REVIEW

Detection of bacterial quorum sensing N-acyl homoserinelactones in clinical samples

Anjali Kumari & Patrizia Pasini & Sylvia Daunert

Received: 21 November 2007 /Revised: 15 February 2008 /Accepted: 20 February 2008 /Published online: 12 April 2008# Springer-Verlag 2008

Abstract Bacteria communicate among themselves usingcertain chemical signaling molecules. These signaling mole-cules generally are N-acyl homoserine lactones (AHLs)in Gram-negative bacteria and oligopeptides in Gram-positive bacteria. In addition, both Gram-positive and Gram-negative bacteria produce a family of signaling moleculesknown as autoinducer-2 that they employ for theircommunications. Bacteria coordinate their behavior byreleasing and responding to the chemical signaling mole-cules present in proportion to their population density. Thisphenomenon is known as quorum sensing. The role ofbacteria in the pathogenesis of several diseases, includinggastrointestinal (GI) disorders, is well established. More-over, rather recently bacterial quorum sensing has beenimplicated in the onset of bacterial pathogenicity. Thus, wehypothesized that the signaling molecules involved inbacterial communication may serve as potential biomarkersfor the diagnosis and management of several bacteria-related diseases. For that, we previously developed amethod based on genetically engineered whole-cell sensingsystems for the rapid, sensitive, cost-effective and quanti-tative detection of AHLs in biological samples, such assaliva and stool, from both healthy and diseased individualswith GI disorders. Although various analytical methods,based on physical-chemical techniques and bacterial whole-cell biosensors, have been developed for the detection ofAHLs in the supernatants of bacterial cultures, only a fewof them have been applied to AHL monitoring in realsamples. In this paper, we report work performed in ourlaboratory and review that from others that describes the

detection of AHLs in biological, clinical samples, and reportsome of our recent experimental results.

Keywords Bacteria . Quorum sensing .N-Acyl homoserinelactones . Pathogenicity . Physiological samples

Introduction

Bacteria monitor their environment for the presence ofother bacteria by producing and responding to certainchemical signaling molecules. The concentration of thesemolecules is in direct proportion to the load or density ofthe bacteria; in other words, there is a direct relationshipbetween the concentration of chemical signaling moleculesand that of the bacteria present in a given environment. Thisphenomenon is termed “quorum sensing” (QS). It allowsbacteria to communicate with each other by synchronizing/regulating the expression of several genes that are necessaryfor bacteria to behave in unison as a group. These bacterialactivities are not useful when expressed by just onebacterium, but they can become beneficial to the bacteriawhen expressed in a group-based manner [1, 2]. Forexample, QS is responsible for bioluminescence in marinebacteria, such as Vibrio harveyi and Vibrio fischeri [3],biofilm formation by both Pseudomonas aeruginosa andBurkholderia cepacia in the lungs of cystic fibrosis (CF)patients [4], swarming and motility in Serratia liquefaciens[5], formation of virulence determinants by Staphylococcusaureus [6], and competence and sporulation in Bacillussubtilis [7].

The chemical signaling molecules in Gram-negativebacteria are N-acyl homoserine lactones (AHLs), while inGram-positive bacteria they are a family of small oligopep-tides. A third class of molecules, autoinducer-2, or AI-2, is

Anal Bioanal Chem (2008) 391:1619–1627DOI 10.1007/s00216-008-2002-3

A. Kumari : P. Pasini : S. Daunert (*)Department of Chemistry, University of Kentucky,Lexington, KY 40506-0055, USAe-mail: daunert@uky.edu

responsible for signaling in both Gram-positive and Gram-negative bacteria [6]. The AHLs produced by Gram-negative bacteria from different species differ in the numberof carbon atoms (four to 18) and substitution at the C-3 ofthe acyl side chain of the AHL molecule (Fig. 1). Forexample, the mechanism of initiation and regulation of QSin Gram-negative bacteria is complicated because itinvolves two types of molecules, namely, AHLs and AI-2.However, recent advances in the understanding of themechanism of regulation by AHLs have made it possible togain insight into the functions and effects that AHLs havenot only in communication within species but also amongspecies. In that regard, it is now well known that two typesof proteins, such as LuxI and LuxR, are involved in AHL-mediated communication. The LuxI/LuxR proteins wereoriginally identified in the marine bacterium V. fischeri,where they regulate the expression of genes responsible forbioluminescence [8]. LuxI-type proteins are the synthasesfor AHLs, whereas LuxR-type ones are the regulatoryproteins that bind to their respective AHL. The AHLsproduced either freely diffuse or are pumped out of the celldepending on the length of the acyl side chain. With theincrease in the bacterial cell density, there is a simultaneousincrease in the concentration of AHLs. At a certain criticalthreshold of AHL concentration, these AHLs bind to their

respective regulatory proteins. For instance, N-(3-oxo)-hexanoyl-DL-homoserine lactone (3-oxo-C6-HSL) inV. fischeri binds to LuxR, resulting in its respective AHL-LuxR complex. The AHL-regulatory protein complex thenbinds to its respective promoter, PluxI. This process resultsin the expression of the genes that are under thetranscriptional control of the specific promoter, PluxI,activated by the AHL-LuxR complex, i.e., the biolumines-cence genes in V. fischeri. These events also lead to theexpression of LuxI in a positive-feedback loop since luxI isalso under the control of the same promoter. A schematic ofthe AHL-dependent LuxI/LuxR regulatory system is shownin Fig. 2.

