- Email: [email protected]
Host defence against Staphylococcus aureus biofilms infection: Phagocytosis of biofilms by polymorphonuclear neutrophils (PMN)
Molecular Immunology 46 (2009) 1805–1813
Contents lists available at ScienceDirect
Molecular Immunology journal homepage: www.elsevier.com/locate/molimm
Host defence against Staphylococcus aureus biofilms infection: Phagocytosis of biofilms by polymorphonuclear neutrophils (PMN) Frank Günther a , Guido H. Wabnitz a , Petra Stroh a , Birgit Prior a , Ursula Obst b , Yvonne Samstag a , Christof Wagner c , G. Maria Hänsch a,∗ a b c
Institut für Immunologie, Universität Heidelberg, Germany Institut für Technische Chemie, Forschungszentrum Karlsruhe, Germany Klinik für Unfallchirurgie und Orthopädie, Berufsgenossenschaftliche Unfallklinik Ludwigshafen, Germany
a r t i c l e
i n f o
Article history: Received 11 November 2008 Received in revised form 14 January 2009 Accepted 25 January 2009 Available online 4 March 2009 Keywords: Biofilm PMN Neutrophils Host defence Phagocytosis
a b s t r a c t Bacteria organised in biofilms are a common cause of relapsing or persistent infections, particularly in patients receiving medical implants such as ventilation tubes, indwelling catheters, artificial heart valves, endoprostheses, or osteosynthesis materials. Bacteria in biofilms are relatively resistant towards antibiotics and biocides, and – according to the current dogma – towards the host defence mechanisms as well. In that context, we addressed the question, how polymorphonuclear neutrophils (PMN), the “first line defence” against bacterial infection, would interact with Staphylococcus aureus biofilms generated in vitro. By time-lapse video microscopy and confocal laser scan microscopy we observed a migration of PMN towards and into the biofilms, as well as clearance of biofilms by phagocytosis. By labelling the bacteria within the biofilm with 3 H thymidine, and by cytofluorometry we could confirm and quantify clearance and phagocytosis of biofilm as well. Of note, the extent of biofilm clearance depended on its maturation state: developing “young” biofilms were more sensitive towards the attack by PMN compared to mature biofilms. In conclusion, contrary to the current dogma, S. aureus biofilms are not inherently protected against the host defence. © 2009 Elsevier Ltd. All rights reserved.
1. Introduction Bacterial biofilms are increasingly recognised as a common cause of persistent infection (Costerton et al., 1999; Parsek and Singh, 2003; Hall-Stoodley et al., 2004; Lynch and Robertson, 2008). Biofilms are defined as microbial communities that colonise surfaces, for example the epithelium, but preferentially implanted medical devices including indwelling catheters, artificial heart valves, orthopaedic prostheses, or osteosynthesis materials (reviewed in Donlan, 2001; Gottenbos et al., 2002; Zimmerli et al., 2004). A hallmark of biofilm formation is the production of a slimy extracellular matrix, in which the bacteria are embedded. Production of the extracellular matrix, also referred to as extracellular polymer substance (EPS), is part of a genetically controlled process that induces a multitude of additional functional and phenotypic alterations, including loss of motility, reduced growth rate, and adhesion to surfaces (Davey and O’Toole, 2000; Watnick and Kolter, 2000; Dunne, 2002 and references therein). Living
∗ Corresponding author at: Institut für Immunologie der Universität Heidelberg, Im Neuenheimer Feld 305, 69120 Heidelberg, Germany. Tel.: +49 6221 564071; fax: +49 6221 565536. E-mail address: [email protected] (G.M. Hänsch). 0161-5890/$ – see front matter © 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.molimm.2009.01.020
as a biofilm appears to be advantageous for the bacteria, especially under unfavourable conditions. Among others, the bacteria acquire a relative resistance towards antibiotics and biocides, which complicates the management of biofilm infections and limits the therapeutic options (Mah and O’Toole, 2001; Stewart, 2002; Davies, 2003 and references therein). Moreover, bacterial biofilms appear to be a challenge for the host defence (Leid et al., 2002; Jesaitis et al., 2002). In our previous work, we addressed the question of the host defence in patients with implant-associated osteomyelitis, a persistent and destructive inflammatory disease, caused by the colonisation of endoprostheses or of osteosynthesis materials by staphylococci species. In those patients, the therapy of choice is the removal of the infected implant. Thus, bacteria and infiltrated immune cells can be collected from the infected site for analysis ex vivo. Among the infiltrated cells, we found predominantly polymorphonuclear neutrophils (PMN) and T lymphocytes in co-existence with the bacterial biofilm (Wagner et al., 2003, 2006), leading to the question whether or not PMN would recognise and attack the bacterial biofilm. Therefore, in the present study we tested the interaction of PMN with Staphylococcus aureus biofilms generated in vitro. We chose S. aureus because it is one of the prevalent pathogens in implantassociated biofilms infections (Arciola et al., 2005). We found that
1806
F. Günther et al. / Molecular Immunology 46 (2009) 1805–1813
S. aureus biofilms are not entirely protected against the attack by PMN, and that especially immature, developing biofilms are cleared by phagocytosis. 2. Materials and methods 2.1. Cultivation of bacteria and generation of biofilms S. aureus isolates from patients were used, as was the commercially available strain “Seattle 1945” (ATCC 25923). The bacteria (1 × 104 ) were diluted in 1 ml Difco Micro Inoculum Broth (Becton Dickinson, Heidelberg, Germany) and incubated as specified for the respective experiment. Polystyrol culture dishes were used (either 96- or 24 well dishes from), chamber slides, or Lab-tek II Chambered Coverglass (all obtained from Nunc, Roskilde, Denmark). For phagocytosis assays, bacteria were grown in 5 ml polystyrene round-bottom tubes (BD Biosciences, Erembrodegem, Belgium). Bacteria were cultivated for 2, 6 or 15 days, respectively, at 37 ◦ C with continuous shaking. 2.2. Isolation of PMN Peripheral blood from healthy human volunteers was obtained by venipuncture and collected heparin-NH4 coated tubes (Sarstedt, Nürnbrecht, Germany). PMN were isolated by centrifugation on PolymorphPrep (Axis-Shield PoC AS, Oslo, Norway) which yielded an 85–95% pure PMN population. The PMN were suspended in Hanks balanced salt solution (HBSS) and used within 1 h. 2.3. Time-lapse video microscopy Biofilms were cultivated in chambered cover slides (8 chambers, Nunc) in the presence of BacLight red (Molecular Probes, Leiden, The Netherlands) (100 nmol final conc.). Isolated PMN, incubated for 15 min with 5 nM Calcein AM (Sigma–Aldrich, Taufkirchen, Germany), were added (2 × 105 per chamber). For up to 60 min, microscopic images were taken every 60 s (Zeiss Axiovert 200 M; Zeiss, Göttingen, Germany). The images were analysed with the supplied Visitron Metamorph 6.2 Software. 2.4. Confocal laser-scanning microscopy S. aureus were cultivated in Lab-Tek II chamber slides (Nunc) for 2, 6 or 15 days at 37 ◦ C. PMN were added for up to 60 min at 37 ◦ C or 4 ◦ C. The slides were then washed carefully, and fixed with 4% PFA (37 ◦ C, 5 min). PMN were labelled with anti-CD66b (Immunotech, Marseille, France) and Cy3-conjugated antibody to mouse IgG from goat (Dako, Glostrup, Denmark). Biofilm and ingested bacteria were stained with SytoBC as nucleic acid stain (500 nM, Molecular Probes, Eugene Oregon, USA). For confocal laser-scanning microscopy, the slides were mounted and analysed with a Leica DMRBE confocal laser-scan microscope using Leica TCS as software (Leica, Wetzlar, Germany). For quantification of the biofilm-free zones, the laserscan-images were converted to black and white. For each time point, four arbitrarily chosen areas of the image were analysed by three investigators in a blinded fashion, and biofilm depleted areas surrounding at least on PMN were estimated using ImageJ Software. The data, given as square units, were averaged and the difference was calculated by ANOVA. To detect S. aureus a monoclonal antibody recognising S. aureus peptidoglycane was used (abcam, Cambridge, UK). 2.5. Phagocytosis of 3 H thymidine labelled biofilms To quantify the effect of PMN on bacterial biofilms, the bacteria were labelled with 3 H thymidine during culture. Following
exposure to PMN, the radioactivity associated with the residual biofilms was used as a measure for biofilm destruction. To distinguish between true phagocytic effects and loss of biofilm material due to excessive washing and handling, experiments were carried out at 37 ◦ C or 4 ◦ C, respectively. The data are expressed as counts per minutes (cpm); calculated was the mean ± S.D. of 12 parallel wells. S. aureus were cultivated in 96-well plates (Packard Bioscience, Groningen, The Netherlands) for 2, 6, and 15 days at 37 ◦ C in Inoculum Broth containing 7.4 × 104 Bq/ml 3 H-thymidine (Hartmann Analytic, Braunschweig, Germany). The biofilms were washed carefully to remove excess radioactivity. PMN suspended in HBSS, containing 0.5% BSA were added (1 × 105 per well) and incubated for 1 h at 37 ◦ C or 4 ◦ C. The non-adherent PMN were removed by careful washing; then aqua bidest. containing 3% acetic acid was added to lyse adherent PMN. The remaining biofilms were washed with PBS containing 0.5% sodium azide and 1% saponin. Finally the biofilms were fixed with 97% aethanol at 4 ◦ C, then desiccated at 50 ◦ C for 30 min Microscint 20 (PerkinElmer, Rodgau-Juegesheim, Germany) scintillation fluid was added (100 l/well and the plates were counted with TopCount Microplate Scintillation & Luminescence Counter (Packard BioScience). 