Environmental Microbiology Volume 13 issue 1 2011 [doi 10.1111%2Fj.1462-2920.2010.02308.x] Jin-Hyung Lee; Moo Hwan Cho; Jintae Lee -- 3-Indolylacetonitrile Decreases Escherichia coli

3-Indolylacetonitrile Decreases Escherichia coliO157:H7 Biofilm Formation and Pseudomonasaeruginosa Virulenceemi_2308 62..73

Jin-Hyung Lee, Moo Hwan Cho and Jintae Lee*School of Display and Chemical Engineering,Yeungnam University, Gyeongsan-si,Gyeongsangbuk-do 712749, Korea.

Summary

Intercellular signal indole and its derivative hydroxy-indoles inhibit Escherichia coli biofilm and diminishPseudomonas aeruginosa virulence. However, indoleand bacterial indole derivatives are unstable in themicrobial community because they are quicklydegraded by diverse bacterial oxygenases. Hence,this work sought to identify novel, non-toxic, stableand potent indole derivatives from plant sources forinhibiting the biofilm formation of E. coli O157:H7 andP. aeruginosa. Here, plant auxin 3-indolylacetonitrile(IAN) was found to inhibit the biofilm formation ofboth E. coli O157:H7 and P. aeruginosa without affect-ing its growth. IAN more effectively inhibited biofilmsthan indole for the two pathogenic bacteria. Addition-ally, IAN decreased the production of virulencefactors including 2-heptyl-3-hydroxy-4(1H)-quinolone(PQS), pyocyanin and pyoverdine in P. aeruginosa.DNA microarray analysis indicated that IAN repressedgenes involved in curli formation and glycerolmetabolism, whereas IAN induced indole-relatedgenes and prophage genes in E. coli O157:H7. Itappeared that IAN inhibited the biofilm formation of E.coli by reducing curli formation and inducing indoleproduction. Also, corroborating phenotypic resultsof P. aeruginosa, whole-transcriptomic data showedthat IAN repressed virulence-related genes andmotility-related genes, while IAN induced severalsmall molecule transport genes. Furthermore, unlikebacterial indole derivatives, plant-originated IAN wasstable in the presence of either E. coli or P. aerugi-nosa. Additionally, indole-3-carboxyaldehyde wasanother natural biofilm inhibitor for both E. coli andP. aeruginosa.

Introduction

Anti-virulence compounds have been suggested as alter-native ways to fight infectious diseases because unlikeantimicrobials, anti-virulence compounds do not affectgrowth and so there is less chance of developing resis-tance (Hentzer et al., 2002; Lesic et al., 2007). Forexample, natural brominated furanones (produced by thered macroalga Delisea pulchra) and their synthetic deriva-tives inhibit the biofilm formation of Escherichia coli (Renet al., 2001) and P. aeruginosa (Hentzer et al., 2003) byinterrupting their quorum sensing (Hentzer et al., 2003;Ren et al., 2004). A quorum-sensing kinase inhibitordecreased the virulence of E. coli O157:H7 without affect-ing its growth (Rasko et al., 2008). Hence, inhibiting thecell signalling is an important way to fight infectious dis-eases, and it is necessary to discover novel compoundsthat inhibit the biofilm formation and the virulence factorproduction of pathogenic bacteria.

Indole is an intercellular signal (Lee and Lee, 2010). Avariety of both Gram-positive and Gram-negative bacteria(Lee and Lee, 2010) produce indole using tryptophanase(TnaA; EC 4.1.99.1) that can reversibly convert tryp-tophan into indole, pyruvate and ammonia (Newton andSnell, 1965). Indole plays diverse biological roles in E. coliand other bacteria. For example, indole controls thevirulence of pathogenic E. coli (Anyanful et al., 2005;Hirakawa et al., 2009) as well as P. aeruginosa, whichcannot produce indole (Lee et al., 2009), and indole con-trols biofilm formation in E. coli (Di Martino et al., 2003;Lee et al., 2007a) and in V. cholerae (Mueller et al., 2009).Additionally, humans maintain a symbiotic relationshipwith indole-producing enteric bacteria over a long periodof time. Hence, recent studies have suggested that indoleand the derivatives that are derived from bacterial indoleplay beneficial roles in the human immune system (Wikoffet al., 2009; Bansal et al., 2010).

In contrast to indole-producing bacteria, many non-indole-producing bacteria, plants and animals producediverse oxygenases, which can oxidize indole andproduce indole derivatives (Ensley et al., 1983; Lee andLee, 2010). Hence, six hydroxyindoles and three indigoidcompounds were previously investigated as biofilminhibitors for E. coli O157:H7 (Lee et al., 2007b) and

Received 20 January, 2010; accepted 11 June, 2010. *For corre-sponding. E-mail [email protected]; Tel. (+82) 53 810 2533; Fax (+82)53 810 4631.

Environmental Microbiology (2011) 13(1), 6273 doi:10.1111/j.1462-2920.2010.02308.x

2010 Society for Applied Microbiology and Blackwell Publishing Ltd

mailto:[email protected]

anti-virulence compounds for P. aeruginosa (Lee et al.,2009). Although the previous results provided a startingpoint for anti-virulence drugs, indole and the bacterialindole derivatives are rapidly degraded by many patho-genic bacteria, such as P. aeruginosa, and increased thebiofilm formation of P. aeruginosa (Lee et al., 2009).Therefore, the use of bacterial indole derivatives is limitedagainst some pathogenic bacteria that have developeddefence systems against these derivatives.

Plants have developed advanced defence mechanismsin order to survive in their ecosystems, and hence, plantsecondary metabolites are major pharmaceutical sources(Zhao et al., 2005; Li and Vederas, 2009). Among thesecompounds, several indole derivatives have shown avariety of biological functions with respect to bacteriaand humans. For example, indole-3-carbinol and 3,3-diindolylmethane (Fig. 1A) that originate from cruciferousvegetables, exhibit antimicrobial, antiviral and anticanceractivities (Higdon et al., 2007; Fan et al., 2009). Likeindole-3-acetic acid, 3-indolylacetonitrile (IAN, Fig. 1A) isa naturally occurring plant growth hormone (auxin) (Joneset al., 1952) that is synthesized from tryptophan in crucif-erous vegetables (Brassica), such as broccoli, cauliflowerand cabbage (Kutcek et al., 1960). Additionally, indole-3-propionic acid (Fig. 1A), which is found in animal blood(Wikoff et al., 2009), is a powerful antioxidant against theAlzheimer b-amyloid (Chyan et al., 1999).

The goal of this study was to identify novel, non-toxic,stable and potent indole derivatives that inhibited thebiofilm formation of E. coli O157:H7 and P. aeruginosa.Five indole derivatives from cruciferous vegetables andindole-3-propionic acid were investigated as biofilm inhibi-tors for E. coli O157:H7 and P. aeruginosa and as anti-virulence compounds for P. aeruginosa. In order tounderstand the molecular basis of the biofilm inhibition ofIAN (the most effective E. coli biofilm inhibitor among thetested chemicals), DNA microarrays, phenotypic assaysand scanning electron microscopy (SEM) were utilized.This is the first report to use IAN in order to inhibit E. coliO157:H7 biofilm formation and to reduce P. aeruginosavirulence.

Results

The main goals of this research were to determinewhether the indole derivatives from plant sources influ-enced E. coli biofilm formation and P. aeruginosa viru-lence and to determine the genetic mechanism usingDNA microarrays and phenotypic assays.