A considerable amount of data has been gathered todemonstrate the role of QS in the onset of virulencedetermination in vitro as well as in vivo. For example,AHL-mediated QS has been proven to be involved in thechronic bacterial infections by P. aeruginosa and B. cepaciathat contribute to lung disease in CF patients [4, 9]. QS hasalso been shown to be responsible for biofilm formation,both on inanimate surfaces and in lung tissue [4, 10, 11].Additionally, QS is thought to play a role in thepathogenesis of several gastrointestinal (GI) tract illnesses.In fact, it is evident that QS is important in the gutcolonization by pathogens. The role of QS in regulating

Structure

N-Butyryl-homoserine lactone (C4-HSL) O

O

NH

O

N-Hexanoyl-homoserine lactone (C6-HSL)

N-3-Oxo-hexanoyl-homoserine lactone (3-oxo-C6-HSL)

N-Heptanoyl-homoserine lactone (C7-HSL)

O

O

NH

O

O

O

NH

O O

O

O

NH

O

N-Octanoyl-homoserine lactone (C8-HSL)

O

O

NH

O

N-3-Oxo-octanoyl-homoserine lactone (3-oxo-C8-HSL)

O

O

NH

O O

N-Decanoyl-homoserine lactone (C10-HSL)

N-Dodecanoyl-homoserine lactone (C12-HSL)

N-Tetradecanoyl-homoserine lactone (C14-HSL)

O

O

NH

O

O

O

NH

O

O

O

NH

O

Fig. 1 Chemical structures ofN-acyl homoserine lactones(AHLs)

1620 Anal Bioanal Chem (2008) 391:1619–1627

virulence genes, such as motility and type III secretion inenteric pathogens, has been extensively proven [12–17]. Ithas also been demonstrated that AHLs inhibit cytokines andchemokines that are immunosuppressive and trigger thosethat are immunostimulators [18]. Both in vitro and in vivostudies, in both animal and human models, have shownsimilar immunomodulatory effects of certain AHLs [19–24]. The role of QS in P. aeruginosa pathogenicity has alsobeen confirmed via both animal and human models[25–28].

All of this evidence points out the importance ofdetecting and quantifying QS signal molecules. Methodsthat can detect these molecules in clinical samples shouldallow for the early diagnosis and monitoring of bacteria-related diseases, both in primary and in secondaryinfections. This should be especially important in caseswhere traditional antibiotic therapy fails. There are severalreports that discuss the development of conventionalphysical-chemical methods, such as high-performanceliquid chromatography (HPLC) with UV detection [29],gas chromatography (GC) with mass spectrometry (MS)detection [30, 31], HPLC-MS [32, 33], ultra performanceliquid chromatography (UPLC), ultra-high-resolution MSand nano liquid chromatography (LC) coupled with MS[34], nuclear magnetic resonance, fast atom bombardmentMS and chemical ionization MS [35], capillary zoneelectrophoresis coupled with MS [36], micellar electroki-netic chromatography with MS detection [37], thin-layerchromatography [38], colorimetry [39], UPLC with diode-array detection [40] and LC coupled with tandem MS (MS-MS) [41, 42], for the identification and quantification ofAHLs in bacterial culture media. Additionally, there are afew types of bacterial whole-cell sensing systems that havebeen engineered for the detection of AHLs. These includereporter strains based on bacteria, such as Escherichia coli,Agrobacterium tumefaciens, and P. aeruginosa [43], whichutilize different QS regulatory systems and incorporatereporter genes, such as lacZ, gfp or lux, coding for β-galactosidase, green fluorescent protein (GFP) and bacterial

luciferase, respectively. Genetically engineered whole-cellsensing systems with the reporter protein β -galactosidasehave been employed to study the mechanism of bacterialcommunication in certain bacteria such as A. tumefaciens,which is responsible for causing crown gall tumors inplants [38, 44]. Shaw et al. demonstrated the presence ofAHLs in bacterial culture supernatants using this sensingsystem. Awhole-cell sensing system employing a gfp-basedgene construct allowed for the study of bacterial chatteringin situ and at the single-cell level, in several bacterialspecies, such as S. liquefaciens, Serratia ficaria, E. coli,P. aeruginosa and Pseudomonas aureofaciens [45]. Thegfp-based whole-cell sensing systems are the mostcommonly used in the study of biofilms [46]. Severalbioluminescence-based whole-cell sensing systems thatincorporate luxAB [47] and luxCDABE reporter geneconstructs [48–50] have been also engineered for thedetection of AHLs. A different approach to the whole-cellsensing of AHLs is based on the use of a violaceum-negativemutant (CV026) of Chromobacterium violaceum [51–53].The wild-type of this Gram-negative bacterium produces apurple pigment, violaceum [51, 52]. The mutant CV026produces the purple color only in the presence of short-chainAHLs (four to eight carbon atoms in the side chain), while itdoes not show any response towards long-chain AHLs (tento 14 carbon atoms in the side chain). However, these long-chain AHLs can be detected indirectly owing to their abilityto inhibit the pigment production in response to activatingshort-chain AHLs, such as N-hexanoyl-DL-homoserinelactone (C6-HSL) [51]. Using this system, McClean et al.[51] detected AHLs in the supernatants of S. liquefaciensand Yersinia enterocolitica cultures.

