Analytical and Bioanalytical Chemistry

, 387:481

The quorum-sensing molecule N-3-oxododecanoyl homoserine lactone (3OC12-HSL) enhances the host defence by activating human polymorphonuclear neutrophils (PMN)

Authors

  • Christof Wagner
    • Institut für Immunologie der Universität Heidelberg
    • Klinik für Unfall- und WiederherstellungschirurgieBerufsgenossenschaftliche Unfallklinik Ludwigshafen
  • Sabine Zimmermann
    • Institut für Immunologie der Universität Heidelberg
  • Gerald Brenner-Weiss
    • Institut für Technische ChemieForschungszentrum Karlsruhe
  • Friederike Hug
    • Institut für Immunologie der Universität Heidelberg
  • Birgit Prior
    • Institut für Immunologie der Universität Heidelberg
  • Ursula Obst
    • Institut für Technische ChemieForschungszentrum Karlsruhe
    • Institut für Immunologie der Universität Heidelberg
Original Paper

DOI: 10.1007/s00216-006-0698-5

Cite this article as:
Wagner, C., Zimmermann, S., Brenner-Weiss, G. et al. Anal Bioanal Chem (2007) 387: 481. doi:10.1007/s00216-006-0698-5

Abstract

The P. aeruginosa quorum-sensing molecule N-3-oxododecanoyl homoserine lactone (3OC12-HSL) interacts not only with bacteria, but also with mammalian cells, among others with those of the immune defence system. We focussed on the possible interaction of 3OC12-HSL with human polymorphonuclear neutrophils (PMN), because these cells are the first to enter an infected site. We found that 3OC12-HSL attracts PMN, and up-regulates expression of receptors known to be involved in host defence, including the adhesion proteins CD11b/CD18 and the immunoglobulin receptors CD16 and CD64. Furthermore, the uptake of bacteria (phagocytosis), which is crucial for an efficient defence against infection, was enhanced. Thus, recognising and responding to 3OC12-HSL not only attracts the PMN to the site of a developing biofilm, but also reinforces their defence mechanisms, and hence could be a means to control the infection in an early stage and to prevent biofilm formation.

Keywords

Quorum sensingPMNFunctional activityHomoserine lactoneHost defence

Introduction

Bacterial biofilms are increasingly recognised as a major cause of persistent and destructive infectious diseases [1, 2]. Among those P. aeruginosa infections are especially devastating in immunocompromised patients [35]. Because the relative resistance of bacteria in biofilms towards antibiotics limits the therapeutic options, the question arises whether or not interference with biofilm formation could be a reasonable alternative [1, 6]. In that regard the quorum-sensing molecules could be suitable targets for an interventional therapy, because they are produced and released in the very early stages of biofilm formation when the bacteria are still more or less planctonic and susceptible to the immune defence mechanisms.

While structural and functional aspects of quorum-sensing molecules are discussed elsewhere in this journal, we like to draw the attention to the fact that quorum-sensing molecules, particularly those of P. aeruginosa, are not only efficient operators within the bacterial community, but also interact with mammalian cells, including epithelial cells, fibroblasts [79] and also cells of the adaptive or innate immune response [1016].

Recognition of bacteria and bacterial products by cells of the immune defence is essential for the activation of the cellular effector functions, such as the efficient uptake, the phagocytosis, and intracellular killing, e.g. by the production of oxygen radicals. Recognition is accomplished by so-called pattern recognition receptors, which are selective for groups of bacterial molecules, including lipopolysaccharides, lipoteichoic acid, or peptidoglycans. Not only phagocytic cells, the major effector cells of the innate, non-adaptive immune response, express these receptors, but also T-lymphocytes, the protagonists of the specific, adaptive immune response. Triggering these receptors is not only essential for the effector function of the immune system (e.g. phagocytosis and killing), but most probably also contribute to the initiation of the adaptive immune response [17, 18].

