Mechanisms of quorum sensing and strategies for quorum sensing disruption in aquaculture pathogens.

doi:10.1111/jfd.12299

Journal of Fish Diseases 2014

Mechanisms of quorum sensing and strategies for quorum sensing disruption in aquaculture pathogens J Zhao1,2, M Chen3, CS Quan1,2 and SD Fan1,2 1 Key Laboratory of Biochemical Engineering State Ethnic Affairs Commission-Ministry of Education, Dalian Nationalities University, Dalian, China 2 College of Life Science, Dalian Nationalities University, Dalian, China 3 College of Bioengineering, Dalian Polytechnic University, Dalian, China

Abstract

In many countries, infectious diseases are a considerable threat to aquaculture. The pathogenicity of micro-organisms that infect aquaculture systems is closely related to the release of virulence factors and the formation of biofilms, both of which are regulated by quorum sensing (QS). Thus, QS disruption is a potential strategy for preventing disease in aquaculture systems. QS inhibitors (QSIs) not only inhibit the expression of virulence-associated genes but also attenuate the virulence of aquaculture pathogens. In this review, we discuss QS systems in important aquaculture pathogens and focus on the relationship between QS mechanisms and bacterial virulence in aquaculture. We further elucidate QS disruption strategies for targeting aquaculture pathogens. Four main types of QSIs that target aquaculture pathogens are discussed based on their mechanisms of action. Keywords: aquaculture pathogen, disease control, disruption, QS mechanism, QSI, virulence.

Introduction

Aquaculture is by far the fastest growing food-production sector in the world. However, infectious diseases caused by pathogens dramatically reduce the productivity of aquaculture and have impeded economic development in many countries (Meyer Correspondence J Zhao, College of Life Science, Dalian Nationalities University, Dalian 116600, China (e-mail: [email protected]) Ó 2014 John Wiley & Sons Ltd

1

1991; Hill 2005; Subasinghe 2009). Global outbreaks of aquaculture disease are estimated to contribute to the loss of several billion US dollars per year (Subasinghe, Bondad-Reantaso & McGladdery 2001). Many different approaches have been used to prevent and treat disease in aquaculture systems. However, each approach has limitations or side effects: residues from chemotherapy may threaten public health with toxicity and also trigger resistance; antibiotics may also result in high residue levels and resistance in microbial pathogens; vaccines, which have specific targets and undergo complicated development before commercialization, are still unavailable for certain pathogens; and probiotics are more useful for prevention than treatment, yet their effectiveness is limited by the environment (Chinabut & Puttinaowarat 2005; Balcazar et al. 2006; Sahu et al. 2008; Subasinghe 2009; Defoirdt, et al. 2011). Therefore, to improve the sustainability of the aquaculture industry, new strategies for disease prevention and treatment need to be explored. In recent years, the discovery of quorum sensing (cell-to-cell communication) in bacteria provided a new perspective for understanding pathogenicity (De Kievit & Iglewski 2000; Deep, Chaudhary & Gupta 2011). Quorum sensing (QS) is the regulation of gene expression in response to fluctuations in cell population density, which correlates with signalling molecule concentration (Miller & Bassler 2001). Basic QS systems in bacteria can be divided into the following four categories: the LuxI/R paradigm system in most Gram-negative bacteria, which uses acyl-homoserine lactones (AHLs) as signalling molecules; the two-component


(membrane-bound sensor histidine kinase) system in Gram-positive bacteria, which uses modified oligopeptides as signals; the LuxS mediated system, which uses autoinducer 2 (AI-2, furanosyl borate ester) as a signal, and the CqsA/S system, which uses cholerae autoinducer 1 (CAI-1, also known as (S)-3-hydroxytridecan-4-one) as a signal (Schauder et al. 2001; Waters & Bassler 2005; Ng & Bassler 2009). Certain Gram-negative bacteria with unknown QS systems may also use diketopiperazines (DKPs) as a new type of signalling molecules (Wang et al. 2010). In addition, complicated QS networks exist. Vibrio harveyi uses a typical parallel QS circuit. Signals from all three channels transfer to the same regulator with superimposed influences on gene expression. QS circuits arranged in series exist in Pseudomonas aeruginosa, where the LasIR circuit activates the RhlIR circuit. In addition, there are competitive QS circuits, as in Bacillus subtilis, and QS systems that respond to host cues, as in Agrobacterium tumefaciens (Waters & Bassler 2005). Interestingly, QS appears to control the expression of virulence factors and biofilm formation (Williams et al. 2000; Jayaraman & Wood 2008; Antunes et al. 2010). Hence, the disruption of pathogen QS provides a new strategy for controlling pathogen infections. Previous studies reported QS in human pathogens, plant pathogens and veterinary pathogens (Pesci et al. 1997; Von Bodman, Bauer & Coplin 2003; Schuster & Greenberg 2006; Sjoblom et al. 2006; Boyen et al. 2009; Antunes et al. 2010; Deep et al. 2011). In this review, we discuss QS mechanisms in aquaculture pathogens, the effect of QS on pathogenicity and QS-related anti-infection strategies.

