Antagonistic interactions amongst bacteriocin-producing enteric bacteria in dual species biofilms

Journal of Applied Microbiology 2002, 93, 345–352

Antagonistic interactions amongst bacteriocin-producing enteric bacteria in dual species biofilms K. Tait and I.W. Sutherland Institute of Cell and Molecular Biology, Kings Buildings, Edinburgh University, Edinburgh, UK 2001 ⁄ 320: received 19 October 2001, revised 15 April 2002 and accepted 25 April 2002

K . T A I T A N D I . W . S U T H E R L A N D . 2002.

Aims: The objective of this study was to investigate the antagonistic interactions between bacteriocin-producing enteric bacteria in dual species biofilms and the interspecies interactions correlated with sensitivity to biocides. Methods and Results: When compared with their single species counterparts, the dual species biofilms formed by bacteriocin-producing strains exhibited a decrease in biofilm size and an increase in sensitivity to the antimicrobial agents hypochlorite, triclosan and benzalkonium chloride. The five dual species biofilms studied all resulted in biofilms containing a mixture of the two strains. This was attributed to the spatial distribution of cells within the biofilm, with each strain forming its own microcolonies. The production of a bacteriocin also gave a strain a competitive advantage when interacting with a bacteriocin-sensitive strain within a biofilm, both in gaining a foothold in a new environment and in preventing the colonization of a potential competitor into a pre-established biofilm. Conclusions: It was concluded that bacteriocins might be used specifically for interacting with competing strains within a biofilm, as opposed to a planktonic, environment. Significance and Impact of the Study: Unlike planktonically grown bacteriocin-producing populations, where one strain will always be out-competed, bacteriocin-producing and bacteriocin-sensitive strains can coexist in biofilm communities, clearly demonstrating major differences between biofilm and planktonic competition. This paper highlights the importance of bacteriocin production in the development of biofilm communities. INTRODUCTION Bacterial biofilms are consistently more resistant to antimicrobial agents when compared with planktonically growing cells and this is a cause of concern in many diverse fields in industry and medicine. The biofilms formed in natural and industrial environments often consist of complex communities of micro-organisms (James et al. 1995). Recent research has indicated that interspecies interactions within biofilms can influence the susceptibility of the biofilm to antimicrobial agents. Bourion and Cerf (1996) demonstrated the importance of an exopolymer-producing strain of Pseudomonas aeruginosa for enhancing the attachment of Listeria innocua. Listeria innocua was protected from disinfectants in the mixed biofilm due to the presence of Ps. aeruginosa and its associated polymers. Skillman et al. (1999) described thicker Correspondence to: Dr K. Tait, Plymouth Marine Laboratory, Prospect Place, Plymouth PL1 3DH, UK.

ª 2002 The Society for Applied Microbiology

biofilm production and enhanced resistance to disinfection by a dual species biofilm of Enterobacter agglomerans and Klebsiella pneumoniae. This was partially attributed to rheological interactions between the extracellular polymers produced, changing the physical properties of the polymers as they interacted. The production of drug-inactivating enzymes may also affect the resistance of a multiple species biofilm to antimicrobial agents. Investigating the interactions between a b-lactamase-producing strain, Moraxella catarrhalis, and a pathogenic strain, Streptococcus pneumoniae, Budhani and Struthers (1998) showed protection of the Strep. pneumoniae by the M. catarrhalis; concentrations of antibiotic that would have ordinarily killed the pneumococcus in monoculture were ineffective. The aim of this study was to determine if the interactions between different species could influence biofilm development and the resistance of the biofilm to antimicrobial agents. Using a two-species system, the interactions between bacteria were compared and correlated with increased or

346 K . T A I T A N D I . W . S U T H E R L A N D

decreased biofilm formation. As the strains used were closely related enteric species, it was not surprising to find that considerable bacteriocin activity occurred. As little is known of the interactions occurring in antagonistic biofilms, this was investigated further. A further aim was to compare the competitive interactions between bacteriocin-producing and bacteriocin-sensitive strains in planktonic and biofilm cultures. The role of bacteriocins in the integration of one strain into a pre-established biofilm and in defending a biofilm from colonization by a potential competitor was also examined.

