ABSTRACT
In young cystic fibrosis (CF) patients, Staphylococcus aureus is typically the most prevalent organism, while in adults, Pseudomonas aeruginosa is the major pathogen. More recently, it was observed that also Streptococcus anginosus plays an important role in exacerbations of respiratory symptoms. These species are often coisolated from CF lungs, yet little is known about whether antibiotic killing of one species is influenced by the presence of others. In the present study, we compared the activities of various antibiotics against S. anginosus, S. aureus, and P. aeruginosa when grown in monospecies biofilms with the activity observed in a multispecies biofilm. Our results show that differences in antibiotic activity against species grown in mono- and multispecies biofilms are species and antibiotic dependent. Fewer S. anginosus cells are killed by antibiotics that interfere with cell wall synthesis (amoxicillin plus sulbactam, cefepime, imipenem, meropenem, and vancomycin) in the presence of S. aureus and P. aeruginosa, while for ciprofloxacin, levofloxacin, and tobramycin, no difference was observed. In addition, we observed that the cell-free supernatant of S. aureus, but not that of P. aeruginosa biofilms, also caused this decrease in killing. Overall, S. aureus was more affected by antibiotic treatment in a multispecies biofilm, while for P. aeruginosa, no differences were observed between growth in mono- or multispecies biofilms. The results of the present study suggest that it is important to take the community composition into account when evaluating the effect of antimicrobial treatments against certain species in mixed biofilms.
INTRODUCTION
Cystic fibrosis (CF) is one of the most common autosomal recessive diseases in Caucasians (1). CF is caused by a mutation in the cystic fibrosis transmembrane conductance regulator (CFTR) protein, eventually leading to thick airway mucus (2). A major predictor of morbidity and mortality in CF patients is respiratory infection due to bacteria that form biofilms in this thick airway mucus (3). In childhood, Staphylococcus aureus is the most prevalent species (4). In adulthood, there is a shift to Pseudomonas aeruginosa being the predominant species. In addition to these common pathogens, the CF lung microbiome contains numerous other microorganisms, including Burkholderia spp., Prevotella spp., Rothia spp., Stenotrophomonas spp., and/or Streptococcus spp. (5, 6) Members of the Streptococcus milleri group (SMG), including Streptococcus anginosus, can be found in children (7, 8) and adults (9, 10) and have recently been linked to acute pulmonary exacerbations in 39% of hospital admissions of adult CF patients (11, 12). In their cross-sectional patient cohort study, Filkins et al. (5) showed that SMG species were also present in clinically stable patients, suggesting that modest levels of SMG species can contribute to patient health, while excessive levels may lead to clinical decline. The biofilm mode of growth of bacteria plays an important role in the persistence of lung infections in CF patients (13). In a biofilm, bacterial cells are protected by a self-produced polymer matrix, show a decreased susceptibility to antibiotics, and are not efficiently cleared by the immune system (14, 15).
Species in a multispecies biofilm can cooperate or compete with each other, and this may impact the course, treatment, and outcome of biofilm-related CF airway infections (16). Most studies thus far have focused on interactions between S. aureus and P. aeruginosa. In early adulthood, S. aureus and P. aeruginosa initially can coexist; however, at a later stage, P. aeruginosa will outcompete S. aureus and benefit from its presence by using S. aureus as an iron source (17). Both 2-heptyl-4-hydroxyquinoline N-oxide (HQNO) and siderophores produced by P. aeruginosa play an important role in this competition. These factors, in combination with oxygen utilization by P. aeruginosa, are known to result in inhibition of the electron transport chain in S. aureus, leading to a growth disadvantage for S. aureus (18). However, suppression of S. aureus respiration by P. aeruginosa has also been reported to protect S. aureus from killing by the aminoglycoside tobramycin (19). In addition, P. aeruginosa is known to form membrane vesicles (MVs) containing cell wall-degrading enzymes. These MVs can attach to the surface of S. aureus, release enzymes, such as LasA protease, and consequently lead to the killing of S. aureus (20).