Although various analytical methods have been used forthe detection of AHLs in bacterial culture media, only a fewof them have been applied to AHL monitoring inphysiological samples. In this review, we will focus onthose articles, by us and others, which describe thedetection of AHLs in biological and clinical samples byemploying both physical-chemical techniques and whole-cell-based biosensors. We will also discuss some of ourexperimental results, based on the detection of AHLs in abiological matrix such as saliva, using LC-MS-MS.

Experimental

Materials

All AHL standard compounds were obtained from Sigma(St. Louis, MO, USA). The AHLs used were homoserinelactone (HSL), N-butyryl-DL-homoserine lactone (C4-HSL),N-hexanoyl-DL-homoserine lactone (C6-HSL), N-(3-oxo)-hexanoyl-DL-homoserine lactone (3-oxo-C6-HSL), N-hepta-

Fig. 2 The AHL-dependent LuxI/LuxR regulatory system

Anal Bioanal Chem (2008) 391:1619–1627 1621

noyl-DL-homoserine lactone (C7-HSL), N-octanoyl-DL-homo-serine lactone (C8-HSL), N-(3-oxo)-octanoyl-L-homoserinelactone (3-oxo-C8-HSL), N-decanoyl-DL-homoserine lactone(C10-HSL), N-dodecanoyl-DL-homoserine lactone (C12-HSL)and N-tetradecanoyl-DL-homoserine lactone (C14-HSL). Ace-tonitrile was of HPLC grade from Fisher Scientific (Pitts-burgh, PA, USA). Methylene chloride (Optima) and methanol(Optima) were also purchased from Fisher Scientific(Pittsburgh, PA, USA). HPLC-grade distilled deionized water(Super-Q Water Purification System, Millipore, Bedford, MA,USA) was used for all the analytical runs. The micro-centrifuge was purchased from Eppendorf (Westbury,NY, USA).

Preparation of standard solutions

Commercially available standard AHLs were dissolved inacetonitrile to obtain 1,000 μg/mL solutions. Twentymicroliters of each of these 1,000 μg/mL solutions of C4-HSL, C6-HSL, 3-oxo-C6-HSL, C8-HSL, 3-oxo-C8-HSL,C10-HSL, C12-HSL and C14-HSL was added to 840 μL ofmethanol-water (35:65, v/v) with 0.1% formic acid, in orderto obtain a stock mixture solution containing 20 μg/mL ofeach of the eight AHLs. From that, AHL standard mixturesolutions, which contained 50, 25, 10, 5, 1 and 0.5 ng/mLof each AHL, respectively, were prepared for the calibra-tion. C7-HSL was used as an internal standard for all thestandard mixtures and samples analyzed. A specific volumeof 1 μg/mL solution of C7-HSL in methanol-water (35:65,v/v) with 0.1% formic acid was added to each of thesestandard mixture solutions to achieve a final internalstandard concentration of 25 ng/mL. These standard sol-utions with the added internal standard served as laboratorycontrol spike solutions. Similarly, 1 μg/mL internal standardsolution was added to water to obtain 25 ng/mL con-centration in the final solution. The resulting solution servedas a blank.

Sample collection and preparation

The saliva samples were obtained from patients with GIdisorders and healthy volunteers. The saliva samples werecollected in the mornings after the patients had brushedtheir teeth to minimize bulk collection of debris and thepresence of oral bacteria. The samples were processed bysimple centrifugation in sterile Eppendorf tubes at13,000 rpm for 2–3 min, at room temperature. The super-natants collected were then stored at −80 °C until they wereanalyzed. For the development and optimization of the LC-MS-MS method for the analysis of real samples, thesupernatants of the individual saliva samples were pooledto obtain four representative pooled saliva samples; spe-cifically, two from the healthy volunteers and two from

patients with GI disorders. The final volumes of the fourpooled saliva samples were 13.5 and 14.25 mL for thehealthy volunteers, and 7.2 and 9 mL for the patients withGI disorders. The four pooled saliva samples from healthyvolunteers and GI patients were extracted three times withmethylene chloride. The extracts were dried over anhydroussodium sulfate, decanted and evaporated to dryness usingnitrogen gas flow. Each of these extracted samples wasreconstituted in 243.75 μL of methanol-water (35:65, v/v)with 0.1% formic acid. A volume of 6.25 μL of 1 μg/mLC7-HSL was added to each sample to achieve 25 ng/mL inthe final solution.

Reversed-phase HPLC

Dual Varian Prostar 210 HPLC pumps and a Varian Prostar210 autosampler were used. These were purchasedfrom Varian (Palo Alto, CA, USA). The HPLC column,SecurityGuard kit and SecurityGuard cartridges (4 mm×2 mm) were purchased from Phenomenex (Torrance, CA,USA). The HPLC column used was Gemini C18, 50 mm×2 mm, 5 μm particle size. Twenty five microliters of AHLstandard mixtures, blanks or samples in mobile phasemethanol-water (35:65, v/v) with 0.1% formic acid wasinjected via microliter pickup in the column, at a flow rateof 0.25 mL/min. The elution method utilized for HPLCseparation of AHLs included an isocratic profile ofmethanol-water (35:65, v/v) for 5 min, followed by a lineargradient from 35 to 95% methanol in water over 33 min. Asubsequent linear gradient from 95 to 35% methanol inwater over 2 min, and an isocratic profile of methanol-water(35:65, v/v) for 10 min were used for flushing the columnfor the following run.