QS mechanisms and their effect on the virulence of aquaculture pathogens

QS mechanisms of aquaculture pathogens QS systems, which include signalling molecules and a series of regulatory proteins, have been identified in many aquaculture pathogens. Three types of signalling molecules (AHLs, AI-2 and CAI-1) have been identified or predicted in V. harveyi, Vibrio anguillarum and Vibrio parahaemolyticus (Milton et al. 2001; Croxatto et al. 2002, 2004; Henke & Bassler 2004a,b; Defoirdt et al. 2008). However, only AHLs and AI-2 have Ó 2014 John Wiley & Sons Ltd

2

J Zhao et al. QS mechanisms and disruption

been discovered in Vibrio vulnificus, Vibrio ichthyoenteri, Aeromonas hydrophila and Edwardsiella tarda (Swift et al. 1997; Kim et al. 2003; Morohoshi et al. 2004; Zhang, Sun & Sun 2008; Valiente et al. 2009; Han et al. 2010; Li et al. 2010). In Vibrio alginolyticus, only AI-2 and CAI1 have been reported (Henke & Bassler 2004b; Ye et al. 2008). In Vibrio mimicus, only AI-2-like activity has been found (Sultan, Miyoshi & Shinoda 2006), whereas in Allivibrio salmonicida, Aeromonas salmonicida, Yersinia ruckeri and Tenacibaculum maritimum, only AHLs have been detected (Swift et al. 1997; Temprano et al. 2001; Bruhn et al. 2005; Romero et al. 2010). Based on the findings summarized above, AHLs are the most common signalling molecules in aquaculture pathogens. AHLs possess acyl side chains that vary from 2 to 8 carbon atoms in length. Certain AHLs also have 3-oxo or 3-hydroxy substituents on the acyl side chains. Most AHLs have an even number of carbon atoms on the acyl chains; acyl chains with an odd number of carbon atoms have only been described in the AHLs from E. tarda and Y. ruckeri. Different combinations of signalling molecules in aquaculture pathogens could perform diverse physiological functions. In addition, the types of signalling molecules used by pathogens and the conditions under which they are expressed need to be explored further different pathogens. QS mechanisms have been comprehensively described in several aquaculture pathogens. A model has been proposed for the V. harveyi threechannel QS system. The three signalling molecules, N-(3-hydroxybutanoyl)-homoserine lactone, AI-2 and CAI-1, are synthesized by the LuxM, LuxS and CqsA enzymes, respectively. These signals are detected at the cell surface by the LuxN, LuxPQ and CqsS receptor proteins, respectively. In the presence of high concentrations of signalling molecules, the receptor proteins switch from acting as kinases to acting as phosphatases, which results in the dephosphorylation of LuxO (a global regulator), leading to the production of the transcriptional regulator protein LuxRVh and expression of QS-regulated genes (Cao & Meighen 1989; Henke & Bassler 2004b; Defoirdt et al. 2008). The QS system model from V. anguillarum shares a certain amount of similarity with that of V. harveyi (Weber et al. 2008). V. anguillarum also possesses three channels that appear to function via the same three types of signalling