2.6. Cytofluorometry Biofilms were grown for 2, 6, and 15 days, then labelled with SytoBC (500 nM final concentration; 30 min at 37 ◦ C with mild shaking in the dark). Then, the biofilm was washed, and isolated PMN (4 × 106 /2 ml) were added and incubated for 45 min at 37 ◦ C in the dark. PMN were then transferred to a new tube; to remove adherent PMN, 2 ml PBS, containing 10% trypsin and 0.01 M EDTA were added. The PMN fractions were combined, pelleted, and resuspended in PBS containing 1% paraformaldehyde. An equal volume of crystal violet (2 g/l in 0, 15 M NaCl) was added to quench the fluorescence due to biofilm material attached to the outside of the PMN. For cytofluorometry, the gate was set for the PMN population and fluorescence associated with the cells was measured as mean fluorescence intensity (MFI). To prevent phagocytosis, experiments were carried out at 4 ◦ C, or with PMN having been preincubated with cytochalasin B (Serva, Heidelberg, Germany). The difference between the MFI obtained at 4 ◦ C and 37 ◦ C was used as a measure for phagocytosis. 2.7. Migration “Transwell” chamber assay Bacteria (1 × 108 in 1 ml) were suspended in DifcoTM Micro Inoculum Broth, diluted 1 to 10 in Hank’s balanced salt solution (HBSS), and placed in a 24 well culture plate (Nunc, Nümbrecht, Germany) for biofilm formation. After 2 days 10 mm tissue culture inserts equipped with a polycarbonate membrane (pore size 3 m) (Nunc) were placed into the wells and PMN (1.5 × 106 /ml) were added. After 24 h, migration of PMN into the lower chamber well was determined by counting the cells microscopically. For comparison, wells without bacteria and with or without interleukin (IL-) 8 (25 ng/ml) (Immunotools, Friesoythe, Germany) as chemoattractant were used. For each experimental condition, six parallel wells were prepared; the values represent the number of PMN (mean ± S.D.). 2.8. Chemotaxis across a membrane filter A modified Boyden chamber assay was used (Brenneis et al., 1993), equipped with a nitrocellulose filter (5 m pore size; 200 m thick; Schleicher & Schuell GmbH Dassel, Germany). As bona fide chemokine IL-8 (8 ng/ml) were used. Random migration was determined using HBSS in place of the chemoattractant. The cells (1 × 106 in 1 ml) were placed into the upper compartment, the chemokines
F. Günther et al. / Molecular Immunology 46 (2009) 1805–1813
1807
Fig. 1. Interaction of PMN with a S. aureus biofilm over time: PMN (green) were placed on the biofilm (red) (time 00), and their action was observed by time-lapse video microscopy for up to 60 min (the time in minutes is depicted in the upper corner of the images). Time zero is shown on the upper left panel, on the right panel the same area after 58 min. A sequence of events of a selected area (marked by the asterisk) is depicted in the lower panel: by 14 min two PMN appear. By 17 min depletion of biofilm next to those PMN is seen; biofilm-depleted areas increase with time. In parallel, an accumulation of red biofilm material is seen on and in the PMN, as well as a yellow staining as the result of the red and green superimposition (a video of these events is supplemented as S1). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
in the lower. After 2 h, the cells migrated into the filters were fixed with propanol, and stained with haematoxylin. The filters were analysed microscopically using an Omnicon Alpha Image Analyzer (Bausch and Lomb, Heidelberg, Germany). Chemotaxis was measured as “leading front”, defined as the distance in m from the top of the filter to a level where at least five cells could still be detected. Two parallel filters were prepared, and on each filter, 10 different areas were evaluated at minimum.
3. Results 3.1. PMN on bacterial biofilms: direct observation by time-lapse video microscopy Biofilms were grown for 6 days in chambered slides in the presence of Baclight RedTM , an efficient fluorescence label for bacteria. Isolated PMN, labelled with calcein, a green fluorescence dye for living cells, were placed on these biofilms and their action was observed directly by time-lapse video microscopy for up to 60 min. The PMN moved on the biofilm. Within minutes, the formation of aggregates containing 5–10 cells was observed. In the wake of the moving PMN, biofilm-depleted areas became apparent (Fig. 1 shows a sequence of these events; a video clip is supplemented as S1). When biofilm fragments were used, a migration of the PMN towards those fragments was seen, followed by a disruption, and eventually uptake of the biofilm, the latter indicated by the yellow staining of the PMN (Fig. 2 shows a sequence of these events; a video clip as supplement S2).
To assess whether PMN migrate into biofilms image, PMN were placed onto the biofilm for 60 min. Following fixation the bacteria were stained with SytoBC, the PMN were labelled with an antibody to CD66b conjugated with Cy3. The images were viewed by laser scan microscopy. For three-dimensional reconstruction, z-stacks were generated. The image of the biofilm alone is shown on the left panel; the biofilm with PMN on the right (Fig. 3). The upper panel shows the reconstruction with a flat angle (approximately 15◦ ), the lower panels a side view, which reveals the typical biofilm morphology with spikes and protrusions. When PMN were added, an infiltration into the biofilm was observed within 20–60 min, as was an uptake of biofilm material, particularly of the spikes. The cells moved as a front, and turned yellow, the latter indicative again of biofilm uptake (Fig. 3). 3.2. Migration of PMN towards bacterial biofilms Because the time-lapse video (S2) suggested a migration of PMN towards bacterial biofilms, we performed experiments to quantify the migration. Biofilms were grown in multiwell chambers and PMN were placed in inserts on top in a transwell assay. After 24 h, the number of PMN having migrated to the lower chamber was determined. When biofilm was present – or as bona fide chemokine – IL-8, PMN were found in the lower chamber, but not in the absence of biofilms or IL-8. On average, 7 × 104 PMN were found on the biofilm, and 5.5 × 105 PMN in the IL-8 containing compartment (Fig. 4A). In a further set of experiments, we tested whether the supernatants of biofilms contained the chemotactic activity by the use
1808
F. Günther et al. / Molecular Immunology 46 (2009) 1805–1813
Fig. 2. Interaction of PMN with a S. aureus biofilm fragment over time: As described above, PMN were placed on the biofilm – here large fragments were used – and the action was observed. The kinetics is shown for two selected areas (a and b). (a) After 8 min a PMN has moved towards a biofilm fragment, it snatches a piece of the film (10 min) and attaches it to its surface (12 min), and finally tears it off (14 min). Then the cell moves towards another fragment (37 min). (b) shows an analogue sequence of events. The intense yellow staining of the cells is indicative of uptake of bacteria (the video is supplemented as S2). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
Fig. 3. PMN migrate into the bacterial biofilm: PMN were placed on a S. aureus biofilm for 30 min and than viewed by confocal laser scan microscopy. The three-dimensional reconstruction shows the migration of the PMN (red) into the biofilm (green). On the left, the biofilm alone is shown, on the right, the biofilm with PMN added. The side view (lower panel) shows the spikes and protrusion of the biofilm and the PMN attacking them (> < marks the section plane; the arrows give the orientation of the image).
F. Günther et al. / Molecular Immunology 46 (2009) 1805–1813
1809
Fig. 4. (A) PMN migrate towards and into bacterial biofilms. (a) Biofilms were grown in tissue culture plates for 2 days. Then PMN were added to an insert (5 m pore size), and placed onto the cultured biofilm. After 24 h, migration of PMN into the lower chamber was determined microscopically. (b) For comparison interleukin 8 (IL-8) as bona fide chemoattractant was added to the culture dish in the absence of a biofilm, (c) shows the experiments without biofilm. In (d), the mean values of four parallel experiments are summarised. (B) Supernatants of biofilms were harvested at the times indicated and used as chemoattractant for PMN in a Boyden chamber assay. On the left, random migration (i.e. migration without chemoattractant) is shown, and migration towards IL-8 as “positive” control. Chemotaxis was measured in m and shown is the summary of 2 parallel filters, where 10 different areas were measured (the box contains 50% of the values; the whiskers show the highest and the lowest values, respectively, the dot the mean value and the horizontal bar the median). Compared to the time 0 (culture medium removed from the biofilm without incubation time) the mean values obtained for the migration towards the supernatants were significantly different (p < 10−5 ).
of a Boyden chamber assay. We used supernatants of biofilms from different maturation states, i.e. from 2 h to 72 h, based on the assumption that the properties of the biofilms change with time. PMN were clearly attracted to the biofilm supernatants, indicating that a soluble chemoattractant is released by the biofilm. The attractive capacity was in the same range as observed for IL-8. Interestingly, although biofilms mature over days, the maximum
attraction was already achieved with the supernatant of a 2 h old biofilm (Fig. 4B). 3.3. Phagocytosis of bacterial biofilms The time-lapse video films suggested phagocytosis uptake and deterioration of biofilm. To verify these observations, PMN were
1810
F. Günther et al. / Molecular Immunology 46 (2009) 1805–1813
Fig. 5. PMN on a S. aureus biofilm viewed by confocal laser scan microscopy: Bacteria in the biofilm (6-days-old) were labelled with SytoBC that stains the DNA (also that of the PMN). The PMN were identified by an antibody to CD66b, labelled red. When PMN were incubated at 37 ◦ C, areas depleted of biofilm around the PMN appeared (A) (arrows), but not when the cells were incubated at 4 ◦ C (B). (C) Biofilms were grown for 2, 6 and 15 days. PMN were added and after 30 min, the areas depleted around the PMN were quantified by planimetry (shown as arbitrary square units; the values are given as mean ± S.D. of four different areas viewed independently by three investigators).
