Abstract
Biofilms are formed by closely adjacent microorganisms embedded into an extracellular matrix this way providing them with strong protection from antimicrobials, which is often further reinforced in polymicrobial biofilms. Despite of the well-known antagonistic interactions between S. aureus and P. aeruginosa, the most common pathogens causing various nosocomial infections, they often form mixed consortia characterized by increased pathogenicity and delayed recovery in comparison with single species infections. Here we show that, while S. aureus could successfully avoid a number of antimicrobials by embedding into the biofilm matrix of P. aeruginosa despite of their antagonism, the very same consortium was characterized by 10–fold higher susceptibility to broad-spectrum antimicrobials compared to monocultures. Moreover, quantitatively similar increase in antimicrobials susceptibility could be achieved when P. aeruginosa was introduced into S. aureus biofilm, compared to S. aureus monoculture. In a reverse experiment, intervention of S. aureus into the mature P. aeruginosa biofilm significantly increased the efficacy of ciprofloxacin against P. aeruginosa. A broader perspective is provided by antagonistic bacteria intervention into already preformed monoculture biofilms leading to the considerable enhancement of their antibiotic susceptibility. We believe that this approach has a strong potential of further development towards innovative treatment of biofilm-associated infections such as transplantation of the skin residential microflora to the wounds and ulcers infected with nosocomial pathogens to speed up their microbial decontamination.
Author summary Biofilms formation is one of the key mechanisms providing pathogenic bacteria with extreme resistance to antimicrobials. On the S. aureus and P. aeruginosa mixed culture model we show explicitly that antimicrobials efficacy against bacteria in mixed biofilms differs considerably from monoculture biofilms. From the one hand, S. aureus avoids vancomycin and ampicillin by the rearrangements to the lower layers of the P. aeruginosa biofilm matrix. On the other hand, in the same consortium susceptibility to ciprofloxacin and aminoglycosides increases nearly 10– fold compared to monocultures. This finding allowed suggesting that intervention of antagonistic bacteria into already preformed monoculture biofilms could be used as an innovative approach to their treatment by increasing their antibiotic susceptibility. Thus, by introducing P. aeruginosa into preformed S. aureus biofilm, susceptibility of S. aureus to aminoglycosides was increased 4-fold, compared to monoculture. The intervention of S. aureus into the mature P. aeruginosa biofilm significantly increased the efficacy of ciprofloxacin against P. aeruginosa. We believe that this approach has a strong potential of further development towards innovative treatment of biofilm-associated infections such as introduction of the skin residential microflora to the wounds and ulcers infected with nosocomial pathogens to speed up their microbial decontamination.
Introduction
Bacterial fouling is an important factor that strongly affects acute and chronic wounds healing. Recent reports indicate that bacterial biofilms prevent wound scratch closure [1]. Besides the physical obstruction of the cells, pathogenic bacteria produce various virulence factors including toxins and proteases that also affect cytokine production by keratinocytes, induce apoptosis of the host cells and cause inflammation [2-6].
S. aureus and P. aeruginosa are one of the most widespread pathogenic agents causing various nosocomial infections, including pneumonia on the cystic fibrosis background, healthcare associated pneumonia and chronic wounds [7-10]. During infection, bacterial cells are embedded into a self-produced extracellular matrix of organic polymers this way forming either mono-or polymicrobial biofilms [11, 12] which drastically reduce their susceptibility to both antimicrobials and the immune system of the host [13, 14]. Accordingly, interspecies interactions between S. aureus and P. aeruginosa within mixed biofilms attracted major attention in recent years including both in vitro [15] and in vivo studies [16]. Current data suggests that bacterial pathogenicity is promoted during polymicrobial infections and recovery is delayed in comparison with monoculture infections [15-17]. Recently, P. aeruginosa was reported to be the dominant pathogen in S. aureus-P. aeruginosa mixed infections [16]. P. aeruginosa is known as a common dominator in polymicrobial biofilm-associated infections due to multiple mechanisms allowing its rapid adaptation to the specific conditions of the host. In particular, P. aeruginosa produces multiple molecules to compete with other microorganisms for space and nutrients. Therefore, it either strongly reduces or even completely outperforms S. aureus during co-culture in vitro in both planktonic and biofilm forms [18-21].
While it is known that S. aureus and P. aeruginosa exhibit rather antagonistic relationship [22, 23], several studies reported their mutual association in acute and chronic wounds embedded in a mixed biofilm [8, 24-28], with S. aureus typically residing on the wound surface, whereas P. aeruginosa being rather observed in the deep layers [15, 28-31]. Interestingly, in mixed P. aeruginosa - S. aureus biofilms from cystic fibrosis patients S. aureus was shown to be dominating during childhood, with P. aeruginosa prevalence increasing with aging and worsening patient prognosis [32-34].
