OUP user menu

Synergistic phage-antibiotic combinations for the control of Escherichia coli biofilms in vitro

Elizabeth M. Ryan, Mahmoud Y. Alkawareek, Ryan F. Donnelly, Brendan F. Gilmore
DOI: http://dx.doi.org/10.1111/j.1574-695X.2012.00977.x 395-398 First published online: 1 July 2012


The potential application of phage therapy for the control of bacterial biofilms has received increasing attention as resistance to conventional antibiotic agents continues to increase. The present study identifies antimicrobial synergy between bacteriophage T4 and a conventional antibiotic, cefotaxime, via standard plaque assay and, importantly, in the in vitro eradication of biofilms of the T4 host strain Escherichia coli 11303. Phage-antibiotic synergy (PAS) is defined as the phenomenon whereby sub-lethal concentrations of certain antibiotics can substantially stimulate the host bacteria's production of virulent phage. Increasing sub-lethal concentrations of cefotaxime resulted in an observed increase in T4 plaque size and T4 concentration. The application of PAS to the T4 one-step growth curve also resulted in an increased burst size and reduced latent period. Combinations of T4 bacteriophage and cefotaxime significantly enhanced the eradication of bacterial biofilms when compared to treatment with cefotaxime alone. The addition of medium (104PFUmL−1) and high (107PFUmL−1) phage titres reduced the minimum biofilm eradication concentration value of cefotaxime against E.coli ATCC 11303 biofilms from 256 to 128 and 32µgmL−1, respectively. Although further investigation is needed to confirm PAS, this study demonstrates, for the first time, that synergy between bacteriophage and conventional antibiotics can significantly improve biofilm control in vitro.

  • biofilm
  • biofilm eradication
  • Escherichia coli
  • phage therapy
  • cefotaxime
  • phage-antibiotic synergy
  • T4 bacteriophage

The renewed interest in bacteriophage therapy as a potential adjunct to conventional antibiotics has led to the evaluation of bacteriophages in the control of bacterial biofilm formation (Abedon, 2006; Hanlon, 2007; Donlan, 2009; Ryan, 2011). Several groups have reported significant biofilm reductions (ranging from 1 log to 6 log), depending upon the constituents of the biofilm, biofilm age, phage selection and length of challenge (Corbin et al., 2001; Sharma et al., 2005). These studies highlight, in part, the potential challenges in using bacteriophages in biofilm control: the presence of extracellular polysaccharide (EPS) shields the phage binding sites on the bacterial surface and retards the penetration of virus into the biofilm matrix; the heterogeneous microenvironment of the biofilm results in bacterial populations exhibiting reduced metabolic activity or dormancy, robbing bacteriophages of the metabolic capacity necessary for infection and replication to occur (Hanlon, 2007). Thus, complete biofilm eradication by phage alone cannot be expected. Several studies have, however, demonstrated the utility of bacteriophages in the control of bacterial biofilms at biomaterial surfaces (Curtin & Donlan, 2006; Carson et al., 2010).

Phage-antibiotic synergy (PAS) describes the phenomenon whereby sub-lethal concentrations of certain antibiotics can substantially stimulate the host bacteria's production of virulent phage. PAS has been previously described in a number of species including Escherichia coli (phage ФMFP) (Comeau et al., 2007) and Klebsiella pneumonia (Bedi et al., 2009) but PAS in biofilm eradication has not been reported in the literature so far. The aim of this study was to assess the application of PAS in the eradication of E. coli ATCC 11303 biofilms, in vitro, using a combination of varying bacteriophage T4 titres and cefotaxime concentrations, compared to antibiotic or phage therapy alone. The effect of PAS on the lytic cycle of T4 phage (burst size and latent period) was also investigated alongside the ability of PAS to reduce the effective minimum biofilm eradication concentration (MBEC) of cefotaxime against E. coli biofilms.

Commercially available T4 bacteriophage ATTC 11303-B4 (LGC Standards, Middlesex, UK) was used in this study and propagated according to manufacturer's instructions using host strain E. coli ATCC 11303 (LGC Standards). Antibiotic stocks were prepared using cefotaxime sodium salt, C7912-1G (Sigma-Aldrich, Dorset, UK) and stored in 1-mL aliquots at −20 °C until required. MICs were determined according to CLSI guidelines (Andrews, 2001) over a range of 1–0.0039 µg mL−1 using the agar dilution method.