molecules and corresponding regulatory proteins as in V. harveyi. Signal transduction from the three channels likely converges on VanU. After a series of signal transduction events similar to those of V. harveyi, high cell density finally activates VanT expression, which regulates downstream gene expression. However, this QS system appears to be even more complex. First, a hierarchical QS system may exist. N-(3-oxodecanoyl)-L-homoserine lactone (ODHL) is synthesized by the enzyme VanI and binds to VanR, a transcriptional activator, whereas VanMN, homologues of V. harveyi LuxMN, acts as regulatory proteins for two additional AHLs, N-hexanoyl homoserine lactone (HHL) and N-(3-hydroxyhexanoyl) homoserine lactone (OhHHL). VanMN regulates the production of ODHL via VanIR. Another unique feature of QS in V. anguillarum is the existence of specific regulatory proteins. As part of the known QS regulatory cascade, sigma factor RpoS independently represses expression of Hfq (an RNA chaperone) and further induces VanT expression. In addition to the above two models, part of the V. vulnificus and V. parahaemolyticus QS systems have also been described (Milton 2006; GodePotratz & McCarter 2011) and is quite similar to that of V. harveyi. However, QS mechanisms in other aquaculture pathogens are largely unknown, and further research is needed. Additional complex QS networks may exist in Gram-negative aquaculture pathogens. Effect of QS on the virulence of aquaculture pathogens As shown in Table 1, in addition to virulence factors (proteases, etc.), other factors that facilitate the function of virulence factors (Type III or Type VI secretion proteins and associated effectors, for example) are regulated by QS. Several interesting phenomena have been discovered. First, different virulence-associated factors are subject to opposite regulation by QS in the same pathogen. For example, in V. vulnificus, the protease VvpE and the haemolysin VvhA are positively and negatively regulated by LuxS/AI-2, respectively. Because expression of virulence-associated factors increases the virulence of aquaculture pathogens, it is unclear how the joint action of virulence factors with different expression profiles functions. It appears that pathogens would become virulent only when there is a net positive effect of Ó 2014 John Wiley & Sons Ltd

3


QS-regulated virulence-associated factors. Mok, Wingreen & Bassler (2003) showed that, to successfully infect a host, a pathogen should express genes encoding secretion systems at a low cell density and express virulence genes at a high cell density. These two types of genes are likely regulated in response to different conditions by QS in aquaculture pathogens to maximize virulence. However, this rule is not absolute, as can be seen in Table 1. Secondly, same virulence-associated factor could be positively and negatively regulated by QS in different pathogens. For example, Type III secretion proteins, which may help inject virulence factors into the cytoplasm of host cells, are negatively regulated by LuxS in V. harveyi. However, these factors appeared to be positively regulated by LuxS in E. tarda. This observation reflects the diversity of QS regulation among pathogens because certain virulence-associated factors were negatively regulated; the total virulence of aquaculture pathogens is represented by the comprehensive expression and effects of multiple virulence-associated genes. In a broad sense, the diversity of QS regulation among pathogens leads to a complex situation, in which it is difficult to judge the net outcome of QS-regulated virulenceassociated factors from different pathogens that coexist in the same aquaculture environment. This increases the difficulty of finding a feasible QS-quenching strategy to control disease in aquaculture systems. Third, certain virulence-associated factors are not regulated by QS. For example, there is no difference in metalloprotease production between vanIR mutants and wild-type strains of V. anguillarum (Milton et al. 1997). Similarly, lipase and haemolysin expression are not dependent on QS (Natrah et al. 2011a), and expression of the vhh haemolysin gene does not appear to be regulated by LuxO in V. harveyi (Ruwandeepika et al. 2011). Challenge tests directly reflect the general effects of QS signalling molecules and regulatory proteins on the virulence of pathogens (Table 2). QS influences virulence in aquaculture pathogens, including Vibrios and Aeromonas species, and QS mutations can reduce or even abolish the virulence of aquaculture pathogens towards certain hosts. In a diarrhoeal isolate of A. hydrophila, QS interference showed reduced virulence in a mouse model (Khajanchi et al. 2009). Biofilm formation was also reduced by the disruption of QS pathway regulatory proteins (Lee et al. 2007;



Table 1 Virulence-associated factors controlled by QS regulatory proteins or signalling molecules in different aquaculture pathogens Pathogens

Regulatory proteins or signalling molecules

V. harveyi

V. anguillarum V. vulnificus

Virulence-associated factors regulated by QS

References

LuxS/AI-2, LuxM/AHL

Extracellular toxin↑, metalloprotease↑, T3SS protein↓, chitinase↓

LuxO

Caseinase↑, gelatinase↑, phospholipase↓, vhp metalloprotease↓

AI-2 RpoS, VanT (LuxR homologue) SmcR (LuxR homologue)

Haemolysin (encoded by vhh) ↑ Extracellular protease (EmpA and PrtV) ↑, haemolysin coregulated-like protein (Hcp) ↓ Protease (encoded by gene vvp) ↑, cytolysin (VvhA) ↓, biofilm↑ Protease (VvpE) ↑, haemolysin (VvhA) ↓

Manefield et al. (2000), Mok et al. (2003), Henke & Bassler (2004a) and Defoirdt et al. (2010) Natrah et al. (2011a,b) and Ruwandeepika et al. (2011) Ruwandeepika et al. (2011) Weber et al. (2009)