During the biofilm formation P. aeruginosa produces three main exopolysaccharides, namely alginate, Pel, and Psl, which form an extracellular matrix in the biofilm exhibiting both structural and protective functions [35-38]. Under prevalent Pel secretion, loose biofilm structures are formed [39] and thus S. aureus is able to penetrate into the biofilm [39]. When growing in consortium with P. aeruginosa, S. aureus switches to the small colony variants (SCVs), a well-characterized phenotype detected in various diseases, including cystic fibrosis and device-related infections [40-43]. SCVs appear as small, smooth colonies on a culture plate and grow significantly slower compared to wild type colonies. Remarkably, switch to the SCV phenotype improves the survival of S. aureus under unfavorable conditions, as it exhibits increased aminoglycoside resistance, biofilm formation, and intracellular survival [40, 43-45]. Prolonged co-culture with P. aeruginosa leads to higher proportions of stable S. aureus SCVs that is further increased in the presence of aminoglycosides [43]. In has been recently suggested that rare observation of S. aureus and P. aeruginosa together in diagnostic cultures of sputum of cystic fibrosis patients could be attributed to the existence of S. aureus as SCVs that are more difficult to detect due to their small size and fastidious growth requirements [40, 45].
S. aureus is a common opportunistic pathogen responsible for the majority of skin infections resulting in increased morbidity, mortality, and exhibiting increased rise of antibiotic-resistant strains in the last decades. Investigations on alternative treatment options against biofilm-associated infections are largely based upon using specialized agents (such as quaternary ammonium compounds, curcumin or chlorquinaldol) or enzymatic treatment that in combinations with antibiotics provide high local drug concentrations avoiding systemic adverse effects [46-52]. While many approaches to targeting staphylococcal biofilms were reported [50, 53-57], only few successive ways of targeting P. aeruginosa are known [52, 58-60]. Among various compounds exhibiting anti-biofilm activities, the derivatives of 2(5H)-furanone have been reported to inhibit biofilm formation by Staphylococci [61-65]. While many of these approaches exhibited promising results against staphylococcal monocultures, their efficiency against polymicrobial biofilms remains questionable.
Only few investigations indicated that extracellular polymeric substances forming the biofilm matrix provide protection against antibiotics to all inhabitants of the biofilm, including the non-producers, although the biofilm as a whole is weakened [39, 66], this way proposing that S. aureus could potentially survive in the presence of P. aeruginosa and even co-exist with it in a polymicrobial biofilm, benefiting from the antimicrobial barrier formed by the P. aeruginosa matrix components.
Here we demonstrate explicitly that S. aureus successfully incorporates into the P. aeruginosa biofilm matrix under conditions of staphylococcus-specific treatment and survives there in presumably SCV-like form. In contrast, the efficiency of broad-spectrum antimicrobials like ciprofloxacine and aminoglycosides active against both bacterial species in mixed biofilms increased nearly 10-fold in comparison with corresponding monocultures. These data suggest that interspecies interactions appear a key determinant that strongly governs the antibiotic susceptibility in mixed biofilms, the fact that should be taken in account when considering an optimized strategy of polymicrobial infections treatment.
Results
Modeling the S. aureus – P. aeruginosa mixed biofilm
Despite of known antagonistic interactions between S. aureus and P. aeruginosa [23], they are still the most common pathogens evoking wound infections and forming mixed biofilms on their surfaces [25, 26, 28]. We have simulated in vitro different situations where either S. aureus in a fresh broth was added to the preformed 24-h old biofilm of P. aeruginosa or, vice versa, P. aeruginosa was added to the preformed 24-h old biofilm of S. aureus, with cultivation continued for the next 24 h. As a control, both strains were inoculated simultaneously and grown for 48 h with the broth exchange after 24 h of cultivation. Both S. aureus and P. aeruginosa were able to penetrate into the preformed biofilm of the other bacterium (Fig 1). Irrespective of which bacterium initially preformed the biofilm and which one was added later, the ratio of their CFUs in the biofilm after 24 h cultivation remained around 1:10 with the prevalence of the first biofilm former (Fig 1 A and B), and was 1:1 when both bacteria were inoculated simultaneously (Fig 1 C). Therefore in the following experiments simultaneous inoculation of both bacteria was used to obtain their mixed biofilm.
