• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Arch Microbiol. Author manuscript; available in PMC Mar 1, 2013.
Published in final edited form as:
PMCID: PMC3408838

Isolation and identification of a bacteriocin with antibacterial and antibiofilm activity from Citrobacter freundii


Multi- and pan-antibiotic-resistant bacteria are a major health challenge in hospital settings. Furthermore, when susceptible bacteria establish surface-attached bio-film populations, they become recalcitrant to antimicrobial therapy. Therefore, there is a need for novel antimicrobials that are effective against multi-drug-resistant and surface-attached bacteria. A screen to identify prokaryote-derived antimicrobials from a panel of over 100 bacterial strains was performed. One compound isolated from Citrobacter freundii exhibited antimicrobial activity against a wide range of Gram-negative bacteria and was effective against biofilms. Random transposon mutagenesis was performed to find mutants unable to produce the antimicrobial compound. Transposons mapped to a bacteriocin gene located on a small plasmid capable of replication in Escherichia coli. The plasmid was sequenced and found to be highly similar to a previously described colicinogenic plasmid. Expression of the predicted bacteriocin immunity gene conferred bacteriocin immunity to E. coli. The predicted bacteriocin gene, colA-43864, expressed in E. coli was sufficient to generate anti-microbial activity, and purified recombinant ColA-43864 was highly effective in killing E. coli, Citrobacter species, and Klebsiella pneumoniae cells in a planktonic and biofilm state. This study suggests that bacteriocins can be an effective way to control surface-attached pathogenic bacteria.

Keywords: Citrobacter freundii, Bacteriocin, Antimicrobial, Biofilm


Members of the Enterobacteriaciae family of Gram-negative bacteria are among the most common food-borne and hospital-acquired pathogens. These organisms have had an immense impact upon human history through devastating diseases, such as the black plague and typhoid fever as well as more common urinary tract infections, dysentery, and hospital-acquired infections (Janda and Abbott 2006). Many members of the Enterobacteriaciae are well known for acquiring antibiotic resistance, including pan-resistant strains of Klebsiella pneumoniae, and carbapenem-resistant Escherichia coli and Serratia marcescens (Nordmann 1998; Bush 2010). The recent discovery of the ndm-1-coded carbapenemase has underscored the need to identify novel antibiotics (Bush 2010; Pillai et al. 2011).

Whereas bacteria can gain resistance through acquisition of additional DNA from a plasmid or other source, mutation of an antibiotic target site or a transporter protein, or other genetic mechanism, most bacteria can also gain antibiotic tolerance though formation of a biofilm (Mah and O’Toole 2001; Anderson and O’Toole 2008; Stewart and Costerton 2001). The mechanisms by which bacteria in biofilms resist killing by antibiotics are not completely elucidated. Proposed biofilm antibiotic resistance mechanisms include the presence of persister cells, microenvironments within the biofilm that prevent the antibiotic efficacy, and reduced antibiotic access to bacteria within a biofilm (Stewart and Costerton 2001; Hoiby et al. 2010; Lewis 2010). Regardless of the mechanism of resistance, the biofilm-associated infections are difficult to treat with antibiotics and tend to be persistent (Donlan 2001; Donlan and Costerton 2002). An antimicrobial capable of efficiently killing bacteria in biofilms would be of considerable use in treating chronic, biofilm-associated infections.

Bacteria have been competing for niches for hundreds of millions of years and have developed elaborate systems to inhibit competitors. The enormous genetic potential of microorganisms can be mined for useful compounds. Because of the pressing need for new antimicrobials that are effective against biofilms, we performed a screen for bacteria-produced antimicrobials. Here, we describe the identification of a bacteriocin from Citrobacter freundii and its characterization as an antimicrobial effective against planktonic and surface-attached bacteria. This is, to our knowledge, the first demonstration of established biofilm control by a bacteriocin.

Materials and methods

Bacterial strains, media, and culture conditions

The bacteria used in this study are listed in Table 1. Bacteria were grown routinely in lysogeny broth (LB) medium at 37°C. E. coli strain WM3064 was grown in medium supplemented with 0.3 mM diaminopimelic acid (DAP). Cells were enumerated as colony-forming units (CFU) on LB agar plates, when appropriate gentamicin was used at 10 μg ml−1 and kanamycin at 50 μg ml−1.

Table 1
Strains used in the study

Microbial inhibition assay

To examine the ability of the tested bacteria to produce antimicrobial compounds, bacteria were grown for 18 h in liquid broth. Thereafter, 20 μl of the overnight culture (~108 CFU ml−1) was spotted on a lawn of microbial cells. Microbial lawns were prepared by spreading 100 μl of an overnight culture on an LB agar plate and incubated at 37°C. Positive production of a diffusible antimicrobial compound was visualized by the inhibition of the susceptible microbial lawn and a clear zone surrounding the examined bacteria colony.

Crude extraction and biochemical analysis of antimicrobial compound from C. freundii

Citrobacter freundii was grown for 24 h in broth at 37°C. One milliliter of the overnight culture was centrifuged for 3 min at 12,000×g, and the supernatant was passed through a 0.2-μm pore-size filter to remove bacteria. Ten microliters of the filter-sterilized supernatant was spotted on a lawn of sensitive microbial cells and incubated for 24 h at 37°C. Positive antimicrobial activity was visualized by the development of a zone of inhibition where the filter-sterilized solution was spotted.

