant increase in permeability (Table 2).

As discussed above, the most prominent AHL detected in the human intestinal ecosystem is unsaturated 3-oxo-C12:2-HSL[39]. Despite a high structural homology with the P. aeruginosa 3-oxo-C12-HSL, this intestinal AHL has recently been found to have opposite properties regarding the barrier function (Table 2). In contrast to 3-oxo-C12-HSL, 3-oxo-C12:2-HSL does not increase paracellular permeability in Caco-2/TC7 enterocytic cells[39,61]. Most importantly, 3-oxo-C12:2-HSL can limit TJ disruption induced by the proinflammatory cytokines interferon-gamma (IFN-γ) and tumor necrosis factor-α (TNF-α)[61] (Figure 3). In these conditions mimicking intestinal inflammation encountered, for example, in IBD, 3-oxo-C12:2-HSL maintains the interaction of the TJ transmembrane proteins occludin and tricellulin with their main cytoplasmic partner ZO-1. It limits cytokine-induced occludin and tricellulin ubiquitination and the interaction of these TJ proteins with the E3 ubiquitin ligase itch, suggesting stabilization of TJ complexes at the plasma membrane in inflammatory conditions. Altogether, these results show that “commensal” intestinal 3-oxo-C12:2-HSL mitigates the deleterious effects of the inflammatory environment on TJs, which are key actors in epithelial barrier function[61].

Epithelial barrier disruption may combine TJ alteration and an unrestricted passage, which occurs following epithelial damage, generated, for example, by cell apoptosis upon exposure to harmful molecules such as high doses of proinflammatory cytokines[63-65]. This TJ-independent breaking of the barrier allows translocation of large particles such as large proteins, entire bacteria, and viruses, which a priori cannot cross the epithelium through the paracellular route even in conditions where TJs are “open”[66]. The 3-oxo-C12-HSL produced by P. aeruginosa exerts cytotoxic effects, particularly through apoptosis induction, in numerous cell types, including the intestinal and colonic epithelial cell lines LS174T[67-69], Caco-2[70], and CT26[62] (Table 2 and Figure 3). Apoptosis triggered by 3-oxo-C12-HSL relies on oxidative stress and caspase-dependent processes[62,69], whereas short-chain C4-HSL does not exert any apoptotic effects[67]. Interestingly, the increase in paracellular permeability to macromolecules induced by 3-oxo-C12-HSL was dramatically exacerbated in Caco-2/TC7 cells cultured in the presence of IFN-γ and TNF-α or cocultured with THP-1 activated monocytic cells[61]. These synergistic effects on barrier disruption probably rely on epithelial cell apoptosis, as they are abolished by a caspase inhibitor (unpublished results). In contrast, intestinal AHL 3-oxo-C12:2-HSL neither exerts cytotoxic effects nor synergizes with proinflammatory cytokines to disrupt the epithelial barrier[61].

Epithelial injury accompanying acute inflammatory conditions is followed by a re-epithelialization phase, during which cell migration plays an important role[71]. Interestingly, 3-oxo-C12-HSL has been shown to dose-dependently modulate Caco-2 cell migration in a wound-healing assay and interact directly with the GTPase activating protein IQGAP1, stressing a potential role of AHL in cytoskeletal reorganization[72] (Table 2).

Another key actor of the intestinal physical barrier is the mucus layer, which is essential to maintain segregation between luminal microorganisms and the epithelium[49]. 3-oxo-C12-HSL induces reduced expression and production of Mucin2 in LS174T cells[67], as well as a decrease in the levels of MUC3 mRNA in the Caco-2 cell line cultivated in an undifferentiated state[70] (Table 2 and Figure 3). Interestingly, differentiated Caco-2 cells, which express higher levels of mucin 3, showed less sensitivity to 3-oxo-C12-HSL-induced apoptosis in the latter study, and the addition of mucin dose-dependently protected cells from apoptosis induced by this AHL[70].

It must be specified that all these studies on barrier function were carried out with high concentrations of AHLs (100-400 μmol/L) (Table 2), knowing that the concentration of 3-oxo-C12-HSL has been estimated to reach 600 μmol/L in biofilms of P. aeruginosa[73].