The majority of studies have so far focussed on membrane-derived bacterial products such as lipopolysaccharides, peptidoglycans or on bacterial DNA and their receptors in immunocompetent cells [19, 20]. Only more recently, an immunmodulatory activity of the quorum-sensing molecules of P. aeruginosa was proposed, particularly of N-3-oxododecanoyl homoserine lactone (3OC12-HSL). This proposition is based on the observation that 3OC12-HSL inhibits the proliferation of T-lymphocytes and modulates the cytokine synthesis in various in vitro models [1014]. Moreover, enhancement of phagocytosis by macrophages of yeast cells was shown [15].

In light of the facts that quorum-sensing molecules are central to the development of bacterial biofilms, and that polymorphonuclear neutrophils (PMN) represent the first line defence against bacterial infections, the interaction of PMN with developing biofilms or their products is of utmost interest.

Indeed, we observed induction of directed migration, chemotaxis, of human PMN by 3OC12-HSL [16]. These findings are in apparent contrast to previous data by Tadeta et al. [21], who showed induction of apoptosis in mouse macrophages and in neutrophils. Apparently, species differences account for the discrepant findings. This assumption is supported by the work by Vikström et al. [15], who were also using human cells and did not observe apoptosis either.

In the present study we further examined the effect on PMN of 3OC12-HSL and of other acyl homoserine lactones (AHL). We found that in addition to induction of chemotaxis, expression of adhesion molecules was enhanced as were the expression of immunoglobulin receptors and the phagocytosis of antibody-coated (opsonised) bacteria.

Materials and methods

Acyl homoserine lactones (AHL)

The 3-deoxo-isomer of 3OC12-HSL was purchased from Sigma-Aldrich (München, Germany). C4-HSL was purchased from Sigma-Aldrich; 2-amino-4-butyrolactone was from Fluka (Buchs, Switzerland). Stock solutions (0.1 M) were prepared in dimethyl sulfoxide (DMSO); these were diluted in Hank’s balanced salt solution (HBSS) immediately before use.

3OC12-HSL and C8-HSL were synthesised according to the method described by Chhabra et al. [13]. In brief, equivalent amounts of Meldrums’s acid (Aldrich, 1 mmol) and the corresponding fatty acids (Fluka) were used, together with 1.1 equivalents of 4-(dimethylamino)pyridine (Fluka) and N,N-dicyclohexylcarbodiimide (Aldrich). Each compound was purified to homogeneity by preparative liquid chromatography.

Bacteria

P. aeruginosa were obtained from patients with implant-associated osteomyelitis. The bacteria were expanded in Difco Micro Inoculum Broth (purchased from Becton, Dickinson and Co, Sparks, MD, USA) diluted 1- to 10-fold in HBSS. For formation of biofilms, the bacteria were seeded into plastic culture dishes (6 well) 1×107 per mL in HBSS and cultivated at 37°C with vigorous shaking. Supernatants were taken after various times.

Isolation of PMN from peripheral blood

Heparin blood was drawn from healthy volunteers, mainly laboratory personal and students, observing the institutional guidelines. PMN were isolated using PolymorphPrep (Nycomed, Oslo, Norway), following the supplier’s recommendations. The PMN fraction was harvested, washed repeatedly in phosphate-buffered saline (PBS, pH 7.4) and was suspended in HBSS containing 0.1% bovine serum albumin (BSA), at a final concentration of 1×106 cells/mL. Cells were identified by expression of CD66b using cytofluorometry (see below). The purification method yielded about 90% PMN.

Evaluation of viability of PMN

Intercalation of propidium iodide into DNA was measured by cytofluorometry as described by Belloc et al. [22]; in parallel, PMN were examined by light microscopy, following staining of the cells by trypan blue or haematoxyllin, respectively.