To gain insight into the nature of the antimicrobial compound, the following treatments were used: storage at −20°C for 16 weeks, heating for 15 min at 80°C, DNase-I treatment for 3 h (120 μg ml−1), proteinase-K or trypsin treatment for 3 h (100 μg ml−1), and filtration through several size-exclusion Microcon Centrifugal Filter Devices (Millipore, Billerica, MA).

Construction C. freundii transposon mutant library

Transposon mutagenesis and mapping were performed as previously described (Medina et al. 2008), except that C. freundii ATCC 43864 was used as the recipient strain. The mariner-based transposon delivery plasmid pBT20 (Kulasekara et al. 2005) was used to create a library of ~4,000 mutants.

Screening for genes involved in the production of C. freundii antimicrobial compound

To screen for mutants that are impaired in their ability to produce the antimicrobial compound, the C. freundii ATCC 43864 transposon mutant library was grown in LB medium for 24 h. A 96-prong multi-well transfer device (Dan-Kar MC96) was used to transfer aliquots of mutant libraries onto plates containing lawns of sensitive C. freundii NCTC 9750. The plates were incubated at 37°C for 24 h. Positive or negative production of the antimicrobial compound was assessed by the formation of a zone of inhibition surrounding each mutant colony. C. freundii ATCC 43864 wild-type and phosphate-buffered saline (PBS) was used as positive and negative controls.

Molecular techniques

The DNA sequence flanking transposon mutants were determined using arbitrary PCR, as described previously (Medina et al. 2008). The PCR products were sequenced using the TnM Int primer at the Molecular Resource Facility, New Jersey Medical School and compared with the GenBank DNA sequence database using the BLASTX program.

The bacteriocin and immunity gene from plasmid pCfc1, to be described later in the text, was cloned using a recombineering technique using Saccharomyces cerevisiae (Shanks et al. 2006). All plasmids used in this study are listed in Table 2. The bacteriocin gene was amplified using primers 2450, accgcttctgcgttctgatttaatctgtatcaTTAGTGATGGTGGTGATGGTGGTGATGTGCAGGTCGGATTAT TTC, and 2451, ctctctactgtttctccatacccgtaggaggaaaaagaATGCCTGGATTTAATTATGGTG that include an in-frame C-terminal poly-histidine tag (underlined), sequence to target recombination with expression vector pMQ124 (Shanks et al. 2009) (lower-case), and sequence to amplify the bacteriocin gene (upper-case). The bacteriocin immunity gene was amplified using primers 2446 (cgttgtaaaacgacggccagtgccaagcttgcatgcctgcGTTTGATTAAAAGGCAGTGT) and 2447 (gaattgtgagcggataacaatttcacacaggaaacatATGAATGAACACTCAATAGATAC), and primers sequences annotated as above. DNA was amplified with a high-fidelity polymerase (Phusion, New England Biolabs), using the manufacturers directions. The recombination reactions place the amino-terminus tagged histidine-tagged under transcriptional control of the E. coli PBAD promoter on the ColE1-based pMQ124 vector and place the immunity gene under transcriptional control of the E. coli Plac promoter on the pBBR1-based pMQ131 vector (Shanks et al. 2009). Plasmid constructs were verified by sequencing (University of Pittsburgh Genomics and Proteomics Core).

Table 2
Plasmids used in the study

Purification of polyhistidine-tagged bacteriocin from E. coli

Escherichia coli S17-1 harboring pMQ348 was grown for 18 h in LB supplemented with 10 μg ml−1 gentamicin to reach a final concentration of OD600 = 0.2. One milliliter of the overnight culture was subcultured in 5 ml of fresh LB and left to grow for 2 h, arabinose, which induces expression of PBAD promoter in pMQ348, was added to the culture (0.2% w/v final concentration) and the tubes were incubated for an additional 3 h. To obtain crude cell proteins, the bacteria were pelleted by centrifugation, washed twice in PBS and resuspended in fresh 500 μl of PBS. Cells were lysed by sonication on ice for 50 s using a VC505 sonicator set on 80% strength (Sonics and Materials Inc., Newtown, CT, USA). Cell debris was pelleted by centrifugation, and the supernatant was removed and passed through a 0.2-μm pore-size filter (hereafter referred to as crude cell extract).

Immobilized metal ion affinity chromatography (IMAC) was used to further purify the His8-tagged bacteriocin, and 1 ml of the crude cell extract was mixed with 200 μl of Talon Metal Affinity Resin (Clontech Laboratories, Inc. Mountain View, CA) suspended washing-buffer (50 mM sodium phosphate, 300 mM NaCl and 10 mM imidazole pH-7). The mixture was stirred for 30 min and then centrifuged for 2 min at 1,000×g. The pelleted resin was collected in a 2-ml tube and the unbound supernatant discarded. The resin was washed twice with washing-buffer containing 60 mM imidazole. Elution of the tagged protein was performed by the addition of 1 ml of washing-buffer supplemented with 200 mM imidazole (elution buffer). The eluent was incubated for 2 min, centrifuged for 2 min, and the supernatant containing the His8-tagged protein was collected in a new tube. The sample was then filtered through a 0.2-μm pore-size filter (hereafter referred to as IMAC-purified bacteriocin). Protein concentration was determined using Bio-Rad Quick start Bradford protein assay.