Gram-positive QS peptides: The effects of AIP (found in Gram-positive bacteria) on intestinal barrier function are much less documented than those of Gram-negative QS molecules. Whereas most of the studies on the effects of AIP on host inflammation describe the indirect effects of AI through the modulation of bacterial metabolism, one article reported the direct effects of AIP. Fujiya et al[74] reported that Bacillus subtilis AIP, named competence and sporulation factor (CSF), induces HSP27 expression and the p38/MAPK pathway and reduces cell death and the loss of the epithelial barrier induced by oxidative stress (Table 2 and Figure 3). Inducible HSPs are needed under stress and help stabilize proteins to prevent denaturation[75]. B. subtilis is part of the normal microbiota; it is also used as a commercial probiotic and is beneficial to the host. Moreover, CSF seems to signal through a receptor named OCTN2 (organic cation/carnitine transporter 2), and polymorphisms of the gene encoding this receptor are part of the susceptibility locus of Crohn’s disease[74].

Indole: Indole exerts a beneficial role on TJ protein expression in several intestinal epithelial cells[76-79] (Table 2). Oral administration of indole to germ-free mice, which display very low indole fecal levels, increased the expression of TJ and adherens junction-associated proteins in the colonic epithelium and improved their resistance to dextran sulfate sodium (DSS)-induced colitis[77]. Indole has been identified as an endogenous agonist of aryl hydrocarbon receptor (AhR), which can compete for receptor binding with well-known AhR ligands[80]; several studies have also stressed the key role of the AhR pathway in indole derivative protective effects[81-83]. Accordingly, several studies have shown that AhR activation strengthens the epithelial barrier by protecting TJs[82,84-87] (Figure 3) or by stimulating antimicrobial peptide production via interleukin (IL)-22[88,89].

Effects of QS molecules on immune response

In addition to their effects on the intestinal barrier, QS molecules were analyzed on different actors of the immune compartment of the intestine, which is involved in a complex crosstalk with the epithelial compartment to maintain an appropriate immune response toward the content of the intestinal lumen.

AHLs: Our group described that 3-oxo-C12:2-HSL, an AHL recently discovered in the human gut[39], exerts anti-inflammatory effects on intestinal epithelial cells. During inflammation, intestinal epithelial cells can secrete some cytokines, among which the chemokine IL-8 promotes the recruitment of neutrophils in the mucosa and participates in the acute-phase response[90,91]. In a study comparing the effect of 3-oxo-C12:2-HSL to 3-oxo-C12-HSL produced by P. aeruginosa, our group demonstrated in the human enterocytic Caco-2/TC7 cell line that 3-oxo-C12:2-HSL, but not 3-oxo-C12-HSL, attenuated the induction of IL-8 secretion induced by the proinflammatory cytokine IL-1β[39,92] (Table 3 and Figure 3). This potential anti-inflammatory effect of 3-oxo-C12:2-HSL is consistent with the hypothesis of a beneficial role of this AHL in gut ecosystems[39], as are its protective effects on TJ integrity. The impact of intestinal 3-oxo-C12:2-HSL on immune cells remains largely unknown.