Next to analyze the biofilm structure and cells distribution in the matrix, the S. aureus - P. aeruginosa mixed biofilm was grown in imaging cover slips, stained with ViaGram™ Red+ to differentiate between S. aureus and P. aeruginosa followed by their analysis with confocal laser scanning microscopy. To estimate the viability of the cells, SYTO9/propidium iodide staining was also performed, as the ViaGram™ Red+ staining requires buffer change that disturbs the biofilm structure. Both S. aureus and P. aeruginosa formed 20-25 μm-thick biofilms when growing as monocultures (Fig 2A, B). While the mixed biofilm was of similar thickness, it appeared more rigid in comparison with monoculture ones (Fig 2C). Interestingly, in the mixed biofilm, S. aureus was distributed unevenly and appeared as cell clumps, apparently as so-called small colony variants (SCV) embedded in the biofilm matrix (see white arrow in Fig 2C). By using differential staining of S. aureus and P. aeruginosa (Fig 3) we have also analyzed the distributions of S. aureus (red-stained) and P. aeruginosa (blue-stained) over the biofilm layers and evaluated their relative fractions in each layer. In agreement with earlier data, S. aureus tended to distribute in the upper layers of the biofilm, while P. aeruginosa dominated in its lower layers (see Fig 3A, C). The fraction of non-viable cells in the mixed biofilm was just slightly higher than in corresponding monoculture biofilms (compare Fig 2 A, B and C and Fig S1), suggesting stability of S. aureus - P. aeruginosa consortium under the conditions used.
In the last decades different approaches to inhibit the biofilm formation by various bacteria were developed [47, 48, 50], appearing nowadays more successful in prevention of S. aureus biofilm formation [53, 54, 56]. Therefore we simulated the S. aureus - P. aeruginosa mixed biofilm formation under the conditions of biofilm-preventing treatment. For that, bacteria were cultivated in the presence of a derivative of 2(5H)-furanone denoted as F105, identified recently study as an efficient inhibitor of growth and biofilm formation by S. aureus [65, 67], while exhibiting no significant effect against P. aeruginosa (Table 1).
When S. aureus was grown in the presence of 2.5 μg/ml of F105, no biofilm was formed, while most of the cells remained viable (Fig 2 D). As expected, no significant effect of F105 on cell viability of P. aeruginosa could be observed (Fig 2 E, Table 1). Moreover, the biofilm formation was slightly increased as determined by crystal violet staining (Fig S2) and CLSM (compare Fig 2 B and E). Therefore, we next used F105 to obtain a model of S. aureus -P. aeruginosa mixed biofilm where the biofilm formation by S. aureus is repressed and the matrix is produced predominantly by P. aeruginosa.
When S. aureus and P. aeruginosa were grown together in the presence of F105, S. aureus clumps were also observed, similarly to the control (compare Fig 2 C and F), suggesting that S. aureus cells are able to form clusters inside the biofilm of P. aeruginosa, despite of its antagonistic pressure (see white arrows on Fig 2F). In marked contrast to the control, the cells were observed only in the bottom layers of the biofilm (compare Fig 3 C and D) suggesting that under conditions of anti-biofilm pressure S. aureus is apparently able to hide in the biofilm formed by P. aeruginosa and survive there.
The microscopic data were further validated by direct CFU counting in the biofilm; by using mannitol salt agar plates and cetrimide agar plates the bacterial species were differentiated and their CFUs were counted separately (Fig S3). In the presence of F105 the amount of adherent viable S. aureus cells decreased by 6 orders of magnitude in monoculture, suggesting complete inhibition of the biofilm formation, while no significant differences in CFUs of P. aeruginosa could be observed (Fig S3). In a mixed biofilm, the S. aureus to P. aeruginosa ratio remained unchanged in the control, while the fraction of viable S. aureus cells decreased slightly in the presence of F105, this way confirming CLSM data and supporting the hypothesis that S. aureus is able to survive in the P. aeruginosa biofilm when its own biofilm formation is repressed.
Atomic force microscopy
The atomic force microscopy of both monocultures and mixed biofilms of S. aureus - P. aeruginosa confirmed the CLSM data. Thus, in control wells the biofilms of monocultures of both strains formed a typical confluent multilayer biofilm (Fig 4, A, B), in mixed biofilm S. aureus was prevalently distributed in the upper layers (Fig 4 C). Interestingly, the adhesion force of the mixed biofilm was 3-fold lower compared to S. aureus monoculture biofilm and 2-fold lower compared to P. aeruginosa monoculture biofilm (Table 2), suggesting more irregular structure of the mixed biofilm [39]. When growing with F105, only P. aeruginosa could be observed on the biofilm surface in the mixed culture, suggesting that S. aureus was hidden into the lower biofilm layers. Since the adhesion force of the mixed biofilm in the presence of F105 was similar to that one in the monoculture P. aeruginosa (Table 2, Fig 7 F), we assumed that the biofilm matrix under these conditions was presumably formed by P. aeruginosa.