Chemotaxis across a membrane filter

A modified Boyden chamber assay was used [23], equipped with a nitrocellulose filter (5-μm pore size; 200-μm thick; Schleicher and Schuell GmbH Dassel, Germany). As bona fide chemokines, recombinant human complement C5a (5 ng/mL) (Sigma-Aldrich) or interleukin 8 (IL-8) (8 ng/mL) (Immunotools, Friesoythe, Germany) were used. Random migration was assessed using HBSS. The cells (1×106 in 1 mL) were placed into the upper compartment, the chemokines in the lower. After 90 min the cells migrated into the filters were fixed with propanol and stained with haematoxylin and evaluated using an Omnicon Alpha Image Analyzer (Bausch and Lomb, Heidelberg, Germany). Chemotaxis was measured as the “leading front”, defined as the distance in μm from the top of the filter to a level where at least five cells could still be seen. Two parallel filters were prepared and 10 different areas were evaluated on each filter. Preincubation of cells with butyrolactone was done at room temperature for 10 min. SB203580 was purchased from Calbiochem (purchased through Merck Biosciences Darmstadt, Germany) and used in a final concentration of 50 μM.

For the experiments using inserts, tissue culture inserts (10 mm, polycarbonate membrane with 3μm pore size) were obtained from Nunc (Nümbrecht, Germany). The bacteria were cultured in 24 well plates (7×106 per mL in HBSS) for various times. The inserts were then placed and isolated PMN (1×106 per mL, 500μL per insert in HBSS) were added. After 24 h at 37°C, PMN having reached the lower chamber were counted.

Cytofluorometry

To determine the expression of surface receptors on PMN, an FITC-labelled antibody to CD66b (Coulter Immunotech, Marseille, France) was used to identify PMN, and phycoerythrin (PE)-labelled antibodies to either anti-CD11b (BD Biosciences, Heidelberg, Germany), anti-CD18 (Serotec GmbH, Düsseldorf, Germany), anti-CD14 (Serotec), anti-CD16 (Becton Dickinson Biosciences) or anti-CD64 (Serotec). Whole blood (300 μL) was incubated with anti-CD66b-FITC and the respective PE-labelled antibodies (0.1–5 μg) or for comparison with mouse IgG (either PE- or FITC-labelled; Beckman Coulter Marseille, France ) for 20 min. The erythrocytes were then lysed using BD FacsLysing solution (BD Biosciences) and subjected to cytofluorometry using FACSCalibur and CellQuest software (Becton Dickinson, Heidelberg, Germany). Ten thousand events were counted. The percentage of double positive cells was used as a measure for up-regulation of CD64; enhancement of expression was determined by the increase in the mean fluorescence intensity (MFI).

Phagocytosis

A commercially available kit (Orpegen, Heidelberg, Germany) was used. In brief, FITC-labelled E. coli were incubated with whole heparinised blood. Uptake into PMN of bacteria was quantified by cytofluorometry.

Production of oxygen radicals

Basically, the method described by Babior et al. [24] was used. In brief, PMN (1×106) were suspended in cytochrome c solution (Sigma-Aldrich, Munich Germany, 1 mg/mL HBSS,) and stimulated with phorbol ester (PMA, Sigma-Aldrich) or with AHL. Reduction of cytochrome c was measured photometrically after 30 min and calculated as ΔOD (550 nm). For priming, the cells were exposed to tumour necrosis factor (TNF) α (2 ng) (Sigma-Aldrich) for 20 min, and then exposed to AHL or exposed to AHL (1 to 10 μM) and then stimulated with TNFα.

Results

Effect of 3OC12-HSL on the viability of PMN

In a first set of experiments the effect of 3OC12-HSL on the survival of PMN in vitro was tested. The cells were isolated from the peripheral blood of healthy donors and incubated with 3OC12-HSL in two concentrations. After various times, viability was assessed by propidium iodide incorporation and by light microscopy. At concentrations of 10 μM or 100 μM 3OC12-HSL did not induce cell death, nor did it affect the constitutive apoptosis of PMN occurring by 24 h or 48 h.