Table 3

Effects of quorum sensing molecules on inflammation in different cell types

Cell type	QS molecule	Effects	Ref.
Effects on innate immune cells
Macrophages	3-oxo-C12-HSL	Anti-inflammatory effects on IL-12 and TNF-α (0.1-100 μmol/L)	Telford et al[94]
		Increased TLR2 and TLR4 expression and decreased TNF-α production (1-100 μmol/L)	Bao et al[180]
		Pro-apoptotic effects (12-50 μmol/L)	Tateda et al[102]
		Increased phagocytosis (100 μmol/L)	Vikström et al[107]
		NF-κB inhibition (4.7 μmol/L)	Kravchenko et al[104]
		Dose-dependent anti-inflammatory effects (1-50 μmol/L)	Kravchenko et al[105]
		Involvement in p38/MAPK signaling (1-100 μmol/L)	Kravchenko et al[105], Vikström et al[107], Glucksam-Galnoy et al[181]
		Activation of the Unfolded Protein Response (6.25-100 μmol/L)	Zhang et al[182]
		Change in cell volume and shape (10-50 μmol/L)	Holm et al[183]
	Indole derivatives	Prevents the induction of pro-inflammatory cytokines	Krishnan et al[184]
	AI-2	Induction of the expression of cytokines, chemokines and TNFSF9	Li et al[41]
Monocytes	AI-3 and analogues	Increase in IL-8 secretion	Kim et al[22]
Dendritic cells	3-oxo-C12-HSL	Pro-apoptotic effects (100 μmol/L)	Boontham et al[185]
		No effect on IL-10 secretion (5-30 μmol/L)	Skindersoe et al[100]
		Increased IL-10 production (5-100 μmol/L)	Li et al[99]
		Decreased IL-12 secretion (5-100 μmol/L)	Li et al[99] and Skindersoe et al[100]
		Increased induction of Treg (5-100 μmol/L)	Li et al[99]
Neutrophils	3-oxo-C12-HSL	Chemoattraction (0.01-100 μmol/L)	Karlsson et al[186] and Zimmermann et al[187]
		Activation of MAPK signaling (12-50 μmol/L)	Tateda et al[102] and Singh et al[188]
		Increased phagocytosis (10 μmol/L)	Wagner et al[189]
		Pro-apoptotic effects (12-50 μmol/L)	Tateda et al[102]
Effects on adaptive immune cells
T cells	3-oxo-C12-HSL	Inhibition of proliferation and activation (0.1-100 μmol/L)	Telford et al[94], Boontham et al[185], Gupta et al[190], and Hooi et al[191]
		Activation of naïve T cells towards Th1 phenotype (5 μmol/L)	Smith et al[95]
		Decreased secretion of IL-4 and IFN-γ (5 μmol/L)	Ritchie et al[96]
		Induction of apoptosis via the mitochondria pathway (100 μmol/L)	Jacobi et al[101]
		Induction of Treg (1-50 μmol/L)	Li et al[99]
	Indole derivatives	Re-programming into tolerogenic T cells	Cervantes-Barragan et al[192]
		Promotion of differentiation towards a regulatory type 1 phenotype	Aoki et al[193]
B cells	3-oxo-C12-HSL	Modulation of immunoglobulin production (0.1-100 μmol/L)	Telford et al[94] and Ritchie et al[194]
ILC	Indole derivatives	Promotion of IL-22 production	Zelante et al[83]
Effects on epithelial cells
Pulmonary tract epithelial cells	3-oxo-C12-HSL	Induction of IL-8 production and NF-B activation (100 μmol/L)	Smith et al[195]
		Increased expression levels of pro-inflammatory cytokines	Jahoor et al[115]
Intestinal epithelial cells	3-oxo-C12-HSL	Mitigation (1-10 μmol/L) or aggravation (> 50 μmol/L) of IL-8 expression induction	Peyrottes et al[92]
	3-oxo-C12:2-HSL	Attenuation of the induction of IL-8 expression (5-50 μmol/L)	Landman et al[39]

AI: Autoinducer; B cells: Lymphocytes B; HSL: Homoserine lactones; IFN-γ: Interferon-γ; IL: Interleukin; ILC: Innate lymphoid cells; MAPK: Mitogen-activated protein kinase; NF-κB: Nuclear factor-kappa B; QS: Quorum sensing; T cells: Lymphocytes T; Th: T helper; TLR: Toll like receptors; TNF-α: Tumor necrosis factor-α; Treg: Regulatory T cells.

The effects of 3-oxo-C12-HSL depend on the concentration and cell type studied[6,93]. Telford et al[94] showed that 3-oxo-C12-HSL inhibits the production of TNF-α and IL-12 [a cytokine involved in the T helper cell-1 type response (Th1-type response)] by lipopolysaccharide-activated macrophages at high concentrations and stimulates the production of antibodies, particularly immunoglobulin G1, which is an indicator of a Th2-type response at lower concentrations (Table 3). Conversely, Smith et al[95] showed that 3-oxo-C12-HSL activates and promotes the differentiation of naive T lymphocytes toward a Th1-like phenotype, while Ritchie et al[96] observed that 3-oxo-C12-HSL inhibits the differentiation of both Th1 and Th2 T lymphocytes (Table 3). Altogether, these results demonstrated that 3-oxo-C12-HSL is an immunomodulator of the Th1/Th2 response. 3-oxo-C12-HSL and two other QS molecules from P. aeruginosa, PQS (Pseudomonas quinolone signal), and HHQ (4-hydroxy-2-heptylquinoline), suppress both innate and adaptive immune responses acting on lymphoid cells, dendritic cells, and neutrophil monocytes/macrophages[22,97,98]. 3-oxo-C12-HSL and PQS decreased the production of the cytokines IL-12 and IFN-γ by activated dendritic cells, which in turn decreased T-cell proliferation and activity[98-100] while promoting the induction of regulatory T-cells[99] (Table 3). 3-oxo-C12-HSL provoked apoptosis of macrophages, neutrophils, and T lymphocytes through activation of caspases and the mitochondrial apoptosis pathway[101,102]. Several reports described inhibition of the nuclear factor-kappa B (NF-κB) pathway by QS molecules from P. aeruginosa[93,103-106] and/or the activation of signaling pathways such as p38 MAPK[105,107] (Table 3). It has been recently demonstrated that only long-chain AHLs such as 3-oxo-C12-HSL modulate the phenotype of dendritic cells and the type 2 immune response through mechanisms involving retinoic acid signaling and the protein kinase AKT[106].