3-Indolylacetonitrile and indole-3-carboxyaldehydeare potent biofilm inhibitors of E. coli O157:H7and P. aeruginosa

Five indole derivatives derived from plant sources andindole-3-propioninc acid found in animal sources were

screened for their ability to inhibit the biofilm formation ofE. coli O157:H7 in LuriaBertani (LB) medium in 96-wellplates for 24 h at 37C. Indole 3-acetic acid (IAA), 3,3-methylene bisindole (MI), indole-3-propioninc acid (I3PA),indole-3-carbinol (I3C), indole-3-carboxyaldehyde (I3CA)and IAN were dissolved in dimethylsulfoxide (DMSO) andscreened at a concentration of 100 mg ml-1. Dimethylsul-foxide [0.1% (v/v)] and indole (100 mg ml-1) were used asthe negative and positive controls respectively. Dimethyl-sulfoxide [0.1% (v/v)] alone did not affect cell growth,biofilm formation, and virulence production in both E. coliO157:H7 and P. aeruginosa. In the absence of the indolederivatives, E. coli O157:H7 formed robust biofilms(Fig. 1). I3CA and IAN most effectively reduced E. coli

Bio

film

fo

rmat

ion

(O

D57

0/62

0 )

0.0

0.5

1.0

1.5

2.0

2.5

3.0

NH

NH

None Indole 7HI IAA MI I3PA I3C I3CA IAN

NH

HO

NH

OHNH

OHO

IAN (mg ml-1)

0 25 50 100 150

Bio

film

fo

rmat

ion

(O

D57

0/60

0)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

A

B

NH

OH

O

NH

NH

OH

E.coli O157:H7

E.coli O157:H7

NH

CN

Fig. 1. A. Effect of indole derivatives [indole, 7-hydroxyindole (7HI),3-indoleacetic acid (IAA), 3,3-methylene bisindole (MI),indole-3-propioninc acid (I3PA), indole-3-carbinol (I3C),indole-3-carboxyaldehyde (I3CA) and 3-indolylacetonitrile (IAN)] at100 mg ml-1 on normalized (OD570/OD620) biofilm formation of E. coliO157:H7 in LB medium at 37C after 24 h in 96-well plates. All ofthe indole derivatives were dissolved in DMSO. DMSO and indolewere used as the negative and positive controls respectively. Thestructures of indole, 7HI, IAA, MI, I3PA, I3C, I3CA and IAN areshown.B. Dose-dependent effect of IAN (0, 25, 50, 100 and 150 mg ml-1)on E. coli O157:H7 biofilm formation. At least two independentexperiments were conducted (total 12 wells), and the error barsindicate one standard deviation.

3-Indolylacetonitrile inhibits E. coli O157:H7 biofilm 63

2010 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 13, 6273

O157:H7 biofilm formation (11-fold for I3CA and 24-foldfor IAN), whereas IAA did not greatly affect E. coliO157:H7 biofilm formation (Fig. 1A). Under the same con-ditions, indole inhibited E. coli O157:H7 biofilm formationby threefold. Therefore, IAN was an eight times moreeffective biofilm inhibitor than indole for E. coli O157:H7 at37C. Also, the addition of IAN (100 mg ml-1) was still morepotent for E. coli O157:H7 biofilm reduction than that(0.125 0.05 OD570/620) of the higher concentration ofindole (200 mg ml-1). Furthermore, IAN decreased E. coliO157:H7 biofilm in a dose-dependent manner from 0 to150 mg ml-1 (Fig. 1B).

To investigate if IAN and I3CA are biofilm inhibitors forpseudomonads, too, IAN and I3CA were tested with P.aeruginosa PAO1. IAN and I3CA caused a significantdecrease in the biofilm formation for P. aeruginosa (2.3-foldfor IAN and 1.9-fold for I3CA, Fig. 2), whereas indole and7-hydroxyindole stimulated P. aeruginosa biofilm formation(Lee et al., 2009). Therefore, IAN and I3CA were biofilminhibitors for both E. coli O157:H7 and P. aeruginosa.

To confirm the lack of metabolism of IAN in E. coliO157:H7 and P. aeruginosa, the extracellular IAN concen-tration was measured in the supernatant of cell cultureusing HPLC. The IAN concentration (110 mg ml-1 for E. coliO157:H7 and 70 mg ml-1 for P. aeruginosa) was stablymaintained over 24 h without any degradation or evapora-tion in the presence of E. coli O157:H7 and P. aeruginosa(Fig. 3A). This result suggests that IAN could not bemetabolized and transported into E. coli O157:H7 and P.aeruginosa. It is notable that neither pathogenic bacteriacould metabolize IAN, whereas P. aeruginosa rapidlydegraded indole and 7-hydroxyindole (Lee et al., 2009).Hence, IAN from plants was a relatively stable and potentbiofilm inhibitor for both E. coli O157:H7 and P. aeruginosa.

Toxicity

To test the toxicity of indole derivatives, the specificgrowth rate was measured in LB medium. The growth rateof E. coli O157:H7 was 1.32 0.03 h-1 in the absenceof the indole derivatives, whereas the growth ratewas 1.28 0.01 h-1 with 100 mg ml-1 IAN (Fig. 3B)and 1.08 0.02 h-1 with 100 mg ml-1 indole-3-carboxyaldehyde (I3CA). The growth rate of P. aeruginosawas 1.45 0.12 h-1 in the absence of the indole deriva-tives, whereas the growth rate was 1.38 0.13 h-1 with100 mg ml-1 IAN (Fig. 3C) and 1.42 0.10 h-1 with100 mg ml-1 I3CA. For IAN, a concentration as high as150 mg ml-1 did not affect the cell growth of E. coliO157:H7 (1.29 0.08 h-1). Therefore, IAN was not toxicto E. coli O157:H7 and P. aeruginosa.

Differential gene expression of E. coli O157:H7cells with IAN

To investigate the global genetic basis of IAN regulationof E. coli O157:H7 biofilm reduction, DNA microarrayswere first used to determine differential gene expression

Bio

film

fo

rmat

ion

(O

D57

0/62

0 )

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

None Indole 7HI IAA I3CA IAN

Fig. 2. Effect of indole derivatives [indole, 7-hydroxyindole (7HI),3-indoleacetic acid (IAA), indole-3-carboxyaldehyde (I3CA) and3-indolylacetonitrile (IAN)] at 100 mg ml-1 on normalized(OD570/OD620) biofilm formation of P. aeruginosa PAO1 in LBmedium at 37C after 7 h in 96-well plates. Indole and the indolederivatives were dissolved in DMSO. DMSO and indole were usedas the negative and positive controls respectively. At least twoindependent experiments were conducted (total 12 wells), and theerror bars indicate one standard deviation.

Culture time (h)0 5 10 15 20 25

IAN

(mg

ml-1

)

0

20

40

60

80

100

120

140

E. coli O157:H7

P. aeruginosa PAO1

A

B E. coli O157:H7 C P. aeruginosa PAO1

Culture time (h)0 4 8 12

Cel

l gro

wth

(O

D60

0)

0123456

Culture time (h)0 4 8 12

Cel

l gro

wth

(O

D60

0)

0

1

2

3

4

5

None IAN None IAN

Fig. 3. Stability of 3-indolylacetonitrile (IAN) in the presence ofE. coli O157:H7 and P. aeruginosa PAO1 (A), and cell growth ofE. coli O157:H7 (B) and P. aeruginosa PAO1 (C) in the presenceof IAN. The initial turbidity of cells was 0.05 at 600 nm, and thecells were cultured at 37C and 250 r.p.m. in LB medium. Eachexperiment was performed using two independent cultures, andone representative data set is shown.

64 J.-H. Lee, M. H. Cho and J. Lee


for E. coli O157:H7 cells with and without IAN(100 mg ml-1). It was found that 25 genes were signifi-cantly regulated (more than twofold); 8 genes wereinduced and 17 were repressed by IAN (Table 1). Mostnoticeably, two curli genes (csgA and csgB), a fimbriagene (z2200) and glycerol-related genes (glpD, glpF,

glpK and glpT) were most repressed (2.0- to 3.6-fold),while two prophage genes (z2978 and z3345) weremost induced (3.0- to 3.7-fold) (Table 1). Additionally,tnaC, which encoded the leader peptide of tna operon,and TrpR binding protein WrbA were also induced. Theimpact of IAN was probably a chronic process ratherthan an acute process because only 25 genes were dif-ferentially (above twofold) changed, which was a smallnumber compared with the biofilm reduction (Fig. 1). Thefull data are available using GEO accession numberGSE19842.

IAN reduces the curli formation in E. coli O157:H7

Curli formation is important for the biofilm formation of E.coli O157:H7 (Ryu and Beuchat, 2005; Uhlich et al.,2006). Since the microarray data showed that two curligenes (csgA and csgB) were most repressed by IAN(Table 1), curli production was measured using a Congored plate and SEM. The Congo red plate was used toobserve curli production of colonies that were grownunder static conditions, whereas SEM were used toobserve curli production from the planktonic cells thatwere used for the DNA microarrays and biofilm cellsgrown in a nylon filter.