Few studies have focused on the interactions between other bacterial species present in the CF lung microbiome. Recently, SMG species were found to increase the expression of P. aeruginosa virulence factors, e.g., elastase and pyocyanin, contributing to CF lung disease progression (21). In contrast, the generation of reactive nitrogenous intermediates by oral streptococci can inhibit P. aeruginosa growth (22, 23). It has also been demonstrated that penicillinases produced by S. aureus protect Streptococcus spp. against penicillins (24, 25). Still, little is known about the effect of interspecies interactions in multispecies biofilms on antibiotic killing of biofilm cells. Therefore, in the present study, we compared the killing of three commonly coisolated bacteria, S. anginosus, S. aureus, and P. aeruginosa, by various antibiotics when grown as a mono- or multispecies biofilm.
RESULTS
P. aeruginosa influences biofilm formation of S. aureus and S. anginosus in a multispecies biofilm. S. anginosus, S. aureus, and P. aeruginosa were grown as mono- or multispecies biofilms in a 96-well microtiter plate (MTP). Medium containing bovine serum albumin (BSA) was used to allow for better growth of S. aureus in the presence of P. aeruginosa, as previously published (26). S. anginosus LMG 14502 grew to a significantly lower cell number when cocultured with S. aureus LMG 10147 and P. aeruginosa DK2 (difference of 0.61 ± 0.19 log, P ≤ 0.05) (Fig. 1). Similar results were obtained for another strain of S. anginosus (LMG 14696) (decrease in cell number of 0.50 ± 0.36 log, P ≤ 0.05). Also, for S. aureus LMG 10147, a reduction in cell number was observed when grown in a multispecies biofilm with S. anginosus LMG 14502 and P. aeruginosa DK2 (1.12 ± 0.40 log, P ≤ 0.05). For P. aeruginosa DK2, no difference in cell numbers was observed.
Average number of CFU of the different strains recovered from mono- and multispecies biofilms. Multispecies biofilms contained S. anginosus (LMG 14502 or LMG 14696), S. aureus LMG 10147, and P. aeruginosa DK2. Error bars represent standard deviations. n = 3 for P. aeruginosa, n = 6 for S. aureus and S. anginosus. *, significantly different from monospecies biofilm (P ≤ 0.05). Inoculum suspensions contained approximately 106 CFU/ml of P. aeruginosa, 106 CFU/ml of S. aureus, and/or 5 × 106 CFU/ml of S. anginosus.
When S. aureus LMG 10147 and S. anginosus (LMG 14502 or LMG 14696) were grown together in a dual-species biofilm, no significant difference in cell numbers could be observed compared to a monospecies biofilm (for S. aureus, differences of 0.02 ± 0.07 log and 0.31 ± 0.47 log when grown with S. anginosus LMG 14502 and LMG 14696, respectively, P > 0.05; for S. anginosus, differences of 0.01 ± 0.75 log and 0.24 ± 0.42 log for LMG 14502 and LMG 14696, respectively, P > 0.05). These data confirm that P. aeruginosa is responsible for the decrease in cell numbers observed in a multispecies biofilm.
More S. aureus cells are killed by antibiotics in a multispecies than monospecies biofilm, while antibiotic killing of P. aeruginosa is not affected.Antibiotic-mediated killing of P. aeruginosa DK2 and S. aureus LMG 10147 in mono- and multispecies biofilms was determined, and the results are shown in Table 1. Our data show that for S. aureus LMG 10147, there was significantly more antibiotic killing in a multispecies biofilm with S. anginosus LMG 14502 and P. aeruginosa DK2 (P ≤ 0.05) than in a monospecies biofilm when exposed to amoxicillin plus sulbactam, ceftazidime, ciprofloxacin, imipenem, levofloxacin, meropenem, tobramycin, or vancomycin. On the other hand, for P. aeruginosa DK2, no significant difference in killing between mono- and multispecies biofilms could be observed for any of the antibiotics tested.