S. aureus and P. aeruginosa susceptibility to antibiotics in mixed biofilms
Our data suggest that S. aureus under anti-biofilm treatment conditions is able to form cell clumps in the biofilm of P. aeruginosa, thereby apparently changing their tolerance to antimicrobials. To further verify this assumption, the effect of various conventional antibiotics on preformed mono-and polymicrobial biofilms was studied. The 48-h old monoculture and mixed biofilms were prepared in 24-well adhesive plates in either absence or presence of F105 to repress the biofilm formation by S. aureus itself. Then the biofilms were washed with sterile 0.9% NaCl and wells were loaded with fresh broth supplemented with antibiotics at wide range of final concentrations to fill the range of their 1-16 fold MBCs (see Table 1 for MBC values). After 24h incubation the amount of CFUs of both S. aureus and P. aeruginosa in the biofilm was determined by the drop plate assay and the distribution of cells in the mixed biofilm was assessed by CLSM.
First, the biofilm-eradicating activity was investigated for the antibiotics conventionally used for S. aureus treatment but typically inefficient against P. aeruginosa including vancomycin, tetracycline, ampicillin and ceftriaxone (Fig 5, S4). In monoculture, vancomycin reduced the amount of viable S. aureus cells in the biofilm by 3 orders of magnitude at 16×MBC (Fig 5 A). Expectedly, when S. aureus cells were grown in the presence of F105 (2.5 μg/ml) and therefore no biofilm could be formed, bacteria were found completely dead after 24-h exposition to the antibiotic at 1-2×MBC (Fig 5 C). Irrespective of either presence or absence of F105 P. aeruginosa remained resistant to the antibiotic (Fig 5 B, D).
In a mixed culture, irrespective of the S. aureus biofilm formation repression by F105, viable S. aureus cells were identified within the biofilm and the efficiency of antibiotics reduced drastically (Fig 5 A, C, compare reds and violets). Statistical significance of this discrepancy was confirmed by the Kruskal-Wallis statistical test at p < 0.05.
For a deeper understanding of localization and viability of bacteria in mixed biofilms under vancomycin treatment also CLSM analysis was performed. In the presence of F105 no biofilm of S. aureus could be observed resulting in significant decrease of viable cells fraction after vancomycin treatment, in contrast to the biofilm-embedded cells (compare Fig 5 E and H). In the mixed biofilm coccal cell clusters were formed in the biofilm matrix similarly to the control (compare Fig 2 C, F and 5, G, J), suggesting that staphylococci are able to escape the antimicrobials and survive by embedding itself into the polymicrobial biofilm.
The distribution of bacteria in the mixed biofilm layers was also assessed by differential staining of S. aureus and P. aeruginosa by the ViaGram™ Red+ (Fig 6). In marked contrast to the control where S. aureus was mostly located in the top layers of the biofilm, under vancomycin treatment most of the cells appeared in the lower and middle layers of the biofilm (compare Fig 3 and Fig 6) suggesting that vancomycin-resistant P. aeruginosa cells in the upper layers of the biofilm apparently prevented the penetration of the antibiotic into the matrix this way reducing the susceptibility of S. aureus to antibiotics considerably. Of note, S. aureus cells remained presumably viable in bottom layers apparently because of protection by P. aeruginosa cells (Fig S5).
Similarly to vancomycin, treatment by ampicillin, tetracycline and ceftriaxone was almost inefficient against biofilm-embedded S. aureus, while under conditions of biofilm formation repression by F105, the 1-2×MBC of antimicrobials led to the complete death of cells in 24 h (Fig S4). Again, in the mixed culture, despite of the S. aureus biofilm formation repression, viable S. aureus cells could be identified within the biofilm. CLSM analysis indicated considerable redistribution of S. aureus from upper to the bottom layers of the biofilm (Fig S6, cells distribution patterns) leading to reduced antibiotic efficacy. Under double treatment by F105 and antimicrobials, the prevalence of P. aeruginosa in the biofilm was observed in agreement with the CFU count data (Fig S4 and S6). These data suggest that under anti-biofilm or antimicrobial treatment conditions S. aureus changes its preferred topical localizations by hiding in the lower layers of mixed biofilm formed by another bacterium like P. aeruginosa insensitive to most antimicrobials thereby increasing its resistance to the treatment.