The molecular mechanisms involved in the effects of QS molecules from P. aeruginosa on immune cells are independent of the Toll-like receptor pathway[105], which is a classical cell process involved in recognizing pathogen fragments. Some reports have indicated that the perception of AHL by mammalian cells involves the bitter taste receptor T2R38[108-110], which is widely expressed in the human digestive tract from the tongue to the colon[111]. Polymorphisms in the TAS2R38 gene may increase susceptibility to infections and colorectal cancer (CRC)[112]. It has been shown that these receptors use inflammatory pathways, which differ according to the cell type and their localization[113]. 3-oxo-C12-HSL binds to the transcription factor peroxisome proliferator-activated receptor γ[114], which has been proposed as a potential receptor for AHL and seems to be involved in AHL proinflammatory effects[115]. Recently, Moura-Alves et al[116] showed that QS molecules produced by P. aeruginosa modulated the activity of the transcription factor AhR, which plays an important role in regulating innate and adaptive immunity[117,118].

Overall, it has been demonstrated that QS molecules from P. aeruginosa have an immunosuppressive effect, allowing the pathogen to evade the immune system during infection. It remains to be determined whether endogenous intestinal 3-oxo-C12:2-HSL participates in controlling intestinal immunity in health and diseases and to decipher the underlying mechanisms.

AI-2 and AI-3: AI-2 is produced by both Gram-negative and Gram-positive bacteria and is mainly studied for its role in bacteria-bacteria communication and the virulence of pathogenic strains[8]. However, little is known about the effect of AI-2 on immune cells. In mice, AI-2 administration has no effect by itself on cytokine expression but aggravates lung inflammation during P. aeruginosa infection by interfering with QS molecules produced by this pathogen[119]. In cultured macrophages, AI-2 induces the expression of several cytokines and chemokines as well as the expression of TNF superfamily member 9 (TNFSF9), a protein involved in the immune response[41] (Figure 3).

The AI-3 system is mainly described in enterohemorrhagic E. coli and is therefore linked to the development of intestinal epithelial lesions, suggesting its proinflammatory activity[120] (Table 3 and Figure 3). Indeed, AI-3 and its analogs increase IL-8 secretion by THP-1 monocytes[22]. Given that the AI-3 structure has only been uncovered recently[22], the direct effect of this molecule on the host is poorly known so far. In addition, since the AI-3 bacterial receptor can recognize host-synthesized epinephrine/norepinephrine (Figure 3), one could suggest that adrenergic receptors could recognize AI-3[19]. However, it has been shown that AI-3 and its analogs do not activate or modulate adrenergic signaling[22].

Gram-positive AIP: The effects of Gram-positive AIP bacteria on inflammation are far less documented than those of Gram-negative QS. Moreover, most of the studies on the effects of AIP on host inflammation describe the indirect effects of AIs through the modulation of bacterial metabolism.

A study described that AIP could selectively cross intestinal epithelial cell Caco-2 cell monolayers[121]. Additionally, it has been reported that AIP can cross the highly selective blood-brain barrier in vivo[122]. These processes seem to be peptide-specific: Clostridium acetobutylicum AIP easily penetrates the blood-brain barrier, while Streptococcus pneumonia’s AIP crosses it poorly[122]. This shows that small molecules such as AIs affect the host's physiology beyond the gastrointestinal tract. For instance, it has been described that AIP has various effects on muscle inflammation as part of the gut-muscle axis. De Spiegeleer et al[121] performed an extensive screening of 75 QS molecules on muscle cells. They demonstrated both pro- and anti-inflammatory effects of four peptides from the genera Staphylococcus, Streptococcus, Lactobacillus, and Bacillus, and some of those peptides have been described in the gut.