LuxS/AI-2 LuxS, SmcR

V. parahaemolyticus

LuxO (represses SmcR) OpaR (LuxR homologue)

V. mimicus V. alginolyticus

LuxO homologue LuxR homologue

LuxO homologue

LuxT LuxS Hfq A. hydrophila

AHL AhyR AhyI AhyI, AhyR

A. salmonicida E. tarda

AsaI EdwI LuxS RpoS QseC QseB

Zn-dependent protease↑, phosphomannomutase (required for synthesis of LPS)↑, Cytotoxins VvhA and RtxA1↑ T3SS↓, T6SS1↓, T6SS2↑

Protease↓ Total extracellular protease activity (especially extracellular alkaline serine protease A) ↑, extracellular polysaccharide↑, a haemolysin coregulated protein (Hcp1) ↓ Extracelluar protease↓, haemolytic products↓, siderophore production↑, a putative virulence factor MviN↓, a haemolysin coregulated protein (Hcp1, hallmark of T6SS ↑ Extracellular protease↑ Protease↑, extracellular polysaccharide↓, biofilm↓ Motility↑, biofilm↑, stress resistance↑,RpoS↑, alkaline serine protease Asp↓, LuxR↓ Biofilm↑ Proteases↑, amylase↑, Dnase↑, haemolysin↑, S layer↑ Biofilm↑ T6SS-associated effectors (haemolysin coregulated protein and valine–glycine repeat family of proteins) ↓, protease↓, biofilm↓ Protease AsaP1↑, a cytotoxic factor↑ Flagellin FliC↓ T3SS↑, biofilm↑, serine protease (encoded by degPEt) ↑ LuxS↓, EdwI↓, biofilm↓ Hemagglutination↑, T3SS elements EseB and EsaC↑, flagellar motilities↑ Hemagglutination↓, flagellar motilities↑, T3SS elements EseB and EsaC↑

Shao & Hor (2001) Kim et al. (2003) and Kawase et al. (2004) Shin, Lee & Yoo (2007)

Shao et al. (2011) Henke & Bassler (2004a) and Gode-Potratz & McCarter (2011) Sultan et al. (2006) Rui et al. (2008) and Sheng et al. (2012)

Wang et al. (2007), Cao et al. (2010) and Sheng et al. (2012) Liu et al. (2012) Ye et al. (2008) Liu et al. (2011) Lynch et al. (2002) Bi et al. (2007) Chu et al. (2011) Khajanchi et al. (2009)

Schwenteita et al. (2011) Morohoshi et al. (2009) Zhang et al. (2008) and Jiao et al. (2010) Xiao et al. (2009) Wang et al. (2011)

↑ indicates that virulence-associated factors are upregulated in the presence of regulatory proteins or signalling molecules. ↓ indicates that virulence-associated factors are downregulated in the presence of regulatory proteins or signalling molecules.

Chu et al. 2011). Interestingly, several studies have speculated that QS has no correlation with virulence: disruption of an AHL-regulated QS channel, an AI-2-regulated QS channel and QS master regulator proteins did not reduce the virulence of certain pathogens (Milton et al. 1997; Defoirdt et al. 2005; Li et al. 2010; Liu et al. 2012). Ó 2014 John Wiley & Sons Ltd

4

As can be seen from the results discussed above, QS-related mutants can exhibit similar or opposite virulence levels compared with wildtype strains. For example, although the disruption of a single channel (e.g., activated by LuxS in V. alginolyticus) effectively reduced virulence, the disruption of global regulators that control



Table 2 Challenge tests to determine the effect of QS mutations on the virulence of aquaculture pathogens Challenge model

Pathogens V. harveyi

A.hydrophila V. anguillarum

V. vulnificus

Gene mutation luxS, luxP, luxO luxM, luxN ahyI,ahyR vanI,vanR, vanM, vanN, vanQ, vanO smcR

Challenge method

Bacterial density

Brine shrimp Artemia franciscana

Bath challenge

104 CFU mL

INT-407 human intestinal epithelial cells Mice

Coculture

MOI ≤50

Intraperitoneal injection with 0.1 mL bacterial suspensions Intraperitoneal injection with 0.2 mL bacterial suspensions Coculture 5 h

101–106 CFU mL

1

104–107 CFU mL

1

Virulencea

Host

Abolished Not reduced Not reduced Not reduced

Reduced

V. alginolyticus

luxS

Reduced

Pagrus major

V. parahaemolyticus

opaR

Increased

Chinese hamster ovary cells

A. hydrophila

ahyR

Reduced

Epithelioma papillosum of carp (Cyprinus carpio) cells Xiphophophorus helleri Hecke