PAGE and Western blot analysis were performed on each purification fraction using standard techniques. The PAGE gel (4–20% gradient) was a precast mini-format gel (Precise Protein Gel, Pierce), and a Bio-Rad Protean 3 device was used for electrophoresis and transfer to an Immobilon transfer membrane (Millipore). The blot was probed with a mouse-anti-polyhistine antibody (Covance product number MMS-156P), and the secondary antibody was a goat-anti-mouse HRP-conjugated antibody (Pierce, product number 32430).

Bacteriocin anti-microbial activity

To assess the antimicrobial activity of bacteriocin on planktonically grown bacteria, tested bacteria were grown for 18 h at 37°C. Thereafter, 1 ml of the cells was pelleted by centrifugation, washed, and resuspended in 1 ml PBS. One hundred-microliter cell aliquots were placed in a 2-ml microfuge tube, and an equal volume of crude or IMAC-purified bacteriocin was added to the tube. Alternatively, as a control, 100 μl of sterile PBS was added to each tube. The tube was incubated at 37°C for the duration of the experiment. Quantification of viable bacteria before and following treatment was performed by CFU enumeration. Each experiment was carried out at least three times.

Bacteriocin (ColA-43864-His8) anti-biofilm activity

Biofilms were formed in a non-tissue culture treated, 96-well polyvinyl chloride microtiter dishes (Becton–Dickinson, Franklin Lakes, NJ, USA) as previously described (O’Toole and Kolter 1998; Merritt et al. 2005). Briefly, microtiter wells were inoculated (100 μl per well) with bacteria that had been grown cell culture to stationary phase in LB medium and diluted 1:100 in fresh LB media. K. pneumoniae biofilms were developed in M63 minimal salts supplemented with 1 mM MgSO4·7H2O, 14 mM sodium citrate, and 34 mM L-proline (Kadouri et al. 2007). The plates were incubated for 18 h at 30°C to generate biofilms. To assess the antimicrobial activity of bacteriocin, the preformed biofilms were washed twice with PBS to remove planktonic cells, and 100 μl of tested bacteriocin sample was added to each well. Alternatively, as a control, 100 μl of sterile PBS was added to the wells. The microtiter dishes were incubated at 30°C for the duration of the experiment. Quantification of biofilm bacteria before and following treatment was performed by washing the microtiter plates with PBS, to remove non-adhering cells, 100 μl of fresh PBS was added to each well, and the samples were sonicated for 8 s using a VC505 sonicator, set on 40% strength, followed by dilution plating and CFU enumeration (Kadouri and O’Toole 2005; Kadouri et al. 2007). Each experiment was carried out at least three times.

Microscopy of biofilms was performed by first establishing biofilms on PVC cover slips in LB medium using the air liquid interface (ALI) method, described by Merritt et al. (2005). Biofilms of C. freundii ATCC 8090 were formed for 20 h, non-adherent bacteria were removed by washing with PBS and then incubated in PBS with either crude lysates from S17-1 + pMQ124 (empty vector) or S17-1 + pMQ348 (colicin expressing plasmid) to a final concentration of 11.6 μg ml−1 total protein. After 60 min, biofilms were again washed with PBS and stained with a commercial live-dead stain (Bac-Light, Invitrogen) according to the manufacturers specifications and visualized with an Olympus Fluoview 1000 confocal laser scanning microscope (CLSM) and Fluoview 2.1 software.


Identification of an antimicrobial compound produced by C. freundii ATCC 43864

In a screen aimed at isolating new antimicrobial compounds, 105 bacteria, representing 42 species and 26 different genera, were cross-examined in a microbial inhibition assay [for a full list of bacteria tested, see Dashiff et al. (2011)].

One bacterial strain, C. freundii ATCC 43864, was found to produce a diffusible antimicrobial compound that inhibited the growth of other C. freundii strains (Fig. 1a, I), and other bacteria described below. Filter-sterilized supernatants of C. freundii ATCC 43864 grown in liquid broth (crude extraction method) also yielded the antimicrobial compound, suggesting that the compound was extracellular and did not require a competing organism to induce its production (Fig. 1a, II).

Fig. 1
Identification of an antimicrobial compound from C. freundii ATCC 43864. a Microbial inhibition assay. C. freundii ATCC 43864 was grown in liquid broth for 24 h, and cells (I) or filter-sterilized supernatant (II) was spotted on a lawn of C. freundii ...

Microbial inhibition assays using both colonies and filter-sterilized supernatants were used to investigate whether additional members of the Citrobacter genus produce a similar antimicrobial compound and to evaluate their sensitivity to the antimicrobial compound produced by C. freundii ATCC 43864. As described in Table 3, an antimicrobial compound was produced only by C. freundii ATCC 43864. The compound was observed to inhibit the growth of C. braakii ATCC 43162, C. freundii NCTC 9750, and C. freundii ATCC 8090, but was not active in inhibiting the producing strain C. freundii ATCC 43864.