added to bacterial biofilms for 30 min at 37 ◦ C. Then the specimens were fixed, SytoBC was added to label the DNA, and anti-CD66b to identify the PMN. As exemplified in Fig. 5A, uptake of the bacteria by the PMN was seen, and in addition depletion of the biofilm around the PMN. When the experiments were carried out at 4 ◦ C, a condition known to prevent phagocytosis of planktonic bacteria, neither uptake of bacteria nor biofilm-depleted areas around the PMN were observed, an example is shown in Fig. 5B. In other experiments, bacteria were identified with an antibody directed to S. aureus peptidoglycane instead of DNA-labelling. Essentially similar data were obtained (not shown). 3.4. Uptake and deterioration of biofilms depends on the maturation stage The observation that protrusions and spikes were readily taken up by the PMN suggested that newly added biofilm material might be especially sensitive to clearance by PMN. To assess the sensitivity of biofilms towards PMN systematically, S. aureus were cultivated on cover slips for 2, 6, and 15 days. By day 2 biofilms were clearly visible, they were, however, more fragile than the films grown for 6 or 15 days. When PMN were added to the biofilms, bacteriadepleted zones around the PMN became apparent after 30 min. The biofilm-depleted areas around the PMN of the 2- and 6-daysold biofilms appeared larger compared to those of the 15-days-old biofilms. By planimetry, a trend towards greater biofilm depletion in less mature biofilms could be confirmed, with statistically significant differences only between 2 and 15 days biofilms (Fig. 5C). The experiments shown were all performed with S. aureus derived from a patient. With a commercially available S. aureus strain, essentially similar data were obtained. 3.5. Quantification by 3 H labelling of the bacteria Because planimetry of the bacteria depleted areas is semiquantitative at best, and does not take into account threedimensional structures, we established a new method for the
quantification of the bacteria within the biofilm. The bacteria were cultured for 2, 6 or 15 days in the presence of 3 H-thymidine, which incorporates into the bacterial DNA. Then the films were carefully washed, and freshly isolated PMN were added for 30 min to allow phagocytosis. Then PMN, and hence phagocytosed bacteria, were removed, and the radioactivity of the residual biofilms was counted. Decline of radioactivity was interpreted as a measure for biofilm deterioration. Compared were always biofilms that had been exposed to PMN with those which had not (an example is shown Fig. 6A). In a parallel experiment, PMN were incubated with the biofilm at 4 ◦ C, which prevents phagocytosis. In that experiment, no loss of biofilm was seen. With use of this method, we now compared phagocytosis by PMN of biofilms that had been cultivated for 2, 6 and 16 days (Fig. 6B). On average, loss of biofilm amounted to 51.5% for 2-daysold films, to 43.7% for 6-days-old films, and to 21.4% for 15-days-old films (mean of three independent experiments with either patients’ derived S. aureus or S. aureus type Seattle 1945; the differences between 2-days-old and 15-days-old films were statistically significant with p = 0.02 using ANOVA). Again, incubation of biofilms with PMN at 4 ◦ C did not result in loss of biofilm.
3.6. Quantification of bacterial biofilm uptake by cytofluorometry As a further method to quantify uptake of bacterial biofilms by the PMN and to verify phagocytosis, as opposed to killing by exocytosed bactericidal compounds, cytofluorometry was performed using SytoBC labelled biofilms. In these experiments, PMN were placed on the biofilms for 30 min at 37 ◦ C or at 4 ◦ C, respectively, then harvested and subjected to cytofluorometry. To differentiate between bacteria that were attached to the outside of the PMN and those inside, crystal violet was used for quenching the outside-fluorescence. When the gate was set for the PMN, green fluorescence was seen in the samples incubated at 37 ◦ C, but not in those incubated at 4 ◦ C (example in Fig. 7A). Compared to 15day-old biofilms, the phagocytosis of 2- and 6-day-old biofilms was considerably higher compared (Fig. 7B and C). Again, with S.
F. Günther et al. / Molecular Immunology 46 (2009) 1805–1813
1811
Fig. 6. Deterioation of 3 H thymidine-labelled biofilms. (A) Bacteria were grown for 2 days in the presence of 3 H thymidine under conditions allowing biofilm formation. PMN were placed on the biofilms for 30 min at 37 ◦ C, then removed and radioactivity associated with the residual biofilms was determined (the data of 12 parallel experiments are shown as box blots with the box containing 50% of the values, the horizontal bar shows the median value, and the “whiskers” the highest and the lowest values, respectively). Compared to biofilm alone, significant loss of radioactivity was seen after incubation of the biofilm with PMN at 37 ◦ C (calculated by ANOVA). When the incubation with PMN was carried out at 4 ◦ C, there was no loss of biofilm. (B) The percentage of loss of biofilm was determined for 2-, 6-, and 15-days-old biofilms (here data of three independent experiments are evaluated; the mean values are given).
Fig. 7. Phagocytosis of bacterial biofilms determined by cytofluoromertry. (A) To parallel tubes containing SytoBC labelled biofilm (grown for 6 days) PMN were added and incubated at either 4 ◦ C (green line) or 37 ◦ C (blue line) for 45 min. Fluorescence outside the cells was quenched with crystal violet and fluorescence associated with the cells was determined. (B) PMN were added to biofilms (for 45 min at 37 ◦ C) that had been grown for 2 (red line), 6 (dark blue line), or 15 days (light blue line), respectively, and fluorescence was measured. (C) Data of 8 independent experiments are summarised for 2, 6, and 15 days old biofilms, respectively. Shown is the mean fluorescence intensity as statistical box blot with the box containing 50% of the values, the horizontal bar shows the median value, and the “whiskers” the highest and the lowest values, respectively). Differences between the mean values were calculated by ANOVA.
1812
F. Günther et al. / Molecular Immunology 46 (2009) 1805–1813
aureus from the patients and with the strain Seattle 1945 essentially similar data were obtained.
4. Discussion Bacterial biofilms are increasingly recognised as a common cause of chronic persistent infections. Thus, the question arises, how the host defence mechanisms react to bacterial biofilms, and whether or not biofilms can be cleared. Based on the facts that PMN are the primary cells in the host defence against bacteria, and are equipped with numerous bactericidal and cytotoxic entities, and that they infiltrate sites of biofilm infections, we assessed the interactions of isolated human PMN with in vitro-generated S. aureus biofilms by a variety of methods. By directly observing the interaction by time-lapse video microscopy, we came to the following conclusions: PMN move on intact biofilms, and biofilm-free areas appeared at the sites the PMN were vacating. Apparently, PMN take up biofilm as they move along. When biofilm fragments were used, a migration of PMN towards the biofilm was seen, as was tearing up of biofilm material, and eventually removal of the biofilm. Some of the PMN turned yellow, most probably due an overlap of the green and the red staining, which indicates close vicinity of biofilm and PMN, presumably attachment and uptake. Of note, the PMN remained viable during the observation time (60 min), and were not immobilised, as had been reported earlier for PMN on Pseudomonas aeruginosa biofilms (Jesaitis et al., 2002). By quantitative methods, we characterised the interaction of PMN with the biofilms. Firstly, we confirmed migration of PMN towards S. aureus biofilms and towards supernatants of the biofilms as well. The latter is in line with the fact that numerous bacteria, including S. aureus, produce and release into the environment substance(s) that attract PMN (Dürr et al., 2005; Rot et al., 1987). Moreover, we demonstrated a migration of PMN into the biofilm. The three-dimensional reconstruction suggested a massive infiltration of the cells that was not confined to water channels, as demonstrated before by Leid et al. (2002). Moreover, the images show that especially spikes and protrusions on the surface of the biofilm were removed. Because these data suggested that the newly formed biofilm material might be especially sensitive to phagocytosis by PMN, we performed a series of experiments with biofilms grown for different times. To quantify the effect of the PMN, areas devoid of biofilms around PMN were measured by planimetry. Compared to mature biofilms, young, immature biofilms (2- to 6-day-old) appeared to be more sensitive compared to 15-day-old biofilms. Because planimetry is two-dimensional only, and thus does not take into account holes in the biofilm, a further method, based on the DNA-labelling of the bacteria was developed. After exposure to PMN and removal of the cells to exclude phagocytosed bacteria, the radioactivity of the remaining biofilm was counted as a measure for biofilm loss. By this method, we could essentially confirm our previous data: loss of biofilm was most extensive in immature biofilms. The method, however, did not allow differentiating between phagocytosis and other means of biofilm degradation. Therefore, we used a cytofluorometry-based method to quantify phagocytosis of planktonic bacteria. We could confirm uptake of bacteria, and again, we found that immature biofilms were more sensitive to phagocytosis compared to mature films, but that also mature biofilms are not entirely protected against the attack by PMN. Taken together, our data show phagocytosis of bacterial biofilms by the PMN. Additional means of biofilm deterioration, such as exocytosis of bactericidal substances, however, cannot be ruled out. Our data seem to contradict an earlier report by Leid et al. (2002), who described adherence of leukocytes to S. aureus biofilms, their
migration into the film, but a failure to phagocytose the bacteria. In these studies, the mononuclear cell fraction was used, that contains predominantly monocytes, T- and B-cells, but not the PMN, which could account for the differences seen. The “halos” described by Leid et al., however, are reminiscent of the biofilm-depleted areas seen in our studies, and could be due to cytotoxic substances released from the PMN. Why immature biofilms are more sensitive to the PMN attack compared to mature biofilms is still a matter of speculation. Under our experimental conditions, the number of bacteria incorporated into the biofilm did not increase during biofilm formation after day 2 (data not shown). Moreover, also, the thickness of the biofilm did not drastically increase with maturation, but the mechanical stability did. We assume that the architecture and the composition of the extracellular polymer matrix changes with time. To date, methods to characterise and to analyse the extracellular polymer substances (EPS) are still being developed. The EPS is a highly intricate and dynamic structure, consisting of complex carbohydrates, proteins, lipids, nucleic acid, and exoenzymes as well (Branda et al., 2005; Flemming et al., 2007). Whether or not the EPS of S. aureus interacts with PMN has not yet been analysed. Data derived from experiments with P. aeruginosa EPS, however, show effects of EPS or EPS-derived compounds such as quorum-sensing molecules, alginates or rhamnolipids on PMN (Pedersen et al., 1990; Zimmermann et al., 2006; Jensen et al., 2007; Hänsch et al., 2008). Accepting the fact that PMN can phagocytose and clear biofilms the question arises how biofilms can persist in vivo. The paradigm “too many–too late” might also apply here, and we presume that the host defence against bacterial biofilms is subject to the same restriction as the defence against planktonic bacteria: the within-host-population dynamics, the health status of the host, and conditions favouring the colonisation with bacteria. The latter is particularly relevant for the so-called implant-associated infections, when biofilms form on medical devices such as indwelling catheters, artificial heart valves, osteosynthesis materials or endoprostheses (Donlan, 2001; Gottenbos et al., 2002). Presumably, these foreign materials provide a readily colonisable surface, and the “the race for the surface”, as it was proposed by Gristina et al. (1989), might be a decisive factor in the formation of biofilms. An attractive notion is that artificial surfaces – in contrast to epithelial cells – are without defence against colonisation and biofilm formation. While data on S. aureus biofilms are lacking, for P. aeruginosa the neutralisation by epithelial cells of quorum sensing molecules required for biofilm formation has been described, which in turn might result in the prevention of biofilm formation on the epithelium (Chun et al., 2004; Hastings, 2004). Another observation that might explain the persistence of biofilms was made when we analysed the infiltrated cells. The PMN at the site of biofilm infections were highly activated, as seen by the up-regulation of surface receptors and priming for superoxide generation, their chemotactic activity, however, was reduced. Thus, it is possible that the PMN cannot infiltrate into the biofilm, and consequently, phagocytosis and clearing of biofilms is restricted to the surface, and hence remains insufficient (Wagner et al., 2004). In conclusion, S. aureus biofilms can be invaded and phagocytosed by PMN. Therefore, they are not inherently protected against the host defence. Whether or not an infection becomes clinically apparent depends on numerous factors, especially the health status of the host, and conditions favouring biofilm formation.