The potential synergistic effects of cefotaxime and T4 phage were assessed by a method adapted from Comeau et al. (2007). Following the determination of cefotaxime MIC for E. coli ATCC 11303, plaque size and phage concentration were determined in the presence of cefotaxime at a concentration range of 0–125 µg mL−1. Crude T4 phage stocks of 7.5 × 104 PFU mL−1 were used in these experiments. Varying concentrations of cefotaxime were prepared in molten 0.6% LB agar and 3 mL of antibiotic-containing molten agar plus 100 µL of an overnight culture of E. coli ATCC 11303, and 100 µL of T4 phage stock were mixed, overlaid on a 1.2% LB agar plate and incubated at 37 °C for 18–24 h. This method used for each antibiotic concentration and samples were plated in triplicate. Plaque diameters (mm) were measured using digital callipers. Following plaque size analysis, the agar overlays from each plate were removed using a sterile cell scraper and added to 10-mL LB broth. The contents were vortexed thoroughly and plaque assays were performed to assess whether antibiotic concentration had an effect on the concentration of phages produced. The mean value from five replicates was taken.

The effect of PAS on the burst size of T4 phage was examined using a standardized method from Weld et al. (2004) with some slight modifications. A 1-mL volume of a mid-log growth E. coli culture (109 CFU mL−1) was added to 1 mL of a 1 × 109 PFU mL−1 stock of T4 phage in a sterile bijoux bottle and incubated for 1 min to allow attachment of the phages to the bacterial cells. Following incubation, the phage/bacteria solution was diluted such that the concentration of bacteria was approximately 10 cells mL−1. This concentration was separated into three aliquots at 2 mL per aliquot in sterile bijoux bottles. No antibiotic was added to one container, 100 µL of a 0.0039 µg mL−1 cefotaxime stock was added to the second container and 100 µL of a 0.0156 µg mL−1 cefotaxime stock added to the third. The resulting dilution factor of 21 therefore gave final cefotaxime concentrations of 0, 1.86 × 10−4 and 7.43 × 10−4 µg mL−1. These final concentrations were again incubated at 37 °C for 3 min to begin the growth cycle of T4 phage. An aliquot of 50 µL of bacteria/phage or bacteria/phage/antibiotic solution was dispensed into 50 sterile tubes. This was completed within 20 min for each sample as literature sources state that the lytic life cycle of T4 phage is approximately 25 min (Abedon, 1992; Weld et al., 2004). Overlays were conducted as described with 3 mL of 0.6% LB agar plus 0.5% NaCl, plus 60 µL of E. coli 11303 (overnight culture) added to each container, vortexed and overlaid onto LB agar. Fifty plates were overlaid for each of the three samples and plates were incubated at 37 °C for 24 h. The average burst size for each sample was calculated by counting the number of plaques formed on each plate and dividing the number of plaques by the number of plates with plaques.

The effect of PAS on the latent period of T4 phage was examined using a standardized one-step growth curve to analyse free phages. One hour prior to commencement of the experiment, the overnight culture of E. coli (100 mL) was added to a sterile 2-L flask containing 900-mL LB broth, 10-mL MgSO4 and 10-mL 1 M CaCl2, which had been prewarmed at 37 °C. This was carried out in triplicate. To start, 1 mL of the culture was added to 1010 PFU of T4 stock. This gave a multiplicity of infection (MoI) of 10. The bacteria-phage suspension was incubated for 1 min at 37 °C to allow attachment of the phages to the bacterial cell. The cells were then centrifuged at 1730 g for 3 min on a Sigma 3-16P centrifuge, and the supernatant was removed. The phage-bacterium pellet was resuspended in 1-mL resuspension buffer and cefotaxime concentrations of 0.0156 and 0.0625 µg mL−1 were added. This sample was added back into the exponential phase culture. This step synchronized the infection of the cells, as the phages that had attached to the bacterial cells within the 1-min incubation period are at the same attachment point in the lytic cycle. This is time zero in the experiment. Samples were taken at 6-min intervals from the bacteriophage-infected culture and plaque assays were carried out in the usual way. The mean values from at least 10 replicates were obtained.

MBEC values were determined for cefotaxime against E. coli ATCC 11303 with and without the addition of T4 phage using the Calgary Biofilm Device as described previously by Carson et al. (2010). Biofilms of E. coli ATCC 11303 were cultivated for 24 h and challenged with cefotaxime over a range of concentrations of 2–1024 µg mL−1 for a further 24 h. The same experiment was repeated with the addition of low (1 × 102 PFU mL−1), medium (1 × 104 PFU mL−1) and high (1 × 107 PFU mL−1) phage titre to determine whether combination of phage-cefotaxime could reduce the previously determined MBEC of cefotaxime alone. The mean value of three replicates was taken in each case.