A. hydrophila

ahyI

Reduced

Epithelioma papillosum cyprini (EPC) cells Carassius auratus gibelio

A. hydrophila

ahyI, ahyR

Reduced

Burbot larvae

V. anguillarum

vanI, vanR

Not reduced

Rainbow trout (Oncorhynchus mykiss)

V. alginolyticus

luxT

V. ichthyoenteri

luxS

Slightly reduced Not reduced

Zebra fish Turbot (Scophthalmus maximus)

References 1

Lee et al. (2007)

1.5 9 106/mL, MOI 15

Coculture 30 min

MOI 1

Intraperitoneal injection with 0.1 mL bacterial suspensions Coculture 30 min

104–1010 CFU mL

Intraperitoneal injection with 0.1 mL bacterial suspensions Bath challenge

104–1010 CFU mL

Intraperitoneal injection with 0.1 mL bacterial suspensions Bath challenge 30 min Intramuscular injection Intraperitoneal injection

Defoirdt et al. (2005)

Gode-Potratz & McCarter (2011) Bi et al. (2007)

1

MOI 1

106 CFU mL

Ye et al. (2008)

Chu et al. (2011)

1

1

Natrah et al. (2012) Milton et al. (1997)

103–104/mL

103–2 9 104/mL 104–107 CFU per fish 105–108 CFU mL

Liu et al. (2012) 1

Li et al. (2010)

MOI, multiplicity of infection. a Compared with wide types.

multichannels (e.g., LuxT in V. alginolyticus) had little effect on virulence (Ye et al. 2008; Liu et al. 2012). In V. harveyi, the disruption of a single channel did not reduce virulence, whereas the disruption of global regulators abolished Ó 2014 John Wiley & Sons Ltd

5

virulence towards brine shrimp. There are several possible explanations for these contradictions. The first and most important explanation is that the coordination of multichannels in QS systems could lead to different results; furthermore,


different QS channels could promote the expression of different virulence-associated genes. This hypothesis is supported by recent studies. In V. harveyi, three types of signalling molecules/QS systems act synergistically and have different contributions to light production and the expression of virulence-associated genes. The intensity of the effect of three signalling molecules on light production in liquid medium is as follows: AHL>AI-2>CAI-1. The relative strength of these three signalling molecules changes in different environmental conditions (Henke & Bassler 2004b). In challenge tests with V. harveyi, AI-2 promoted Artemia survival more strongly than AHL (Defoirdt et al. 2005), which implies that AHL contributes more than AI-2 to the expression of virulence factors. Additionally, it has been suggested that virulence genes respond differently to different combinations of signalling molecules (Mok et al. 2003). This proposal was supported by Waters & Bassler (2006), who showed that V. harveyi responds differently to different signalling molecule input states. QS-controlled promoters can be divided into three types depending on the level of response to signalling molecules, or specifically, due to different binding affinities for LuxR: (i) AHL+AI-2≫AHL or AI-2 alone (little effect), (ii) AHL+AI-2>AHL or AI-2 alone (significant effect) and (iii) AHL+AI-2 have an effect similar to that AHL or AI-2 alone. In addition, the timing of the three inputs was also different. The three types of signalling molecules reached a critical concentration that induced sensor proteins to switch from kinase mode to phosphatase mode in the following order: CAI-1>AHL>AI-2 (Henke & Bassler 2004b). In conclusion, the intensity and timing of QS circuits of aquaculture pathogens may vary depending on the organism and environmental conditions, resulting in different gene expression profiles. Thus, effective QS interference should successfully curb all parallel QS circuits and operate continuously. Aquaculture pathogens may take full advantage of QS complexity and coordinate virulence gene expression dynamics during infection to improve their own survival. Aquaculture pathogens could sense signalling molecules from other micro-organisms in the vicinity, and signal interference from various populations in the environment could force a specific pathogen to gradually change its own gene expression pattern to adapt to its surroundings, further affecting the extent of infection (Jayaraman & Wood 2008). Ó 2014 John Wiley & Sons Ltd