Clearly, IAN decreased the curli production on theCongo red plate (Fig. 4A) and in the SEM analysis forplanktonic cells (Fig. 4B) and biofilm cells (Fig. 4C).Notably, curli formation in the absence of IAN was moredistinctive in biofilm cells than planktonic cells. The resultcan be explained by the previous results that curli produc-tion is regulated via the level of c-di-GMP (Weber et al.,2006) and that a high level of c-di-GMP results in higherbiofilm formation in E. coli (Mndez-Ortiz et al., 2006).Since Congo red binds both curli and cellulose (Zogajet al., 2003), a cellulose-specific assay using calcofluorwas also used to investigate the cellulose productionupon the addition of IAN. Unlike the curli productionwith colonies, the cellulose production with planktoniccells slightly increased upon the addition of IAN(1.6 0.1-fold). Therefore, these results indicated thatthe reduction of Congo red binding was caused by adecrease in the curli production (Fig. 4A).

Under SEM, polymerized curli appeared as 4- to7-nm-wide fibres of varying lengths, and a large amountof curli appeared as a tangled and amorphous matrixsurrounding bacterial cells (Chapman et al., 2002). IANclearly decreased the curli production from planktoniccells (Fig. 4B). The results were in good agreementwith the microarray data (Table 1). Therefore, IAN inhib-ited curli production, which further decreased E. coliO157:H7 biofilm formation. Additionally, IAN alsoreduced the production of polymeric matrix in P. aerugi-nosa (Fig. 4D).

Table 1. Partial list of the genes that were repressed and inducedfrom E. coli O157:H7 cells upon addition of 100 mg ml-13-indolylacetonitrile (IAN).

Gene nameFoldchange Description

Curli genescsgB -3.6 Curlin minor subunitcsgA -2.6 Cryptic curlin major subunitz2200 -2.2 Type-1 fimbrial protein, A

chain precursor

Glycerol-related genesglpD -2.0 Glycerol-3-phosphate

dehydrogenaseglpF -2.3 Glycerol facilitator proteinglpT -2.3 Sn-glycerol-3-phosphate

transporterglpK -3.2 Glycerol kinase

Prophage genesz2978 3.7 Putative replication protein

for prophagez3345 3.0 Antitermination protein Q for

prophage

Indole-related genestnaC 2.3 Tryptophanase leader

peptidewrbA 2.7 TrpR binding protein WrbA

OthersflaB 2.4 Fructose-bisphosphate

aldolasez2254 2.4 H-repeat-containing Rhs

elemententE 2.1 Enterobactin synthase

subunit EmarR 2.0 Repressor of multiple

antibiotic resistanceynfK -2.0 Predicted dethiobiotin

synthetaseafuA -2.0 Periplasmic ferric

iron-binding proteinfruB -2.0 Bifunctional fructose-specific

PTS IIA/HPr proteinyhcI -2.0 N-acetylmannosamine

kinasefocA -2.0 Formate transporterilvL -2.1 ilvG operon leader peptidenapD -2.1 Assembly protein for

periplasmic nitratereductase

uhpT -2.2 Sugar phosphate antiporternarK -3.1 Nitrite extrusion proteinyfiD -3.5 Autonomous glycyl radical

cofactor GrcAz3921 2.9 Hypothetical proteinz2148 2.2 Hypothetical protein

The cells were grown until the absorbance of 1.0 in LB medium at37C at 250 r.p.m. with and without IAN. The full data are availableusing GEO accession number GSE19842.



IAN increases indole production in E. coli O157:H7

Indole decreases the biofilm formation of E. coliO157:H7 (Bansal et al., 2007; Lee et al., 2007b). Thewhole-transcriptome analysis here showed that IANinduced tnaC (Table 1), which positively regulates theindole production (Gong and Yanofsky, 2002; Lee et al.,2007a). Hence, the extracellular indole concentrationswere measured in the presence of IAA, IAN and I3CA.Confirming the microarray data, IAN produced twofoldmore extracellular indole than the control, whereas IAAdid not significantly affect indole production (Fig. 5A).These results were distinctive because isatin (doubleoxidized indole) decreased indole production andincreased biofilm formation (Lee et al., 2007b). Hence,the results partially explained the biofilm reduction of E.coli O157:H7 for IAN.

Quantitative real-time reverse transcription polymerasechain reaction (qRT-PCR) was used to investigate thegene expression for csgA (encoding curlin major subunit)and tnaA (encoding indole-synthesizing tryptophanase)with and without IAN. The qRT-PCR result was compa-rable to the DNA microarray data in expression of csgAwith IAN (-4.6 0.7-fold in qRT-PCR versus -2.6-fold inmicroarrays at a cell absorbance of 1.0). Additionally, theqRT-PCR results showed 3.4 2.1-fold induction of tnaAwith IAN at a cell absorbance of 2.0, which corroboratedthe increased indole production with IAN.

Differential gene expression of P. aeruginosa cellswith IAN

To investigate the global genetic basis of IAN regulation ofP. aeruginosa biofilm reduction, DNA microarrays wereused to determine differential gene expression for P.aeruginosa cells with and without IAN (100 mg ml-1). Itwas found that 50 genes were significantly regulated(more than 1.7-fold); 29 genes were induced and 21 wererepressed by IAN (Table 2). Most noticeably, twovirulence-related genes (pqsE and pvcC), a fimbrial gene(z2200) and three motility-related genes (type IV twitchingmotility protein pilI, flagellar biosynthesis protein flhF andflagellar motor protein motD) were most repressed, while11 small molecule transport genes were most induced(Table 2). The full data are available using GEO accessionnumber GSE21508.

Flagellar-mediated motility and type IV pili protein areconditionally necessary for P. aeruginosa biofilm develop-ment (OToole and Kolter, 1998; Klausen et al., 2003).Hence, IAN probably reduces P. aeruginosa biofilm devel-opment partially by repressing these motility-relatedgenes (pilI, flhF and motD) (Table 2). Also, the PQSsystem plays a role in P. aeruginosa biofilm formationunder diverse conditions (Diggle et al., 2003; Yang et al.,2009). Recently, another trasncriptomic analysis revealedthat PqsE is required for swarming motility, biofilm devel-opment and virulence, and PqsE is a key regulator infacilitating the environmental adaptation of P. aeruginosato plant and animal hosts (Rampioni et al., 2010). Sinceplant auxin IAN most repressed pqsE (Table 2), our datasupport the role of PqsE in controlling the adaptive behav-iour of P. aeruginosa to plant.

qRT-PCR was also used to investigate the geneexpression for pqsE, pvcC and pilI with and without IANtreatment in P. aeruginosa. The qRT-PCR result wascomparable to the DNA microarray data in expression ofpqsE (-2.9 0.5-fold in qRT-PCR versus -2.1-fold inmicroarrays), pvcC (-1.5 0.2-fold in qRT-PCR versus-1.8-fold in microarrays), and pilI (-3.6 0.6-foldin qRT-PCR versus -1.8-fold in microarrays) with IANtreatment.

Fig. 4. Effect of 3-indolylacetonitrile (IAN) on polymeric matrixproduction in E. coli O157:H7 and P. aeruginosa PAO1. Curliproduction in E. coli O157:H7 observed by (A) a Congo red plateafter 24 h, (B) SEM for planktonic cells and (C) SEM for biofilmcells grown in a nylon filter. Polymeric matrix production in P.aeruginosa PAO1 was also observed by SEM for biofilm cellsgrown in a nylon filter (D). For the SEM analysis, planktonic cellswere grown in LB medium at a cell absorbance of 1.0 and weredirectly fixed through the addition of glutaraldehyde andformaldehyde and filtered with a 0.45 mm Nylon filter. Biofilm cellswere cultured on nylon filter in 96-well plate at 37C for 24 h.



IAN reduces virulence factor production in P. aeruginosa

Since the microarray data showed that IAN repressed twovirulence-related genes (pqsE and pvcC) (Table 2), weassayed the production of four virulence factors, pyocya-nin, PQS, siderophore pyoverdine and rhamnolipid, in P.aeruginosa for 24 h. Most distinctively, indole decreasedthe pyocyanin production by 29-fold, which was similarto the previous results (Lee et al., 2009), and IANdecreased the pyocyanin production by sevenfold,whereas the same concentration of plant auxin IAA (as astructural control) did not affect the pyocyanin production(Fig. 5B). Similarly, indole decreased the PQS productionsignificantly, which was similar to the previous results (Leeet al., 2009), and IAN decreased the pyocyanin produc-tion by threefold (Fig. 5C). Hence, IAN exhibited a smallerimpact than indole on the production of pyocyanine andPQS in P. aeruginosa. However, IAN decreased morepyoverdine production than that of indole (Fig. 5D). Addi-tionally, both indole and IAN did not significantly affectrhamnolipid production after 24 h growth (data notshown).