M A T E R I A LS A N D M E T H O D S Bacterial strains The Ent. agglomerans strains, Ent and 53b, were isolated from biofilms on industrial surfaces (Dr M.V. Jones, Unilever Research Laboratory, Port Sunlight, UK). Enterobacter gergoviae ⁄ 1.15 was isolated from a biofilm developed on a glass surface in a small stream adjacent to the laboratory. These strains were typed using the API 20E identification systems (bioMerieux, Marcy-I’Etoile, France). Other strains used were Ent. cloacae NCTC 5920 and Escherichia coli ATCC 11229. To distinguish microscopically between the strains, a variant of Ent containing a GFP plasmid was also used (Skillman et al. 1998). The culture medium routinely used was yeast extract (YE) medium (Sutherland and Wilkinson 1965). Growth of biofilm material Flask and bead method. The culture system comprised 250-ml Erlenmeyer flasks containing 100 ml YE media and 15 g 4-mm glass beads. Cultures were incubated at 30C with shaking at 80 rev min)1. Viable counts of the cells attached to the beads were estimated by removal of the bead from the flask and gentle washing in three changes of sterile phosphate-buffered saline (PBS). The bead was placed in a sterile microcentrifuge tube containing 1 ml sterile distilled water and vortexed for 1 min to remove the adhered cells. A viable count was made of the suspension of the mechanically removed cells and the number of cfu cm)2 estimated. Yellow box method. The batch culture system comprised a stainless steel stand capable of supporting glass coverslips (22 · 22 mm; VWR International, Lutterworth, UK) and a culture vessel (yellow tip box; Greiner, Labortechnik, Stonehouse, UK) in which the stand was immersed (Hughes et al. 1998). The tip box, holding 250 ml media, was incubated at 30C. The medium was agitated using a magnetic stirrer (120 rev min)1).

Microtitre plate method. Plastic 96-well microtitre plates (VWR International, Lutterworth, UK) were used for biofilm cultivation. Volumes of 100 ll were dispensed into each well and the plates incubated for 72 h at 30C. Culture liquid was then removed and the wells washed with sterile PBS to leave a monolayer of adhered cells. Microscopy Glass coverslips were immersed in 10 mmol l)1 cetyl pyridinium chloride for 5 min and then immersed in 25 lg ml)1 propidium iodide (PI; Sigma, Gillingham, UK). Pre-treatment with the detergent caused cell membrane damage and access of PI to the cell interior. GFP was not masked by the presence of PI in the cell and, therefore, EntGFP cells appeared green (or yellow) and other bacteria appeared red under u.v. illumination. The biofilms were observed under a Polyvar microscope with tungsten bulb attachment (Reichert-Jung Leica Microsystems, Milton Keynes, UK) using a violet-blue excitation filter (395–446 nm). Bacteriocins Testing for the presence of bacteriocin activity. A 250-ml flask containing 100 ml YE medium was inoculated with the strain to be tested and incubated overnight at 30C. The culture was then centrifuged at 10 000 g for 20 min to separate the cells from the supernatant fluid. The supernatant fluid was filter sterilized (0Æ22 lm pore size; Millipore, Walford, UK) and stored at 4C. The activity of the spent media was estimated as follows. Firstly, lawns of a wide selection of bacteria were spotted with 10 ll spent media solution. The plates were incubated overnight at 30C and the lawns examined for any effects produced by addition of the spent media. Secondly, a 250-ml flask containing 90 ml YE medium was inoculated with bacteria. During the mid-logarithmic phase, 10 ml of the filter-sterilized spent media was added to the culture and the O.D.600 nm monitored over a period of 24 h. Cultures were incubated at 30C. A drop in the O.D. of the culture was indicative of cell lysis and possible bacteriocin production. Proteinase activity. The solution to be tested (0Æ3 ml) was added to a test tube with 1Æ7 ml distilled water and 1Æ0 ml azoalbumin solution (5 mg ml)1 azoalbumin (Sigma) in 0Æ1 mol l)1 Tris buffer, pH 7Æ5). A blank contained 2Æ0 ml distilled water and 1Æ0 ml azoalbumin solution. The tests were incubated for 1 h in a water-bath at 30C. To stop the reaction and precipitate unhydrolysed azoalbumin, 2Æ0 ml 8% trichloroacetic acid were added and the mixture centrifuged at 10 000 g. The clear supernatant fluid (2Æ0 ml)