Acknowledgment The work was supported by a grant from the Deutsche Forschungsgemeinschaft (WA1623/1-5).
F. Günther et al. / Molecular Immunology 46 (2009) 1805–1813
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.molimm.2009.01.020. References Arciola, C.R., An, Y.H., Campoccia, M.E., Donati, M.E., Montanaro, L., 2005. Etiology of implant orthopedic infections: a survey on 1027 clinical isolates. Int. J. Artif. Org. 28, 1091–1100. Branda, S.S., Vik, A., Freidman, L., Kolter, R., 2005. Biofilms: the matrix revisited. Trends Microbiol. 13, 20–26. Brenneis, H., Schmidt, A., Blaas-Mautner, P., Wörner, I., Ludwig, R., Hänsch, G.M., 1993. Chemotaxis of polymorphonuclear neutrophils (PMN) in patients suffering from recurrent infection. Eur. J. Clin. Invest. 23, 693–698. Chun, C.K., Ozer, E.A., Welsh, M.J., Zabner, J., Greenberg, E.P., 2004. Inactivation of a Pseudomonas aeruginosa quorum-sensing signal by human airway epithelia. Proc. Natl. Acad. Sci. 101, 3587–3590. Costerton, J.W., Stewart, P.S., Greenberg, E.P., 1999. Bacterial biofilms: a common cause of persistent infections. Science 284, 1318–1322. Davey, M.E., O’Toole, G.A., 2000. Microbial biofilms: from ecology to molecular genetics. Microbiol. Mol. Biol. Rev. 64, 847–867. Davies, D., 2003. Understanding biofilm resistance to antibacterial agents. Nat. Rev. Drug Discov. 2, 114–122. Donlan, R.M., 2001. Biofilms and device-associated infections. Emerg. Infect. Dis. 7, 277–281. Dunne, M.W., 2002. Bacterial adhesion: seen any good biofilm lately. Clin. Microbiol. Rev. 15, 155–166. Dürr, M.C., Kristian, S.A., Otto, M., Matteoli, G., Margolis, P.J., Trias, J., van Kessel, K.P., van Strijp, J.A., Bohn, E., Landmann, R., Peschel, A., 2005. Neutrophil chemotaxis by molecular pattern-recognition formylated peptides are crucial but not the sole neutrophil chemoattractant produced by Staphylococcus aureus. Cell. Microbiol. 8, 207–217. Flemming, H.-C., Neu, T.R., Wozniak, D.J., 2007. The EPS Matrix: the house of biofilm cells. J. Bacteriol. 189, 7945–7947. Gottenbos, B.G., Busscher, H.J., van der Mei, H.C., Nieuwenhuis, P., 2002. Pathogenesis and prevention of biomaterial centered infections. J. Mat. Sci. Med. 18, 717–722. Gristina, A.G., Nayler, P., Myrvik, Q., 1989. Infections from biomaterials and implants. Med. Prog. Technol. 14, 205–224. Hall-Stoodley, L., Costerton, J.W., Stoodley, P., 2004. Bacterial biofilms: from the environment to infectious disease. Nat. Rev. Microbiol. 2, 95–108. Hänsch, G.M., Brenner-Weiss, G., Prior, B., Wagner, C., Obst, U., 2008. The extracellular polymer substance of Pseudomonas aeruginosa: too slippery for neutrophils to migrate on? Int J. Artif. Organs 31, 796–809.
1813
Hastings, J.W., 2004. Bacterial quorum-sensing signals are inactivated by mammalian cells. Proc. Natl. Acad. Sci. 101, 3993–3994. Jensen, P.Ø., Bjarnsholt, T., Rasmussen, T.B., Calum, H., Moser, C., Pressler, T., Givskov, M., Hoiby, N., 2007. Rapid necrotic killing of PMN is caused by quorum sensing controlled production of rhamnolipids by Pseudomonas aeruginosa. Microbiology 153, 1329–1338. Jesaitis, A.J., Franklin, M.J., Berglund, D., Sasaki, M., Lord, C.I., Bleazard, J., Duffy, J.E., Beyenal, H., Lewandowski, Z., 2002. Compromised host defense on Pseudomonas aeruginosa biofilms: characterisation of neutrophil and biofilm interactions. J. Immunol. 171, 4329–4339. Leid, J.G., Shirtliff, H.C., Costerton, J.W., Stoodley, P., 2002. Human leukocytes adhere to, penetrate, and respond to Staphylococcus aureus biofilms. Infect. Immun. 70, 6339–6345. Lynch, S.A., Robertson, G.T., 2008. Bacterial and fungal biofilm infections. Ann. Rev. Med. 59, 415–428. Mah, T.F., O’Toole, G.A., 2001. Mechanisms of biofilm resistance to antimicrobial agents. Trends Microbiol. 9, 34–39. Parsek, M.R., Singh, P.K., 2003. Bacterial Biofilms: an emerging link to disease pathogenesis. Ann. Rev. Microbiol. 57, 677–701. Pedersen, S.S., Kharazmi, A., Espersen, F., Hoiby, N., 1990. Pseudomonas aeruginosa alginate in cystic fibrosis and the inflammatory response. Infect. Immun. 58, 3363–3368. Rot, A., Henderson, L.E., Copeland, T.D., Leonhard, E.J., 1987. A series of six ligands for the human formyl peptide receptor: tetrapeptides with high chemotactic potency and efficacy. PNAS 84, 7967–7971. Stewart, P.S., 2002. Mechanisms of antibiotic resistance in bacterial biofilms. Int. J. Med. Microbiol. 292, 107–113. Wagner, C., Heck, D., Lautenschläger, K., Iking-Konert, C., Heppert, V., Wentzensen, A., Hänsch, G.M., 2006. T-lymphocytes in implant-associated posttraumatic osteomyelitis: identification of cytotoxic T-effector cells at the site of infection. Shock 25, 241–246. Wagner, C., Kaksa, A., Müller, W., Denefleh, B., Heppert, V., Wentzensen, A., Hänsch, G.M., 2004. Polymorphonuclear neutrophils in posttraumatic osteomyelitis: cells recovered from the inflamed site lack chemotactic activitiy, but generate superoxides. Shock 22, 108–115. Wagner, C., Kondella, K., Bernschneider, T., Heppert, V., Wentzensen, A., Hänsch, G.M., 2003. Post-traumatic osteitis: analysis of inflammatory cells recruited into the site of infection. Shock 20, 503–510. Watnick, P., Kolter, R., 2000. Biofilm, city of microbes. J. Bacteriol. 182, 2675–2679. Zimmerli, W., Trampuz, A., Ochsner, P.E., 2004. Prosthetic-joint infections. N. Engl. J. Med. 351, 1645–1654. Zimmermann, S., Wagner, C., Müller, W., Brenner-Weiss, G., Hug, F., Prior, B., Obst, U., Hänsch, G.M., 2006. Induction of neutrophil chemotaxis by the quorum-sensing molecule N-3-oxododecanoyl homoserine lactone (3OC12-HSL). Infect. Immun. 74, 5687–5692.
