Biofilm induced tolerance towards antimicrobial peptides

Abstract

Increased tolerance to antimicrobial agents is thought to be an important feature of microbes growing in biofilms. We address the question of how biofilm organization affects antibiotic susceptibility. We established Escherichia coli biofilms with differential structural organization due to the presence of IncF plasmids expressing altered forms of the transfer pili in two different biofilm model systems. The mature biofilms were subsequently treated with two antibiotics with different molecular targets, the peptide antibiotic colistin and the fluoroquinolone ciprofloxacin. The dynamics of microbial killing were monitored by viable count determination, and confocal laser microscopy. Strains forming structurally organized biofilms show an increased bacterial survival when challenged with colistin, compared to strains forming unstructured biofilms. The increased survival is due to genetically regulated tolerant subpopulation formation and not caused by a general biofilm property. No significant difference in survival was detected when the strains were challenged with ciprofloxacin. Our data show that biofilm formation confers increased colistin tolerance to cells within the biofilm structure, but the protection is conditional being dependent on the structural organization of the biofilm, and the induction of specific tolerance mechanisms.

Figure 1. E.coli cells in biofilms differing in organization display differential survival after colistin treatment.

A). Microtitre plate biofilm assay with the parental E. coli SAR18 in the presence and absence of the de-repressed IncF plasmid and mutants in the IncF plasmid (traD, traX, and traQ,) stained with crystal violet. Values are normalized with respect to the mean of SAR18 F⁺ (traD) set as 100% B). A microtitre plate biofilm antibiotic tolerance assay, in which the indicated strains were challenged with 10 µg/ml colistin for 24h and bacterial survival determined by viable count as indicated by white bars. Untreated control is indicted by black bars. C). Microtitre plate biofilm antibiotic tolerance assay, in which the indicated strains were challenged with 0.3 µg/ml ciprofloxacin for 24h and bacterial survival determined by viable count as indicted by white bars. Untreated control is indicated by black bars. D). The survival of planktonically cultured stationary phase E. coli SAR18 in the presence and absence of the de-repressed IncF plasmid and mutants in the IncF plasmid (traD, traX, and traQ,) after a 24 h 10 µg/ml colistin or 0.3 µg/ml ciprofloxacin challenge. Black bars indicate CFU in untreated culture while white bars indicate treated cells as specified in the figure.

Figure 2. The increased survival of cells in biofilms is dependent on a functional basS gene.

A). Microtitre plate biofilm assay with the parental E. coli AF504 in the presence and absence of the de-repressed IncF plasmid and mutants in the IncF plasmid (traD,), insertion deletions of the basS and basR genes, and the curli overproducing opmR strain stained with crystal violet. Values are normalized with respect to the mean of AF504 F⁺ (traD) set as 100%. B). A microtitre plate biofilm antibiotic tolerance assay in which the indicated strains were challenged with 10 µg/ml colistin for 24 h and bacterial survival determined. The fraction untreated/treated is indicated for each strain.

Figure 3. A colistin tolerant subpopulation is formed in mature E .coli flow-chamber biofilms.

CLSM micrographs of a mature AF504 gfp/F⁺ biofilm after A) control, B) 6.5h; and C) 24h 25 µg/ml of colistin challenge. The biofilm was stained with PI at the specified time points before image acquisition. Scale bars represent 50 µm D) CLSM micrograph showing the location of live (green) and dead (red) cells in E. coli AF504 gfp/F⁺ biofilm grown for 48 h and then exposed to 10 µg/ml of colistin for 24 h, an overview image taken with a 10× objective. Scale bar represent 300 µm. E) CLSM 3D micrograph image taken with a 40× objective of a single microcolony in the AF504 gfp/F⁺ biofilm. F) CLSM micrograph 3D structure showing the spatial distribution of living (green) and dead (red) cells of E. coli AF504 gfp/F⁺ in a 6 days old biofilm, that has been grown in antibiotic free media for 2 days and then exposed to 25 µg/ml colistin for 24 h and then switched back to antibiotic free medium.

Figure 4. The colistin tolerance is influenced by biofilm structural organization and tolerance induction is basS dependent.

A). CLSM micrographs showing the distribution of live and dead cells in biofilms of the SAR18/F⁺ (top left panel), SAR18/F⁺(traD) (top right panel), SAR18/F⁺(traX) (bottom left panel), and PHL628 gfp ompR234 (bottom right panel) after 24h challenge by 10 µg/ml colistin. Scale bars represent 50 µm. B). CLSM micrographs showing the distribution of live and dead cells in biofilms of AF504 gfp/F⁺ (left panel) and AF574 gfp basS::Kan/F⁺ (right panel) mutant showing the basS dependence of subpopulation formation after 24h challenge by 10 µg/ml colistin. Overview images taken with a 40× objective. Scale bars represent 100 µm.

Figure 5. The yfbE gene is induced within structured biofilms prior to colistin challenge.

A). Fluorescence of yfbE-gfp transcriptional fusion expressed in bacteria grown in N-minimal media, pH 7.7, containing 38 mM glycerol with either 10 mM Mg²⁺ (black bars), or 10 µM Mg²⁺ and 100 µM Fe³⁺ (white bars) Data shown are the average of three experiments. Mean fluorescence units were detected by using a Becton Dickingson FACS Vantage SE cell sorter. Strains harboring the attB::bla-rrnBP1::gfpmut3b or the pJBA113 vector alone demonstrated constitutive fluorescence and no fluorescence in all growth conditions, respectively. B). Fraction of fluorescent cells in the microtitre plate biofilm assay with the parental E. coli AF504 in the presence and absence of the de-repressed IncF plasmid and mutants in the IncF plasmid (traD,), and the reciprocal insertion deletions of the basS and basR genes, harboring the yfbE-gfp transcriptional fusion. The threshold for GFP fluorescence was cells with a value above 600 arbitrary units. 99% of the constitutively expressing cells were above this threshold while 96% of the negative control was below. C). CLSM micrographs showing the expression of the yfbE-gfp transcriptional fusion in flow chamber biofilms of AF504/F⁺ (left panel) and AF574 basS/F⁺ (right panel). Green indicates GFP expressing cells all other cells appear red from treatment with SYTO62.