ª 2002 The Society for Applied Microbiology, Journal of Applied Microbiology, 93, 345–352

BACTERIOCIN PRODUCTION IN BIOFILMS

was transferred to a clean tube and 2Æ0 ml 0Æ5 mol l)1 NaOH added to intensify the colour. The absorbancy of the solution was then read at 400 nm. Assays of bacteriocin activity. Overnight cultures of a susceptible strain were centrifuged at 10 000 g, the cells washed in PBS and resuspended in PBS containing 1% glucose. The O.D. of the culture was adjusted to 1Æ0 at 600 nm. Washed cell suspension (0Æ9 ml) and 100 ll of the filter-sterilized spent media sample to be tested were added to a sterile microcentrifuge tube and vortexed. Samples were incubated in a water-bath at 30C for 1 h, microfuged at 10 000 g for 5 min and the absorbency read at 260 nm. The release of nucleic acid provided an ideal method of estimating quantities of smaller bacteriocins. U.v. irradiation of cultures. Planktonic cultures were centrifuged at 10 000 g, the cells washed in PBS and placed in a u.v. cabinet (ca 265 nm) for 5 min. Cells were left to recover for 30 min and bacteriocin activities assayed. Estimation of biofilm-eradicating concentration The appropriate concentration of antimicrobial agent was added to 72 h microtitre plate biofilms and left at room temperature for 5 min. The wells were rinsed with PBS and re-filled with fresh YE medium (100 ll). Any surviving bacteria in the biofilm could re-grow into the liquid phase. After incubation at 30C for 24 h, the absorbency of the microtitre plate cultures at 570 nm was read. The biofilmeradicating concentration (BEC) was taken as the lowest concentration of disinfectant to prevent any regrowth.

RESULTS

347

Investigation of growth-inhibiting agents in spent media of the strains indicated that all strains had the ability to inhibit the growth of a number of strains used in this study. The exception to this was the pairing of 11229 and Ent, where the antagonistic agents produced by the bacteria had no inhibitory effect. Enteric bacteria produce two major groups of bacteriocin: colicins and microcins. Due to their small size, microcins can be easily distinguished. The agents produced by the strains were passed through 10 kDa cut-off filters (Millipore). The filtrates were added to susceptible cultures and monitored for a reduction in O.D.600 nm. From this, 11229, Ent and 5920 were each thought to produce microcins and 1.15 was thought to produce two microcins and also a larger bacteriocin which had protease activity (1 ml of a 24-h culture of 1.15 hydrolysed 52 (± 4) mg protein in 1 h). Strain 53b also produced a larger agent with unknown activity. The smaller bacteriocins of 1.15 were produced during the late exponential (M-4) and stationary (M-8) growth phases. The protease was only active against Ent, M-4 was active against 5920 and M-8 inhibited the growth of strains 11229, Ent, 5920 and 53b. To clarify whether the agents produced by 11229, Ent, 1.15 and 5920 were microcins, the effect of u.v. radiation on bacteriocin production was investigated. Microcins, unlike colicins, are not amplified by agents that elicit the SOS response (Walker 1984). No increase in bacteriocin production was measured, indicating that the smaller agents produced by these strains were likely to be microcins. The agents produced by 11229, 1.15 and 53b were bacteriolytic (i.e. the filtrate released nucleic acid upon addition to susceptible cells), whereas the agents produced by Ent and 5920 were bacteriostatic.

Bacteriocin production

Comparison of competitive interactions in planktonic and biofilm cultures

The growth of Ent and 1.15 decreased in a dual species biofilm when compared with single species biofilms of the strains (Fig. 1). This suggested that antagonism was occurring between the strains and possible bacteriocin activity.