CFU of S. aureus LMG 10147, P. aeruginosa DK2, and S. anginosus LMG 14502 and LMG 14696 killed per biofilm after treatment with antibiotics when grown in a mono- or multispecies biofilma
Antibiotic killing of S. anginosus in mono- versus multispecies biofilms.The log CFU values of S. anginosus LMG 14502 killed per biofilm after treatment with several antibiotics are shown in Table 1. A significantly decreased killing of S. anginosus LMG 14502 in a multispecies biofilm was observed for amoxicillin plus sulbactam, imipenem, and vancomycin (P ≤ 0.05). To determine whether or not this was dependent on the P. aeruginosa strain used, antibiotic killing of S. anginosus in the presence of S. aureus LMG 10147 and P. aeruginosa MH340 was evaluated as well (Table 1). While the presence of P. aeruginosa MH340 had no significant effect on the S. anginosus cell number in the absence of treatment (6.73 ± 0.53 log CFU/biofilm in a multispecies biofilm compared to 7.29 ± 0.28 log CFU/biofilm in a monospecies biofilm; P > 0.05), after treatment, a significant decrease in killing by amoxicillin plus sulbactam, imipenem, and vancomycin was again observed in the multispecies biofilm. Furthermore, when using MH340, significantly decreased killing was also observed for cefepime, ceftazidime, levofloxacin, and meropenem but not for ciprofloxacin or tobramycin.
In order to determine if the decreased antibiotic killing is due to an extracellular factor produced by S. aureus and/or P. aeruginosa, antibiotic killing of S. anginosus LMG 14502 was assessed in the presence of supernatant (1:1 diluted in medium) obtained from a monospecies biofilm of P. aeruginosa DK2 or MH340 or S. aureus LMG 10147 (Fig. 2). First, we assessed the biofilm formation of the strains under these conditions in the absence of antimicrobial agents. When grown in the supernatant of P. aeruginosa DK2, a significant reduction in biofilm formation was observed (0.52 ± 0.14 log, P ≤ 0.05), while the diluted supernatant of P. aeruginosa MH340 or S. aureus LMG 10147 had no effect. Second, the growth curves of S. anginosus LMG 14502 were evaluated. The results show that S. anginosus grew to a lower optical density (OD) in biofilm medium 1:1 diluted with S. aureus supernatant than in undiluted biofilm medium. Growing S. anginosus in biofilm medium diluted 1:3 with Milli-Q (MQ) water could mimic growth profile in supernatant, whereas almost no growth could be observed in a 1:10 dilution in MQ (Fig. S1). Therefore, to evaluate if an alteration in the growth curve of S. anginosus in the presence of supernatant could be responsible for altered antibiotic killing, killing of S. anginosus LMG 14502 was evaluated in biofilm medium 1:3 diluted with MQ. Biofilm cell numbers were assessed in the absence of treatment and after treatment with cefepime, imipenem, and vancomycin. No difference could be observed compared to growth in undiluted biofilm medium (see Table S2 in the supplemental material).
Log S. anginosus biofilm cells killed after treatment with antibiotic solutions (concentrations in micrograms per milliliter) under several conditions: (i) monospecies biofilm in undiluted medium (BHI supplemented with 5% BSA, 0.5% mucin type II, and 0.3% agar), (ii) monospecies biofilm in diluted biofilm supernatant of a monospecies biofilm of P. aeruginosa DK2, (iii) monospecies biofilm in diluted biofilm supernatant of a monospecies biofilm of P. aeruginosa MH340, (iv) monospecies biofilm in diluted biofilm supernatant of a monospecies biofilm of S. aureus LMG10147. Error bars represent standard deviations. n ≥ 3. *, significantly different from monospecies biofilm (P ≤ 0.05).
However, a significant decrease in killing by amoxicillin plus sulbactam, cefepime, imipenem, meropenem, or vancomycin could be observed when S. anginosus was grown in the supernatant of S. aureus LMG 10147 but not in the supernatant of P. aeruginosa. These data suggest that S. aureus is responsible for the observed decreased antibiotic killing of S. anginosus in a multispecies biofilm.