Mass-spectrometric detection

The HPLC column was connected to an electrosprayionization (ESI) chamber which was used in positive ionmode. Nitrogen was used as the drying gas at 20 psi and300 °C. The needle and shield voltage were maintained at5,000 and 600 V, respectively. The pressure for thenebulizing gas, nitrogen, and the temperature for the ESIhousing were kept at 50 psi and 50 °C, respectively. Thepositive ion ESI chamber was interfaced to a Varian 1200 Lquadrupole mass spectrometer. The [M + H]+ ions weremonitored for each AHL.

MS-MS detection

The [M + H]+ ions were subjected to collision-induceddissociation using argon as the collision gas at 2 mTorrpressure. The voltages for the capillary and the electronmultiplier were 40 and 1,600 V, respectively. Specific

1622 Anal Bioanal Chem (2008) 391:1619–1627

daughter ions of each AHL, that is, [M + H-101]+ and [M-H2O]

+ ions, were monitored.

Results and discussion

Detection of AHLs in biological and clinical samples

As an example of AHL detection in biological samples, wereport the work of Taylor et al. [54], who analyzed differenttypes of marine organisms, such as sponges, macroalgae,sea grasses, bryozoans, ascidians, and corals, in a studyaimed at showing the presence of AHL-producing bacteriaassociated with these organisms. Chemical extracts ofsponge tissue were analyzed by a combination of methods,such as a bacterial cell sensing system, solid-phase extrac-tion and GC-MS. AHLs were found to be present in all ofthe abovementioned organisms. Wu et al. [11] analyzed thelung tissues of mice that were experimentally infected withP. aeruginosa. A GFP-based bacterial sensing system wasused for in vivo studies of these infected lung tissues.AHLs were detected mainly in the tissues with severepathological conditions, thus confirming the associationbetween QS and P. aeruginosa pathogenicity. In a similartype of experiment, Sokol et al. [55] utilized a bacterialcell sensing system for the detection of AHLs in lunghomogenates of mice infected with B. cenocepacia.They used a traI-luxCDABE reporter gene system tomainly detect C8-HSL. The levels of this AHL werefound to correlate with those of inflammatory mediators,thus demonstrating the role of QS in the severity of B.cenocepacia infections. Jacobi et al. [56] performed studiesboth in vitro and in vivo in liver and spleen homogenates ofmice that were infected with Y. enterocolitica serotype O8,which is highly pathogenic for mice. They utilized abioluminescent bacterial cell sensing system for thedetection of AHLs. Middleton et al. [4] detected AHLsdirectly in sputum samples from CF patients colonized byP. aeruginosa and B. cepacia. Bacterial cell sensingsystems were employed for the detection of the presenceof AHLs. Subsequently, LC coupled with high-resolutionMS was used for the identification studies of the AHLsdetected. The AHLs identified in the sputum samples fromCF patients infected by P. aeruginosa were C6-HSL and N-(3-oxo)-dodecanoyl-DL-homoserine lactone, which are twoof the major cognate molecules produced by P. aeruginosa.

Erickson et al. [57] also detected AHLs directly in thesputum samples of CF patients. Bacterial whole-cell-basedbiosensing systems were employed for the detection ofAHLs. The results obtained indicated that QS systems areactive in the lungs of CF patients.

In another report, AHLs were detected in mucopurulentrespiratory secretions from CF patients [9]. Extracts of

these samples were prefractionated by fast-pressure LC, andthen a bacterial cell sensing system was used for thedetection of AHLs.

Favre-Bonte et al. [58] detected AHLs directly in thelung tissue of two CF patients employing bacterial cellsensing systems. Infections by P. aeruginosa have beenreported to be common in lung allograft recipients,especially when transplants are chronically rejected. Bron-choalveolar lavage (BAL) fluid samples taken from stablelung transplant recipients were analyzed for the presence ofAHLs using a bacterial cell sensing system. Long-chainAHLs (C10–C14) were detected in these BAL fluidsamples, even in the absence of rejection or apparentinfection [59].

As reported above, various analytical approaches havebeen employed for the detection of AHLs in clinicalsamples as well as in samples from animal models ofdisease. However, data on the pathophysiological concen-trations of AHLs in clinical samples are scant. Specifically,AHLs have been detected at picomolar to micromolarconcentrations in clinical samples such as sputum, muco-purulent respiratory secretions and lung tissues of CFpatients [4, 9, 57, 58], and in saliva and stools of healthyand GI disease subjects [60]. The high sensitivity andselectivity of whole-cell sensing systems allow for thedetection of very low levels of analytes, down tosubattomoles [61]. Whole-cell sensing systems can beused on samples directly, or upon simple, minimal samplepreparation, thus allowing accurate analyte quantificationand short analysis times. The fast responses of whole-cell sensing systems along with multiplexing featuresalso enhance the rapidity of the analytical process. Inaddition, they bear the added advantage, over conven-tional analytical methods, of providing information onthe bioavailability of the analyte. Although the limits ofdetection attainable with physical-chemical methods maybe sufficient for the detection of AHLs in certain clinicalsamples, and multiple analytes can be identified simul-taneously in one sample, some drawbacks exist. Themain problems lie in the lengthy and tedious proceduresnecessary for sample preparation, and in the overallduration of the analytical process, which is affected bythe long analysis time required for each sample and theinability to analyze multiple samples in one analytical run.Finally, it is important to note that while physical-chemicalmethods are long-established techniques that are oftenchosen as reference methods, whole-cell sensing systemshave improved significantly since their first reporteddevelopment and use for the detection of naphthalene byKing et al. [62] more than 15 years ago. At present, thereliability of whole-cell sensing systems, at least in con-trolled laboratory environments, has been demonstrated inwork by us and others [63–67].