Next, we investigated the effect of broad-spectrum antimicrobials such as ciprofloxacin, amikacin and gentamycin which are active against both S. aureus and P. aeruginosa (see Table 1). In contrast to the previous group of antimicrobials, in monoculture high concentrations of ciprofloxacin efficiently eradicated even the biofilm-embedded P. aeruginosa (Fig 7 B). Interestingly, when the mixed biofilm was treated, nearly 10-fold lower concentration of antimicrobial was required to obtain similar reduction of P. aeruginosa CFUs in the biofilm. Moreover, in the mixed biofilm complete death of both P. aeruginosa and S. aureus could be observed at 8×MBC of ciprofloxacin, in marked contrast to monocultures. Similarly, 1-2×MBC of aminoglycosides (amikacin or gentamicin) led to the complete death of both P. aeruginosa and S. aureus in mixed biofilm (Fig 8) while reducing their CFUs in monocultures only by 2-3 orders of magnitude at 8×MBCs.
In the presence of F105, just 1×MBC of any tested antimicrobial was already sufficient for the complete eradication of S. aureus biofilm (see Fig 7 C and Fig 8 C), similarly to the previous group of antibiotics such as vancomycin, tetracycline, ampicillin and ceftriaxone (see Fig 5 and Fig S4). The presence of F105 did not affect the susceptibility of monoculture P. aeruginosa biofilm to antibiotics. In contrast, in mixed biofilms inhibition of S. aureus by F105 restored the susceptibility of P. aeruginosa back to the monoculture level, suppressing the observed high efficiency of antimicrobials against this bacterium in the mixed biofilm (compare Fig 7 B and D, Fig 8 B and D). In contrast to S. aureus-specific antibiotics, the efficiency of ciprofloxacin and aminoglycosides against S. aureus in mixed biofilm in the presence of F105 was similar to the monoculture level.
The CLSM analysis of S. aureus and P. aeruginosa monoculture and mixed biofilms treated with Ciprofloxacin confirmed the CFUs counting data. In particular, while 8×MBC did not affect either S. aureus or P. aeruginosa cells in monoculture biofilms (Fig 7E, F), in the mixed biofilm P. aeruginosa was identified as non-viable, although S. aureus remained partially alive (Fig 7 G). In marked contrast, repression of the S. aureus biofilm production by F105 led to a reversal with most P. aeruginosa cells green-stained while S. aureus identified as non-viable in mixed culture (Fig 7 J).
The distribution of bacteria in the mixed biofilm layers under treatment with ciprofloxacin was also assessed by differential staining of S. aureus and P. aeruginosa using ViaGram™ Red+ (Fig 9). In contrast to vancomycin treatment, here S. aureus dominated in the upper layers of the mixed biofilm (compare Fig 6 and 9 A and C) and remain alive, while P. aeruginosa were presumably dead (See Fig S5, S7, S8) suggesting no reversal protection of P. aeruginosa by S. aureus biofilm. On the other hand, double treatment by ciprofloxacin combined with F105 resulted in hiding of S. aureus in the bottom layers of the biofilm and increased resistance of P. aeruginosa. Treatment by amikacin and gentamycin led to considerably different distributions of bacteria over the biofilm layers with the prevalence of S. aureus in the bottom layers irrespective of its biofilm repression by F105 (Fig S8, cells distribution patterns) but qualitatively similar bacterial survival patterns (see Fig S6). Moreover, under single antibiotic treatment P. aeruginosa were presumably dead, while S. aureus remained viable (Fig S8). In the presence of F105 P. aeruginosa remained alive and much less S. aureus cells could be observed in the biofilm, as almost all of them were identified as non-viable.
Taken together these data suggest complex interspecies interactions between S. aureus and P. aeruginosa in mixed biofilm under treatment by antimicrobials with different specificity.
Intervention of P. aeruginosa into S. aureus biofilm and vice versa as a possible way to enhance antimicrobial susceptibility
Our results indicate that under appropriate conditions both S. aureus and P. aeruginosa due to their antagonistic interactions appear more susceptible to broad-spectrum antimicrobials in polymicrobial biofilms, compared to their monoculture counterparts. Based on these data, we have suggested that also the susceptibility of monoculture biofilms could be increased by deliberate intervention of P. aeruginosa into preformed S. aureus biofilm, and vice versa.