Table 3
Production of antimicrobial compound by Citrobacter spp

To further characterize the antimicrobial compound, crude extracts were isolated from C. freundii ATCC 43864 and subjected to a number of challenges. Treated extracts were spotted on a lawn of sensitive C. freundii NCTC 9750 to evaluate their antimicrobial activity. The antimicrobial compound was found to be resistant to DNase-I and freezing and sensitive to protease activity and heat. The compound was also found to have a molecular mass between 50 and 100 kDa based upon fractionation with size-exclusion filtration columns (data not shown). From this analysis, it was concluded that the antimicrobial compound was likely to be a protein.

Construction of C. freundii ATCC 43864 transposon mutant library, and isolation of mutants defective in the synthesis of the antimicrobial compound

To isolate mutants defective in the synthesis of the antimicrobial compound, a mariner-based transposon was used to mutagenize C. freundii ATCC 43864. Mutants were transferred onto lawn of C. freundii NCTC 9750, which was found to be sensitive to the compound. The plates were incubated for 24 h until a zone of inhibition was seen surrounding each mutant (Fig. 1b). Using this approach from ~4,000 mutant colonies, a mutant unable to produce a zone of inhibition was isolated (Fig. 1c). This mutant was designated Cf-8A. No difference in growth rate was observed between Cf-8A mutants and the C. freundii recipient when grown in LB medium (data not shown).

Mapping of mutation Cf-8A to a predicted bacteriocin gene, colA-43864

The Cf-8A transposon mutation was mapped using arbitrary PCR to base pair 456 of a predicted bacteriocin gene, that is ~99% identical at the DNA level to the colA gene of C. freundii strain CA31, also known as colA-CA31. The resulting predicted protein has one amino acid different from the previously identified ColA-CA31, specifically a leucine to serine change at amino acid 328. Because the newly isolated bacteriocin gene is not identical to colA-CA31 and is derived from a different strain, we named the gene colA-43864. Bacteriocins are bacterial-produced antimicrobial proteins, of which colicins are a well-characterized subgroup (Cascales et al. 2007).

Since the colA-CA31 gene was found on a small plasmid that could replicate in E. coli (Morlon et al. 1982), we tested whether the Cf-8A mutation was also on a plasmid. To this end, total DNA was isolated from mutant strain Cf-8A and used to electroporate E. coli. If the Cf-8A mutation was on a plasmid capable of replication in E. coli, then we predicted that the plasmid could be selected for using the gentamicin resistance marker on the transposon. We were able to isolate gentamicin-resistant E. coli colonies harboring a plasmid isolated from the Cf-8A (data not shown), supporting that the Cf-8A was on a plasmid.

A small plasmid, named pCfc1, was isolated from the WT ATCC 43864 strain and sequenced (Genbank JF795024). Sequence analysis reveals that the plasmid is 6.72 kb in length, is ~99% identical to the pColA-CA31 plasmid of C. freundii strain CA31 (Morlon et al. 1988b), and has high similarity to bacteriocin-bearing plasmids from other Citrobacter species. The gene organization of pColA-CA31 (Morlon et al. 1988b) and the pCfc1 is identical and includes predicted genes for a bacteriocin, a bacteriocin immunity protein, a lysis protein, entry exclusion proteins, and plasmid mobility genes.

Antimicrobial activity of E. coli harboring C. freundii ATCC 43864 bacteriocin plasmid (pCfc1)

In order to establish that the genes involved in the synthesis of the antimicrobial compound are plasmid-borne and that no additional genes are required for the production of the bacteriocin, the pCfc1 was mobilized into E. coli. To accomplish this, we moved a selectable marker onto the plasmid, by mutagenesis with the pBT20-derived mariner transposon. C. freundii ATCC 43864 cells containing pCfc1 were mutagenized with the mariner transposon. The prediction was that, in a small subset of the mutant colonies, the transposon would localize to the pCfc1 plasmid rather than the chromosome, thereby adding a selectable marker to the pCfc1 plasmid. Mutant colonies were pooled, and plasmid DNA was isolated. The harvested plasmids were mobilized by electroporation into E. coli S17-1 cells, and the S17-1 cells were plated on LB agar supplemented with gentamicin. One Gmr isolate (E. coli-21pA) was selected for further analysis. DNA sequencing of purified plasmid from E. coli-21pA confirmed that the cell harbors pCfc1 containing a transposon insertion at position 441 relative to the arbitrarily determined origin in an intergenic region between two predicted entry exclusion genes. This mutation is not predicted to interfere with transcription of the bacteriocin gene. The marked plasmid was designated as pCfc1-21A.

Wild-type E. coli S17-1, E. coli S17-21pA (E. coli S17-1-bearing pCfc1-21A in which the pCfc1 plasmid has transposon inserted in an intergenic region), and E. coli S17-8A (E. coli S17-1-bearing pCfc1-8A in which the pCfc1 plasmid has a transposon mutation in the colA-43864 gene) were all spotted on a lawn of C. freundii NCTC 9750. As seen in Fig. 2, a clear zone of inhibition was seen around E. coli S17-21pA, supporting that the genes necessary for antimicrobial production are all present on pCfc1. As anticipated, no antimicrobial activity was seen on lawns that were inoculated with S17-1 wild-type or bearing the pCfc1-8A plasmid (Fig. 2. E. coli S17-1 wild-type and E. coli S17-8A, respectively).