Indole: Several studies have reported that indole exerts anti-inflammatory effects in the intestine and protects against pathogenic infection[123] (Table 3). AhR, an important contributor to the maintenance of innate and adaptive immunity, drives most of these effects, particularly in the intestinal mucosa[117,118,124].

Several reports have shown altered Trp metabolism in gut inflammation in humans and mice [81,125-127]. A decrease in endogenous indole was observed in human feces from subjects with celiac disease or IBD[127]. This was associated with a decrease in AhR activity in the intestinal mucosa. In parallel, an increase in Trp levels in the same samples suggested that the gut microbiota-dependent metabolism of Trp was altered[127]. In mouse models of celiac disease and IBD, the implantation of indole-producing bacteria increases AhR activity and protects them from gut inflammation[81]. Interestingly, Moura-Alves et al[116] showed that 3-oxo-C12-HSL and HHQ had an inhibitory effect on AhR activity and could compete with well-known activators of AhR. This observation raises the question of potential competition between several QS molecules for AhR-dependent modulation of innate and adaptive immunity.

Effects of QS molecules on carcinogenesis

There is growing evidence that gut microbiota dysbiosis plays a major role in CRC development[128]. Indeed, modifications of commensal gut microbiota in favor of opportunist bacteria promote intestinal inflammation, which is well known as a driver event in CRC onset[129,130]. Thus, the concept of the “bacterial driver-passenger model” highlights the crosstalk between host immunity and colonic microbiota[131]. For example, some driver pathogens, such as Bacteroides fragilis, have been proposed to promote a strong Th17 inflammatory response[132]. This proinflammatory microenvironment might favor colonization by opportunist pathogens such as Fusobacterium spp. Accordingly, Fusobacteria-dominant biofilms were associated with human CRC[133-135]. Altogether, these findings support that polymicrobial interactions and intercellular communications might play an important role in CRC development[136]. Nevertheless, how bacteria communicate with themselves and with the host during CRC remains poorly understood.

Recent findings provide new insights into the role of the QS molecule AI-2 in intercellular communication during CRC. First, the AI-2 concentration is increased in tumors compared to the surrounding normal tissue in human CRC[41]. These levels also correlate with the progression of the disease according to the CRC TNM (tumor node and metastasis) score[41]. Regarding the tumor immune microenvironment, the AI-2 concentration positively correlates with TNFSF9 expression, which is mainly expressed by tumor-associated macrophages, and negatively correlates with the CD4/CD8 ratio, suggesting that AI-2 associates with the antitumor response[41]. At the molecular level, it was demonstrated that AI-2 induces in vitro M1 polarization of U987-derived macrophages through the TNFSF9 signaling pathway[137]. These findings reveal that AI-2 could be an important factor linked to the immune tumor microenvironment and shed light on the role of the quorum-sensing system during CRC development and progression. Interestingly, mammalian epithelial cells are able to produce AI-2 analog molecules that mimic AI-2 effects (Figure 3), illustrating the complexity of bacteria-host crosstalk[45]. Thus, a better characterization of QS molecules involved in tumorigenesis might be an opportunity to improve our knowledge of the mechanisms underlying CRC development.

The host strikes back to QS

Interkingdom signaling works in two ways, as host cells are able to counterattack the QS system using several strategies.

As described above, as part of AI-3 signaling, host hormones such as epinephrine and norepinephrine can be recognized by EHEC and lead to the expression of virulence genes[19]. This AI-3/epinephrine/norepinephrine signaling is not restricted to EHEC, and the receptor QseC is also expressed, for example, by the intestinal pathogenic Salmonella enterica serovar Typhimurium[138] (Table 1). Recently, in silico analysis suggested that another catecholamine neurotransmitter, dopamine, can bind to QseC. However, no effect in vitro was measured[139]. This study of interkingdom signaling through hormones has been named “microbial endocrinology”[140].

Interestingly, there is evidence that human epithelial cells can produce AI-2 mimicking molecules. The study was conducted on Caco-2 intestinal epithelial cells, and the authors showed that an AI-2 mimic is produced not only when cells are in contact with bacteria but also after TJ disruption by calcium deprivation or DSS treatment[45]. This emphasizes how much AI-2 is a universal language between Gram-positive and Gram-negative bacteria and the host (Figure 3).