Discussion

In this study, we demonstrated that IAN inhibited thebiofilm formation of both E. coli O157:H7 and P. aerugi-nosa PAO1 and reduced the production of virulencefactors from P. aeruginosa without affecting their growth.We have sought the genetic mechanisms of IAN in both E.coli O157:H7 and P. aeruginosa using the transcriptomicassays and the phenotypic assays.

Cruciferous (or Brassica) vegetables, includingbroccoli, cauliflower, brussels and cabbage, are richsources of glucosinolates including indole derivatives,and a high intake of cruciferous vegetables lowered riskof the lung and colorectal cancer in some epidemiologi-cal studies (Kutcek et al., 1960; Higdon et al., 2007).Previous studies have shown that I3C and MI (Fig. 1)from cruciferous vegetables have antimicrobial,antiviral and anticancer activities (Higdon et al., 2007;Fan et al., 2009). Cruciferous vegetables also produceascorbigen and IAN (Kutcek et al., 1960). Like the plantauxin IAA, IAN is synthesized from tryptophan in crucif-erous vegetables (Kutcek et al., 1960) and is a natu-rally occurring plant growth hormone (auxin) (Joneset al., 1952). IAN can be converted into another plantauxin IAA by nitrilases in plants (Kobayashi et al., 1993).Although the function of IAN as an auxin is obvious inplant cells (Kutcek et al., 1960), a few studies of IANhave examined in microbial ecology. Only a few plantpathogenic bacteria, such as Agrobacterium tumefa-ciens, Rhizobium sp. (Kobayashi et al., 1995) andPseudomonas syringae, contain nitrilase to use IAN as anitrogen source (Howden et al., 2009). However, P.aeruginosa and E. coli do not have nitrilase (Howdenet al., 2009), so they do not degrade IAN (Fig. 3A). Addi-tionally, unlike other glucosinolates, IAN shows no anti-microbial activity for P. aeruginosa (Aires et al., 2009).Interestingly, plant auxin IAA did not show any activity onE. coli O157:H7 and P. aeruginosa, whereas anotherplant auxin IAN showed significant effects on bothstrains (Figs 1, 2, 4 and 5).

Fig. 5. Effect of indole and indole derivativeson (A) the extracellular indole concentration inE. coli O157:H7 and (B) pyocyaninproduction, (C) PQS production (a picture ofthin layer chromatography is also shown) and(D) pyoverdine production in P. aeruginosaPAO1. Indole (I), 7-hydroxyindole (7HI),3-indoleacetic acid (IAA),indole-3-carboxyaldehyde (I3CA) and3-indolylacetonitrile (IAN) at 100 mg ml-1 weredissolved in DMSO. DMSO and indole wereused as the negative and positive controlsrespectively. The extracellular indoleconcentration was measured at E. coliO157:H7 cell absorbance of 2.0 in LBmedium. Production of pyocyanin and PQSwas measured in LB medium and productionof pyoverdine was measured in minimalsuccinate medium from planktonic P.aeruginosa PAO1 cells after 24 h. At least twoindependent experiments were conducted,and the error bars indicate one standarddeviation.

A. E. coli O157:H7 B. P. aeruginosaIn

do

le p

rod

uct

ion

(mg

ml-1

)

0

10

20

30

40

50

None IAA I3CA IAN Pyo

cyan

ine

pro

du

ctio

n

(

fold

ch

ang

e)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

None IAA I 7HI I3CA IAN

PQ

S p

rod

uct

ion

(fo

ld c

han

ge)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

None I IAA 7HI I3CA IAN

C. P. aeruginosa D. P. aeruginosaP

yove

rdin

e p

rod

uct

ion

(fo

ld c

han

ge)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

None I IAA 7HI I3CA IAN



Current study demonstrated for the first time that IANacted as a biofilm inhibitor for E. coli O157:H7, as well asan antivirulence compound for P. aeruginosa withoutaffecting their growth. Importantly, while indole and

7-hydroxyindole were degraded by P. aeruginosa (Leeet al., 2009), IAN was stable in the presence of either E.coli O157:H7 or P. aeruginosa (Fig. 3), probably becauseIAN was less prevalent in the microbial community, so that

Table 2. Partial list of the genes that were induced and repressed from P. aeruginosa PAO1 cells upon addition of 100 mg ml-1 3-indolylacetonitrile(IAN).

Gene name Fold change Description

Virulence genesPA1000 -2.1 Quinolone signal response protein PqsEPA2256 -1.8 Pyoverdine biosynthesis protein PvcC

Motility-related genesPA0410 -1.8 Type IV twitching motility protein PilIPA1453 -1.8 Flagellar biosynthesis protein FlhFPA1461 -1.7 Flagellar motor protein MotD

Other repressed genesPA0525 -1.7 Probable dinitrification protein NorDPA1060 -1.8 Hypothetical proteinPA1063 -1.8 Hypothetical proteinPA1178 -1.7 PhoP/Q and low Mg2+ inducible outer membrane protein H1 precursorPA1632 -1.7 K+-transporting ATPase KdpFPA1797 -1.8 Hypothetical proteinPA1832 -1.7 Probable proteasePA1877 -1.8 Probable secretion proteinPA1882 -1.8 Probable transporterPA2345 -1.9 Hypothetical proteinPA2655 -1.7 Hypothetical proteinPA2778 -1.7 Hypothetical proteinPA3075 -1.8 Hypothetical proteinPA3595 -1.7 Major facilitator superfamily (MFS) transporterPA3715 -1.8 Hypothetical proteinPA4377 -1.9 Hypothetical protein

Induced small molecule transport genesPA0119 1.7 Dicarboxylate transporterPA0138 1.8 Permease of ABC transporterPA0196 2.1 Pyridine nucleotide transhydrogenase PntBPA0203 1.7 Binding protein component of ABC transporterPA0204 2.1 Permease of ABC transporterPA0206 1.7 ATP-binding component of ABC transporterPA2135 2.0 Probable transporterPA2914 1.7 Probable transporterPA3376 1.7 ATP-binding component of ABC transporterPA3407 1.9 Haem acquisition protein HasApPA3448 2.3 Permease of ABC transporter

Other induced genesPA0193 1.7 Hypothetical proteinPA0238 2.1 Hypothetical proteinPA0239 1.8 Hypothetical proteinPA0422 1.7 Hypothetical proteinPA0522 1.7 Hypothetical proteinPA0653 1.7 Hypothetical proteinPA1235 1.8 Probable transcriptional regulatorPA1360 1.7 Hypothetical proteinPA1489 2.0 Hypothetical proteinPA1929 1.9 Hypothetical proteinPA2903 1.8 Precorrin-3 methylase CobJPA3413 1.9 Hypothetical proteinPA3492 1.8 Hypothetical proteinPA3501 2.4 Hypothetical proteinPA4886 1.9 Probable two-component sensorPA5144 2.1 Hypothetical proteinPA5417 1.7 Sarcosine oxidase delta subunit, SoxDPA5420 2.7 Formyltetrahydrofolate deformylase PurU2

The cells were grown until the absorbance of 2.0 in LB medium at 37C at 250 r.p.m. with and without IAN. The full data are available using GEOaccession number GSE21508.



two pathogenic bacteria did not possess a defencesystem to degrade IAN. Furthermore, since IAN is nottoxic to bacteria, it has a smaller selection pressure fordrug resistance than antibiotics that are toxic to bacteria.

Since humans contain a large amount of indole in theintestine, the addition of indole (1 mM corresponding to117 mg ml-1) into human epithelial (HCT-8) cells was nottoxic (Bansal et al., 2010). While toxicity data of IAN inhuman cells is not yet available, IAN (up to 200 mg kg-1

dose) was not teratogenic to the rat foetus, but IAN in ahigh dose (300 mg kg-1) induced sedation and ataxia inrats (Nishie and Daxenbichler, 1980).

