Developing selective media for quantification of multispecies biofilms following antibiotic treatment

Abstract

The lungs of cystic fibrosis (CF) patients are chronically colonized by a polymicrobial biofilm community, leading to difficult-to-treat infections. To combat these infections, CF patients are commonly treated with a variety of antibiotics. Understanding the dynamics of polymicrobial community composition in response to antibiotic therapy is essential in the search for novel therapies. Culture-dependent quantification of individual bacteria from defined multispecies biofilms is frequently carried out by plating on selective media. However, the influence of the selective agents in these media on quantitative recovery before or after antibiotic treatment is often unknown. In the present study we developed selective media for six bacterial species that are frequently co-isolated from the CF lung, i.e. Pseudomonas aeruginosa, Staphylococcus aureus, Streptococcus anginosus, Achromobacter xylosoxidans, Rothia mucilaginosa, and Gemella haemolysans. We show that certain supplementations to selective media strongly influence quantitative recovery of (un)treated biofilms. Hence, the developed media were optimized for selectivity and quantitative recovery before or after treatment with antibiotics of four major classes, i.e. ceftazidime, ciprofloxacin, colistin, or tobramycin. Finally, in a proof of concept experiment the novel selective media were applied to determine the community composition of multispecies biofilms before and after treatment with tobramycin.

Fig 1. Overview of selectivity experiments.

The first column shows the intended species and an abbreviated name of the medium, the second column (O₂) shows aerobic (A) or anaerobic (AN) incubation, and the third column (t) shows incubation times. Strains belonging to the same species showed the same results. Pa: P. aeruginosa AA2, PAO1, and AA44; Sa: S. aureus SP123 and Mu50; Sag: S. anginosus LMG 14696 and LMG 14502; Ax: A. xylosoxidans LMG 26680 and LMG 14980; Rm: R. mucilaginosa DSM 20746 and ATCC 49042; Gh: G. haemolysans LMG 18984 and LMG 1068. Results are shown for all tested media: Pseudomonas Isolation Agar (PIA), LB + triclosan (LB tricl), LB + 7.5% NaCl (LB NaCl), nalidixic acid sulphamethazine agar (NAS), Mitis-salivarius agar (MSA), McConkey + 5 mg/mL xylose + 20 μg/mL vancomycin (McXV), McConkey + 5 mg/mL xylose + 20 μg/mL vancomycin (McXV) + 20 μg/mL aztreonam (McXVA), McConkey + aztreonam (McA), Rothia mucilaginosa selective medium (RMSM), nutrient agar + 5 μg/mL mupirocin + 10 μg/mL colistin sulphate (NMC), Edwards medium + 10 μg/mL colistin sulphate (Edwards + col), Columbia blood agar + 32/6.4 μg/mL co-trimoxazole (CBA + co32), Columbia agar + 32/6.4 μg/mL co-trimoxazole (CA + co32).

Fig 2. Quantification of untreated biofilms of the six bacterial species on candidate selective media compared to a general medium.

(A) P. aeruginosa AA2 (B) S. aureus SP 123 (C) S. anginosus LMG 14696 (D) A. xylosoxidans LMG 26680 (E) R. mucilaginosa DSM 20476 (F) G. haemolysans LMG 18984. PIA = Pseudomonas Isolation Agar; LB tricl = LB agar supplemented with triclosan in various concentrations (μg/mL); LB NaCl = LB supplemented with 7.5% NaCl; McA5 = McConkey agar supplemented with 5 μg/mL aztreonam; NMC = nutrient agar supplemented with 5 μg/mL mupirocin and 10 μg/mL colistin sulphate; CA + co32 = Columbia agar with 32/6.4 μg/mL co-trimoxazole. Graphs show mean recovery and standard deviations. * p < 0.05, n ≥ 3.

Fig 3. Quantification of biofilms after antibiotic treatment on selective media compared to a general medium.

(A) P. aeruginosa AA2 (B) S. aureus SP 123 (C) S. anginosus LMG 14696 (D) A. xylosoxidans LMG 26680 (E) R. mucilaginosa DSM 20476 (F) G. haemolysans LMG 18984. PIA = Pseudomonas Isolation Agar; LB tricl = LB agar supplemented with triclosan in various concentrations (μg/mL); LB NaCl = LB supplemented with 7.5% NaCl; McA5 = McConkey agar supplemented with 5 μg/mL aztreonam; NMC = nutrient agar supplemented with 5 μg/mL mupirocin and 10 μg/mL colistin sulphate; CA + co32 = Columbia agar with 32/6.4 μg/mL co-trimoxazole. Graphs show mean recovery and standard deviations. * p < 0.05, n ≥ 3.

Fig 4. Proof of concept: Quantification of multispecies biofilms.

A multispecies biofilm of the six bacterial species was grown for 24 hours and subsequently incubated with fresh medium as an untreated control or treated with 100 μg/mL tobramycin in medium for an additional 24 hours in microaerophilic conditions. Graphs show mean recovery and standard deviations. The detection limit of 10² CFU/mL is represented by a dashed line. * p < 0.05, n ≥ 3.