PAS was initially assessed by measurement of plaque diameter on a bacterial lawn upon the addition of doubling concentrations of cefotaxime over a range of 0–0.125 µg mL−1. The data in Table 1 illustrate that phage-cefotaxime combination has a profound effect on bacteriophage T4 plaque size. Increasing the concentration of cefotaxime substantially increases the diameter of T4 plaques. Once a threshold concentration of 0.125 µg mL−1 of the antibiotic was used, no plaques could be detected as this concentration of cefotaxime was sufficient to inhibit bacterial lawn formation. The first significant rise in plaque size occurred at the MIC concentration (0.0156 µg mL−1). The optimum PAS antibiotic concentration is the highest possible to produce the maximum increase in plaque size but not sufficiently high to inhibit bacterial lawn formation (Santos et al., 2009); thus, the optimal cefotaxime concentration for effective PAS was deduced to be 0.0625 µg mL−1. Plaque size is generally accepted to be proportional to the efficiency of adsorption, the burst size of the phage and diffusion of phages in the medium and is inversely proportional to the latent period (Santos et al., 2009). As expected, PAS also increased the burst size of T4 phage with increasing cefotaxime concentration. The addition of 0.000186 and 0.00743 µg mL−1 cefotaxime resulted in an increase in the T4 phage burst size from 8 to 80 and 163 PFU mL−1, respectively.

View this table:
Table 1

The effect of cefotaxime concentration on plaque size (mm) and phage concentration (PFU mL−1) of T4 phage. (Each value is the mean of three measurements)

Cefotaxime concentration (µg mL−1)Plaque diameter (mm)Phage titre (PFU mL−1)
0.0156 (MIC)3.55.60E+11

Table 1 also shows that with increasing concentrations of cefotaxime, phage concentration continuously increases to reach a maximum, with more than 5-log increase in phage concentration, at 0.0625 µg mL−1, which was determined to be the optimal synergistic antibiotic concentration. Increase in phage production at sub-lethal concentrations is generally accepted to be owing to the filamentation of bacterial cells (Comeau et al., 2007; Santos et al., 2009). Figure 1 shows a standard one-step growth curve of T4 phage, plus the growth curve in the presence of 0.0625 µg mL−1 cefotaxime. In the standard curve, T4 phage increased in concentration from approximately 5 × 106 PFU mL−1 at 24 min to 1 × 108 PFU mL−1 at 42 min, which follows the expected trend. T4 phage has a reported latent period of 25 min (Doermann, 1952). Generally, T4 E. coli systems exhibit an ‘eclipse period’ of 12 min, during which no infective phage particles can be recovered from infected cells, only the original experimental phages can be detected. The addition of a sub-lethal concentration of cefotaxime increased the initial concentration of phages during this eclipse period, approximately 5 × 107 PFU mL−1 compared to approximately 5 × 106 PFU mL−1 for T4 phage alone after 12 min. This suggests that the rate of adsorption of phage-bacterial cell was increased by the addition of cefotaxime, although further experimental studies must be conducted to support this. The latent period is defined as the period which begins at virion attachment and terminates at lysis; furthermore, the above figure shows that the addition of cefotaxime reduced latent period to 18 min compared to 24 min for T4 E. coli only. Lysis time was also reduced compared to standard T4 E. coli system, with the cefotaxime sample reaching full lysis concentration (5 × 108 PFU mL−1) at 24 min, while the T4 system reached full lysis at 42 min. This may be explained by the increased production of phages and increased speed of phage production owing to bacterial cell filamentation.

Figure 1

Effect of cefotaxime addition on the one-step growth curve of T4 phage.

As expected, the MBEC for cefotaxime against 24-h E. coli biofilm (256 µg mL−1, in agreement with previously published MBEC data (Ceri et al., 1999)) was significantly higher (approximately 10 000 times) than the measured MIC value (0.0156 µg mL−1). As T4 phage has proven to exert a synergistic effect when combined with cefotaxime against E. coli in its planktonic mode of growth, the effect of adding T4 bacteriophage on the antibiofilm activity of cefotaxime, by measuring reduction in MBEC value, was investigated in the presence of low (102 PFU mL−1), medium (104 PFU mL−1) and high (107 PFU mL−1) T4 phage titres. As shown in Table 2, although the addition of low phage concentration to cefotaxime resulted in no change in the MBEC, medium and high phage titres resulted in a reduction in MBEC from 256 µg mL−1 (as previously determined for cefotaxime alone) to 128 and 32 µg mL−1, respectively. On the other hand, none of the phage titres could result in complete biofilm eradication when used without cefotaxime. In fact, even the challenge with the highest phage titre (107 PFU mL−1) resulted in no more than 0.9-log reduction in biofilm surviving cells.

View this table:
Table 2

Effect of T4 phage addition, at three different titres, on MBEC of cefotaxime against Escherchia coli biofilm

T4 phage concentration (PFU mL−1)Cefotaxime MBEC (µg mL−1)

The marked decrease in MBEC value upon the addition of T4 phage, especially in high titre, suggests the involvement of PAS in the complete eradication of E. coli biofilms in vitro. This study proves for the first time that the additive effects of phage and antibiotics can also be useful in treating bacteria within the biofilm matrix and that such combinations used in synergy significantly enhance biofilm eradication. Clinically, phage-antibiotic combinations in synergy could potentially revive the use of some antibiotic agents not normally considered treatment of choice for certain infections by augmenting their activity in vivo.


  • Editor: Thomas Bjarnsholt


View Abstract