6


There could be other possible explanations for the relationship between QS and the expression of virulence-associated genes. Certain regulatory proteins other than QS regulatory proteins may indirectly regulate QS and the expression of virulenceassociated genes in aquaculture pathogens. Virulence-associated factors might in turn regulate QS. Sha et al. (2005) found that Type III secretion systems (T3SSs) were positively correlated with QS in A. hydrophila. The authors showed that disruption of the T3SS aopB gene inhibited signalling molecule production. A recent study also found that Type VI secretion systems (T6SSs) regulated QS in V. anguillarum. T6SS positively regulated the expression of RpoS and VanT, which further regulated the expression of extracellular proteases (EmpA and PrtV) and a haemolysin coregulated-like protein (Hcp) (Weber et al. 2009). Certain regulatory proteins may work together with core elements of QS systems and induce virulence indirectly (Fig. 1). Examples of this type of regulatory proteins are V. anguillarum RpoS and V. alginolyticus Hfq (Weber et al. 2008, 2009; Liu et al. 2011). In this case, certain QSrelated mutants may not be effective at curbing virulence. Interestingly, Liu et al. (2011) found that Hfq negatively regulated LuxR of QS system in V. alginolyticus, which indicated the complicated relationship between QS regulatory proteins. Further research is needed to explore complex signal transduction networks so that effective QS disruption could be carried out. Recently developed

Figure 1 Model of virulence induction by core elements of a QS system and other regulatory proteins.


techniques such as transcriptomics and proteomics will be helpful in elucidating the entire QS network in different pathogens (Dagkessamanskaia et al. 2004; Di Cagno et al. 2011). Finally, the differences in virulence reported in QS-related mutants could be due to the choice of challenge models, including hosts, challenge methods, bacterial density and the species of the aquaculture pathogens (Table 2). It is interesting to note that the same QS-related mutations in A. hydrophila led to different results in different virulence tests. The virulence was not reduced towards brine shrimp in a bath challenge but was significantly reduced towards several other aquaculture animals or cells either by intraperitoneal injection, coculture or bath challenge (Defoirdt et al. 2005; Bi, Liu & Lu 2007; Chu et al. 2011; Natrah et al. 2012). This indicates that there could be differences in host resistance. The effect of QS on virulence might also depend on the life stage of the host; however, this is rarely mentioned. In several other notable pathogens, including V. anguillarum, V. alginolyticus and V. ichthyoenteri, the virulence of QS-related mutants was only slightly reduced (Milton et al. 1997; Defoirdt et al. 2005; Li et al. 2010; Liu et al. 2012). This was possibly due to the limited choice of challenge models, and additional experiments using different challenge models could further indicate whether QS has any influence on virulence. Additionally, the presence of limited QS mutants might be the reason why QS appears to be ineffective at inducing virulence. This is consistent with the above two interpretations: multichannels in QS systems may function together to induce virulence, or other regulatory proteins in addition to QS could effectively induce virulence. Strategies for disrupting QS in aquaculture pathogens: are these valid approaches for combatting infection?

To date, research has shown that QS regulates virulence factors, T3SSs, T6SSs and biofilm formation, all of which are involved in bacterial pathogenicity. Thus, QS interference may be an effective method for inhibiting the virulence of pathogens, further reducing disease in aquaculture systems and increasing productivity. Quorum quenching is performed primarily by three mechanisms (Yates et al. 2002; Defoirdt Ó 2014 John Wiley & Sons Ltd

7


et al. 2004; Ren et al. 2008). The first method involves blocking the synthesis of signalling molecules. The biosynthetic pathways for AHL and AI-2 have been elucidated. AHL is produced by the catalysis of AHL synthase using acyl-ACP and S-adenosyl methionine (SAM) as substrates (Zhang 2003). The synthetic route of AI-2 is as follows: as an intermediate of S-adenosyl methionine (SAM), S-adenosylhomocysteine (SAH) is catalysed by S-adenosylhomocysteine nucleosidase to produce S-ribosylhomocysteine (SRH). Next, 4,5-dihydroxy 2,3-pentanedione (DPD) is produced from SRH with LuxS as a catalyst. DPD is unstable and finally forms AI-2 (Schauder et al. 2001). Inhibition of substrate formation or enzyme activity, either by the reconstruction of metabolic pathways or by the exogenous addition of substrate analogues as QS inhibitors (QSIs), could block the biosynthesis of signalling molecules (Li et al. 2003; Alfaro et al. 2004). The second method involves deactivation or degradation of signalling molecules. Signalling molecules can be degraded three ways. First, chemical degradation can be induced by increasing the pH to alkaline conditions or increasing the temperature. Second, signalling molecules can be degraded through enzymatic degradation. AHL-degrading enzymes (mainly acylases, lactonases and oxidoreductases) have been widely identified in prokaryotes and have a wide substrate spectrum (Dong & Zhang 2005; Czajkowski & Jafra 2009). Third, micro-organisms can metabolize signalling molecules. For example, Variovorax paradoxus can use OHHL as a sole source of energy and nitrogen (Leadbetter & Greenberg 2000). The third method of quorum quenching involves impeding the binding of signalling molecules to their receptors by competitive inhibition or by decreasing the DNA-binding activity of the receptor, thus affecting downstream gene expression (Zhang & Dong 2004). However, the effectiveness of the above three strategies for quorum quenching in controlling disease in aquaculture systems has not been demonstrated. In the following section, we summarize the four types of QSIs targeting aquaculture pathogens, based on the quorum-quenching pathways described above. The names and structures of QSIs that have been shown to either decrease production of QS-regulated pathogen virulence factors or to protect aquaculture animals from infection are shown in Fig. 2.