Anal Bioanal Chem (2008) 391:1619–1627 1623

Detection of AHLs in saliva and stool using whole-cellsensing systems

Since the involvement of QS in several bacterial infectionsis well supported, we previously hypothesized that AHLsproduced by bacteria could serve as potential biomarkers inthe management of bacterial diseases and, thus, monitoringthem in biological samples may be a significant analyticaltool for the investigation of such diseases [60]. Specifically,altered bacteria-host interaction has been implicated inseveral health imbalances and disorders. For instance, thereare several reports in the literature that support the role ofbacteria in GI disorders such as inflammatory bowel diseaseand irritable bowel syndrome [68–70]. However, themechanisms underlying bacterial involvement in thesediseases have not been fully elucidated. In these disorderscommensal bacteria may become pathogenic. The roles ofhost genetics and immune response are also being investi-gated. Variations in the host may also affect bacterialbehavior. To investigate whether bacterial behavior plays arole in these disorders, we developed a method employingwhole-cell biosensing systems for the detection of AHLs inbiological matrixes such as saliva and stool [60]. Asreported in the previous section of this review, whole-cellbiosensing systems have been employed to study QS.However, there is a paucity of rigorous studies thatthoroughly assess the analytical performance of thesebacterial whole-cell sensing systems, especially whenapplied to physiological samples. To that end, we analyt-ically characterized bacterial whole-cell sensing systems in

terms of limit of detection, dynamic range, reproducibilityand matrix effects when applied to physiological samples.In our studies, saliva and stool samples were chosen foranalysis since the AHL concentration in these biologicalmatrixes may reflect their systemic and GI concentration,respectively [71, 72]. Also, the additional benefit ofchoosing these biological matrixes is that these samplesare collected noninvasively and can be analyzed withminimal sample pretreatment. The whole-cell sensingsystems employed for the detection of long- and short-chain AHLs were based on genetically engineered bacteriaharboring plasmids that bear the genes of the lasRI andrhlRI regulatory systems along with those of the luciferasereporter. Using our whole-cell sensing systems, we suc-cessfully demonstrated the presence of AHLs in saliva andstool samples from both healthy subjects and individualswith GI disorders. The method developed was sensitive,rapid and cost-effective. We also detected AHLs in salivaby an LC-MS-MS method, as discussed in the next section.

Table 2 LC- tandem MS detection of AHLs

AHLs [M + H-101]+ (m/z) [M - H2O]+ (m/z)

C4-HSL 71 (100) 154 (38.6)3-Oxo-C6-HSL 113 (100) 196 (7.6)C6-HSL 99 (100) 182 (14.1)3-Oxo-C8-HSL 141 (100) 224 (3.1)C7-HSL 113 (100) 196 (10.6)C8-HSL 127 (100) 210 (10.9)C10-HSL 155 (100) 238 (20.2)C12-HSL 183 (100) 266 (55.0)C14-HSL 211 (100) 294 (85.4)

The relative abundance (percentage) is shown in parentheses

Fig. 3 Liquid chromatography-tandem mass spectrometry (LC-MS-MS) chromatogram of standard AHLs, each at 50 ppb concentration,in a mixture that was spiked with 25 ppb C7-HSL internal standard

Table 1 Liquid chromatography (LC)-mass spectrometry (MS)detection of N-acyl homoserine lactones (AHLs)

AHLs [M + H]+ (m/z) Retention time (min)

Standardmixture

Spikedsaliva

C4-HSL 172 1.12 1.223-Oxo-C6-HSL 214 1.29 1.35C6-HSL 200 3.04 3.053-Oxo-C8-HSL 242 4.15 4.12C7-HSL 214 6.99 6.91C8-HSL 228 12.81 12.77C10-HSL 256 20.37 20.33C12-HSL 284 25.55 25.49C14-HSL 312 29.44 29.37

C4-HSL N-butyryl-DL-homoserine lactone, 3-oxo-C6-HSL N-(3-oxo)-hexanoyl-DL-homoserine lactone, C6-HSL N-hexanoyl-DL-homoserinelactone, 3-oxo-C8-HSL N-(3-oxo)-octanoyl-L-homoserine lactone,C7-HSL N-heptanoyl-DL-homoserine lactone, C8-HSL N-octanoyl-DL-homoserine lactone, C10-HSL N-decanoyl-DL-homoserine lactone,C12-HSL N-dodecanoyl-DL-homoserine lactone, C14-HSL N-tetradecanoyl-DL-homoserine lactone