Fig. 2
Transformation of E. coli with pCfc1. Overnight cultures of wild-type E. coli S17-1, E. coli S17-21pA, and E. coli-8A were spotted on a lawn of sensitive C. freundii NCTC 9750. Antimicrobial activity is seen by the formation of a zone of inhibition around ...

Production and purification of bacteriocin (ColA-43864) in E. coli

To determine whether the colA-43864 gene was sufficient for the observed antimicrobial phenotype, we cloned a poly-histidine tagged version of the gene and placed it under control of an arabinose-inducible promoter. The His8-tag was placed separately on the C- and N-termini of the predicted ColA-43864 protein, and while both conferred antimicrobial activity, the C-terminal version exhibited higher relative antimicrobial activity (data not shown), and thus, subsequent work was performed with the carboxy-terminal tagged version. To express colA-43864 in E. coli, pMQ348 (pMQ124-colA-43864-His8) was mobilized by electroporation into E. coli S17-1 cells, to make E. coli S17-pMQ348. As a control, E. coli S17-1 was also transformed with empty vector control pMQ124, to generate E. coli S17-pMQ124.

When grown in the presence of 0.2% (w/v) glucose, no difference in growth was seen between E. coli S17-pMQ348 and E. coli S17-pMQ124 (Fig. 3). However, under promoter-inducing conditions, when 0.2% (w/v) L-arabinose was added to the media, a clear inhibition of E. coli S17-pMQ348 growth was measured compared to the empty vector control (Fig. 3). These data suggest that expressing the colA-43864 gene in E. coli has a toxic effect.

Fig. 3
Growth of E. coli S17-1 expressing ColA-43864-His8. Cultures of E. coli S17-pMQ348 (solid line) and empty vector control E. coli S17-pMQ124 (broken line) were grown at 37°C in LB supplemented with 0.2% glucose (colA-43864-inhibiting condition, ...

To isolate crude extracts and pure recombinant ColA-43864 from E. coli, overnight cultures of E. coli S17-pMQ348 and empty vector control E. coli S17-pMQ124 were grown in the presence of arabinose and crude cell extracts were prepared, as described in materials and methods. ColA-43864 was isolated by IMAC, as described in materials and methods. PAGE analysis shows the purification of a single protein of the predicted mass in the elution fraction of E. coli S17-pMQ348 lysates, whereas no band was observed in the elution fraction of the empty vector control (Fig. 4a). To further confirm that the single band was the ColA-43864-His8 construct, Western blotting was performed. We observed that the single eluted band seen on the PAGE gel was detected by an anti-polyhistidine antibody, whereas that band was absent in the empty vector control (Fig. 4b).

Fig. 4
Purification of ColA-43864-His8. a ColA-43864-His8 purification fractions analyzed by PAGE. Crude lysates with empty vector negative control (pMQ124) or pMQ124 + His8-colA-43864-His8 (pMQ348) were generated identically and purified by IMAC. Fractions ...

Antimicrobial activity of crude extract and purified ColA-43864-His8

To characterize the bacteriocin’s antimicrobial specificity, crude cell extracts from E. coli S17-pMQ348 and control E. coli S17-pMQ124 were prepared. Samples of the crude extracts (20 μl, containing 15 and 35 μg total protein from E. coli S17-pMQ348 and E. coli S17-pMQ124, respectively) were spotted on lawns of examined bacteria. As seen in Table 4, the crude ColA-43864-containing extract was able to inhibit the growth of several bacteria from the family Enterobacteriaceae, including C. braakii, C. freundii, Enterobacter gergoviae, E. coli, Klebsiella pneumoniae, and Yersini pseudotuberculosis. Several other tested species were immune to the bacteriocin-containing extracts (Table 4). No inhibition was seen when crude protein extract from empty vector control E. coli S17-pMQ124 was spotted (Table 4). Similar inhibition patterns were seen when C. freundii ATCC 43864 colonies were spotted on the lawns (data not shown).

Table 4
Antimicrobial activity of bacteriocin

To assess the antimicrobial activity of bacteriocin on planktonic bacteria grown in broth, crude extracts, isolated from E. coli S17-pMQ348 (37 μg total protein) and control E. coli S17-pMQ124 (87 μg total protein), were added to overnight cultures. Incubating C. freundii and E. coli for 30 min with crude cell bacteriocin extracts had resulted in a 8 and 6 log reduction in cell counts. A more moderate 2–3 log reduction was measured for K. pneumoniae and Y. pseudotuberculosis. No reduction was seen after incubation with PBS or crude cell extracts isolated from E. coli S17-pMQ124 empty vector control (Fig. 5a).

Fig. 5
Antimicrobial activity of crude ColA-43864 extract. a Effect of ColA-43864 on planktonic bacteria. Tested bacteria (~109 CFU ml−1) were incubated for 30 min with PBS (black bars), 87 μg protein extracted from empty vector control E. coli ...

Whereas many antimicrobial agents are effective against planktonic cells, fewer are active against biofilms. Furthermore, while the results of one study suggest that bacteriocins can inhibit the development of biofilms (Hancock et al. 2010), it is not known whether bacteriocins have any effect upon cells in a biofilm. When placed on biofilms, crude cell extracts, isolated from E. coli S17-pMQ348 (75 μg total protein), were able to significantly reduce biofilm cell viability by up to 7 logs within 2 h of incubation (Fig. 5b).