Figure 2 Structures of QSI compounds targeting aquaculture pathogens.

Ó 2014 John Wiley & Sons Ltd

8



Figure 2 Continued.

Ó 2014 John Wiley & Sons Ltd

9



QSIs that bind to signalling molecule receptor proteins in aquaculture pathogens Chemical analogues of signalling molecules compete for receptor protein binding sites because of their structural similarities. Recently, a series of QSIs, most of which are chemical analogues of AHL and AI-2, have been shown to be effective. Swift et al. found that the exogenous addition of 10 lM ODHL effectively delayed the appearance of serine protease and reduced its production in A. salmonicida (Swift et al. 1997). A potent QSI, N-(heptylsulphanylacetyl)-L-homoserine lactone (HepS-AHL), competitively binds to LuxR and LasR and interferes with QS in V. fischeri and P. aeruginosa. Dozens of other QSIs, including compounds from garlic extract, homoserine lactone (HSL) derivatives and acetamide, positively inhibit LuxR (Persson et al. 2005). N-(propylsulphanylacetyl)-L-homoserine lactone (ProSAHL), N-(pentylsulphanylacetyl)-L-homoserine lactone (PenS-AHL) and HepS-AHL effectively reduce the expression of toxic proteases (e.g., AsaP1) by A. salmonicida at a dose of 10 lM (Rasch et al. 2007; Schwenteita et al. 2011). Natrah et al. (2012) found that N-tetradecanoyl-L-homoserine lactone effectively decreased the virulence of both A. hydrophila and A. salmonicida towards burbot larvae at a dose of 10 lM. However, Kastbjerg et al. (2007) found that three sulphur-containing AHL analogues did not inhibit protease production in Y. ruckeri. The authors hypothesized that protease production was either not regulated by QS or not regulated by the AHL-mediated QS system. AI-2-mediated QS can be blocked in a similar manner. QSIs were screened from nucleoside analogues, and an active 3-(methoxyphenylpropionamido) ribofuranosyl derivative was obtained that most likely interfered with signal transduction at the level of LuxPQ in V. harveyi. This QSI decreases protease activity in V. anguillarum by 23 3% and decreased biofilm formation in V. anguillarum and V. vulnificus by 35 11% and 17 15%, respectively. This QSI successfully protects Artemia shrimp against V. harveyi at a dose of 40 lM with no effect on the survival of the shrimp or V. harveyi. Pyrogallol and 4-methoxycarbonyl-phenylboronic acid (MCPBA, an AI-2 homologue) both decrease protease activity and biofilm formation in V. anguillarum and V. vulnificus at a dose of 40 lM, possibly because Ó 2014 John Wiley & Sons Ltd