Induction of chemotaxis by 3OC12-HSL and supernatants of P. aeruginosa

Chemotaxis, the active, directed migration of cells towards an infected site, is central to an efficient host defence. Chemotaxis is elicited by numerous inflammatory cytokines, by bacterial products, and as shown before, 3OC12-HSL, but not the commercially available 3-deoxo-isomer. AHL with shorter fatty acids or 2-amino-4-butyrolactone did not induce migration (Table 1). Of note was that butyrolactone, but not the AHL with short fatty acids, inhibited the 3OC12-HSL-induced chemotaxis, but not the chemotaxis induced by other chemokines, such as IL-8. Inhibition of mitogen-activating protein kinase by SB203580 reduced the chemotaxis mediated by 3OC12-HSL or IL-8 (data summarised in Table 1).
Table 1

Induction of chemotaxis by 3OC12-HSL

 

Chemotactic stimulus

 

Chemotaxis (μM)a

PMN preincubated with

Without

8-HSL

4-Amino-2-butyrolactone

SB203580

No stimulus

34±12

35±7

32±12

17±14

3OC12-HSLb

110±23

124±38

69±23

55±23

3-Deoxo-isomer

44±17

54±19

54±24

35±12

8-HSL

39±12

44±16

40±11

31±14

4-Amino-2-butyrolactone

29±12

34±16

35±7

29±12

Interleukin 8 (2 ng)

140±34

128±29

110±27

59±43

aAll values are given as mean±SD of at least 4 independent experiments using cells of different donors

bAHL were used in a final concentration of 10 μM

In light of the finding that AHL are produced during biofilm formation, supernatants of P. aeruginosa were harvested various times after seeding the bacteria. Supernatants harvested within the first 24 h of culture induced chemotaxis of PMN. In the presence of butyrolactone the chemotaxis was inhibited (Fig. 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs00216-006-0698-5/MediaObjects/216_2006_698_Fig1_HTML.gif
Fig. 1

Induction of chemotaxis of PMN. Left chemotaxis was measured as migration (μm) in a Boyden chamber in response to culture medium (spontaneous, random migration), complement C5a or supernatants of P. aeruginosa cultivated for the times indicated. Shown are the data of a representative experiment with two parallel filters where 10 areas were evaluated. The data are shown as statistical box and whiskers blots. Asterisks indicate that the mean values were significantly different from those obtained for the random migration (ANOVA one way; p < 0.001). Right migration induced by the P. aeruginosa supernatant (open boxes data for the supernatant harvested after 6 h are shown) could be inhibited by butyrolactone (striped boxes), but not the migration induced by C5a or IL-8 (open boxes)

In another set of experiments, PMN were added into inserts with a porous bottom and then placed into wells containing P. aeruginosa, grown for 2 h to 72 h. Migration of the PMN out of the inserts towards the bacteria was seen when P. aeruginosa were grown for up to 24 h; thereafter, PMN were only attracted when in addition to the bacteria IL-8 as a chemokine was present (Fig. 2).
https://static-content.springer.com/image/art%3A10.1007%2Fs00216-006-0698-5/MediaObjects/216_2006_698_Fig2_HTML.gif
Fig. 2

Attraction of PMN to P. aeruginosa. Left P. aeruginosa were cultivated in a plastic culture dish for various times. Isolated PMN were added in an insert with a porous bottom (3-μm pore size) and migration into the lower chamber was determined by counting the PMN. Attraction of PMN to the lower chamber was seen when P. aeruginosa were cultivated for up to 24 h; thereafter, the migration declined. IL-8 was used as positive stimulus in the culture medium to determine the spontaneous, random migration (shown are the mean values of 2 independent experiments)