Previous studies have shown that long-term coexistence of bacteriocin-producing and bacteriocin-sensitive strains in liquid cultures cannot be achieved but that one strain will always be out-competed (Riley and Gordon 1999). Results

-2

6

Cfu cm (x10 )

2·5

Fig. 1 Colonization of glass beads by the strains Ent (h) and 1.15 (j) as single and dual species biofilms. Viable counts of biofilm material were estimated at 6, 24 and 72 h. Results are the outcome of four replicate experiments and bars represent S.E.

2 1·5 1 0·5 0 6h

24 h

72 h

6h

Ent

ª 2002 The Society for Applied Microbiology, Journal of Applied Microbiology, 93, 345–352

24 h 1·15

72 h

6h

24 h Dual

72 h

348 K . T A I T A N D I . W . S U T H E R L A N D

in Fig. 1 suggest that the situation may be different for biofilm competition. The competitive interactions between Ent and 1.15 in planktonic and biofilm cultures were, therefore, compared (Fig. 2). Biofilm cultures were grown on glass beads and, to increase the competition between the strains, an inoculum containing a 20 : 1 ratio of cells was used, with Ent being the dominant strain in each case. Results show that there were considerable differences between the interactions amongst bacteriocin-producing strains in planktonic and biofilm cultures. In the planktonic culture, 1.15 was gradually out-competed whereas in the biofilm culture, 1.15 remained established within the

6

biofilm. This suggested that bacteriocin-producing strains could coexist in biofilms. To test this theory further, four more competitive pairs of bacteria were selected, 5920 ⁄ 1.15, 11229 ⁄ 1.15, Ent ⁄ 53b and the pairing of 1.15 and Ent where 1.15 was the dominant strain in the inoculum. The previous pairing of 1.15 and Ent (Fig. 2) contained Ent as the dominant strain in the initial inoculum. Table 1 demonstrates the outcome of competition between these pairs in biofilm and planktonic cultures. In the biofilm cultures, the subordinate strain became established within the biofilm, sometimes even eventually dominating the biofilm as in the case of Ent ⁄ 1.15. As in

(a)

5

-1

9

Cfu ml (x10 )

4

3

2

1

0 -1 5

(b)

4

2 6 Cfu cm- (x10 )

3

2

1

0

-1 0

20

40 Time (h)

60

80

Fig. 2 Comparison of the competitive interactions between (a) planktonic and (b) biofilm cultures. Single species Ent (n) and 1.15 (m) cultures are compared with dual species cultures (Ent (h) and 1.15 (j)). Results are the outcome of four replicate experiments and bars represent S.E.

ª 2002 The Society for Applied Microbiology, Journal of Applied Microbiology, 93, 345–352

BACTERIOCIN PRODUCTION IN BIOFILMS

Table 1 Comparison of the competitive interactions between planktonic and biofilm cultures Strains

Biofilm

Planktonic

Ent ⁄ 1.15 1.15 ⁄ Ent 5920 ⁄ 1.15 11229 ⁄ 1.15 Ent ⁄ 53b

0Æ6 2 5 15 3Æ5

All Ent 125 All 5920 All 11229 50

The initial inoculum was adjusted to a 20 : 1 ratio of cells, with the underlined strain being the dominant strain in each pair. Ratios of each strain in each pair were monitored over a 3-d period and the figures shown are the ratios after 72 h growth.