Contents lists available at ScienceDirect
Molecular Immunology journal homepage: www.elsevier.com/locate/molimm
Host defence against Staphylococcus aureus biofilms infection: Phagocytosis of biofilms by polymorphonuclear neutrophils (PMN) Frank Günther a , Guido H. Wabnitz a , Petra Stroh a , Birgit Prior a , Ursula Obst b , Yvonne Samstag a , Christof Wagner c , G. Maria Hänsch a,∗ a b c
Institut für Immunologie, Universität Heidelberg, Germany Institut für Technische Chemie, Forschungszentrum Karlsruhe, Germany Klinik für Unfallchirurgie und Orthopädie, Berufsgenossenschaftliche Unfallklinik Ludwigshafen, Germany
a r t i c l e
i n f o
Article history: Received 11 November 2008 Received in revised form 14 January 2009 Accepted 25 January 2009 Available online 4 March 2009 Keywords: Biofilm PMN Neutrophils Host defence Phagocytosis
a b s t r a c t Bacteria organised in biofilms are a common cause of relapsing or persistent infections, particularly in patients receiving medical implants such as ventilation tubes, indwelling catheters, artificial heart valves, endoprostheses, or osteosynthesis materials. Bacteria in biofilms are relatively resistant towards antibiotics and biocides, and – according to the current dogma – towards the host defence mechanisms as well. In that context, we addressed the question, how polymorphonuclear neutrophils (PMN), the “first line defence” against bacterial infection, would interact with Staphylococcus aureus biofilms generated in vitro. By time-lapse video microscopy and confocal laser scan microscopy we observed a migration of PMN towards and into the biofilms, as well as clearance of biofilms by phagocytosis. By labelling the bacteria within the biofilm with 3 H thymidine, and by cytofluorometry we could confirm and quantify clearance and phagocytosis of biofilm as well. Of note, the extent of biofilm clearance depended on its maturation state: developing “young” biofilms were more sensitive towards the attack by PMN compared to mature biofilms. In conclusion, contrary to the current dogma, S. aureus biofilms are not inherently protected against the host defence. © 2009 Elsevier Ltd. All rights reserved.
1. Introduction Bacterial biofilms are increasingly recognised as a common cause of persistent infection (Costerton et al., 1999; Parsek and Singh, 2003; Hall-Stoodley et al., 2004; Lynch and Robertson, 2008). Biofilms are defined as microbial communities that colonise surfaces, for example the epithelium, but preferentially implanted medical devices including indwelling catheters, artificial heart valves, orthopaedic prostheses, or osteosynthesis materials (reviewed in Donlan, 2001; Gottenbos et al., 2002; Zimmerli et al., 2004). A hallmark of biofilm formation is the production of a slimy extracellular matrix, in which the bacteria are embedded. Production of the extracellular matrix, also referred to as extracellular polymer substance (EPS), is part of a genetically controlled process that induces a multitude of additional functional and phenotypic alterations, including loss of motility, reduced growth rate, and adhesion to surfaces (Davey and O’Toole, 2000; Watnick and Kolter, 2000; Dunne, 2002 and references therein). Living
∗ Corresponding author at: Institut für Immunologie der Universität Heidelberg, Im Neuenheimer Feld 305, 69120 Heidelberg, Germany. Tel.: +49 6221 564071; fax: +49 6221 565536. E-mail address: [email protected] (G.M. Hänsch). 0161-5890/$ – see front matter © 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.molimm.2009.01.020
as a biofilm appears to be advantageous for the bacteria, especially under unfavourable conditions. Among others, the bacteria acquire a relative resistance towards antibiotics and biocides, which complicates the management of biofilm infections and limits the therapeutic options (Mah and O’Toole, 2001; Stewart, 2002; Davies, 2003 and references therein). Moreover, bacterial biofilms appear to be a challenge for the host defence (Leid et al., 2002; Jesaitis et al., 2002). In our previous work, we addressed the question of the host defence in patients with implant-associated osteomyelitis, a persistent and destructive inflammatory disease, caused by the colonisation of endoprostheses or of osteosynthesis materials by staphylococci species. In those patients, the therapy of choice is the removal of the infected implant. Thus, bacteria and infiltrated immune cells can be collected from the infected site for analysis ex vivo. Among the infiltrated cells, we found predominantly polymorphonuclear neutrophils (PMN) and T lymphocytes in co-existence with the bacterial biofilm (Wagner et al., 2003, 2006), leading to the question whether or not PMN would recognise and attack the bacterial biofilm. Therefore, in the present study we tested the interaction of PMN with Staphylococcus aureus biofilms generated in vitro. We chose S. aureus because it is one of the prevalent pathogens in implantassociated biofilms infections (Arciola et al., 2005). We found that
1806
F. Günther et al. / Molecular Immunology 46 (2009) 1805–1813
S. aureus biofilms are not entirely protected against the attack by PMN, and that especially immature, developing biofilms are cleared by phagocytosis. 2. Materials and methods 2.1. Cultivation of bacteria and generation of biofilms S. aureus isolates from patients were used, as was the commercially available strain “Seattle 1945” (ATCC 25923). The bacteria (1 × 104 ) were diluted in 1 ml Difco Micro Inoculum Broth (Becton Dickinson, Heidelberg, Germany) and incubated as specified for the respective experiment. Polystyrol culture dishes were used (either 96- or 24 well dishes from), chamber slides, or Lab-tek II Chambered Coverglass (all obtained from Nunc, Roskilde, Denmark). For phagocytosis assays, bacteria were grown in 5 ml polystyrene round-bottom tubes (BD Biosciences, Erembrodegem, Belgium). Bacteria were cultivated for 2, 6 or 15 days, respectively, at 37 ◦ C with continuous shaking. 2.2. Isolation of PMN Peripheral blood from healthy human volunteers was obtained by venipuncture and collected heparin-NH4 coated tubes (Sarstedt, Nürnbrecht, Germany). PMN were isolated by centrifugation on PolymorphPrep (Axis-Shield PoC AS, Oslo, Norway) which yielded an 85–95% pure PMN population. The PMN were suspended in Hanks balanced salt solution (HBSS) and used within 1 h. 2.3. Time-lapse video microscopy Biofilms were cultivated in chambered cover slides (8 chambers, Nunc) in the presence of BacLight red (Molecular Probes, Leiden, The Netherlands) (100 nmol final conc.). Isolated PMN, incubated for 15 min with 5 nM Calcein AM (Sigma–Aldrich, Taufkirchen, Germany), were added (2 × 105 per chamber). For up to 60 min, microscopic images were taken every 60 s (Zeiss Axiovert 200 M; Zeiss, Göttingen, Germany). The images were analysed with the supplied Visitron Metamorph 6.2 Software. 2.4. Confocal laser-scanning microscopy S. aureus were cultivated in Lab-Tek II chamber slides (Nunc) for 2, 6 or 15 days at 37 ◦ C. PMN were added for up to 60 min at 37 ◦ C or 4 ◦ C. The slides were then washed carefully, and fixed with 4% PFA (37 ◦ C, 5 min). PMN were labelled with anti-CD66b (Immunotech, Marseille, France) and Cy3-conjugated antibody to mouse IgG from goat (Dako, Glostrup, Denmark). Biofilm and ingested bacteria were stained with SytoBC as nucleic acid stain (500 nM, Molecular Probes, Eugene Oregon, USA). For confocal laser-scanning microscopy, the slides were mounted and analysed with a Leica DMRBE confocal laser-scan microscope using Leica TCS as software (Leica, Wetzlar, Germany). For quantification of the biofilm-free zones, the laserscan-images were converted to black and white. For each time point, four arbitrarily chosen areas of the image were analysed by three investigators in a blinded fashion, and biofilm depleted areas surrounding at least on PMN were estimated using ImageJ Software. The data, given as square units, were averaged and the difference was calculated by ANOVA. To detect S. aureus a monoclonal antibody recognising S. aureus peptidoglycane was used (abcam, Cambridge, UK). 2.5. Phagocytosis of 3 H thymidine labelled biofilms To quantify the effect of PMN on bacterial biofilms, the bacteria were labelled with 3 H thymidine during culture. Following
exposure to PMN, the radioactivity associated with the residual biofilms was used as a measure for biofilm destruction. To distinguish between true phagocytic effects and loss of biofilm material due to excessive washing and handling, experiments were carried out at 37 ◦ C or 4 ◦ C, respectively. The data are expressed as counts per minutes (cpm); calculated was the mean ± S.D. of 12 parallel wells. S. aureus were cultivated in 96-well plates (Packard Bioscience, Groningen, The Netherlands) for 2, 6, and 15 days at 37 ◦ C in Inoculum Broth containing 7.4 × 104 Bq/ml 3 H-thymidine (Hartmann Analytic, Braunschweig, Germany). The biofilms were washed carefully to remove excess radioactivity. PMN suspended in HBSS, containing 0.5% BSA were added (1 × 105 per well) and incubated for 1 h at 37 ◦ C or 4 ◦ C. The non-adherent PMN were removed by careful washing; then aqua bidest. containing 3% acetic acid was added to lyse adherent PMN. The remaining biofilms were washed with PBS containing 0.5% sodium azide and 1% saponin. Finally the biofilms were fixed with 97% aethanol at 4 ◦ C, then desiccated at 50 ◦ C for 30 min Microscint 20 (PerkinElmer, Rodgau-Juegesheim, Germany) scintillation fluid was added (100 l/well and the plates were counted with TopCount Microplate Scintillation & Luminescence Counter (Packard BioScience). 