To evaluate if the decreased antibiotic killing observed in biofilm medium and in supernatant of S. aureus LMG 10147 is strain dependent, a second S. anginosus strain was tested (LMG 14696). Again, a significant decrease in killing of S. anginosus by amoxicillin plus sulbactam, cefepime, imipenem, meropenem, or vancomycin, but not by ciprofloxacin, levofloxacin, or tobramycin, could be observed when S. anginosus was grown in a multispecies biofilm (Table 1). When grown in a supernatant of S. aureus LMG 10147, decreased killing by amoxicillin plus sulbactam, cefepime, imipenem, meropenem, and vancomycin was observed as well (Fig. 2).
For vancomycin (512 μg/ml), experiments were also carried out using the supernatant of a planktonic overnight S. aureus LMG 10147 culture in biofilm medium. Again, protection of S. anginosus biofilm cells against vancomycin could be observed (decrease of 0.55 ± 0.66 log), indicating that the effect is not specific for a biofilm supernatant of S. aureus LMG 10147 but that it is also observed with a supernatant of planktonic S. aureus cultures.
Furthermore, experiments with vancomycin (2 times the MIC) were also carried out using planktonic cultures of S. anginosus instead of biofilms, grown in undiluted biofilm medium, or grown in medium 1:1 diluted with biofilm supernatant of S. aureus. No viable planktonic cells could be recovered under any of the conditions, indicating that the protective effect of S. aureus biofilm supernatant is specific for S. anginosus biofilm cells.
Other S. aureus strains also decrease susceptibility of S. anginosus.To evaluate whether other S. aureus strains also cause a decrease in antibiotic killing, S. anginosus was grown and treated in the supernatants of several other strains (Table 2). There was significantly less killing of S. anginosus (for all antibiotics tested) when it was grown and treated in the supernatants of S. aureus W8, W1, W22, and Mu50 (P ≤ 0.05). When grown in the supernatants of USA300 and ATCC 25923, there was also less killing of S. anginosus, except for cefepime (with the supernatant of USA300) and meropenem and vancomycin (with the supernatant of ATCC 25923). Killing by ciprofloxacin, levofloxacin, and tobramycin was also evaluated in the supernatants of S. aureus strains LMG 10147 and W8 (Table 3), but no difference could be observed. Similar results were obtained with S. anginosus LMG 14696 (Table 4).
CFU of S. anginosus LMG 14502 killed per biofilm after treatment with amoxicillin plus sulbactam, cefepime, imipenem, meropenem, or vancomycin when grown in biofilm medium or in diluted biofilm supernatant of several S. aureus strainsa
CFU of S. anginosus LMG 14502 killed per biofilm after treatment with ciprofloxacin, levofloxacin, or tobramycin when grown in biofilm medium or in diluted biofilm supernatant of S. aureus LMG 10147 or W8a
CFU S. anginosus LMG 14696 killed per biofilm after treatment with amoxicillin plus sulbactam, cefepime, imipenem, meropenem, or vancomycin when grown in biofilm medium or in diluted biofilm supernatant of S. aureus W22, USA300, ATCC 25923, or Mu50a
For both S. anginosus LMG 14502 and LMG 14696, experiments were carried out using dual-species biofilms of S. anginosus and several S. aureus strains. The results show that both S. anginosus strains were also protected against antibiotic killing when grown together with S. aureus (Tables S3 and S4).
Cell-free culture supernatant of S. aureus did not alter MICs of S. anginosus LMG 14502.To evaluate if a cell-free supernatant of S. aureus biofilms could lead to a decrease in MIC value, the MIC values of amoxicillin, amoxicillin plus sulbactam, meropenem, cefepime, imipenem, and vancomycin toward S. anginosus LMG 14502 were determined in biofilm medium and in a supernatant of S. aureus LMG 10147 (diluted 1:1 in medium) (Table 5). No difference in MICs were observed under the two conditions, except for amoxicillin. In supernatant, the MIC of amoxicillin was 0.5 μg/ml, whereas the MIC in undiluted biofilm medium was 0.0625 μg/ml. The addition of a β-lactamase inhibitor again reduced the MIC in supernatant to the same value as in undiluted biofilm medium. The MIC values of S. anginosus LMG 14696, S. aureus LMG 10147, and P. aeruginosa DK2 were also determined (Table S5).