By studying the whole transcriptomic profiling, it wasshown that IAN repressed curli genes and induced indoleand prophage genes in E. coli O157:H7 (Table 1). Curliare fibres that are produced by E. coli and have the samebiochemical and biophysical properties of human amyloidthat are associated with Alzheimers and prion disease(Chapman et al., 2002). Additionally, curli formation isimportant for the biofilm formation of pathogenic E. coliO157:H7 (Ryu and Beuchat, 2005; Uhlich et al., 2006) aswell as non-pathogenic E. coli (Prigent-Combaret et al.,2000; Reisner et al., 2006). Using a Congo red assay andSEM, this study confirmed that IAN inhibited the biofilmformation of E. coli O157:H7 by reducing curli production(Fig. 4A, B and C). This result suggested that the inhibi-tion of curli formation could be a way to eradicate thebiofilm formation of E. coli O157:H7.

The genetic mechanism of the biofilm formation of E.coli O157:H7 is a complex process that is now beingunveiled. In addition to curli formation mentioned above,intercellular signal molecules, such as autoinducer-2(Yoon and Sofos, 2008) and indole (Bansal et al., 2007;Lee et al., 2007b), are also involved in the biofilm forma-tion of E. coli O157:H7. Since IAN increased indole pro-duction and decreased the biofilm formation of E. coliO157:H7, IAN partially interfered with the indole signal-ling. Therefore, it was confirmed that the elevated indoleproduction was responsible for the reduction of biofilmformation, and IAN triggered the indole production.

Recently, a report showed that prophage genes playeda role in E. coli biofilm formation (Wang et al., 2009). Forexample, the deletion of 35 prophage genes, such asCP4-57 and DLP12, increased biofilm formation up to17-fold in E. coli. The current study showed that IAN mostinduced two putative prophage genes, z2978 and z3345(Table 1). Although it is highly speculative, these proph-age genes may also be involved in controlling the biofilmformation of E. coli O157:H7, which must be furtherstudied.

Previously, we reported that indole affects biofilm for-mation of non-pathogenic E. coli and pathogenic E. coliO157:H7 via different mechanisms; for example, indoledecreased E. coli K-12 biofilm formation mediated by

SdiA (Lee et al., 2007a; Lee et al., 2008), and indoledecreased E. coli O157:H7 biofilm by reducing cell motilityand attachment (Bansal et al., 2007). Hence, a role ofSdiA in IAN-mediated reduction of biofilm formation of E.coli was explored by using E. coli K-12 BW25113 wild-type and its sdiA knockout mutant. As a result, IANdecreased the biofilm formation of both wild-type and sdiAknockout mutant in a same dose-dependent manner (datanot shown). Therefore, it appears IAN does not workthrough SdiA in E. coli K-12.

Plants are the richest sources of bioactive moleculesincluding antimicrobial compounds. Several non-toxic,plant-derived biofilm inhibitors exist for pathogenic bacte-ria, such as furanones that originate from the red mac-roalga Delisea pulchra by blocking the quorum sensing ofbacteria (Ren et al., 2001; Hentzer et al., 2003), garlicextract as a quorum-sensing inhibitor (Rasmussen et al.,2005; Shuford et al., 2005), both ursolic acid (Ren et al.,2005) and corosolic acid (Garo et al., 2007) that are pro-duced by the ebony tree Diospyros dendo, and asiaticacid that is produced by the tropical medicinal plant Cen-tella asiatica (Garo et al., 2007). Unlike other plant biofilminhibitors, IAN inhibited E. coli biofilm formation throughcurli formation and indole production. Therefore, diverseplants may have developed different means of coping withenvironmental bacteria. Since various indole derivativesfrom plants and numerous synthetic indole derivatives arecommercially available, work is currently in progress toidentify stronger and safer biofilm inhibitors.

Experimental procedures

Bacterial strains, materials and growthrate measurements

Serotype enterohemorrhagic E. coli O157:H7 [ATCC43895,EDL933 strain (Strockbine et al., 1986)] and P. aeruginosaPAO1 [the sequenced Holloway strain (Stover et al., 2000)]were used. LuriaBertani medium (Sambrook et al., 1989)was used as the medium for the growth of E. coli and P.aeruginosa. Indole, IAA, MI, I3PA, I3C, I3CA, IAN, crystalviolet, Congo red, Coomasie brilliant blue, b-mercaptoethanol, sodium phosphate and trifluoroacetic acid (TFA)were purchased from Sigma-Aldrich Co. (Missouri, USA). 7HIwas purchased from Fisher Scientific Co. (Pittsburg, USA).The other chemicals (acetonitrile, amyl alcohol, formalde-hyde, glutaraldehyde, ethyl alcohol, DMSO, hydrochloricacid, OsO4 and p-dimethylamino-benzaldehyde) were pur-chased from Duksan Pure Chemical (Ansan, Korea). Thestrains were initially streaked from -80C glycerol stock on aLB plate and a fresh single colony was inoculated in LB(25 ml) in 250 ml flasks and cultured at 37C and 250 r.p.m.Human body temperature (37C) was chosen here becausethis study focused on investigating the effects of indolederivatives on human pathogenic bacteria, although theimpact of indole in E. coli was less significant at 37C than atlow temperatures (Lee et al., 2008). Overnight cultures were



re-inoculated at 1:100 dilution in the medium. For the cellgrowth measurements, the optical density was measured at600 nm using a spectrophotometer (UV-160, Shimadzu,Japan). Each experiment was performed with at least twoindependent cultures.

Crystal-violet biofilm assay

A static biofilm formation assay was performed in 96-wellpolystyrene plates (SPL life sciences, Korea) as previouslyreported (Pratt and Kolter, 1998). Briefly, cells were inocu-lated with an initial turbidity of 0.05 at 600 nm and cultured for24 h without shaking at 37C. Cell growth and total biofilmwere measured using crystal violet staining. Each data pointwas averaged from at least 12 replicate wells (six wells fromeach of at least two independent cultures).

Virulence factor assays of P. aeruginosa

Except for the pyoverdine assay, overnight cultures werediluted 1:100 and contacted with indole and indole derivativesor diluent DMSO (negative control). The pyocyanin assay ofP. aeruginosa was adapted (Essar et al., 1990); after growthfor 24 h, supernatants were extracted with chloroform andanalysed spectrophotometrically. The PQS assay wasadapted (Attila et al., 2008); after growth for 24 h, superna-tants were extracted with acidified ethyl acetate and analysedby thin layer chromatography. The pyoverdine assay wasadapted (Anyanful et al., 2005); after growth in minimal suc-cinate medium, cells were diluted to a turbidity of 0.05 at600 nm in fresh minimal succinate medium and were grownfor 24 h. The pyoverdine concentration was measured spec-trophotometrically at 405 nm (Stintzi et al., 1998). The rham-nolipid assay was adapted (Wilhelm et al., 2007); aftergrowth for 24 h, supernatants were assayed for rhamnolipidsusing the orcinol colorimetric assay. At least two independentexperiments were conducted.

Assay of indole and IAN

To measure the concentration of extracellular indole, E. coliO157:H7 was grown with and without IAN in LB medium at37C and 250 r.p.m. to an absorbance of 2.0 at 600 nm. Theabsorbance of 2.0 represented the time point when E. colisecreted the stationary signal indole at a concentration ofapproximately 25 mg ml-1. The extracellular indole concentra-tion was measured with the protocol using the Kovacsreagent (Kawamura-Sato et al., 1999) and was corroboratedwith HPLC. To measure the concentration of IAN, bacteriawere grown with and without IAN (110 mg ml-1 for E. coli and70 mg ml-1 for P. aeruginosa) at 37C and 250 r.p.m. Theconcentrations of indole or IAN were measured with reverse-phase HPLC using a 100 4.6 mm Chromolith PerformanceRP-18e column (Merck KGaA, Darmstadt, Germany) (Leeet al., 2007b) and elution with H2O-0.1% TFA and acetonitrileas the mobile phases at a flow rate of 0.5 ml min-1 (50:50).Under these conditions, the retention times and the absor-bance maxima were 5.1 min/271 nm for indole and4.3 min/298 nm for IAN. Each experiment was performedwith two independent cultures.