10


the two QSIs bind to a LuxP homologue (Ni et al. 2008a; Brackman et al. 2009). Several phenylboronic acids also inhibit AI-2. However, these QS antagonists are only effective against V. harveyi luminescence at an IC50 of 3–14 lM (Ni et al. 2008b). These compounds require further study. QSIs that decrease the DNA-binding activity of signalling molecule receptor proteins in aquaculture pathogens Cinnamaldehyde and its derivatives are potentially effective QSIs, as they decrease the binding of LuxR to promoter sequences. Treatment with cinnamaldehyde at concentrations of 100–150 lM decreases protease activity in V. anguillarum and biofilm formation in V. anguillarum and V. vulnificus. Cinnamaldehyde and 2-NO2-cinnamaldehyde completely protect Artemia against V. harveyi without affecting its growth. Treatment with 4-methoxy-cinnamaldehyde also decreases protease activity and biofilm formation in V. anguillarum (Brackman et al. 2008). As Natrah et al. (2012) reported, cinnamaldehyde effectively decreases the virulence of A. hydrophila and A. salmonicida towards burbot larvae at a dose of 0.01 lM. Furanone derivatives, which are structurally similar to AHL molecules, are potential QSIs that function similarly to cinnamaldehyde. Treatment with (5Z)-4-Bromo-5-(bromomethylene) -3-butyl2(5H)-furanone (100 lM) from the alga Dellsea pulchra inhibits toxin production in V. harveyi and protects prawns from V. harveyi infection (Manefield et al. 2000). Four furanone derivatives (50 lM) reduce the LuxR concentration, further decreasing AHL concentration and the expression of downstream genes (Manefield et al. 2002). The QSI furanone C-30 is effective against V. anguillarum. An extremely small amount (0.01 or 0.1 lM) of furanone C-30 significantly reduces mortality in rainbow trout infected by V. anguillarum without affecting growth or survival of the pathogen. However, a higher dose of furanone C-30 was toxic to rainbow trout (Rasch et al. 2004). Recently, it was shown that the natural (5Z)-4-bromo-5-(bromomethylene)-3-butyl-2(5H)furanone blocks all three channels of the QS system in V. harveyi. By decreasing the DNAbinding activity of the QS transcriptional regulator LuxRVh, the furanone inhibits the expression of QS-regulated genes (Defoirdt et al. 2007).


Recently, a new type of compound, brominated thiophenones, was found to play the same role as furanones in QS disruption. The thiophenone compound TF310, (Z)-4-((5-(bromomethylene)2-oxo-2,5-dihydrothiophen-3-yl) methoxy)-4-oxobutanoic acid, completely protects brine shrimp larvae from V. harveyi at a dose of 2.5 lM. The ratio of the toxic and therapeutic concentrations of this compound was 40, which indicates a relatively low toxicity compared with furanone compounds (Defoirdt et al. 2012). QSIs that interact with AI-2 synthase in aquaculture pathogens Recently, Zhang et al. (2009) identified two small peptides with a sequence similar to the C site of LuxS. These peptides, or genetic strains expressing the peptides, could be promising antimicrobial agents based on QS interference. The peptides interact specifically with LuxS and inhibit AI-2 activity. Consequently, the peptides curb biofilm growth, decrease the expression of the T3SS genes esrA and orf26 and inhibit the dissemination and survival of E. tarda in infected fish. This discovery provides new strategies for designing QSIs against aquaculture pathogens. Similarly, halogenated furanones covalently modify and inactivate LuxS (Zang et al. 2009). This has been shown in several different microbes. Furanone (60–100 lg mL 1) inhibits AI-2 signalling in Escherichia coli and leads to the differential expression of many genes (Ren et al. 2004). Synthetic bromated furanone inhibits Streptococci biofilm formation via the AI-2-mediated QS pathway (Lonn-Stensrud et al. 2007). Halogenated furanone at doses

Acne, quorum sensing and danger.

Quorum sensing dependent phenotypes and their molecular mechanisms in Campylobacterales.

Quorum sensing in extreme environments.

Interfering with Bacterial Quorum Sensing.

Size of quorum sensing communities.

Optimal census by quorum sensing.

Quorum sensing: a quantum perspective.

Bacterial Quorum Sensing Stabilizes Cooperation by Optimizing Growth Strategies.

Quorum sensing control of Type VI secretion factors restricts the proliferation of quorum-sensing mutants.

Quorum Sensing Communication Modules for Microbial Consortia.

Bacterial quorum sensing: circuits and applications.

Quorum sensing activity in Pandoraea pnomenusa RB38.

Redox-mediated quorum sensing in plants.

Anti-quorum sensing potential of Adenanthera pavonina.

Quorum sensing inhibition, relevance to periodontics.

Where antibiotic resistance mutations meet quorum-sensing.

Why Quorum Sensing Controls Private Goods.

Corrigendum: Why Quorum Sensing Controls Private Goods.

Quorum quenching is responsible for the underestimated quorum sensing effects in biological wastewater treatment reactors.

Chemical composition and disruption of quorum sensing signaling in geographically diverse United States propolis.

Exploiting Quorum Sensing Interfering Strategies in Gram-Negative Bacteria for the Enhancement of Environmental Applications.

Social Evolution Selects for Redundancy in Bacterial Quorum Sensing.

A crucial role for spatial distribution in bacterial quorum sensing.

Quorum sensing and self-quorum quenching in the intracellular pathogen Brucellamelitensis.