To verify the efficacy of this approach, P. aeruginosa suspension (106 CFU/mL) was added to the 24 h-old S. aureus biofilm and bacteria were incubated for the next 24 h. Then the biofilm was washed by sterile saline and fresh broth containing different antimicrobials was added into the wells. After 24 h the number of P. aeruginosa and S. aureus CFUs was counted by using differential media.
The introduction of P. aeruginosa into S. aureus biofilm did not change the efficacy of any antibiotic against P. aeruginosa itself (Fig 10, lane II). In contrast, 1×MBC of ciprofloxacin led to the reduction of viable S. aureus in biofilm by 3 orders of magnitude, while in the monoculture 4-8×MBC was required to achieve the same effect (Fig 10, lane I, compare reds and violets). Amikacin and gentamycin, being almost inefficient against S. aureus monoculture biofilm up to 8×MBC, were able to decrease the S. aureus CFUs in biofilm by 3 orders of magnitude already at 1-2×MBC after introduction of P. aeruginosa with the most pronounced effect observed for gentamycin.
In the reverse experiment, when S. aureus was added to the P. aeruginosa biofilm, a remarkable increase of ciprofloxacin efficacy against P. aeruginosa could be observed (Fig 10, lane IV, compare blues and violets), while the susceptibility of S. aureus itself did not change significantly. The efficacy of aminoglycosides has increased only against S. aureus, while not against P. aeruginosa.
Discussion
Biofilm formation represents an important virulence factor of many bacteria, as the extracellular matrix drastically reduces their susceptibility to antimicrobials resulting in up to 1000-fold higher tolerance to antibiotics of biofilm-embedded cells compared to their planktonic forms [14, 68, 69]. In contrast, polymicrobial communities are often characterized by concurrent interspecies interactions that likely overwhelm the potential benefits from biofilm protection. Here we have shown that the antagonistic interactions between S. aureus and P. aeruginosa, the most common pathogenic agents causing various nosocomial infections [7-9], drastically affect their susceptibility to antibiotics making them significantly more or less vulnerable to treatment than in monoculture biofilms depending on both conditions and chosen antimicrobial agents.
Despite of the antagonistic relationship between S. aureus and P. aeruginosa described in multiple studies [22, 23], these bacteria can be found in close association in acute and chronic wounds being embedded into mixed biofilms [8, 24-28]. Our in vitro data show that the inoculation of S. aureus to the mature P. aeruginosa biofilm or vice versa leads to the formation of mixed biofilm, although with the prevalence of the first biofilm former (Fig 1). The co-cultivation of both bacteria results in the formation of a more rigid biofilm, where S. aureus is located mainly in the upper layers, while P. aeruginosa can be found mostly in the lower layers of the biofilm (Fig 3), in agreement with earlier data [15, 28-31].
Next, we investigated the effect of two groups of antimicrobials on bacterial viability in mixed biofilms. The first group contained vancomycin, tetracycline, ampicillin and ceftriaxone that are known to exhibit specific activity against S. aureus while leaving P. aeruginosa nearly unaffected. The second group included broad-spectrum antibiotics such as Ciprofloxacin, Gentamicin and Amikacin that exhibited comparable MBC values against both studied bacteria (see Table 1). Additionally, we also simulated the biofilm-preventing treatment with earlier described compound F105, specifically affecting only S. aureus biofilm formation[65]. In control experiments with S. aureus monoculture biofilms, none of the antimicrobials exhibited any bactericidal effect at their 8-16×MBCs, while 1×MBC was already sufficient for the complete eradication of all adherent cells under biofilm repression conditions with F105 (compare Figs 5, 7, 8, S4, reds on panels A and C). In addition, ciprofloxacin, gentamicin and amikacin at 8×MBCs significantly reduced the number of CFUs of biofilm-embedded P. aeruginosa (Figs 5, 7, 8, S4, blues on panels B and D).
In mixed biofilms, treatment with antimicrobials active specifically against S. aureus such as vancomycin, tetracycline, ampicillin and ceftriaxone as well as by biofilm repressing agent F105, S. aureus could successfully escape from the treatment by re-localization to the middle and lower layers of the biofilm. Irrespective of the S. aureus biofilm formation repression, S. aureus cells remained viable under these conditions being embedded into the matrix of P. aeruginosa biofilm and were insensitive to antimicrobials (see Figs 6, S6) suggesting that staphylococci are able to escape the antimicrobials by embedding into the biofilm matrix of P. aeruginosa and survive there, despite of their antagonistic interactions. Notably, in mixed biofilms S. aureus formed cell clumps in the biofilm matrix (compare Fig 2 C, F and 5, C, F) presumably in the form of small colonies.