349

Three systems were selected to examine the role of bacteriocins in colonization: 1.15 ⁄ 5920, 11229 ⁄ Ent and Ent ⁄ 1.15. The agent produced by 5920 is ineffective against 1.15, while both microcins produced by 1.15 are active against 5920. This partnership would allow the role of bacteriocin production in invading and defending a biofilm to be evaluated. Ent and 1.15 each produce a bacteriocin capable of inhibiting the growth of the other strain and 11229 and Ent form a non-competitive biofilm. The degree of success of the strain integrating into a biofilm is shown as the percentage cfu cm)2 of the invading strain within the total biofilm population (Table 2). The results indicated that the bacteriocins produced by 1.15 were beneficial in the integration of the strains into 5920 biofilms. It is also likely that the bacteriocin production was useful in the defence against the invasion of strain 5920 into a pre-established biofilm of 1.15. In comparison, in the non-competitive biofilms of 11229 and Ent, where the bacteriocins produced were ineffective, the invading strain comprised a smaller percentage of the population. Interestingly, 1.15 was also more successful in invading an Ent biofilm and in preventing the colonization of Ent cells. Disinfection of competitive biofilms

Fig. 3 Greyscale micrograph of the microcolony formation by a dual species biofilm of 1.15 (darker microcolonies) and EntGFP (lighter microcolony). Bar ¼ 75 lm

Fig. 2a, the subordinate strain was gradually out-competed in the planktonic cultures. Figure 3 depicts a greyscale micrograph of biofilm microcolony formation by a dual species biofilm of Ent and 1.15. This indicated that the two strains formed distinct colonies. The protease and microcin production by 1.15 in a dual species biofilm with Ent was also monitored and compared with a single species control biofilm (Fig. 4). No differences were observed between the single species biofilm and planktonic cultures. The protease production by 1.15 in single and dual culture was also very similar. However, an early induction of microcin synthesis and increased syntheses in the dual species competitive biofilm were seen. Integration of bacteriocin-producing strains into preestablished biofilms. The production of a bacteriocin could give an organism a competitive advantage when interacting with other microbes, both in gaining a foothold in a new environment and also in preventing the colonization of a potential competitor into a pre-established biofilm.

The resistance of biofilm material to the disinfectants sodium hypochlorite (HPC), benzalkonium chloride (BC) and the biocide triclosan (TR) was estimated. Table 3 shows the BEC of BC, HPC and TR against single and dual species biofilms. When comparing the theoretical and actual data, the results show that the competitive biofilms Ent ⁄ 1.15 and 1.15 ⁄ 5920 were actually less resistant to the antimicrobial agents in each case. When compared with the competitive biofilms, the non-competitive biofilm of 11229 ⁄ Ent was only slightly less resistant to the antimicrobial agents. DISCUSSION Most research into interspecies interactions within biofilms has focused on the beneficial aspects of these relationships. However, not all interactions will be beneficial, antagonistic interactions may also play an important role in the development of microbial communities. The production of antimicrobial compounds, including toxins, bacteriolytic enzymes, bacteriophages, antibiotics and bacteriocins (Riley 1998), seems to be a generic phenomenon for most bacteria. Bacteriocin production occurs across all major groups of the Eubacteria and Archaebacteria. Smarda and Smajs (1998) reported that 35% of E. coli strains appearing in the human intestinal tract are colicinogenic. It has also been estimated that 90% of Ps. aeruginosa from environmental and clinical sources produce R- and F-type pyocins (Riley 1998). Taking this scale of antimicrobial agent production into

ª 2002 The Society for Applied Microbiology, Journal of Applied Microbiology, 93, 345–352

350 K . T A I T A N D I . W . S U T H E R L A N D

0·8

(a)

0·7

Iytic activity

0·6 0·5 0·4 0·3 0·2 0·1 0·0

mg protein hydrolysed h-1

80

(b)

60

40

20

0

0

20

40

60

Time (h) Table 2 Integration of a secondary colonizing strain into a pre-established biofilm % integration of invading strain into biofilm

11229 integration into Ent biofilms Ent integration into 11229 biofilms 1.15 integration into 5920 biofilms 5920 integration into 1.15 biofilms Ent integration into 1.15 biofilms 1.15 integration into Ent biofilms

24 h

72 h

6Æ9 5Æ1 5Æ6 18Æ3 5Æ9 11Æ4

7Æ9 8Æ5 32Æ4 2Æ0 8Æ5 18Æ6

Biofilms of the pre-established strain were inoculated with an invading strain and viable counts estimated at 24 and 72 h. The integration is shown as the percentage cfu cm)2 of the invading strain within the total biofilm population.