MIC values of S. anginosus LMG 14502 in biofilm medium compared to growth in diluted biofilm supernatant of a monospecies S. aureus LMG 10147 biofilm
Extracellular DNA concentration is not altered in biofilms grown in the supernatant of S. aureus.To investigate the role of extracellular DNA (eDNA) in the reduced antibiotic efficacy, its concentration in biofilms grown in undiluted medium and in diluted supernatant was determined. No significant differences were observed (Table 6). To confirm the lack of a role for eDNA, we evaluated the activities of a selection of antibiotics (imipenem, meropenem, cefepime, and vancomycin) in the presence of DNase I. No differences in killing efficacy could be observed (Table 7).
eDNA concentration of S. anginosus LMG 14502 and LMG 14696 grown in biofilm medium or in diluted biofilm supernatant of S. aureus LMG 10147a
CFU of S. anginosus LMG 14502 killed per biofilm after treatment with cefepime, imipenem, meropenem, or vancomycin with or without addition of DNase I when grown in biofilm medium diluted 1:1 with supernatant of S. aureus LMG 10147a
DISCUSSION
Research on the activity of antibiotics against biofilms has mainly focused on monospecies biofilms (27–29), while many biofilm-related infections are due to multiple species (30, 31). Therefore, in the present study, we investigated the role of community composition on antibiotic-mediated killing in a multispecies biofilm.
Antibiotic killing of S. anginosus, S. aureus, and P. aeruginosa was compared in mono- and multispecies biofilms. Our results show that the antibiotic killing of P. aeruginosa is not influenced by the presence of S. aureus and S. anginosus, which is in line with results observed by Price et al. (32). They reported no significant difference in killing of P. aeruginosa by tobramycin when grown together with Streptococcus constellatus, also a member of the SMG. DeLeon et al. (33) investigated the killing of P. aeruginosa by gentamicin and ciprofloxacin. Also, in this study, no significant difference in killing of P. aeruginosa was observed when grown together with S. aureus. On the other hand, Michelsen et al. (34) suggested there was protection of P. aeruginosa in the presence of S. aureus on agar plates containing inhibitory levels of tobramycin, gentamicin, and ciprofloxacin, which could be explained by an overexpression of efflux mechanisms and lipopolysaccharide modification, induced by the presence of S. aureus (34).
Furthermore, our results show that more S. aureus cells are killed by antibiotic treatment when grown together with S. anginosus and P. aeruginosa, both by antibiotics that interfere with cell wall synthesis and those that interfere with other cellular processes. For example, we observed an increased killing by ciprofloxacin and tobramycin. However, DeLeon et al. (33) observed no difference in killing by ciprofloxacin and even some protection of S. aureus against gentamicin in the presence of P. aeruginosa. Aminoglycoside-modifying enzymes produced by P. aeruginosa were proposed to be involved. In addition, Hoffman et al. (19) showed that HQNO produced by P. aeruginosa protected S. aureus during coculture from killing by tobramycin. The use of another growth medium and other bacterial strains or the presence of S. anginosus could be responsible for the observed differences in outcome between the results of the present study and theirs. In contrast to S. aureus, fewer S. anginosus cells were killed in a multispecies biofilm, but this was observed only for antibiotics that interfere with cell wall synthesis and not with other antibiotics. However, the results for S. anginosus to some extent also depend on the P. aeruginosa strain present in the multispecies community (DK2 versus MH340).