Remarkably, when the S. aureus – P. aeruginosa mixed biofilms were treated with any of the broad-spectrum antimicrobials such as ciprofloxacin, gentamicin or amikacin, nearly 10–fold lower concentrations were sufficient to achieve the same reduction in the CFUs number of both bacteria, in comparison with monoculture treatment (Figs 7 and 8, compare violets with reds or blues on panels A and B). This effect was more pronounced for aminoglycosides, which at already 1–2×MBC led to the complete eradication of the mixed biofilm, while in monocultures 8×MBC was required to reduce the number of CFUs by 3–5 orders of magnitude (Fig 8). Moreover, tetracycline and ceftriaxone, while being inefficient against P. aeruginosa, at high concentrations significantly reduced the CFUs of this bacterium in the mixed biofilms (Fig S4).
Interestingly, under repression of the S. aureus biofilm formation by F105, the efficiency of antimicrobials against S. aureus did not change significantly, while the sensitivity of P. aeruginosa was restored to the level characteristic for its monoculture biofilm (Figs 7 and 8, compare violets with reds or blues on panels C and D). This effect could be attributed to the redistribution of the S. aureus cells to the bottom levels of the mixed biofilm and significant reduction of their fraction (see Figs 9, cells distributions). On the other hand, the observed reinstatement of the P. aeruginosa sensitivity to antimicrobials could originate from the repression of the antagonistic factors production by S. aureus due to complex changes in its cell metabolism in the presence of F105. Nevertheless, the molecular basis of these complex interbacterial interactions that under certain conditions lead to a clear reversal in the antimicrobials susceptibility requires further investigations.
Taken together, our data clearly indicate that efficient treatment of biofilm-associated mixed infections requires antimicrobials which would be active against dominant pathogens. As we have shown for the S. aureus and P. aeruginosa mixed culture model, in this case the interbacterial antagonism under certain conditions assists antimicrobial treatment. In contrast, treatment by antibiotics with different efficacy against various consortia members leads to the survival of sensitive cells in the matrix formed by the resistant ones.
Finally, we have shown that S. aureus and P. aeruginosa are able to penetrate into each other’s mature biofilms (see Fig 1 A and B) and by this intervention significantly affect the susceptibility of the mixed biofilm to antimicrobials (Fig 10). When P. aeruginosa was introduced into S. aureus biofilm, all antimicrobials reduced the amount of CFUs of both bacteria in the biofilm by 3 orders of magnitude at 1–2×MBC with more pronounced effect observed for gentamicin. In the reverse experiment, the inoculation of S. aureus to the mature P. aeruginosa biofilm significantly increased the efficacy of ciprofloxacin against P. aeruginosa.
From a broader perspective, we believe that artificial intervention of antagonistic bacteria into already preformed monoculture biofilms could be used to enhance their antimicrobial treatment efficacy. We suggest that this approach has a strong potential of further development towards innovative treatment of biofilm-associated infections such as introduction of the skin residential microflora to the wounds and ulcers infected with nosocomial pathogens to speed up their microbial decontamination. While in this work we demonstrated the synergy of interbacterial antagonism with antimicrobials using the well-studied S. aureus - P. aeruginosa model system, we believe that many other bacteria of normal body microflora are available to antagonize with nosocomial pathogens and thus can be used for the enhancement of microbial infections treatment by using microbial transplantation.
Materials and methods
Derivate of 2(5H)-furanone designed as F105 (3-Chloro-5(S)-[(1R,2S,5R)-2-isopropyl-5-methylcyclohexyloxy]-4-[4-methylphenylsulfonyl]-2(5H)-furanone) was described previously [67] and synthesized at the department of Organic Chemistry, A.M. Butlerov Chemical Institute, Kazan Federal University.
Bacterial strains and growth conditions
Staphylococcus aureus subsp. aureus (ATCC® 29213™) and Pseudomonas aeruginosa (ATCC® 27853™) were used in this assay. The bacterial strains were stored in 10 % (V/V) glycerol stocks at –80 °C and freshly streaked on blood agar plates (BD Diagnostics) followed by their overnight growth at 35°C before use. Fresh colony material was used to adjust an optical density to 0.5 McFarland (equivalent to 108 cells/mL) in 0.9 % NaCl solution that was used as a working suspension. For the biofilm assay the previously developed BM broth (glucose 5g, peptone 7g, MgSO4× 7H2O 2.0g and CaCl2× 2H2O 0.05g in 1.0 liter tap water) [49, 63, 70] where both S. aureus and P. aeruginosa formed rigid biofilms in 2 days was used. The mannitol salt agar (peptones 10g, meat extract 1g, NaCl 75g, D-mannitol 10g, agar-agar 12g in 1.0 liter tap water, Oxoid) and cetrimide agar (Sigma) were used to distinguish S. aureus and P. aeruginosa, respectively, in mixed cultures. Bacteria were grown under static conditions at 35°C for 24–72 hours as indicated.