80

Fig. 4 Protease and microcin production by strain 1.15. Protease production (a) is compared with microcin production (b) in single species planktonic cultures (m), single species biofilms (d) and dual species biofilms (j). Protease production was measured as mg protein hydrolysed h)1 and microcin production by the lysis of strain 53b cells. Results are the outcome of three replicate experiments and bars represent S.E.

consideration, it is likely that many of the interactions occurring between bacteria in biofilms will be antagonistic in nature. In their review of the ecological role of bacteriocins in bacterial planktonic cultures, Riley and Gordon (1999) stated that bacteriocin-producing and bacteriocin-sensitive strains cannot coexist; one strain will always be outcompeted. The results in this paper indicate that the same may not be true of biofilms. Mathematical modelling has shown that it is possible to achieve a dynamic equilibrium between bacteriocin-producing and bacteriocin-sensitive strains when the cells are growing on a solid surface (Frank 1994; Durrett and Levin 1997). This is thought to be due to the spatial distribution of cells within the biofilm; each strain will form its own microcolonies. An examination of the structure of a competitive biofilm of strains Ent and 1.15

ª 2002 The Society for Applied Microbiology, Journal of Applied Microbiology, 93, 345–352

BACTERIOCIN PRODUCTION IN BIOFILMS

351

Table 3 Biofilm-eradicating concentration of benzalkonium chloride (BC), sodium hypochlorite (HPC) and triclosan (TR) against single and dual species biofilms Strains

BC (mg ml)1)

HPC (%)

TR (lg ml)1)

Single sp. 11229 Ent 1.15 5920

0Æ4 0Æ4 0Æ4 0Æ45

0Æ018 0Æ018 0Æ03 0Æ022

50 45 55 55

Dual sp.

Theoretical

Experiment

Theoretical

Experiment

Theoretical

Experiment

11229 ⁄ Ent Ent ⁄ 1.15 1.15 ⁄ 5920

0Æ4 0Æ4 0Æ425

0Æ35 0Æ3 0Æ35

0Æ018 0Æ027 0Æ026

0Æ018 0Æ02 0Æ022

47Æ32 52Æ5 55

45 30 40

Theoretical results are based on the concentrations of disinfectant required to eradicate the single species biofilm and the ratio of cells in the dual species biofilms. This is compared with experimental data.

revealed that distinct microcolonies were formed within the biofilm (Fig. 3). However, other factors, such as the diffusion of the bacteriocin through the biofilm and the subsequent gradients formed, may also affect the equilibrium between the strains. For example, the proteases of bacteria require a strict pH optimum. In a heterogeneous environment, such as a biofilm, pockets of high acidity or alkalinity are likely to occur, restricting the activity of proteases. The coexistence of bacteriocin-producing and bacteriocinsensitive strains was seen to affect the size and sensitivity of the biofilm to antimicrobial agents. This was thought to be associated with the biofilm bacteria having to deal with a double assault, being affected by both bacteriocin and biocide. It has been suggested that production of a bacteriocin would give an organism a competitive advantage when interacting with other microbes. The production of antagonistic compounds may be beneficial in gaining a foothold in a new environment. Bacteriocins may also prevent the colonization of a pre-established microbial community by a competitive species, a phenomenon known as colonization resistance (Marsh and Bowden, 2000). Table 2 shows that the production of bacteriocins allowed 1.15 to invade preestablished microcin-sensitive 5920 biofilms and also protected the 1.15 biofilm population from a secondary colonizing population of 5920. However, at 24 h, the numbers of 5920 cells that had invaded the 1.15 biofilm were high. This suggests that the trigger for the increased 1.15 microcin production seen in Fig. 4 was due to the close proximity of the competitor cells. 1.15 was also more successful in invading an Ent biofilm and in preventing the colonization of Ent cells. This indicates that the range of antimicrobial agents being produced and their effectiveness in debilitating their potential bacterial competitor must also be taken into consideration. These results suggest that