When antibiotic-mediated killing of sessile S. anginosus cells was quantified in the supernatant of several S. aureus strains (supernatant of planktonic cultures and of biofilm of S. aureus), decreased killing of S. anginosus was again observed for antibiotics that interfere with cell wall synthesis, but not for other classes. In contrast, in the supernatant of P. aeruginosa (both DK2 and MH340), no decreased killing of sessile S. anginosus cells could be observed for any of the antibiotics used. These results indicate that a yet-unidentified factor produced by S. aureus is responsible for the reduced killing of S. anginosus. As the effect is less pronounced when grown in a multispecies biofilm than in the supernatant of S. aureus, the negative influence of P. aeruginosa on S. aureus or S. anginosus might lead to a lower protective effect of S. aureus on S. anginosus. These experiments were also carried out with planktonic S. anginosus cultures, but no protection was seen, suggesting that the observed decreased killing of S. anginosus is biofilm specific.
Next, we wanted to investigate which mechanism could be responsible for the decreased antibiotic killing of S. anginosus. Some studies reported that eDNA in the biofilm matrix could chelate cations leading to antibiotic tolerance, e.g., to vancomycin (35–38). However, when quantifying the eDNA concentration in monospecies S. anginosus biofilms, we could not observe a difference between the eDNA concentrations in biofilms grown with and without supernatant. We confirmed that eDNA is not a major player in this regard by showing that the addition of DNase I (described to enhance the effect of antibiotics under some conditions [35, 39]) did not affect the number of surviving S. anginosus cells recovered from biofilms treated with antibiotics. Furthermore, the effect observed is also not due to the presence of β-lactamases of S. aureus, as protection is also seen against vancomycin, against amoxicillin in the presence of a β-lactamase inhibitor, and in the supernatant of a β-lactamase-negative S. aureus strain (ATCC 25923). In addition, the effect also seems to be independent from growth inhibition, as growth is reduced by DK2 but not by MH340, whereas DK2 has less of an effect than MH340 on the reduction in antibiotic susceptibility of S. anginosus.
Our data ambiguously demonstrate that interactions between species in a multispecies biofilm do not always lead to a change in antimicrobial susceptibility, but these changes depend on the antibiotic and the species involved. Interestingly, sessile S. anginosus cells seem to be protected by one or more compounds secreted by S. aureus, as decreased killing is also observed when S. anginosus is grown in the cell-free supernatant of S. aureus planktonic or biofilm cultures. Further experiments will be needed to elucidate the mechanisms involved. As protection was observed only against antibiotics that interfere with cell wall synthesis, the likely target is peptidoglycan synthesis. Finally, the treatment of CF patients is often directed toward the major pathogens P. aeruginosa and/or S. aureus, and our data show that this can lead to increased survival of S. anginosus, which in turn could lead to acute exacerbations and a decline in lung function (10–12, 40).
MATERIALS AND METHODS
Bacterial strains. S. anginosus LMG 14502 (isolated from a human throat) and LMG 14696 (isolated from a respiratory infection), P. aeruginosa DK2 (isolated from CF sputum [41]) and MH340 (PAO1 [42], isolated from a wound), S. aureus LMG 10147 (isolated from a wound), S. aureus W2, W8, and W22 (three CF sputum isolates, kindly provided by Jozef Dingemans), ATCC 25923 (a clinical isolate), Mu50 (isolated from a wound), and USA300 (isolated from a soft tissue infection) were cultured overnight at 37°C in brain heart infusion (BHI) broth (Oxoid, Basingstoke, UK).
Formation of biofilms in medium.Biofilm formation was performed as described previously (43). Briefly, inoculum suspensions containing approximately 106 CFU/ml of P. aeruginosa, 106 CFU/ml of S. aureus, and/or 5 × 106 CFU/ml of S. anginosus were made in biofilm medium (BHI supplemented with 5% [wt/vol] BSA, 0.5% [wt/vol] mucin type II, and 0.3% [wt/vol] agar), BSA was added to allow coculture of all three species together, and mucin and agar were added to reflect aspects of the CF lung (26). The inoculum cell numbers were based on preliminary optimization experiments. We tested several starting inoculum combinations to find out which combination led to 24-h-old biofilms with similar cell numbers for the three species. Ninety-six-well MTPs (TTP, Trasadingen, Switzerland) were filled with the suspensions of all three species alone or together. After 4 h of adhesion at 37°C, the supernatant was removed, and 100 μl of fresh medium was added. After an additional 20 h, cells were rinsed with 100 μl of physiological saline (PS), and 200 μl of an antibiotic solution in biofilm medium was added to the mature biofilm. To the control wells, 200 μl of biofilm medium was added. The plates were incubated for 24 h at 37°C. For each test condition, a minimum of three biological replicates with each two technical replicates was included.