Biofilm assays
Biofilm formation was assessed in 24-well polystirol plates (Eppendorf) by staining with crystal violet as described earlier in [71] with modifications. Bacteria with an initial density of 3×107 CFU/ml were seeded in 2 ml BM at 37°? and cultivated for 48 h under static conditions. Then the culture liquid was removed and the plates were washed once with phosphate-buffered saline (PBS) pH=7.4 and dried for 20 min. Then, 1 ml of a 0.5% crystal violet solution (Sigma-Aldrich) in 96% ethanol was added per well followed by incubation for 20 min. The unbounded dye was washed off with PBS. The bound dye was eluted in 1 ml of 96% ethanol, and the absorbance at 570 nm was measured on a Tecan Infinite 200 Pro microplate reader (Switzerland). Cell-free wells subjected to all staining manipulations were used as control.
The biofilms were additionally analyzed by confocal laser scanning microscopy (CLSM) on Carl Zeiss LSM 780 confocal microscope. Both mono-and mixed cultures of S. aureus and P. aeruginosa were grown on cell imaging cover slips (Eppendorf) under static conditions for 48 h in BM broth. Next one-half of the medium was replaced by the fresh one containing antimicrobials at final concentrations as indicated and cultivation was continued for the next 24 h. The samples were then stained for 5 min with the SYTO® 9 (ThermoFisher Scientific) at final concentration of 0.02 μg/ml (green fluorescence) and propidium iodide (Sigma) at final concentration of 3 μg/ml (red fluorescence) to differentiate between viable and non-viable bacteria. To differentiate between gram-positive and gram-negative bacterial species ViaGram™ Red+ (ThermoFisher Scientific) was used. The microscopic images were obtained with a 1-μm Z-stacks.
Evaluation of antibacterial activity
The minimum inhibitory concentration (MIC) of antimicrobials was determined by the broth microdilution method in 96-well microtiter plates (Eppendorf) according to the recommendation of the European Committee for Antimicrobial Susceptibility Testing (EUCAST) rules for antimicrobial susceptibility testing [72]. Briefly, the 108 cells/mL bacterial suspension was subsequently diluted 1:300 with BM broth supplemented with various concentrations of antimicrobials in microwell plates to obtain a 3×105 cells/mL suspension. The concentrations of antimicrobials ranged from 0.25 to 512 mg/L. Besides the usual double dilutions, additional concentrations were included in between. The cultures were next incubated at 35°C for 24 h. The MIC was determined as the lowest concentration of antimicrobials for which no visible bacterial growth could be observed after 24 h incubation.
To determine the MBC of antimicrobials the CFU/mL were further evaluated in the culture liquid from those wells without visible growth. 10 μl of the culture liquid from the wells with no visible growth were inoculated into 3ml of LB broth followed by cultivation for 24h. The MBC was determined as the lowest concentration of compound for which no visible bacterial growth could be observed according to the EUCAST of the European Society of Clinical Microbiology and Infectious Diseases (ESCMID) [73].
Drop plate assay
To evaluate the viability of both detached and planktonic cells, a series of 10-fold dilutions of liquid culture from each well were prepared in 3 technical repeats and dropped by 5 μl onto LB agar plates. CFUs were counted from the two last drops typically containing 5-15 colonies and further averaged. To evaluate the viability of the biofilm-embedded cells, the wells were washed twice with 0.9% NaCl in order to remove the non-adherent cells. The biofilms were also suspended in 0.9% NaCl by scratching the well bottoms with subsequent treatment in an ultrasonic bath for 2 min to facilitate the disintegration of bacterial clumps [63]. Viable cells were counted by the drop plate method as described above.
Statistical analysis
Experiments were carried out in six biological repeats with newly prepated cultures and medium in each of them. The fraction of non-viable cells in microscopic images was estimated as the relative fraction of the red cells among all cells in the combined images obtained by overlaying of the green and the red fluorescence microphotographs (10 images per each sample) by using BioFilmAnalyzer software[74]. The statistical significance of the discrepancy between monoculture and mixed biofilms treatment efficacy was determined using the Kruskal-Wallis statistical test with significance threshold at p < 0.05.