Total RNA isolation

For the microarray experiments and qRT-PCR experiments,E. coli O157:H7 and P. aeruginosa PAO1 were inoculated in25 ml of LB medium in 250 ml shake flasks with overnightcultures (1:100 dilution). Either IAN (100 mg ml-1) that wasdissolved in 25 ml DMSO was added or 25 ml DMSO wasadded as a control. Cells were cultured in LB at 37C withshaking at 250 r.p.m. until an absorbance of 1.0 for E. coliO157:H7 and 2.0 for P. aeruginosa PAO1 at 600 nm wasreached. Due to the low biofilm formation by IAN, planktoniccells were utilized to investigate the effect of IAN on whole-global transcriptome. The cells were immediately chilled withdry ice and 95% ethanol (to prevent RNA degradation) for30 s before centrifugation at 13 000 g for 2 min. The cellpellets were immediately frozen with dry ice and stored at-80C. Total RNA was isolated using a Qiagen RNeasy miniKit (Valencia, CA, USA). RNA quality was assessed using anAgilent 2100 bioanalyser and a RNA 6000 Nano Chip (AgilentTechnologies, Amstelveen, the Netherlands), and the quan-tity was determined using ND-1000 Spectrophotometer(NanoDrop Technologies, DE, USA).

DNA microarray analysis

The E. coli GeneChip Genome 2.0 Array (Affymetrix, P/N900551, Santa Clara, USA) and the P. aeruginosa GenechipGenome Array (Affymetrix, P/N 900339) were used in orderto study the differential gene expression profile of the cellsafter the addition of IAN. From each sample, 10 mg of totalRNA was converted to cDNA using random primers. Thepurified cDNA was fragmented using 0.6 U/mg of DNase I andend-labelled by terminal transferase reaction incorporating abiotinylated dideoxynucleotide. Hybridization was performedfor 16 h at 45C for the E. coli Genechip and at 50C for P.aeruginosa Genechip and 60 r.p.m. as described in the GeneChip Expression Analysis Technical Manual (Affymetrix).After the hybridization, the chips were stained and washed ina Genechip Fluidics Station 450 (Affymetrix) and scanned byusing a Genechip Array scanner 3000 7G (Affymetrix). Theprobe array images were inspected for any image artifacts.Background values, noise values and scaling factors of botharrays were examined and were comparable. The intensitiesof the polyadenosine RNA controls of Bacillus subtilis (lys,phe, thr and dap) at different concentrations were used inorder to monitor the labelling and scanning process. A genewas considered differentially expressed when the P-value forcomparing two chips was lower than 0.05 (to assure that thechange in gene expression was statistically significant andthat any false positives arise less than 5%) and when theexpression ratio was higher (twofold for the E. coli Genechipand 1.7-fold for P. aeruginosa Genechip). Gene functionswere obtained from the AffymetrixNetAffx Analysis Center(https://www.affymetrix.com/analysis/netaffx/index.affx).

Quantitative real-time reverse transcription polymerasechain reaction

To corroborate the DNA microarray data and curli assay andto confirm the induction of indole production with IAN, qRT-



https://www.affymetrix.com/analysis/netaffx/index.affx

PCR was used to investigate the transcription level of csgA(encoding curlin major subunit) and tnaA (encodingindole-synthesizing tryptophanase), in E. coli O157:H7 pqsE(quinolone signal response protein), pvcC (pyoverdine bio-synthesis protein) and pilI (twitching motility protein) in P.aeruginosa with and without IAN treatment. Two primers forcsgA (forward primer 5-AGATGTTGGTCAGGGCTCAG-3and reverse primer 5-CGTTGTTACCAAAGCCAACC-3),two primers for tnaA (forward primer 5-TACACCATTCCGACTCACCA-3 and reverse primer 5-CCGTATCGAAGGCTTCTTTG-3), two primers for pqsE (forward primer 5-GACATGGAGGCTTACCTGGA-3 and reverse primer 5-CTCAGTTCGTCGAGGGATTC-3), two primers for pvcC (forwardprimer 5-GACCAGGGTATCGTCGTCAG-3 and reverseprimer 5-AGCGGATAGTCGAAGGGACT-3), and two pri-mers for pilI (forward primer 5-ATCATGGACCTCTGCGGTTT-3 and reverse primer 5-GAAGACGCCATGAATGAAGG-3) were used. The expression level of the housekeep-ing gene rrsG (16S rRNA, forward primer 5-TATTGCACAATGGGCGCAAG-3 and reverse primer 5-ACTTAACAAACCGCCTGCGT-3) for E. coli O157:H7 and proC(pyrroline-5-carboxylate reductase, forward primer 5-CAGGCCGGGCAGTTGCTGTC-3 and reverse primer 5-GGTCAGGCGCGAGGCTGTCT-3) for P. aeruginosa were used tonormalize the expression data of the genes of interest. TheqRT-PCR method was adapted from a previous study (Leeet al., 2008). qRT-PCR was performed using a SYBR Greenmaster mix (Applied Biosystems, Foster City, USA) and a ABI7500 Real-Time PCR System (Applied Biosystems) with twoindependent cultures.

Curli assay with a Congo red plate and SEM

To measure the curli production, LB agar medium containing20 mg ml-1 Congo red (Sigma), 10 mg ml-1 Coomassie brilliantblue (Sigma), 15 g l-1 agar was used as previously describedto visualize E. coli curli expression after 24 h incubation at37C (Reisner et al., 2006). IAN (100 mg ml-1) was added tothe Congo red plate. Since Congo red binds both curli andcellulose (Zogaj et al., 2003), a quantitative cellulose assayusing cellulose-specific binding calcofluor (Ma and Wood,2009) was used to determine if curli or cellulose was identi-fied by the Congo red assay. Cells (2 ml) at a turbidity of 4were centrifuged and resuspended in 1 ml of 1% tryptonewith 16 mg ml-1 calcofluor and incubated for 2 h at 250 r.p.m.Bacterial bound calcofluor was removed by centrifugation for5 min at 17 000 g, and the amount of unbound calcofluor tocellulose was determined by measuring the absorbance ofthe supernatant at 350 nm. In order to corroborate the plateassay, SEM was used with a modified protocol (Hossainet al., 1996). Briefly, for planktonic cells, when cell growthreached at optical density of 1.0, the cells were directly fixedthrough the addition of glutaraldehyde (2.5% in the final con-centration) and formaldehyde (2% in the final concentration)and incubated at 4C overnight. Then, the cells were col-lected by filtering with a 0.45 mm Nylon filter (Nalgene, NewYork, USA) under vacuum. The filter, which contained thecells, was cut into 0.5 0.5 mm squares and placed in cleanampoules. For biofilm cells, a nylon filter was cut into0.5 0.5 mm square and placed in 96-well plates with300 ml of cells with an initial turbidity of 0.05 at 600 nm. Cells

and the nylon filter were incubated together to form biofilmcells at 37C for 24 h without shaking. After fixation withglutaraldehyde and formaldehyde, either planktonic cells orbiofilm cells grown on a nylon filter were washed three timeswith a 0.2 M sodium phosphate buffer (pH 7.2) for 20 mineach time at 4C and stored in the same buffer at 4C over-night. Then, the samples were washed twice with a 0.2 Msodium phosphate buffer for 20 min before they were post-fixed for 90 min with an Osmium solution [containing 1.5 mlof a sodium phosphate buffer (0.2 M), 3 ml of 2% OsO4 and3 ml deionized water]. The cells were washed again fourtimes with 1.5 ml of a sodium phosphate buffer (0.2 M) for20 min. In the next step, cells were dehydrated through suc-cessive 20 min incubations in 50%, 70%, 80%, 90% and95% ethanol, and then two successive 20 min incubations in100% ethanol. After the dehydration process, the cells wereincubated twice in isoamyl acetate for 20 min. These filterpieces, which contained the cells, were dried using a critical-point dryer (HCP-2, Hitachi, Japan). Then, the dried filterswere affixed to SEM stubs, and coated with white gold for200 s using an Ion-sputter (E-1030, Hitachi, Japan). Thespecimens were examined using SEM S-4100 (Hitachi,Japan) at a voltage of 15 kV and magnifications rangingfrom 2000 to 25 000.

Acknowledgements

This research was supported by the Yeungnam Universityresearch grant (to J. Lee). J-H. Lee was supported by theBrain Korea 21 Project from the Ministry of Education andHuman Resources, Korea.