2.6. Cytofluorometry Biofilms were grown for 2, 6, and 15 days, then labelled with SytoBC (500 nM final concentration; 30 min at 37 ◦ C with mild shaking in the dark). Then, the biofilm was washed, and isolated PMN (4 × 106 /2 ml) were added and incubated for 45 min at 37 ◦ C in the dark. PMN were then transferred to a new tube; to remove adherent PMN, 2 ml PBS, containing 10% trypsin and 0.01 M EDTA were added. The PMN fractions were combined, pelleted, and resuspended in PBS containing 1% paraformaldehyde. An equal volume of crystal violet (2 g/l in 0, 15 M NaCl) was added to quench the fluorescence due to biofilm material attached to the outside of the PMN. For cytofluorometry, the gate was set for the PMN population and fluorescence associated with the cells was measured as mean fluorescence intensity (MFI). To prevent phagocytosis, experiments were carried out at 4 ◦ C, or with PMN having been preincubated with cytochalasin B (Serva, Heidelberg, Germany). The difference between the MFI obtained at 4 ◦ C and 37 ◦ C was used as a measure for phagocytosis. 2.7. Migration “Transwell” chamber assay Bacteria (1 × 108 in 1 ml) were suspended in DifcoTM Micro Inoculum Broth, diluted 1 to 10 in Hank’s balanced salt solution (HBSS), and placed in a 24 well culture plate (Nunc, Nümbrecht, Germany) for biofilm formation. After 2 days 10 mm tissue culture inserts equipped with a polycarbonate membrane (pore size 3 m) (Nunc) were placed into the wells and PMN (1.5 × 106 /ml) were added. After 24 h, migration of PMN into the lower chamber well was determined by counting the cells microscopically. For comparison, wells without bacteria and with or without interleukin (IL-) 8 (25 ng/ml) (Immunotools, Friesoythe, Germany) as chemoattractant were used. For each experimental condition, six parallel wells were prepared; the values represent the number of PMN (mean ± S.D.). 2.8. Chemotaxis across a membrane filter A modified Boyden chamber assay was used (Brenneis et al., 1993), equipped with a nitrocellulose filter (5 m pore size; 200 m thick; Schleicher & Schuell GmbH Dassel, Germany). As bona fide chemokine IL-8 (8 ng/ml) were used. Random migration was determined using HBSS in place of the chemoattractant. The cells (1 × 106 in 1 ml) were placed into the upper compartment, the chemokines
F. Günther et al. / Molecular Immunology 46 (2009) 1805–1813
1807
Fig. 1. Interaction of PMN with a S. aureus biofilm over time: PMN (green) were placed on the biofilm (red) (time 00), and their action was observed by time-lapse video microscopy for up to 60 min (the time in minutes is depicted in the upper corner of the images). Time zero is shown on the upper left panel, on the right panel the same area after 58 min. A sequence of events of a selected area (marked by the asterisk) is depicted in the lower panel: by 14 min two PMN appear. By 17 min depletion of biofilm next to those PMN is seen; biofilm-depleted areas increase with time. In parallel, an accumulation of red biofilm material is seen on and in the PMN, as well as a yellow staining as the result of the red and green superimposition (a video of these events is supplemented as S1). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
in the lower. After 2 h, the cells migrated into the filters were fixed with propanol, and stained with haematoxylin. The filters were analysed microscopically using an Omnicon Alpha Image Analyzer (Bausch and Lomb, Heidelberg, Germany). Chemotaxis was measured as “leading front”, defined as the distance in m from the top of the filter to a level where at least five cells could still be detected. Two parallel filters were prepared, and on each filter, 10 different areas were evaluated at minimum.
3. Results 3.1. PMN on bacterial biofilms: direct observation by time-lapse video microscopy Biofilms were grown for 6 days in chambered slides in the presence of Baclight RedTM , an efficient fluorescence label for bacteria. Isolated PMN, labelled with calcein, a green fluorescence dye for living cells, were placed on these biofilms and their action was observed directly by time-lapse video microscopy for up to 60 min. The PMN moved on the biofilm. Within minutes, the formation of aggregates containing 5–10 cells was observed. In the wake of the moving PMN, biofilm-depleted areas became apparent (Fig. 1 shows a sequence of these events; a video clip is supplemented as S1). When biofilm fragments were used, a migration of the PMN towards those fragments was seen, followed by a disruption, and eventually uptake of the biofilm, the latter indicated by the yellow staining of the PMN (Fig. 2 shows a sequence of these events; a video clip as supplement S2).
To assess whether PMN migrate into biofilms image, PMN were placed onto the biofilm for 60 min. Following fixation the bacteria were stained with SytoBC, the PMN were labelled with an antibody to CD66b conjugated with Cy3. The images were viewed by laser scan microscopy. For three-dimensional reconstruction, z-stacks were generated. The image of the biofilm alone is shown on the left panel; the biofilm with PMN on the right (Fig. 3). The upper panel shows the reconstruction with a flat angle (approximately 15◦ ), the lower panels a side view, which reveals the typical biofilm morphology with spikes and protrusions. When PMN were added, an infiltration into the biofilm was observed within 20–60 min, as was an uptake of biofilm material, particularly of the spikes. The cells moved as a front, and turned yellow, the latter indicative again of biofilm uptake (Fig. 3). 3.2. Migration of PMN towards bacterial biofilms Because the time-lapse video (S2) suggested a migration of PMN towards bacterial biofilms, we performed experiments to quantify the migration. Biofilms were grown in multiwell chambers and PMN were placed in inserts on top in a transwell assay. After 24 h, the number of PMN having migrated to the lower chamber was determined. When biofilm was present – or as bona fide chemokine – IL-8, PMN were found in the lower chamber, but not in the absence of biofilms or IL-8. On average, 7 × 104 PMN were found on the biofilm, and 5.5 × 105 PMN in the IL-8 containing compartment (Fig. 4A). In a further set of experiments, we tested whether the supernatants of biofilms contained the chemotactic activity by the use
1808
F. Günther et al. / Molecular Immunology 46 (2009) 1805–1813
Fig. 2. Interaction of PMN with a S. aureus biofilm fragment over time: As described above, PMN were placed on the biofilm – here large fragments were used – and the action was observed. The kinetics is shown for two selected areas (a and b). (a) After 8 min a PMN has moved towards a biofilm fragment, it snatches a piece of the film (10 min) and attaches it to its surface (12 min), and finally tears it off (14 min). Then the cell moves towards another fragment (37 min). (b) shows an analogue sequence of events. The intense yellow staining of the cells is indicative of uptake of bacteria (the video is supplemented as S2). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
Fig. 3. PMN migrate into the bacterial biofilm: PMN were placed on a S. aureus biofilm for 30 min and than viewed by confocal laser scan microscopy. The three-dimensional reconstruction shows the migration of the PMN (red) into the biofilm (green). On the left, the biofilm alone is shown, on the right, the biofilm with PMN added. The side view (lower panel) shows the spikes and protrusion of the biofilm and the PMN attacking them (> < marks the section plane; the arrows give the orientation of the image).
F. Günther et al. / Molecular Immunology 46 (2009) 1805–1813
1809
Fig. 4. (A) PMN migrate towards and into bacterial biofilms. (a) Biofilms were grown in tissue culture plates for 2 days. Then PMN were added to an insert (5 m pore size), and placed onto the cultured biofilm. After 24 h, migration of PMN into the lower chamber was determined microscopically. (b) For comparison interleukin 8 (IL-8) as bona fide chemoattractant was added to the culture dish in the absence of a biofilm, (c) shows the experiments without biofilm. In (d), the mean values of four parallel experiments are summarised. (B) Supernatants of biofilms were harvested at the times indicated and used as chemoattractant for PMN in a Boyden chamber assay. On the left, random migration (i.e. migration without chemoattractant) is shown, and migration towards IL-8 as “positive” control. Chemotaxis was measured in m and shown is the summary of 2 parallel filters, where 10 different areas were measured (the box contains 50% of the values; the whiskers show the highest and the lowest values, respectively, the dot the mean value and the horizontal bar the median). Compared to the time 0 (culture medium removed from the biofilm without incubation time) the mean values obtained for the migration towards the supernatants were significantly different (p < 10−5 ).
of a Boyden chamber assay. We used supernatants of biofilms from different maturation states, i.e. from 2 h to 72 h, based on the assumption that the properties of the biofilms change with time. PMN were clearly attracted to the biofilm supernatants, indicating that a soluble chemoattractant is released by the biofilm. The attractive capacity was in the same range as observed for IL-8. Interestingly, although biofilms mature over days, the maximum
attraction was already achieved with the supernatant of a 2 h old biofilm (Fig. 4B). 3.3. Phagocytosis of bacterial biofilms The time-lapse video films suggested phagocytosis uptake and deterioration of biofilm. To verify these observations, PMN were
1810
F. Günther et al. / Molecular Immunology 46 (2009) 1805–1813
Fig. 5. PMN on a S. aureus biofilm viewed by confocal laser scan microscopy: Bacteria in the biofilm (6-days-old) were labelled with SytoBC that stains the DNA (also that of the PMN). The PMN were identified by an antibody to CD66b, labelled red. When PMN were incubated at 37 ◦ C, areas depleted of biofilm around the PMN appeared (A) (arrows), but not when the cells were incubated at 4 ◦ C (B). (C) Biofilms were grown for 2, 6 and 15 days. PMN were added and after 30 min, the areas depleted around the PMN were quantified by planimetry (shown as arbitrary square units; the values are given as mean ± S.D. of four different areas viewed independently by three investigators).