1624 Anal Bioanal Chem (2008) 391:1619–1627

Detection of AHLs in saliva using LC-MS-MS

HPLC separation and MS detection of AHLs

Mixtures containing various concentrations of commercial-ly available AHLs, specifically, C4-HSL, C6-HSL, 3-oxo-C6-HSL, C7-HSL, C8-HSL, 3-oxo-C8-HSL, C10-HSL,C12-HSL and C14-HSL, were injected in the reversed-phase HPLC column to establish the respective retentiontimes. The mobile phase was methanol-water, with 0.1%formic acid that enabled sharpening of peaks, thus in-creasing resolution. Different ratios of methanol to waterwere tried for the initial isocratic profile, from 80 to 35%methanol in water. The chosen mobile phase ratio ofmethanol to water was 35:65, v/v, which reduced thedragging and broadening of the peaks, and increased theirresolution compared with the others. Most of the com-pounds analyzed were clearly separated in the experimentalconditions chosen, as shown by their retention times(Table 1). These retention times were highly reproducible

over all the subsequent runs, demonstrating precise con-ditions and reliability of the instrument. We utilized thespecificity offered by [M + H]+ ions to distinguish AHLs,especially the closely eluted C4-HSL and 3-oxo-C6-HSL(Table 1). Although standard mixtures of AHLs could beseparated and identified in approximately 30 min usingHPLC-MS, it has to be taken into account that saliva is acomplex biological matrix that is more likely to containinterfering molecules. To that end, we employed HPLC-MS-MS to improve the specificity and sensitivity of AHLdetection in the saliva matrix.

MS-MS detection of AHLs

MS-MS fragmentation of AHLs produces three kinds ofdaughter ions for the [M + H]+ ions: [M + H-101]+, [M + H-H2O]

+, and [M + H-CO]+ [41]. The most abundant daughterions, specifically, [M + H-101]+ and [M + H-H2O]

+, weremonitored for the identification of AHLs (Table 2). Theidentification of the AHLs was based on their specificretention times in the HPLC column, their specific daughter

Fig. 4 LC-MS-MS chromatograms of two pooled saliva samplesfrom healthy volunteers: a pooled saliva 1; b pooled saliva 2

Fig. 5 LC-MS-MS chromatograms of two pooled saliva samplesfrom diseased individuals: a pooled saliva 3; b pooled saliva 4

Anal Bioanal Chem (2008) 391:1619–1627 1625

ions and the relative abundance of the daughter ions of eachAHL. As described in “Experimental,” all the AHL standardmixtures and samples analyzed were spiked with the sameconcentration (25 ng/mL) of the internal standard. Theinternal standard chosen was C7-HSL, which is rarelyproduced by bacteria [42]. It has also been reported that itssignal intensity is not generally affected, irrespective of thematrix type in which it is detected [41]. Although the [M +H]+ ions and both of the daughter ions monitored for C7-HSL and 3-oxo-C6-HSL proved to be the same, theirrespective retention times and the segments in which theywere monitored differed significantly (Tables 1 and 2). Thus,C7-HSL could be employed as an internal standard.

MS-MS detection of AHLs in standard mixtures

Standard mixtures containing 0.5, 1, 5, 10, 25 and 50 ng/mLconcentrations of each of the eight AHLs tested in thisstudy were analyzed for calibration purposes. A represen-tative LC-MS-MS chromatogram demonstrating the detec-tion of standard AHLs at 50 ng/mL concentration in amixture, spiked with 25 ng/mL C7-HSL internal standard,is shown in Fig. 3. The method for acquisition of theselected daughter ions of the AHLs in a mixture was seg-mented depending on the retention time of the parentmolecule. Segmentation allowed for the selection anddetection of only the specific daughter ions of the parent[M + H]+ ion(s). All of the AHLs analyzed were quantifieddown to a concentration of 0.5 ng/mL. For each AHL,calibration plots were obtained by plotting the normalizedmeasured peak areas as a function of concentration. Thepeak areas were normalized by dividing the peak area ofeach AHL with that of the internal standard. The measuredsignals correlated with the concentration of the AHL in alinear relationship over the range of concentrations testedfrom 0.5 to 50 ng/mL. The coefficients of regression (r2)for each of the eight AHLs ranged from 0.992 to 0.997.

MS-MS detection of AHLs in saliva samples

After proving the ability of the LC-MS-MS method toseparate and identify the AHLs tested, we applied it to theanalysis of four representative saliva samples that wereprepared by pooling specimens from healthy individuals andsubjects with GI disorders, respectively. The results obtaineddemonstrate the presence of AHLs in saliva of both healthyand diseased individuals (Figs. 4 and 5). Owing to thelimited number of samples analyzed, it proved to bedifficult to establish a trend in the type and concentrationof AHLs that may be present in either healthy or diseasedindividuals. However, the analysis of individual samplesfrom significant numbers of healthy controls and patientswould certainly provide this type of information.

Conclusion

The data reported herein show that the detection of AHLsin biological, clinical samples may have implications ininvestigating the causative mechanisms of bacteria-relateddiseases. Additionally, the levels of AHLs in biologicalmatrixes may correlate with the health status of subjects.Thus, AHLs may serve as biomarkers of bacteria-relateddisorders and could potentially be employed in earlydiagnosis and monitoring of disease activity. Various methodshave been developed for the detection of AHLs inphysiological samples. However, there is still a need formore methods to detect not only AHLs but also the otherchemical signaling molecules, such as the oligopeptides andAI-2. This would allow us to gain a better understanding ofthe role of QS in bacterial disease and would help indesigning better agents to combat diseases that havebacteria playing a primary or a secondary role in theirpathogenesis.

Acknowledgements This work was partly supported by the NationalScience Foundation, grant CHE-0416553 and the Broad Foundation,Broad Medical Research Program (BMRP), grant IBD-0198R. Wethank the Office of the Vice President of Research of the University ofKentucky for a University Research Professorship to S.D. A.K. andS.D. acknowledge support from a Gill Fellowship and a Gill EminentProfessorship, respectively. A.K. also acknowledges support from theResearch Challenge Trust Fund.