References

Aires, A., Mota, V.R., Saavedra, M.J., Rosa, E.A., andBennett, R.N. (2009) The antimicrobial effects of glucosi-nolates and their respective enzymatic hydrolysis productson bacteria isolated from the human intestinal tract. J ApplMicrobiol 106: 20862095.

Anyanful, A., Dolan-Livengood, J.M., Lewis, T., Sheth, S.,Dezalia, M.N., Sherman, M.A., et al. (2005) Paralysis andkilling of Caenorhabditis elegans by enteropathogenicEscherichia coli requires the bacterial tryptophanase gene.Mol Microbiol 57: 9881007.

Attila, C., Ueda, A., Cirillo, S.L.G., Cirillo, J.D., Chen, W., andWood, T.K. (2008) Pseudomonas aeruginosa PAO1 viru-lence factors and poplar tree response in the rhizosphere.Microb Biotechnol 1: 1729.

Bansal, T., Englert, D., Lee, J., Hegde, M., Wood, T.K., andJayaraman, A. (2007) Differential effects of epinephrine,norepinephrine, and indole on Escherichia coli O157:H7chemotaxis, colonization, and gene expression. InfectImmun 75: 45974607.

Bansal, T., Alaniz, R.C., Wood, T.K., and Jayaraman, A.(2010) The bacterial signal indole increases epithelial-celltight-junction resistance and attenuates indicators ofinflammation. Proc Natl Acad Sci USA 107: 228233.

Chapman, M.R., Robinson, L.S., Pinkner, J.S., Roth, R.,Heuser, J., Hammar, M., et al. (2002) Role of Escherichiacoli curli operons in directing amyloid fiber formation.Science 295: 851855.



Chyan, Y.J., Poeggeler, B., Omar, R.A., Chain, D.G., Fran-gione, B., Ghiso, J., and Pappolla, M.A. (1999) Potentneuroprotective properties against the Alzheimer b-amyloidby an endogenous melatonin-related indole structure,indole-3-propionic acid. J Biol Chem 274: 2193721942.

Di Martino, P., Fursy, R., Bret, L., Sundararaju, B., andPhillips, R.S. (2003) Indole can act as an extracellularsignal to regulate biofilm formation of Escherichia coli andother indole-producing bacteria. Can J Microbiol 49: 443449.

Diggle, S.P., Winzer, K., Chhabra, S.R., Worrall, K.E.,Cmara, M., and Williams, P. (2003) The Pseudomonasaeruginosa quinolone signal molecule overcomes the celldensity-dependency of the quorum sensing hierarchy,regulates rhl-dependent genes at the onset of stationaryphase and can be produced in the absence of LasR. MolMicrobiol 50: 2943.

Ensley, B.D., Ratzkin, B.J., Osslund, T.D., Simon, M.J.,Wackett, L.P., and Gibson, D.T. (1983) Expression of naph-thalene oxidation genes in Escherichia coli results in thebiosynthesis of indigo. Science 222: 167169.

Essar, D.W., Eberly, L., Hadero, A., and Crawford, I.P. (1990)Identification and characterization of genes for a secondanthranilate synthase in Pseudomonas aeruginosa: inter-changeability of the two anthranilate synthases and evolu-tionary implications. J Bacteriol 172: 884900.

Fan, S., Meng, Q., Saha, T., Sarkar, F.H., and Rosen, E.M.(2009) Low concentrations of diindolylmethane, a metabo-lite of indole-3-carbinol, protect against oxidative stress in abrca1-dependent manner. Cancer Res 69: 60836091.

Garo, E., Eldridge, G.R., Goering, M.G., DeLancey Pulcini,E., Hamilton, M.A., Costerton, J.W., and James, G.A.(2007) Asiatic acid and corosolic acid enhance the suscep-tibility of Pseudomonas aeruginosa biofilms to tobramycin.Antimicrob Agents Chemother 51: 18131817.

Gong, F., and Yanofsky, C. (2002) Analysis of tryptophanaseoperon expression in vitro: accumulation of TnaC-peptidyl-tRNA in a release factor 2-depleted S-30 extract preventsRho factor action, simulating induction. J Biol Chem 277:1709517100.

Hentzer, M., Riedel, K., Rasmussen, T.B., Heydorn, A.,Andersen, J.B., Parsek, M.R., et al. (2002) Inhibition ofquorum sensing in Pseudomonas aeruginosa biofilm bac-teria by a halogenated furanone compound. Microbiology148: 87102.

Hentzer, M., Wu, H., Andersen, J.B., Riedel, K., Rasmussen,T.B., Bagge, N., et al. (2003) Attenuation of Pseudomonasaeruginosa virulence by quorum sensing inhibitors. EMBOJ 22: 38033815.

Higdon, J.V., Delage, B., Williams, D.E., and Dashwood, R.H.(2007) Cruciferous vegetables and human cancer risk: epi-demiologic evidence and mechanistic basis. PharmacolRes 55: 224236.

Hirakawa, H., Kodama, T., Takumi-Kobayashi, A., Honda, T.,and Yamaguchi, A. (2009) Secreted indole serves as asignal for expression of type III secretion system translo-cators in enterohaemorrhagic Escherichia coli O157:H7.Microbiology 155: 541550.

Hossain, M.M., Nakayama, H., and Goto, N. (1996) In vitroinduction of apoptosis of developing brain cells by5-azacytidine. Int J Dev Neurosci 14: 1117.

Howden, A.J., Rico, A., Mentlak, T., Miguet, L., and Preston,G.M. (2009) Pseudomonas syringae pv. syringae B728ahydrolyses indole-3-acetonitrile to the plant hormoneindole-3-acetic acid. Mol Plant Pathol 10: 857865.

Jones, E.R., Henbest, H.B., Smith, G.F., and Bentley, J.A.(1952) 3-indolylacetonitrile: a naturally occurring plantgrowth hormone. Nature 169: 485487.

Kawamura-Sato, K., Shibayama, K., Horii, T., Iimuma, Y.,Arakawa, Y., and Ohta, M. (1999) Role of multiple effluxpumps in Escherichia coli in indole expulsion. FEMS Micro-biol Lett 179: 345352.

Klausen, M., Heydorn, A., Ragas, P., Lambertsen, L., Aaes-Jrgensen, A., Molin, S., and Tolker-Nielsen, T. (2003)Biofilm formation by Pseudomonas aeruginosa wild type,flagella and type IV pili mutants. Mol Microbiol 48: 15111524.

Kobayashi, M., Izui, H., Nagasawa, T., and Yamada, H.(1993) Nitrilase in biosynthesis of the plant hormoneindole-3-acetic acid from indole-3-acetonitrile: cloning ofthe Alcaligenes gene and site-directed mutagenesis of cys-teine residues. Proc Natl Acad Sci USA 90: 247251.

Kobayashi, M., Suzuki, T., Fujita, T., Masuda, M., andShimizu, S. (1995) Occurrence of enzymes involved inbiosynthesis of indole-3-acetic acid from indole-3-acetonitrile in plant-associated bacteria, Agrobacteriumand Rhizobium. Proc Natl Acad Sci USA 92: 714718.

Kutcek, M., Prochazka, ., and Grenberger, D. (1960)Biogenesis of ascorbigen, 3-indolylacetonitrile and indole-3-carboxylic acid from D, L-tryptophan-3-14C in Brassicaoleracea L. Nature 187: 6162.

Lee, J., Jayaraman, A., and Wood, T.K. (2007a) Indole is aninter-species biofilm signal mediated by SdiA. BMC Micro-biol 7: 42.

Lee, J., Bansal, T., Jayaraman, A., Bentley, W.E., and Wood,T.K. (2007b) Enterohemorrhagic Escherichia coli biofilmsare inhibited by 7-hydroxyindole and stimulated by isatin.Appl Environ Microbiol 73: 41004109.

Lee, J., Zhang, X.S., Hegde, M., Bentley, W.E., Jayaraman,A., and Wood, T.K. (2008) Indole cell signaling occursprimarily at low temperatures in Escherichia coli. ISME J 2:10071023.

Lee, J., Attila, C., Cirillo, S.L.G., Cirillo, J.D., and Wood, T.K.(2009) Indole and 7-hydroxyindole diminish Pseudomonasaeruginosa virulence. Microb Biotechnol 2: 7590.

Lee, J.-H., and Lee, J. (2010) Indole as an intercellularsignal in microbial community. FEMS Microbiol Rev 34:426444.