To further assess the antimicrobial effect of ColA-43864 on pre-formed biofilms, CLSM was used to analyze bacteriocin-treated biofilms treated with live-dead fluorescent stains. Extracts from bacteria expressing recombinant ColA-43864-His8 had a clear bacteriocidal effect on C. freundii biofilm cells (note red staining Fig. 5c, pMQ348 Dead), whereas most the C. freundii biofilm exposed to extracts without the recombinant ColA-43864-His8 was largely composed of live cells (green staining Fig. 5c, pMQ124 Live). Furthermore, the colicin-exposed biofilms were stained red throughout the biofilms (data not shown), suggesting that ColA-43864 is able to penetrate the biofilms.

A range of purified recombinant ColA-43864-His8 (0.02–2.00 μg) was used to test the impact of this protein on both planktonically grown bacteria and biofilms. Two micrograms of ColA-43864-His8 caused a 6–7 log reduction in planktonically grown and biofilm cell viability (Fig. 6a, b). The reduction in cell viability was dose dependent (Fig. 6a, b). In an additional experiment, K. pneumoniae biofilms (composed of 2.5 × 106 CFU ml−1) were incubated with 2 μg of ColA-43864-His8. A 2 log reduction (1 × 104 CFU ml−1) was seen within a 30-min incubation period. No reduction was measured for K. pneumoniae biofilms incubated with control PBS or mock IMAC-purified protein isolated from E. coli S17-pMQ124 (2.4 × 106 CFU ml−1 and 2.5 × 106 CFU ml−1, respectively). Treating the bacteria with the elution buffer, used to elute the IMAC-purified protein, did not cause any reduction in bacterial CFU (data not shown).

Fig. 6
Antimicrobial activity of ColA-43864-His8. Effect of IMAC-purified ColA-43864 on planktonic (a) or biofilm (b) planktonic bacteria. C. freundii NCTC 9750 (6 × 109 and 2 × 106 CFU ml−1 for planktonic and biofilm, respectively) were ...

Expressing C. freundii immunity gene provides resistance from ColA-43864 antimicrobial activity

In order to investigate whether the predicted immunity gene, from C. freundii ATCC 43864 plasmid pCfc1, is capable of providing resistance to the antimicrobial activity of ColA-43864, bacteriocin-sensitive wild-type E. coli S17-1, C. freundii NCTC 9750, and C. freundii ATCC 8090 were transformed by electroporation with pCfc-8A (pCfc1 with a mutated colA-43864 gene). The transformants were spread on an LB agar plate, and 20-μl aliquots of crude ColA-43864-containing extracts were spotted on top of each lawn (15 and 35 μg total protein from E. coli S17-pMQ348 and E. coli S17-pMQ124, respectively). As seen in Fig. 7a, cells harboring the bacteriocin-defective plasmid, with an intact immune gene, were protected from ColA-43864-containing extracts. In another experiment, E. coli S17-1 was transformed with the immune gene, cloned under control of the E. coli Plac promoter on a multicopy plasmid, to yield strain E. coli S17-pMQ345. Indicating that the immune gene was sufficient to produce ColA-43864-resistant E. coli (Fig. 7a, E. coli S17-pMQ345). CFU enumeration of planktonically grown bacteria also confirmed the resistance of E. coli S17-pMQ345 to the antimicrobial effect of crude cell extracts containing ColA-43864 (Fig. 7b).

Fig. 7
Expression of a predicted bacteriocin immunity gene in ColA-43864-sensitive bacteria. a Bacteriocin antimicrobial activity assay. Crude ColA-43864-containing extracts were spotted on microbial lawns of C. freundii NCTC 9750, C. freundii ATCC 8090, and ...


In this study, we have determined the identity of an antimicrobial protein from a C. freundii strain ATCC 43864 using a genetic approach. The gene that codes for the antimicrobial, colA-43864, is on a small plasmid almost identical in sequence to the well-characterized colicin A-producing plasmid from C. freundii strain CA31 (Morlon et al. 1988a). The colA-CA31 gene from C. freundii strain CA31 was previously described as a type-A bacteriocin (Davies and Reeves 1975) and used for several studies to characterize the role of “A” type bacteriocins (Varenne et al. 1981; Crozel et al. 1983). Based on the ~99% sequence identity, we predict that the biofilm-controlling ColA-43864 described in this study is also an “A” type bacteriocin. Bacteriocins are released through the action of lysis proteins rather than through traditional secretion systems (Cascales et al. 2007). The “A” type bacteriocins gain access to bacterial cells through binding to a complex of outer membrane proteins (OmpF and BtuB) and lipopolysaccharides (Chai et al. 1982; Cascales et al. 2007), then move through the periplasmic space with the aid of TolA and TolB (Hecht et al. 2010), and kill bacteria by disrupting their membrane and uncoupling the proton gradient (Schein et al. 1978; Cascales et al. 2007).