References

1. Schauder S, Bassler BL (2001) Genes Dev 15:1468–14802. Bassler BL, Losick R (2006) Cell 125:237–2463. Nealson KH, Platt T, Hastings JW (1970) J Bacteriol 104:313–3224. Middleton B, Rodgers HC, Camara M, Knox AJ, Williams P,

Hardman A (2002) FEMS Microbiol Lett 207:1–75. Eberl L, Winson MK, Sternberg C, Stewart GSAB, Christiansen

G, Chhabra SR, Bycroft B, Williams P, Molin S, Givskov M(1996) Mol Microbiol 20:127–136

6. Miller MB, Bassler BL (2001) Annu Rev Microbiol 55:165–1997. Kleerebezem M, Quadri LEN, Kuipers OP, De Vos WM (1997)

Mol Microbiol 24:895–9048. Engebrecht J, Silverman M (1987) Nucleic Acids Res 15:10455–

104679. Chambers CE, Visser MB, Schwab U, Sokol PA (2005) FEMS

Microbiol Lett 244:297–30410. Kjelleberg S, Molin S (2002) Curr Opin Microbiol 5:254–25811. Wu H, Song Z, Hentzer M, Andersen JB, Heydorn A, Mathee K,

Moser C, Eberl L, Molin S, Hoiby N, Givskov M (2000)Microbiology 146:2481–2493

12. Surette MG, Bassler BL (1998) Proc Natl Acad Sci USA95:7046–7050

13. Sperandio V, Mellies JL, Nguyen W, Shin S, Kaper JB (1999)Proc Natl Acad Sci USA 96:15196–15201

14. Falcao JP, Sharp F, Sperandio V (2004) Curr Issues IntestMicrobiol 5:9–18

15. Sircili MP, Walters M, Trabulsi LR, Sperandio V (2004) InfectImmun 72:2329–2337

16. Kaper JB, Sperandio V (2005) Infect Immun 73:3197–3209

1626 Anal Bioanal Chem (2008) 391:1619–1627

17. Kendall MM, Sperandio V (2007) Curr Opin Gastroenterol23:10–15

18. Smith KM, Bu Y, Suga H (2003) Chem Biol 10:81–8919. DiMango E, Zar HJ, Bryan R, Prince A (1995) J Clin Invest

96:2204–221020. Telford G, Wheeler D, Williams P, Tomkins PT, Appleby P,

Sewell H, Stewart GSAB, Bycroft BW, Pritchard DI (1998) InfectImmun 66:36–42

21. Smith RS, Fedyk ER, Springer TA, Mukaida N, Iglewski BH,Phipps RP (2001) J Immunol 167:366–374

22. Wu H, Song Z, Givskov M, Doring G, Worlitzsch D, Mathee K,Rygaard J, Hoiby N (2001) Microbiology 147:1105–1113

23. Smith RS, Harris SG, Phipps R, Iglewski B (2002) J Bacteriol184:1132–1139

24. Miyairi S, Tateda K, Fuse ET, Ueda C, Saito H, Takabatake T,Ishii Y, Horikawa M, Ishiguro M, Standiford TJ, Yamaguchi K(2006) J Med Microbiol 55:1381–1387

25. Tang HB, DiMango E, Bryan R, Gambello M, Iglewski BH,Goldberg JB, Prince A (1996) Infect Immun 64:37–43

26. Rumbaugh KP, Griswold JA, Iglewski BH, Hamood AN (1999)Infect Immun 67:5854–5862

27. de Kievit TR, Iglewski BH (2000) Infect Immun 68:4839–484928. Rumbaugh KP, Hamood AN, Griswold JA (2004) J Surg Res

116:137–14429. Reimmann C, Beyeler M, Latifi A, Winteler H, Foglino M,

Lazdunski A, Haas D (1997) Mol Microbiol 24:309–31930. Charlton TS, de Nys R, Netting A, Kumar N, Hentzer M, Givskov

M, Kjelleberg S (2000) Environ Microbiol 2:530–54131. Cataldi TRI, Bianco G, Palazzo L, Quaranta V (2007) Anal

Biochem 361:226–23532. Zhang L, Murphy PJ, Kerr A, Tate ME (1993) Nature 362:446–44833. Lithgow JK, Wilkinson A, Hardman A, Rodelas B, Wisniewski-

Dyé F, Williams P, Downie JA (2000) Mol Microbiol 37:8134. Fekete A, Frommberger M, Rothballer M, Li X, Englmann M,

Fekete J, Hartmann A, Eberl L, Schmitt-Kopplin P (2007) AnalBioanal Chem 387:455–467

35. Pearson JP, Gray KM, Passador L, Tucker KD, Eberhard A, IglewskiBH, Greenberg EP (1994) Proc Natl Acad Sci USA 91:197–201

36. Frommberger M, Hertkorn N, Englmann M, Jakoby S, HartmannA, Kettrup A, Schmitt-Kopplin P (2005) Electrophoresis 26:1523–1532

37. Frommberger M, Schmitt-Kopplin P, Menzinger F, Albrecht V,Schmid M, Eberl L, Hartmann A, Kettrup A (2003) Electropho-resis 24:3067–3074