added to bacterial biofilms for 30 min at 37 ◦ C. Then the specimens were fixed, SytoBC was added to label the DNA, and anti-CD66b to identify the PMN. As exemplified in Fig. 5A, uptake of the bacteria by the PMN was seen, and in addition depletion of the biofilm around the PMN. When the experiments were carried out at 4 ◦ C, a condition known to prevent phagocytosis of planktonic bacteria, neither uptake of bacteria nor biofilm-depleted areas around the PMN were observed, an example is shown in Fig. 5B. In other experiments, bacteria were identified with an antibody directed to S. aureus peptidoglycane instead of DNA-labelling. Essentially similar data were obtained (not shown). 3.4. Uptake and deterioration of biofilms depends on the maturation stage The observation that protrusions and spikes were readily taken up by the PMN suggested that newly added biofilm material might be especially sensitive to clearance by PMN. To assess the sensitivity of biofilms towards PMN systematically, S. aureus were cultivated on cover slips for 2, 6, and 15 days. By day 2 biofilms were clearly visible, they were, however, more fragile than the films grown for 6 or 15 days. When PMN were added to the biofilms, bacteriadepleted zones around the PMN became apparent after 30 min. The biofilm-depleted areas around the PMN of the 2- and 6-daysold biofilms appeared larger compared to those of the 15-days-old biofilms. By planimetry, a trend towards greater biofilm depletion in less mature biofilms could be confirmed, with statistically significant differences only between 2 and 15 days biofilms (Fig. 5C). The experiments shown were all performed with S. aureus derived from a patient. With a commercially available S. aureus strain, essentially similar data were obtained. 3.5. Quantification by 3 H labelling of the bacteria Because planimetry of the bacteria depleted areas is semiquantitative at best, and does not take into account threedimensional structures, we established a new method for the
quantification of the bacteria within the biofilm. The bacteria were cultured for 2, 6 or 15 days in the presence of 3 H-thymidine, which incorporates into the bacterial DNA. Then the films were carefully washed, and freshly isolated PMN were added for 30 min to allow phagocytosis. Then PMN, and hence phagocytosed bacteria, were removed, and the radioactivity of the residual biofilms was counted. Decline of radioactivity was interpreted as a measure for biofilm deterioration. Compared were always biofilms that had been exposed to PMN with those which had not (an example is shown Fig. 6A). In a parallel experiment, PMN were incubated with the biofilm at 4 ◦ C, which prevents phagocytosis. In that experiment, no loss of biofilm was seen. With use of this method, we now compared phagocytosis by PMN of biofilms that had been cultivated for 2, 6 and 16 days (Fig. 6B). On average, loss of biofilm amounted to 51.5% for 2-daysold films, to 43.7% for 6-days-old films, and to 21.4% for 15-days-old films (mean of three independent experiments with either patients’ derived S. aureus or S. aureus type Seattle 1945; the differences between 2-days-old and 15-days-old films were statistically significant with p = 0.02 using ANOVA). Again, incubation of biofilms with PMN at 4 ◦ C did not result in loss of biofilm.
3.6. Quantification of bacterial biofilm uptake by cytofluorometry As a further method to quantify uptake of bacterial biofilms by the PMN and to verify phagocytosis, as opposed to killing by exocytosed bactericidal compounds, cytofluorometry was performed using SytoBC labelled biofilms. In these experiments, PMN were placed on the biofilms for 30 min at 37 ◦ C or at 4 ◦ C, respectively, then harvested and subjected to cytofluorometry. To differentiate between bacteria that were attached to the outside of the PMN and those inside, crystal violet was used for quenching the outside-fluorescence. When the gate was set for the PMN, green fluorescence was seen in the samples incubated at 37 ◦ C, but not in those incubated at 4 ◦ C (example in Fig. 7A). Compared to 15day-old biofilms, the phagocytosis of 2- and 6-day-old biofilms was considerably higher compared (Fig. 7B and C). Again, with S.
F. Günther et al. / Molecular Immunology 46 (2009) 1805–1813
1811
Fig. 6. Deterioation of 3 H thymidine-labelled biofilms. (A) Bacteria were grown for 2 days in the presence of 3 H thymidine under conditions allowing biofilm formation. PMN were placed on the biofilms for 30 min at 37 ◦ C, then removed and radioactivity associated with the residual biofilms was determined (the data of 12 parallel experiments are shown as box blots with the box containing 50% of the values, the horizontal bar shows the median value, and the “whiskers” the highest and the lowest values, respectively). Compared to biofilm alone, significant loss of radioactivity was seen after incubation of the biofilm with PMN at 37 ◦ C (calculated by ANOVA). When the incubation with PMN was carried out at 4 ◦ C, there was no loss of biofilm. (B) The percentage of loss of biofilm was determined for 2-, 6-, and 15-days-old biofilms (here data of three independent experiments are evaluated; the mean values are given).
Fig. 7. Phagocytosis of bacterial biofilms determined by cytofluoromertry. (A) To parallel tubes containing SytoBC labelled biofilm (grown for 6 days) PMN were added and incubated at either 4 ◦ C (green line) or 37 ◦ C (blue line) for 45 min. Fluorescence outside the cells was quenched with crystal violet and fluorescence associated with the cells was determined. (B) PMN were added to biofilms (for 45 min at 37 ◦ C) that had been grown for 2 (red line), 6 (dark blue line), or 15 days (light blue line), respectively, and fluorescence was measured. (C) Data of 8 independent experiments are summarised for 2, 6, and 15 days old biofilms, respectively. Shown is the mean fluorescence intensity as statistical box blot with the box containing 50% of the values, the horizontal bar shows the median value, and the “whiskers” the highest and the lowest values, respectively). Differences between the mean values were calculated by ANOVA.
1812
F. Günther et al. / Molecular Immunology 46 (2009) 1805–1813
aureus from the patients and with the strain Seattle 1945 essentially similar data were obtained.
4. Discussion Bacterial biofilms are increasingly recognised as a common cause of chronic persistent infections. Thus, the question arises, how the host defence mechanisms react to bacterial biofilms, and whether or not biofilms can be cleared. Based on the facts that PMN are the primary cells in the host defence against bacteria, and are equipped with numerous bactericidal and cytotoxic entities, and that they infiltrate sites of biofilm infections, we assessed the interactions of isolated human PMN with in vitro-generated S. aureus biofilms by a variety of methods. By directly observing the interaction by time-lapse video microscopy, we came to the following conclusions: PMN move on intact biofilms, and biofilm-free areas appeared at the sites the PMN were vacating. Apparently, PMN take up biofilm as they move along. When biofilm fragments were used, a migration of PMN towards the biofilm was seen, as was tearing up of biofilm material, and eventually removal of the biofilm. Some of the PMN turned yellow, most probably due an overlap of the green and the red staining, which indicates close vicinity of biofilm and PMN, presumably attachment and uptake. Of note, the PMN remained viable during the observation time (60 min), and were not immobilised, as had been reported earlier for PMN on Pseudomonas aeruginosa biofilms (Jesaitis et al., 2002). By quantitative methods, we characterised the interaction of PMN with the biofilms. Firstly, we confirmed migration of PMN towards S. aureus biofilms and towards supernatants of the biofilms as well. The latter is in line with the fact that numerous bacteria, including S. aureus, produce and release into the environment substance(s) that attract PMN (Dürr et al., 2005; Rot et al., 1987). Moreover, we demonstrated a migration of PMN into the biofilm. The three-dimensional reconstruction suggested a massive infiltration of the cells that was not confined to water channels, as demonstrated before by Leid et al. (2002). Moreover, the images show that especially spikes and protrusions on the surface of the biofilm were removed. Because these data suggested that the newly formed biofilm material might be especially sensitive to phagocytosis by PMN, we performed a series of experiments with biofilms grown for different times. To quantify the effect of the PMN, areas devoid of biofilms around PMN were measured by planimetry. Compared to mature biofilms, young, immature biofilms (2- to 6-day-old) appeared to be more sensitive compared to 15-day-old biofilms. Because planimetry is two-dimensional only, and thus does not take into account holes in the biofilm, a further method, based on the DNA-labelling of the bacteria was developed. After exposure to PMN and removal of the cells to exclude phagocytosed bacteria, the radioactivity of the remaining biofilm was counted as a measure for biofilm loss. By this method, we could essentially confirm our previous data: loss of biofilm was most extensive in immature biofilms. The method, however, did not allow differentiating between phagocytosis and other means of biofilm degradation. Therefore, we used a cytofluorometry-based method to quantify phagocytosis of planktonic bacteria. We could confirm uptake of bacteria, and again, we found that immature biofilms were more sensitive to phagocytosis compared to mature films, but that also mature biofilms are not entirely protected against the attack by PMN. Taken together, our data show phagocytosis of bacterial biofilms by the PMN. Additional means of biofilm deterioration, such as exocytosis of bactericidal substances, however, cannot be ruled out. Our data seem to contradict an earlier report by Leid et al. (2002), who described adherence of leukocytes to S. aureus biofilms, their
migration into the film, but a failure to phagocytose the bacteria. In these studies, the mononuclear cell fraction was used, that contains predominantly monocytes, T- and B-cells, but not the PMN, which could account for the differences seen. The “halos” described by Leid et al., however, are reminiscent of the biofilm-depleted areas seen in our studies, and could be due to cytotoxic substances released from the PMN. Why immature biofilms are more sensitive to the PMN attack compared to mature biofilms is still a matter of speculation. Under our experimental conditions, the number of bacteria incorporated into the biofilm did not increase during biofilm formation after day 2 (data not shown). Moreover, also, the thickness of the biofilm did not drastically increase with maturation, but the mechanical stability did. We assume that the architecture and the composition of the extracellular polymer matrix changes with time. To date, methods to characterise and to analyse the extracellular polymer substances (EPS) are still being developed. The EPS is a highly intricate and dynamic structure, consisting of complex carbohydrates, proteins, lipids, nucleic acid, and exoenzymes as well (Branda et al., 2005; Flemming et al., 2007). Whether or not the EPS of S. aureus interacts with PMN has not yet been analysed. Data derived from experiments with P. aeruginosa EPS, however, show effects of EPS or EPS-derived compounds such as quorum-sensing molecules, alginates or rhamnolipids on PMN (Pedersen et al., 1990; Zimmermann et al., 2006; Jensen et al., 2007; Hänsch et al., 2008). Accepting the fact that PMN can phagocytose and clear biofilms the question arises how biofilms can persist in vivo. The paradigm “too many–too late” might also apply here, and we presume that the host defence against bacterial biofilms is subject to the same restriction as the defence against planktonic bacteria: the within-host-population dynamics, the health status of the host, and conditions favouring the colonisation with bacteria. The latter is particularly relevant for the so-called implant-associated infections, when biofilms form on medical devices such as indwelling catheters, artificial heart valves, osteosynthesis materials or endoprostheses (Donlan, 2001; Gottenbos et al., 2002). Presumably, these foreign materials provide a readily colonisable surface, and the “the race for the surface”, as it was proposed by Gristina et al. (1989), might be a decisive factor in the formation of biofilms. An attractive notion is that artificial surfaces – in contrast to epithelial cells – are without defence against colonisation and biofilm formation. While data on S. aureus biofilms are lacking, for P. aeruginosa the neutralisation by epithelial cells of quorum sensing molecules required for biofilm formation has been described, which in turn might result in the prevention of biofilm formation on the epithelium (Chun et al., 2004; Hastings, 2004). Another observation that might explain the persistence of biofilms was made when we analysed the infiltrated cells. The PMN at the site of biofilm infections were highly activated, as seen by the up-regulation of surface receptors and priming for superoxide generation, their chemotactic activity, however, was reduced. Thus, it is possible that the PMN cannot infiltrate into the biofilm, and consequently, phagocytosis and clearing of biofilms is restricted to the surface, and hence remains insufficient (Wagner et al., 2004). In conclusion, S. aureus biofilms can be invaded and phagocytosed by PMN. Therefore, they are not inherently protected against the host defence. Whether or not an infection becomes clinically apparent depends on numerous factors, especially the health status of the host, and conditions favouring biofilm formation.