The plasmid pCfc1-21A was able to replicate in E. coli and was sufficient to confer antimicrobial activity to E. coli; however, when the predicted bacteriocin gene was mutated with a transposon, no antimicrobial activity was observed, suggesting that the antibacterial activity requires the colA-43864 gene. To further confirm that the bacteriocin gene is necessary and sufficient for the antimicrobial activity, the colA-43864 gene was cloned and placed expressed in E. coli under control of an arabinose-inducible promoter. Induction of the colA-43864 gene resulted in a severe growth defect, supporting that expression of the colA-43864 gene alone is toxic to the host cell. Crude fractions from the E. coli strain bearing pMQ348, but not the empty vector control, exhibited antimicrobial activity against species from several genera of Gram-negative bacteria including important pathogens K. pneumoniae and E. coli. Interestingly, only 8 out of 13 K. pneumoniae isolates exhibited sensitivity to the colicin. The source of bacteriocin resistance by the five bacteriocin-resistant K. pneumoniae isolates may be due to an altered or absent bacteriocin receptor, or they may harbor bacteriocin immunity genes. However, we were not able to detect immunity genes by PCR while using primers based on conserved sequences from multiple immunity genes (data not shown).

Plasmid pCfc1-8A or the predicted immunity protein gene alone, expressed on a multicopy plasmid, was able to provide protection from ColA-43864. This result further supports that the antimicrobial effect was specific to the ColA-43864, rather than some unrelated compound induced by production of ColA-43864.

A previous study showed that bacteriocin-producing E. coli strains could inhibit biofilm formation by non-bacteriocin-producing E. coli strains on catheter material (Hancock et al. 2010). While the study provided strong genetic evidence that bacteriocins could inhibit biofilm formation (Hancock et al. 2010), it was not determined whether established biofilms can be treated with bacteriocins or whether bacteriocins separated from other bacterial components could be used. Other studies have explored the influence of bacteriocins on biofilm formation by competing organisms (Kreth et al. 2005; Tait and Sutherland 2002). Tait and Sutherland showed that bacteriocin-producing strains have a selective advantage within a multi-species biofilm and their data suggest that bacteriocins may have evolved to function largely in the biofilms rather than among planktonic cells. Here, we show for the first time that a purified bacteriocin can effectively and rapidly kill cells in biofilms in a dose-dependent manner. The ability of the ColA-43864 to eradicate a large portion of the cells within a biofilm supports the model that diffusion of the ColA-43864 into the biofilm is not a limiting factor; however, it is possible that there are microenvironments within the biofilm where ColA-43864 would not function due to suboptimal pH or other conditions.

This study supports that bacteriocins could be used to effectively treat some bacterial biofilms. Future analysis will focus on determining the effect of bacteriocins in tandem with other antimicrobials and biocontrol agents such as phage and predatory bacteria to optimize methods for destroying established biofilms.


The authors would like to thank Nicholas Stella for technical assistance and Dr. Yohei Doi, Division of Infectious Diseases, University of Pittsburgh School of Medicine for kindly providing bacterial strains. This work was supported by funding from the Foundation of UMDNJ faculty research grant to D.E.K., and NIH AI085570 and a Research to Prevent Blindness Career Development Award to R.M.Q.S. Additional support was provided by NIH grant EY08098 and the Eye and Ear Foundation of Pittsburgh.

Contributor Information

Robert M. Q. Shanks, Campbell Laboratory of Ophthalmic Microbiology, Department of Ophthalmology, University of Pittsburgh, Pittsburgh, PA 15213, USA.

Aliza Dashiff, Department of Oral Biology, University of Medicine and Dentistry of New Jersey, Newark, NJ 07101, USA.

Jason S. Alster, Department of Oral Biology, University of Medicine and Dentistry of New Jersey, Newark, NJ 07101, USA.

Daniel E. Kadouri, Department of Oral Biology, University of Medicine and Dentistry of New Jersey, Newark, NJ 07101, USA.