Antibiotic solutions.The differences in susceptibility of all three species between growth in mono- and multispecies biofilms were determined toward amoxicillin plus sulbactam (both Sigma-Aldrich, Diegem, Belgium), aztreonam (TCI Europe, Zwijndrecht, Belgium), cefepime (Sigma-Aldrich), ceftazidime (Sigma-Aldrich), ciprofloxacin (Sigma-Aldrich), colistin (Sigma-Aldrich), imipenem (Sigma-Aldrich), levofloxacin (Sigma-Aldrich), meropenem (Hospira, IL, USA), tobramycin (TCI Europe), and vancomycin (Sigma-Aldrich) in biofilm medium. The concentrations used (partly based on concentrations that can be reached in serum or sputum) were the following: 5 μg/ml amoxicillin (44), 4 μg/ml sulbactam, 500 μg/ml aztreonam (45), 500 μg/ml cefepime (46), 150 μg/ml ceftazidime (47), 150 μg/ml ciprofloxacin (48), 200 μg/ml colistin (49), 100 μg/ml imipenem (50), 500 μg/ml levofloxacin (51), 500 μg/ml meropenem (52), 200 μg/ml tobramycin (53), and 512 μg/ml vancomycin (54).
Generation of cell-free culture supernatant and formation of S. anginosus biofilms and planktonic cultures in cell-free culture supernatant.Monospecies biofilms of S. aureus or P. aeruginosa were formed in biofilm medium, as described above. After 4 h of adhesion and 20 h of maturation, the culture supernatant was collected and centrifuged at 5,000 rpm for 10 min. The supernatant was subsequently sterilized using 0.22-μm-pore-size filters (Merck Millipore, Billerica, MA, USA), resulting in cell-free culture supernatant. To confirm the complete removal of all microorganisms, 100 μl of the supernatant was spread on a BHI agar plate and incubated for 24 h at 37°C. Supernatant was either used immediately or stored at −20°C for a maximum of 48 h. For the supernatant of planktonic S. aureus cultures, S. aureus was grown overnight in biofilm medium for 16 h, and the supernatant was collected as described above.
To evaluate the effect of growth in supernatant, first, growth curves of S. anginosus LMG 14502 were determined in biofilm medium, biofilm medium diluted 1:1 with the supernatant of an S. aureus LMG 10147 biofilm, or diluted 1:3 or 1:10 with MQ water (Merck Millipore), using an Envision multilabel reader (PerkinElmer LAS, Boston, MA, USA) by plotting the OD at 590 nm (OD590) versus the incubation time. Next, monospecies biofilms of S. anginosus were grown and treated as described above in a 1:1 mixture of biofilm medium and biofilm supernatant of S. aureus or P. aeruginosa, or in biofilm medium diluted 1:3 with MQ. Cell numbers were determined by plate counting.
To determine the activity of a biofilm supernatant on planktonic S. anginosus cultures, overnight S. anginosus cultures in BHI were grown to an OD605 of 0.05 in biofilm medium or in biofilm medium diluted 1:1 in biofilm supernatant of S. aureus. After 24 h at 37°C and 250 rpm, the tubes were centrifuged (5,000 rpm, 5 min), the supernatant was removed, and the pellets were rinsed with PS. Vancomycin (2× the MIC [2 μg/ml]) in biofilm medium or in biofilm medium diluted 1:1 with supernatant was added to the test tubes for another 24 h (37°C, 250 rpm). The tubes were again centrifuged (5,000 rpm, 5 min), the supernatant was removed, and pellets were rinsed using PS. Cell numbers were determined using the plate count method, as described below.