bacteriocin-producing strains do have a competitive advantage over bacteriocin-sensitive strains within a biofilm; these extremely useful tools may have evolved specifically for interacting with competing strains within biofilm communities. In a comparison of protease production in planktonic and biofilm cells, Evans et al. (1994) reported that the production of protease was higher in biofilm populations, also suggesting that the production of antimicrobial agents may be very important for biofilm cultures. In this study, the single species planktonic and single species biofilm cultures of 1.15 produced similar quantities of protease and microcin (Fig. 4). However, this may be associated with the methodologies used; a batch culture method was used to grow cultures to very high cell densities, densities probably similar to those found in biofilms. ACKNOWLEDGEMENTS K.T. gratefully acknowledges receipt of a MAFF studentship. REFERENCES Bourion, F. and Cerf, O. (1996) Disinfection efficiency against pureculture and mixed-population biofilms of Listeria innocua and Pseudomonas aeruginosa on stainless steel, Teflon and rubber. Sciences des Aliments 16, 151–166. Budhani, R.K. and Struthers, J.K. (1998) Interaction of Streptococcus pneumoniae and Moraxella catarrhalis: Investigation of the indirect pathogenic role of b-lactamase producing Moraxallae by use of a continuous culture biofilm system. Antimicrobial Agents and Chemotherapy 42, 2521–2526. Durrett, R. and Levin, S. (1997) Allelopathy in spatially distributed populations. Journal of Theoretical Biology 185, 165–171.

ª 2002 The Society for Applied Microbiology, Journal of Applied Microbiology, 93, 345–352

352 K . T A I T A N D I . W . S U T H E R L A N D

Evans, E., Brown, M.R.W. and Gilbert, P. (1994) Iron chelator, exopolysaccharide and protease production in Staphylococcus epidermidis: a comparative study of the effects of specific growth rate in biofilm and planktonic culture. Microbiology 140, 153–157. Frank, S.A. (1994) Spatial polymorphism of bacteriocins and other allelopathic traits. Evolutionary Ecology 8, 369–386. Hughes, K.A., Sutherland, I.W., Clark, J. and Jones, M.V. (1998) Bacteriophage and associated polysaccharides depolymerases — novel tools for study of bacterial biofilms. Journal of Applied Microbiology 85, 583–590. James, G.A., Beaudette, L. and Costerton, J.W. (1995) Interspecies bacterial interactions in biofilms. Journal of Industrial Microbiology 15, 257–262. Marsh, P.D. and Bowden, G.H.W. (2000) Microbial community interactions in biofilms. In Community Structure and Co-Operation in Biofilms Society for General Microbiology Symposium no. 59. pp. 167–198. ed. Allison, D.G., Gilbert, P., Lappin-Scott, H.M. and Wilson, M. Cambridge: Cambridge University Press.

Riley, M.A. (1998) Molecular mechanism of bacteriocin evolution. Annual Review of Genetics 32, 255–278. Riley, M.A. and Gordon, M.G. (1999) The ecological role of bacteriocins in bacterial competition. Trends in Microbiology 7, 129–133. Skillman, L.C., Sutherland, I.W. and Jones, M.V. (1999) The role of exopolysaccharides in dual species biofilm development. Journal of Applied Microbiology Symposium Supplement 85, 13–18. Skillman, L.S., Sutherland, I.W., Jones, M.V. and Goulsbra, A. (1998) Green fluorescent protein as an novel species specific marker in enteric dual-species biofilms. Microbiology 144, 2095–2101. Smarda, J. and Smajs, D. (1998) Colicins — Exocellular lethal proteins of Escherichia coli. Folia Microbiologica 43, 563–582. Sutherland, I.W. and Wilkinson, J.F. (1965) Depolymerases for bacterial EPS obtained from phage-infected bacteriae. Journal of General Microbiology 39, 373–383. Walker, G.C. (1984) Mutagenesis and inducible responses to deoxyribonucleic acid damage in Escherichia coli. Microbiology Reviews 48, 60–93.

ª 2002 The Society for Applied Microbiology, Journal of Applied Microbiology, 93, 345–352