Up-regulation of adhesion proteins on PMN by 3OC12-HSL

Adhesion proteins are expressed on the surface of various cells, including PMN, and they mediate binding of cells to surfaces, such as the blood vessel walls or the connective tissue, and are thus required for the active migration of cells on surfaces. Furthermore, adhesion proteins also participate in the binding of bacteria, which in turn is a prerequisite for phagocytosis. Expression of these proteins can be modulated by various means thus modifying the PMN function. To assess the effects of AHL, 3OC12-HSL or the 3-deoxo-isomer were added to whole blood for various times; then expression of CD11b and CD18 was measured by cytofluorometry. An up-regulation of CD11b was seen; when using the MFI as a measure for CD11b expression an increase of 160 to 200% was calculated when the cells had been exposed to 3OC12-HSL (10 μM) for 10 to 20 min (example in Fig. 3; data for all experiments are summarised in Table 2). CD18 was also up-regulated; although the effect was marginal, it was seen in all experiments (Table 2). For both receptors, CD11b and CD18, up-regulation was not further enhanced when increasing the concentration of 3OC12-HSL to 100 μM; the 3-deoxo-isomer had no effect.
https://static-content.springer.com/image/art%3A10.1007%2Fs00216-006-0698-5/MediaObjects/216_2006_698_Fig3_HTML.gif
Fig. 3

Effect of 3OC12-HSL on the surface expression of CD11b. Whole blood was incubated with 3OC12-HSL (10 μM) for 20 min (thick line) or for comparison without the stimulus (thin line); then surface expression of CD11b on PMN (identified by expression of CD66b) was measured using a PE-labelled antibody to CD11b. The isotype controls are shown on the left

Table 2

Up-regulation of CD11b and CD18 on PMN following exposure of whole blood to AHL

Stimulus added

CD11ba

CD18a

10 min

20 min

10 min

20 min

None

773.9

520

707.9

523.04

3OC12-HSLc

951.7b

1028.7b

746.9

653.0b

3-Deoxo-isomerc

650.6

560.9

700.1

507.5

Interleukin 8c

1280.6

1400.1b

840.7b

880.8b

aMeasured as mean fluorescence intensity (MFI); values represent the mean of three independent experiments using blood of different donors

bValues are different from baseline values when using t-test (p < 0.05)

cAHL were used in a concentration of 10 μM; concentration of IL-8 2 ng/mL

Effect of 3OC12-HSL on the expression of immunoglobulin receptors or the lipopolysaccharide receptor CD14

PMN are equipped with receptors specific for the constant region of antibodies and are essential for the uptake of antibody-coated opsonised bacteria. Among these receptors, we found that the expression of the immunoglobulin receptor CD16 was enhanced by 3OC12-HSL (10 to 100 μM), as was expression of the high-affinity immunoglobulin receptor CD64, which is not constitutively expressed on PMN. Expression of CD14, the binding receptor for lipopolysaccharides, was not affected (Table 3).
Table 3

Effect of 3OC12-HSL on immunoglobulin receptors, or on CD14

 

Baseline values

Without stimulus (20 min)

3OC12-HSL (10 μM, 20 min)

3-Deoxo-isomer (10 μM, 20 min)

Mean fluorescence intensitya (n=3)

CD16

1832

1856

3489b

1768

% positive cells (mean of n=2)

CD64

2.4

2.5

30.4a

4.5

CD14

12.0

7.9

11.3

9.8

aValues are given as mean of triplicates

bValues are different from baseline values when using t-test (p <0.05)

Effect of 3OC12-HSL on phagocytosis and oxygen radical production

Uptake of antibody-coated opsonised FITC-labelled E. coli by PMN was measured by cytofluorometry. In the presence of 3OC12-HSL the number of bacteria per cell was marginally but reproducibly increased to 130%; in contrast the 3-deoxo-isomer had no effect (Fig. 4).
https://static-content.springer.com/image/art%3A10.1007%2Fs00216-006-0698-5/MediaObjects/216_2006_698_Fig4_HTML.gif
Fig. 4

Phagocytosis of opsonised E. coli. Whole blood was incubated with 3OC12-HSL (10 μM) for 20 min (filled circles) or, for comparison, without any stimulus (open circles) and then opsonised; FITC-labelled E. coli were added and green fluorescence within the PMN gate (identified by DNA staining) was measured. Left a representative experiment using increasing doses of E. coli (mean of duplicates); right data for three independent experiments using cells of different donors and 2×107E. coli are summarised as box as whiskers blot (open box untreated cells; striped box exposure to 10μM 3OC12-HSL for 20 min)

In the range of 0.1 to 100 μM, 3OC12-HSL did not induce the oxygen radical synthesis; it also had no priming effect on PMN, nor did it induce the oxygen radical synthesis of TNFα-primed PMN (data not shown).