Quantification of biofilm cells.After 24 h of treatment, biofilms were washed with PS, and cells were collected by sonication and vortexing, as described previously (26). Cell numbers were determined by the plate count method, using selective agar for S. anginosus (BHI agar supplemented with 1.25 mg/liter triclosan [Sigma-Aldrich]), S. aureus (tryptic soy agar supplemented with 7.5% NaCl), and P. aeruginosa (Pseudomonas isolation agar). S. anginosus plates were incubated anaerobically for 48 h at 37°C, while S. aureus and P. aeruginosa plates were incubated aerobically for 48 h at 37°C. The log CFU/biofilm was calculated by subtracting the log of surviving cells after treatment from the corresponding log control cells (of which the averages are presented in Fig. 1).
Determination of the MIC.The MIC values of amoxicillin (plus sulbactam), cefepime, imipenem, meropenem, and vancomycin toward S. anginosus LMG 14502 were determined in duplicate according to the EUCAST broth microdilution protocol in flat-bottom 96-well MTP (TTP) (55), with concentrations ranging from 0.03125 to 25 μg/ml. The MIC was defined as the lowest concentration for which no significant difference in the OD590 was observed between the blank and inoculated wells after 24 h of growth at 37°C (56). The results obtained in replicate experiments did not differ more than 2-fold. When a 2-fold difference was observed, the lowest concentration was recorded as the MIC.
Effects of DNase I on antibiotic susceptibility of a 24-h-old S. anginosus biofilm.Biofilms were formed as described above in cation-supplemented medium (0.015% [wt/vol] CaCl2, 2.0 mM MgCl2), which is essential for DNase I activity (57, 58). After 24 h of incubation at 37°C, the supernatant was discarded, and biofilms were washed with 100 μl of PS. Two hundred microliters of an antibiotic solution, together with 100 μg/ml DNase I (Sigma-Aldrich), was added to the wells. After 24 h of additional incubation at 37°C, cell numbers were determined by plate counting on BHI agar supplemented with 1.25 mg/liter triclosan.
Quantification of extracellular DNA in the biofilm matrix.eDNA in the biofilm matrix was quantified as previously described (35). Briefly, S. anginosus biofilms were formed in biofilm medium or a supernatant of S. aureus, as described above. Biofilm cells were washed with PS and collected by pipetting up and down in Eppendorf protein LoBind microcentrifuge tubes (1.5 ml) (Eppendorf AG, Hamburg, Germany). One hundred microliters of this solution was used for plate counting to determine the number of biofilm cells. Subsequently, biofilm cells were separated from the matrix by centrifugation at 5,000 rpm for 10 min at 4°C. The supernatant was aspirated and filtered through a 0.2-μm-pore-size cellulose acetate filter (Whatman GmbH, Dassel, Germany). The amount of eDNA was quantified using the QuantiFluor double-stranded DNA (dsDNA) system kit (Promega, Madison, WI, USA) and normalized to the number of biofilm cells (determined by plate counting). Five biological replicates were included.
Statistical data analysis.Statistical data analysis was performed using SPSS software, version 24 (SPSS, Chicago, IL, USA). The normal distribution of the data was verified using the Shapiro-Wilk test. Normally distributed data were analyzed using an independent sample t test. Nonnormally distributed data were analyzed using a Mann-Whitney test. Differences with a P value of ≤ 0.05 were considered significant.
ACKNOWLEDGMENTS
This research has been funded by the Interuniversity Attraction Poles Programme initiated by the Belgian Science Policy Office and by FWO Vlaanderen.
FOOTNOTES
- Received 9 February 2017.
- Returned for modification 8 March 2017.
- Accepted 2 July 2017.
- Accepted manuscript posted online 10 July 2017.
Supplemental material for this article may be found at https://doi.org/10.1128/AAC.00302-17 .
- Copyright © 2017 American Society for Microbiology.