Acknowledgment The work was supported by a grant from the Deutsche Forschungsgemeinschaft (WA1623/1-5).
F. Günther et al. / Molecular Immunology 46 (2009) 1805–1813
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.molimm.2009.01.020. References Arciola, C.R., An, Y.H., Campoccia, M.E., Donati, M.E., Montanaro, L., 2005. Etiology of implant orthopedic infections: a survey on 1027 clinical isolates. Int. J. Artif. Org. 28, 1091–1100. Branda, S.S., Vik, A., Freidman, L., Kolter, R., 2005. Biofilms: the matrix revisited. Trends Microbiol. 13, 20–26. Brenneis, H., Schmidt, A., Blaas-Mautner, P., Wörner, I., Ludwig, R., Hänsch, G.M., 1993. Chemotaxis of polymorphonuclear neutrophils (PMN) in patients suffering from recurrent infection. Eur. J. Clin. Invest. 23, 693–698. Chun, C.K., Ozer, E.A., Welsh, M.J., Zabner, J., Greenberg, E.P., 2004. Inactivation of a Pseudomonas aeruginosa quorum-sensing signal by human airway epithelia. Proc. Natl. Acad. Sci. 101, 3587–3590. Costerton, J.W., Stewart, P.S., Greenberg, E.P., 1999. Bacterial biofilms: a common cause of persistent infections. Science 284, 1318–1322. Davey, M.E., O’Toole, G.A., 2000. Microbial biofilms: from ecology to molecular genetics. Microbiol. Mol. Biol. Rev. 64, 847–867. Davies, D., 2003. Understanding biofilm resistance to antibacterial agents. Nat. Rev. Drug Discov. 2, 114–122. Donlan, R.M., 2001. Biofilms and device-associated infections. Emerg. Infect. Dis. 7, 277–281. Dunne, M.W., 2002. Bacterial adhesion: seen any good biofilm lately. Clin. Microbiol. Rev. 15, 155–166. Dürr, M.C., Kristian, S.A., Otto, M., Matteoli, G., Margolis, P.J., Trias, J., van Kessel, K.P., van Strijp, J.A., Bohn, E., Landmann, R., Peschel, A., 2005. Neutrophil chemotaxis by molecular pattern-recognition formylated peptides are crucial but not the sole neutrophil chemoattractant produced by Staphylococcus aureus. Cell. Microbiol. 8, 207–217. Flemming, H.-C., Neu, T.R., Wozniak, D.J., 2007. The EPS Matrix: the house of biofilm cells. J. Bacteriol. 189, 7945–7947. Gottenbos, B.G., Busscher, H.J., van der Mei, H.C., Nieuwenhuis, P., 2002. Pathogenesis and prevention of biomaterial centered infections. J. Mat. Sci. Med. 18, 717–722. Gristina, A.G., Nayler, P., Myrvik, Q., 1989. Infections from biomaterials and implants. Med. Prog. Technol. 14, 205–224. Hall-Stoodley, L., Costerton, J.W., Stoodley, P., 2004. Bacterial biofilms: from the environment to infectious disease. Nat. Rev. Microbiol. 2, 95–108. Hänsch, G.M., Brenner-Weiss, G., Prior, B., Wagner, C., Obst, U., 2008. The extracellular polymer substance of Pseudomonas aeruginosa: too slippery for neutrophils to migrate on? Int J. Artif. Organs 31, 796–809.
1813
Hastings, J.W., 2004. Bacterial quorum-sensing signals are inactivated by mammalian cells. Proc. Natl. Acad. Sci. 101, 3993–3994. Jensen, P.Ø., Bjarnsholt, T., Rasmussen, T.B., Calum, H., Moser, C., Pressler, T., Givskov, M., Hoiby, N., 2007. Rapid necrotic killing of PMN is caused by quorum sensing controlled production of rhamnolipids by Pseudomonas aeruginosa. Microbiology 153, 1329–1338. Jesaitis, A.J., Franklin, M.J., Berglund, D., Sasaki, M., Lord, C.I., Bleazard, J., Duffy, J.E., Beyenal, H., Lewandowski, Z., 2002. Compromised host defense on Pseudomonas aeruginosa biofilms: characterisation of neutrophil and biofilm interactions. J. Immunol. 171, 4329–4339. Leid, J.G., Shirtliff, H.C., Costerton, J.W., Stoodley, P., 2002. Human leukocytes adhere to, penetrate, and respond to Staphylococcus aureus biofilms. Infect. Immun. 70, 6339–6345. Lynch, S.A., Robertson, G.T., 2008. Bacterial and fungal biofilm infections. Ann. Rev. Med. 59, 415–428. Mah, T.F., O’Toole, G.A., 2001. Mechanisms of biofilm resistance to antimicrobial agents. Trends Microbiol. 9, 34–39. Parsek, M.R., Singh, P.K., 2003. Bacterial Biofilms: an emerging link to disease pathogenesis. Ann. Rev. Microbiol. 57, 677–701. Pedersen, S.S., Kharazmi, A., Espersen, F., Hoiby, N., 1990. Pseudomonas aeruginosa alginate in cystic fibrosis and the inflammatory response. Infect. Immun. 58, 3363–3368. Rot, A., Henderson, L.E., Copeland, T.D., Leonhard, E.J., 1987. A series of six ligands for the human formyl peptide receptor: tetrapeptides with high chemotactic potency and efficacy. PNAS 84, 7967–7971. Stewart, P.S., 2002. Mechanisms of antibiotic resistance in bacterial biofilms. Int. J. Med. Microbiol. 292, 107–113. Wagner, C., Heck, D., Lautenschläger, K., Iking-Konert, C., Heppert, V., Wentzensen, A., Hänsch, G.M., 2006. T-lymphocytes in implant-associated posttraumatic osteomyelitis: identification of cytotoxic T-effector cells at the site of infection. Shock 25, 241–246. Wagner, C., Kaksa, A., Müller, W., Denefleh, B., Heppert, V., Wentzensen, A., Hänsch, G.M., 2004. Polymorphonuclear neutrophils in posttraumatic osteomyelitis: cells recovered from the inflamed site lack chemotactic activitiy, but generate superoxides. Shock 22, 108–115. Wagner, C., Kondella, K., Bernschneider, T., Heppert, V., Wentzensen, A., Hänsch, G.M., 2003. Post-traumatic osteitis: analysis of inflammatory cells recruited into the site of infection. Shock 20, 503–510. Watnick, P., Kolter, R., 2000. Biofilm, city of microbes. J. Bacteriol. 182, 2675–2679. Zimmerli, W., Trampuz, A., Ochsner, P.E., 2004. Prosthetic-joint infections. N. Engl. J. Med. 351, 1645–1654. Zimmermann, S., Wagner, C., Müller, W., Brenner-Weiss, G., Hug, F., Prior, B., Obst, U., Hänsch, G.M., 2006. Induction of neutrophil chemotaxis by the quorum-sensing molecule N-3-oxododecanoyl homoserine lactone (3OC12-HSL). Infect. Immun. 74, 5687–5692.