38. Shaw PD, Ping G, Daly SL, Cha C, Cronan JE Jr, Rinehart KL,Farrand SK (1997) Proc Natl Acad Sci USA 94:6036–6041

39. Yang Y-H, Lee T-H, Kim JH, Kim EJ, Joo H-S, Lee C-S, KimB-G (2006) Anal Biochem 356:297–299

40. Englmann M, Fekete A, Kuttler C, Frommberger M, Li X,Gebefugi I, Fekete J, Schmitt-Kopplin P (2007) J Chromatogr A1160:184–193

41. Morin D, Grasland B, Vallee-Rehel K, Dufau C, Haras D (2003) JChromatogr A 1002:79–92

42. Ortori C, Atkinson S, Chhabra S, Cámara M, Williams P, BarrettD (2007) Anal Bioanal Chem 387:497–511

43. Brelles-Marino G, Bedmar EJ (2001) J Biotechnol 91:197–20944. Fuqua C, Winans SC (1996) J Bacteriol 178:435–44045. Andersen JB, Heydorn A, Hentzer M, Eberl L, Geisenberger O,

Christensen BB, Molin S, Givskov M (2001) Appl EnvironMicrobiol 67:575–585

46. De Kievit TR, Gillis R, Marx S, Brown C, Iglewski BH (2001)Appl Environ Microbiol 67:1865–1873

47. Swift S, Winson MK, Chan PF, Bainton NJ, Birdsall M, ReevesPJ, Rees CED, Chhabra SR, Hill PJ, Throup JP, Bycroft BW,Salmond GPC, Williams P, Stewart GSAB (1993) Mol Microbiol10:511–520

48. Swift S, Karlyshev AV, Fish L, Durant EL, Winson MK, ChhabraSR, Williams P, Macintyre S, Stewart GS (1997) J Bacteriol179:5271–5281

49. Winson MK, Camara M, Latifi A, Foglino M, Chhabra SR,Daykin M, Bally M, Chapon V, Salmond GPC et al (1995) ProcNatl Acad Sci USA 92:9427–9431

50. Winson MK, Swift S, Fish L, Throup JP, Jorgensen F, ChhabraSR, Bycroft BW, Williams P, Stewart GSAB (1998) FEMSMicrobiol Lett 163:185–192

51. McClean KH, Winson MK, Fish L, Taylor A, Chhabra SR,Camara M, Daykin M, Lamb JH, Swift S, Bycroft BW, StewartGS, Williams P (1997) Microbiology 143:3703–3711

52. McLean RJC, Pierson LS III, Fuqua C (2004) J MicrobiolMethods 58:351–360

53. Burmolle M, Hansen LH, Oregaard G, Sorensen SJ (2003) MicrobEcol 45:226–236

54. Taylor MW, Schupp PJ, Baillie HJ, Charlton TS, de Nys R,Kjelleberg S, Steinberg PD (2004) Appl Environ Microbiol70:4387–4389

55. Sokol PA, Sajjan U, Visser MB, Gingues S, Forstner J, Kooi C(2003) Microbiology 149:3649–3658

56. Jacobi CA, Bach A, Eberl L, Steidle A, Heesemann J (2003)Infect Immun 71:6624–6626

57. Erickson DL, Endersby R, Kirkham A, Stuber K, Vollman DD,Rabin HR, Mitchell I, Storey DG (2002) Infect Immun 70:1783–1790

58. Favre-Bonte S, Pache J-C, Robert J, Blanc D, Pechere J-C, vanDelden C (2002) Microb Pathog 32:2002143–147

59. Ward C, Camara M, Forrest I, Rutherford R, Pritchard G, DaykinM, Hardman A, de Soyza A, Fisher AJ, Williams P, Corris PA(2003) Thorax 58:444–446

60. Kumari A, Pasini P, Deo SK, Flomenhoft D, Shashidhar H,Daunert S (2006) Anal Chem 78:7603–7609

61. Feliciano J, Pasini P, Deo SK, Daunert S (2006) Photoproteins asreporters in whole-cell sensing. Wiley-VCH, Weinheim

62. King JMH, DiGrazia PM, Applegate B, Burlage R, Sanseverino J,Dunbar P, Larimer F, Sayler GS (1990) Science 249:778–781

63. Daunert S, Barrett G, Feliciano JS, Shetty RS, Shrestha S, Smith-Spencer W (2000) Chem Rev 100:2705–2738

64. Belkin S (2003) Curr Opin Microbiol 6:206–21265. Gu MB, Mitchell RJ, Kim BC (2004) Adv Biochem Eng

Biotechnol 87:269–30566. Galluzzi L, Karp M (2006) Comb Chem High Throughput Screen

9:501–51467. Yagi K (2007) Appl Microbiol Biotechnol 73:1251–125868. Sartor RB (2004) Gastroenterology 126:1620–163369. Chandran P, Satthaporn S, Eremin O, Robins A (2003) Surgeon

1:12570. Baumgart M, Dogan B, Rishniw M, Weitzman G, Bosworth B,

Yantiss R, Orsi RH, Wiedmann M, McDonough P, Kim SG, BergD, Schukken Y, Scherl E, Simpson KW (2007) ISME J 1:403–418

71. Kaufman E, Lamster IB (2002) Crit Rev Oral Biol Med 13:197–21272. Crama-Bohbouth G, Kan PL-V, Weterman IT, Biemond I, Peña

AS (1984) Dig Dis Sci 29:1089–1092

Anal Bioanal Chem (2008) 391:1619–1627 1627