  • Anderson GG, O’Toole GA. Innate and induced resistance mechanisms of bacterial biofilms. Curr Top Microbiol Immunol. 2008;322:85–105. [PubMed]
  • Bush K. Alarming beta-lactamase-mediated resistance in multidrug-resistant Enterobacteriaceae. Curr Opin Microbiol. 2010;13:558–564. [PubMed]
  • Cascales E, et al. Colicin biology. Microbiol Mol Biol Rev. 2007;71:158–229. [PMC free article] [PubMed]
  • Chai T, Wu V, Foulds J. Colicin A receptor: role of two Escherichia coli outer membrane proteins (OmpF protein and btuB gene product) and lipopolysaccharide. J Bacteriol. 1982;151:983–988. [PMC free article] [PubMed]
  • Croal LR, Jiao Y, Newman DK. The fox operon from Rhodobacter strain SW2 promotes phototrophic Fe(II) oxidation in Rhodobacter capsulatus SB1003. J Bacteriol. 2007;189:1774–1782. [PMC free article] [PubMed]
  • Crozel V, Lazdunski C, Cavard D. Localization of genes responsible for replication and immunity to colicin A on plasmid ColA-CA31. Mol Gen Genet. 1983;192:500–505. [PubMed]
  • Dashiff A, Junka RA, Libera M, Kadouri DE. Predation of human pathogens by the predatory bacteria Micavibrio aeruginosavorus and Bdellovibrio bacteriovorus. J Appl Microbiol. 2011;110:431–444. [PubMed]
  • Davies JK, Reeves P. Genetics of resistance to colicins in Escherichia coli K-12: cross-resistance among colicins of group A. J Bacteriol. 1975;123:102–117. [PMC free article] [PubMed]
  • Donlan RM. Biofilm formation: a clinically relevant microbiological process. Clin Infect Dis. 2001;33:1387–1392. [PubMed]
  • Donlan RM, Costerton JW. Biofilms: survival mechanisms of clinically relevant microorganisms. Clin Microbiol Rev. 2002;15:167–193. [PMC free article] [PubMed]
  • Hancock V, Dahl M, Klemm P. Probiotic Escherichia coli strain Nissle 1917 outcompetes intestinal pathogens during biofilm formation. J Med Microbiol. 2010;59:392–399. [PubMed]
  • Hecht O, Zhang Y, Li C, Penfold CN, James R, Moore GR. Characterisation of the interaction of colicin A with its co-receptor TolA. FEBS Lett. 2010;584:2249–2252. [PubMed]
  • Hoiby N, Bjarnsholt T, Givskov M, Molin S, Ciofu O. Antibiotic resistance of bacterial biofilms. Int J Antimicrob Agents. 2010;35:322–332. [PubMed]
  • Janda MJ, Abbott SL. The Enterobacteria. 2. American Society of Microbiology; Washington, DC: 2006.
  • Kadouri D, O’Toole GA. Susceptibility of biofilms to Bdellovibrio bacteriovorus attack. Appl Environ Microbiol. 2005;71:4044–4051. [PMC free article] [PubMed]
  • Kadouri D, Venzon NC, O’Toole GA. Vulnerability of pathogenic biofilms to Micavibrio aeruginosavorus. Appl Environ Microbiol. 2007;73:605–614. [PMC free article] [PubMed]
  • Kreth J, Merritt J, Shi W, Qi F. Competition and coexistence between Streptococcus mutans and Streptococcus sanguinis in the dental biofilm. J Bacteriol. 2005;187:7193–7203. [PMC free article] [PubMed]
  • Kulasekara HD, Ventre I, Kulasekara BR, Lazdunski A, Filloux A, Lory S. A novel two-component system controls the expression of Pseudomonas aeruginosa fimbrial cup genes. Mol Microbiol. 2005;55:368–380. [PubMed]
  • Lewis K. Persister cells. Annu Rev Microbiol. 2010;64:357–372. [PubMed]
  • Mah TF, O’Toole GA. Mechanisms of biofilm resistance to antimicrobial agents. Trends Microbiol. 2001;9:34–39. [PubMed]
  • Medina AA, Shanks RM, Kadouri DE. Development of a novel system for isolating genes involved in predator-prey interactions using host independent derivatives of Bdellovibrio bacteriovorus 109J. BMC Microbiol. 2008;8:33. [PMC free article] [PubMed]
  • Merritt JH, Kadouri DE, O’Toole GA. Growing and analyzing static biofilms. Current protocols in microbiology. 2005:1B.1.1–1B.1.17. [PubMed]
  • Morlon J, Cavard D, Lazdunski C. Physical map of pColA-CA31, an amplifiable plasmid, and location of colicin A structural gene. Gene. 1982;17:317–321. [PubMed]
  • Morlon J, Chartier M, Bidaud M, Lazdunski C. The complete nucleotide sequence of the colicinogenic plasmid ColA. High extent of homology with ColE1. Mol Gen Genet. 1988a;211:231–243. [PubMed]
  • Morlon J, Sherratt D, Lazdunski C. Identification of functional regions of the colicinogenic plasmid ColA. Mol Gen Genet. 1988b;211:223–230. [PubMed]
  • Nordmann P. Trends in beta-lactam resistance among Enter-obacteriaceae. Clin Infect Dis. 1998;27(Suppl 1):S100–S106. [PubMed]
  • O’Toole GA, Kolter R. Flagellar and twitching motility are necessary for Pseudomonas aeruginosa biofilm development. Mol Microbiol. 1998;30:295–304. [PubMed]
  • Pillai DR, McGeer A, Low DE. New Delhi metallo-beta-lactamase-1 in Enterobacteriaceae: emerging resistance. CMAJ. 2011;183:59–64. [PMC free article] [PubMed]
  • Schein SJ, Kagan BL, Finkelstein A. Colicin K acts by forming voltage-dependent channels in phospholipid bilayer membranes. Nature. 1978;276:159–163. [PubMed]
  • Shanks RM, Caiazza NC, Hinsa SM, Toutain CM, O’Toole GA. Saccharomyces cerevisiae-based molecular tool kit for manipulation of genes from gram-negative bacteria. Appl Environ Microbiol. 2006;72:5027–5036. [PMC free article] [PubMed]
  • Shanks RM, Kadouri DE, MacEachran DP, O’Toole GA. New yeast recombineering tools for bacteria. Plasmid. 2009;62:88–97. [PMC free article] [PubMed]
  • Stewart PS, Costerton JW. Antibiotic resistance of bacteria in biofilms. Lancet. 2001;358:135–138. [PubMed]
  • Tait K, Sutherland IW. Antagonistic interactions amongst bacteriocin-producing enteric bacteria in dual species biofilms. J Appl Microbiol. 2002;93:345–352. [PubMed]
  • Varenne S, Cavard D, Lazdunski C. Biosynthesis and export of colicin A in Citrobacter freundii CA31. Eur J Biochem. 1981;116:615–620. [PubMed]
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...