Discussion

P. aeruginosa produce and release quorum-sensing molecules, among others the acyl homoserine lactones (AHL), as they form biofilms. Here, we addressed the question whether or not these molecules are recognised by polymorphonuclear neutrophils (PMN), because these are the first cells of the innate immune response to enter an infected site. We found that supernatants of P. aeruginosa, collected within 2 to 24 h after cultivating the bacteria under conditions leading to biofilm formation, contained a chemotactic activity for PMN. This chemotactic activity was most probably due to 3OC12-HSL because as shown before 3OC12-HSL can induce the chemotaxis of PMN [16]; moreover, the chemotaxis could be inhibited by butyrolactone which selectively inhibited the 3OC12-HSL-induced chemotaxis.

Because chemotaxis and phagocytosis of bacteria is critically dependent on adhesion proteins, the effect of 3OC12-HSL on the surface expression of the ß2-integrin CD11b/CD18 on PMN was tested. An up-regulation was seen within 20 min after exposure, compatible with a transport to the membrane of preformed molecules, as it occurs also in response to bona fide stimuli such as IL-8. Furthermore, an up-regulation of the low-affinity immunoglobulin (IgG) receptor CD16 was seen, as was induction of CD64, the latter representing the high-affinity receptor for IgG. Corresponding to the enhanced expression of the immunoglobulin receptors, an increased uptake of opsonised bacteria by 3OC12-HSL-pretreated PMN was seen.

Our findings are in line with a previous report by Vickström et al. [15] who also described enhancement of phagocytosis in an experimental setting employing in vitro differentiated human macrophages and yeast particles. Also in line with the data by Vickström et al., we could not demonstrate induction of oxygen radical synthesis; moreover, we did not see a priming or modulation of the oxidative burst by 3OC12-HSL. Thus, although macrophages and PMN are different in many aspects, they obviously share the capacity to recognise and respond to 3OC12-HSL.

It is also noteworthy that in both macrophages and PMN, the activation by 3OC12-HSL is critically dependent on the activity of the p38 mitogen activated protein (MAP) kinase. Together with our previous data, where specific and saturable binding of 14C-3OC12-HSL to PMN was demonstrated, as was the dependency on a signalling cascade involving tyrosine kinase, phospholipase C, protein kinase C and again to MAP kinase [16], a receptor for 3OC12-HSL is rather likely. In keeping with a receptor for 3OC12-HSL is also the requirement of a definite isomeric conformation. Our data correspond to earlier findings by Chhabra et al. [13], who described that the immunmodulatory activity of the AHL, determined in a splenocyte proliferation assay, depended critically on the chemical structures, particularly on the length of the fatty acids, the lactone ring and the L-configuration [13].

Conclusions

As pointed out above, recognition of bacterial products is essential for the host defence response. Because bacteria embedded in biofilms are supposedly rather resistant to the host defence mechanisms [25, 26], only recognition at an early stage would allow interference with biofilm formation. In that regard, recognising quorum-sensing molecules would be of utmost importance. Given that 3OC12-HSL is released by the bacteria in the initial phase of biofilm formation, the sensing of a 3OC12-HSL gradient would attract PMN to the site of a developing biofilm. At that early stage phagocytosis and killing of the bacteria could still be possible, particularly since the receptors critically involved in the bactericidal activity are also up-regulated by 3OC12-HSL.

Thus, recognition of and attraction by 3OC12-HSL could be regarded as a means of PMN to control biofilm development.

Copyright information

© Springer-Verlag 2006