Lesic, B., Lpine, F., Dziel, E., Zhang, J., Zhang, Q., Pad-field, K., et al. (2007) Inhibitors of pathogen intercellularsignals as selective anti-infective compounds. PLoSPathog 3: 12291239.

Li, J.W.H., and Vederas, J.C. (2009) Drug discovery andnatural products: end of an era or an endless frontier?Science 325: 161165.

Ma, Q., and Wood, T.K. (2009) OmpA influences Escherichiacoli biofilm formation by repressing cellulose productionthrough the CpxRA two-component system. Environ Micro-biol 11: 27352746.

Mndez-Ortiz, M.M., Hyodo, M., Hayakawa, Y., andMembrillo-Hernndez, J. (2006) Genome-wide transcrip-tional profile of Escherichia coli in response to high levels



of the second messenger 3,5-cyclic diguanylic acid. J BiolChem 281: 80908099.

Mueller, R.S., Beyhan, S., Saini, S.G., Yildiz, F.H., and Bar-tlett, D.H. (2009) Indole acts as an extracellular cue regu-lating gene expression in Vibrio cholerae. J Bacteriol 191:35043516.

Newton, W.A., and Snell, E.E. (1965) Formation and interre-lationships of tryptophanase and tryptophan synthetases inEscherichia coli. J Bacteriol 89: 355364.

Nishie, K., and Daxenbichler, M.E. (1980) Toxicology of glu-cosinolates, related compounds (nitriles, R-goitrin, isothio-cyanates) and vitamin U found in Cruciferae. Food CosmetToxicol 18: 159172.

OToole, G.A., and Kolter, R. (1998) Flagellar and twitchingmotility are necessary for Pseudomonas aeruginosabiofilm development. Mol Microbiol 30: 295304.

Pratt, L.A., and Kolter, R. (1998) Genetic analysis of Escheri-chia coli biofilm formation: roles of flagella, motility, chemo-taxis and type I pili. Mol Microbiol 30: 285293.

Prigent-Combaret, C., Prensier, G., Le Thi, T.T., Vidal, O.,Lejeune, P., and Dorel, C. (2000) Developmental pathwayfor biofilm formation in curli-producing Escherichia colistrains: role of flagella, curli and colanic acid. EnvironMicrobiol 2: 450464.

Rampioni, G., Pustelny, C., Fletcher, M.P., Wright, V.J.,Bruce, M., Rumbaugh, K.P., et al. (2010) Transcriptomicanalysis reveals a global alkyl-quinolone-independentregulatory role for PqsE in facilitating the environmentaladaptation of Pseudomonas aeruginosa to plant andanimal hosts. Environ Microbiol 12: 16591673.

Rasko, D.A., Moreira, C.G., Li de, R., Reading, N.C., Ritchie,J.M., Waldor, M.K., et al. (2008) Targeting QseC signalingand virulence for antibiotic development. Science 321:10781080.

Rasmussen, T.B., Bjarnsholt, T., Skindersoe, M.E., Hentzer,M., Kristoffersen, P., Kte, M., et al. (2005) Screening forquorum-sensing inhibitors (QSI) by use of a novel geneticsystem, the QSI selector. J Bacteriol 187: 17991814.

Reisner, A., Krogfelt, K.A., Klein, B.M., Zechner, E.L., andMolin, S. (2006) In vitro biofilm formation of commensaland pathogenic Escherichia coli strains: impact of envi-ronmental and genetic factors. J Bacteriol 188: 35723581.

Ren, D., Sims, J.J., and Wood, T.K. (2001) Inhibition of biofilmformation and swarming of Escherichia coli by (5Z)-4-bromo-5-(bromomethylene)-3-butyl-2(5H)-furanone.Environ Microbiol 3: 731736.

Ren, D., Bedzyk, L.A., Ye, R.W., Thomas, S.M., and Wood,T.K. (2004) Differential gene expression shows natural bro-minated furanones interfere with the autoinducer-2 bacte-rial signaling system of Escherichia coli. Biotechnol Bioeng88: 630642.

Ren, D., Zuo, R., Gonzlez Barrios, A.F., Bedzyk, L.A., Eld-ridge, G.R., Pasmore, M.E., and Wood, T.K. (2005) Differ-ential gene expression for investigation of Escherichia colibiofilm inhibition by plant extract ursolic acid. Appl EnvironMicrobiol 71: 40224034.

Ryu, J.H., and Beuchat, L.R. (2005) Biofilm formation byEscherichia coli O157:H7 on stainless steel: effect of

exopolysaccharide and curli production on its resistance tochlorine. Appl Environ Microbiol 71: 247254.

Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989) MolecularCloning: A Laboratory Manual. Cold Spring Harbor, NY,USA: Cold Spring Harbor Laboratory Press.

Shuford, J.A., Steckelberg, J.M., and Patel, R. (2005) Effectsof fresh garlic extract on Candida albicans biofilms. Anti-microb Agents Chemother 49: 473.

Stintzi, A., Evans, K., Meyer, J.M., and Poole, K. (1998)Quorum-sensing and siderophore biosynthesis inPseudomonas aeruginosa: lasR/lasI mutants exhibitreduced pyoverdine biosynthesis. FEMS Microbiol Lett166: 341345.

Stover, C.K., Pham, X.Q., Erwin, A.L., Mizoguchi, S.D., War-rener, P., Hickey, M.J., et al. (2000) Complete genomesequence of Pseudomonas aeruginosa PAO1, an opportu-nistic pathogen. Nature 406: 959964.

Strockbine, N.A., Marques, L.R., Newland, J.W., Smith, H.W.,Holmes, R.K., and OBrien, A.D. (1986) Two toxin-converting phages from Escherichia coli O157:H7 strain933 encode antigenically distinct toxins with similar biologicactivities. Infect Immun 53: 135140.

Uhlich, G.A., Cooke, P.H., and Solomon, E.B. (2006) Analy-ses of the red-dry-rough phenotype of an Escherichia coliO157:H7 strain and its role in biofilm formation and resis-tance to antibacterial agents. Appl Environ Microbiol 72:25642572.

Wang, X., Kim, Y., and Wood, T.K. (2009) Control and ben-efits of CP4-57 prophage excision in Escherichia coli bio-films. ISME J 3: 11641179.

Weber, H., Pesavento, C., Possling, A., Tischendorf, G., andHengge, R. (2006) Cyclic-di-GMP-mediated signallingwithin the sigma network of Escherichia coli. Mol Microbiol62: 10141034.

Wikoff, W.R., Anfora, A.T., Liu, J., Schultz, P.G., Lesley, S.A.,Peters, E.C., and Siuzdak, G. (2009) Metabolomics analy-sis reveals large effects of gut microflora on mammalianblood metabolites. Proc Natl Acad Sci USA 106: 36983703.

Wilhelm, S., Gdynia, A., Tielen, P., Rosenau, F., and Jaeger,K.-E. (2007) The autotransporter esterase EstA ofPseudomonas aeruginosa is required for rhamnolipid pro-duction, cell motility, and biofilm formation. J Bacteriol 189:66956703.

Yang, L., Nilsson, M., Gjermansen, M., Givskov, M., andTolker-Nielsen, T. (2009) Pyoverdine and PQS mediatedsubpopulation interactions involved in Pseudomonasaeruginosa biofilm formation. Mol Microbiol 74: 13801392.

Yoon, Y., and Sofos, J.N. (2008) Autoinducer-2 activityof gram-negative foodborne pathogenic bacteria and itsinfluence on biofilm formation. J Food Sci 73: M140M147.

Zhao, J., Davis, L.C., and Verpoorte, R. (2005) Elicitor signaltransduction leading to production of plant secondarymetabolites. Biotechnol Adv 23: 283333.

Zogaj, X., Bokranz, W., Nimtz, M., and Rmling, U. (2003)Production of cellulose and curli fimbriae by members ofthe family Enterobacteriaceae isolated from the humangastrointestinal tract. Infect Immun 71: 41514158.



emi_2308 62..73

Environmental Microbiology Volume 13 issue 1 2011 [doi 10.1111%2Fj.1462-2920.2010.02308.x] Jin-Hyung Lee; Moo Hwan Cho; Jintae Lee -- 3-Indolylacetonitrile Decreases Escherichia coli

Documents