Hosts have also developed defense tools against QS, leading to a mechanism named quorum quenching. Mammals can synthesize enzymes named paraoxonases (PONs) that hydrolyze the lactone ring of long-chain AHLs[141] (Figure 3). There are three types of PONs (PON 1, PON2, PON3) that are highly conserved across species, and PON2 has greater activity on AHLs[142]. It has been demonstrated that PON2 is more highly expressed in the human jejunum than in other parts of the intestine[143]. Interestingly, PON1 and PON3 are expressed at lower levels in patients with Crohn’s disease and ulcerative colitis than in healthy subjects[144]. A case-control study has also shown that carriage of the PON1 R192 allele in Ashkenazi Jewish may confer protection against the development of IBD. This allele was significantly less common among IBD Ashkenazi patients, with a significant odds ratio of 0.61[145].

QS IN THE GUT: FUTURE DIRECTIONS FOR THIS NEW PLAYER

Using QS to modulate gut microbiota: Application in gut ecosystem disorders

Gut dysbiosis is an imbalance in the composition of microorganisms inside the digestive tract, especially described with bacteria. This dysregulation has been shown to be a preponderant risk factor in several digestive and extra digestive diseases[146-149]. For example, in IBD and recurrent C. difficile infections, it is well known that the over- and underrepresentation of certain phyla can lead to a pathologic state[150,151]. Modulating the gut microbiota may be the key to treating or even preventing such diseases by restoring normobiosis. Fecal microbiota transplantation is now commonly used in the setting of C. difficile infections[152,153]. However, the lack of standardization and the safety and quality issues of this procedure call for the development of new strategies.

Theoretically, AHLs remain good candidates in this approach using natural molecules from QS to modulate microbiota composition and gut inflammation. As seen above, AHL signaling may involve different pathways that contribute to controlling intestinal inflammation, such as inhibition of NF-κB, modulation, inhibition of MAPK activation, increase in regulatory T cell induction, decrease in proinflammatory cytokines, and modulation of junctional complexes in the epithelial barrier. Indeed, using QS molecules could play a role in both components (gut microbiota and host responses) of gut ecosystem disorders observed in metabolic and inflammatory diseases. AHL-based QS devices already exist as therapeutic applications for the dynamic control of Gram-negative bacterial populations, especially in infectious diseases. Other QS molecules could be extended as potential clinical therapies for diseases related to the gut microbiota that involve biofilm formation and antibiotic resistance[154]. Research efforts must investigate the potential of this new trial.

In addition to therapeutic applications, one could consider QS molecules as reliable biomarkers for dysbiosis-related chronic diseases such as IBD or CRC. Indeed, it has been shown that the presence of some AI-1 QS molecules in the gut ecosystem directly correlates with bacterial group size[39]. AHLs could represent a biomarker of the bacterial level population acting as a magnifying glass for dysbiosis. In addition, AI-2 concentration increased during adenomas to colorectal transition and CRC progression[41]. This opens the perspective for using the QS system as a biomarker for the prevention and follow-up of chronic diseases.

For the future

Knowing which commensal bacteria carry QS systems, their site of production, their ability to be mobilized during dysbiosis, and their effect on the luminal or mucosal microenvironment are as many unresolved questions. The scientific community, together with gastroenterologists, needs to tackle these issues to pave the way for translation into clinical use. Future directions also involve designing dedicated QS derivatives targeting either the host cells or the bacterial compartment. Such QS derivatives have already been reported to control the epithelial cell inflammation pathway with a wider effect than natural 3-oxoC12:2 without bacterial-activating properties[92]. Using QS molecules as an approach to tackle the gut microbiota compartment has already been proven to be a successful strategy[42], thus leading to interesting perspectives. Considering the QS system as a new player in the gut ecosystem, it represents a control platform to shape the host's gut microbiota population and/or major physiological pathways.

CONCLUSION

In conclusion, the intestinal microbiota interacts mutually with epithelial and immune cells of the coevolved host in a beneficial, reciprocal relationship. The QS signaling of bacteria probably contributes substantially to establishing symbiotic interactions in certain dynamics of interaction between the different kingdoms. A better understanding of QS systems by researchers and gastroenterologists involved in describing and managing ecological disorders of the intestinal ecosystem is a new approach that opens up fascinating therapeutic opportunities.

ACKNOWLEDGEMENTS

We thank Rainteau D for his support and knowledge on quorum sensing. We acknowledge the Association François Aupetit for its unwavering support.