Biofilms and antibiotic susceptibility of multidrug-resistant bacteria from wild animals

Sample collection and bacterial isolation

Bacterial isolates were collected aseptically from freshly voided faecal samples of different species of wild animals, including birds, reptiles and mammals, at the north of Portugal (Dias et al., 2014). Collection was done as soon as the animal entered the Centre for Treatment of Wild Animals (Veterinary Hospital of the University of Trás-os-Montes and Alto Douro). MacConkey agar (Merck, Germany) was used for bacterial isolation and the plates were incubated at 37 °C for 24 h. Afterwards, the pure cultures were checked for oxidase activity (1% tetramethyl phenylenediamine; Merck, Darmstadt, Germany) before cryopreservation in Brain Heart Infusion broth (BHI; Oxoid, Cheshire, UK) with 30% (v/v) glycerol at −70 °C.

Antibiotic susceptibility testing

The antimicrobial sensitivity profile of recovered bacteria was determined using the disk diffusion assay on Muller-Hinton agar (MH; Oxoid, Cheshire, UK) according to the standards and interpretive criteria described by the Clinical and Laboratory Standards Institute guidelines (CLSI, 2014). The strains were tested for susceptibility to a panel of antibiotics (Oxoid, UK): amoxicillin (AML₁₀), amoxicillin/clavulanic acid (AMC₃₀), ticarcillin (TIC₇₅), ticarcillin/clavulanic acid (TIM₈₅), piperacillin (PRL₁₀₀), piperacillin/tazobactam (TZP₁₁₀), cephalothin (KF₃₀), cefoxitin (FOX₃₀), ceftriaxone (CRO₃₀), cefoperazone (CFP₃₀), ceftazidime (CAZ₃₀), cefotaxime (CTX₃₀), cefepime (FEP₃₀), imipenem (IPM₁₀), aztreonam (ATM₃₀), streptomycin (S₁₀), kanamycin (K₃₀), amikacin (AK₃₀), gentamicin (CN₁₀), tobramycin (TOB₁₀), nalidixic acid (NA₃₀), ciprofloxacin (CIP₅), erythromycin (E₁₅), tetracycline (TE₃₀), trimethoprim-sulfamethoxazole (SXT₂₅), chloramphenicol (C₃₀) and fosfomycin (FOS₅₀).

Identification of bacterial resistant isolates

Identification was performed by 16S rRNA gene sequence analysis. For that, general bacterial 5′ AGA GTT TGA TCA TGG CTC AG 3′ forward primer and 5′ GGT TAC CTT GTT ACG ACT T 3′ reverse primer were used to amplify nearly full-length 16S rRNA gene as described previously (Macfarlane et al., 2004; Soler et al., 2004). Sequencing analysis were performed using the BigDye Terminator V3.1 cycle sequencing kit ant the ABI Prism® 3100 Avant Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). The nucleotide sequences were edited using the software Chromas 2 and compared with published sequences in the Nacional Center for Biotechnology Information (NCBI) databases using BLAST. Phylogenetic tree was produced by the neighbor-joining method (Saitou & Nei, 1987) with Kimura’s 2-parameter method (Kimura, 1980) using the MEGA4-molecular evolutionary genetic analysis program (Tamura et al., 2007).

Minimal inhibitory concentration (MIC) determination

Standardized planktonic MIC’s of ciprofloxacin (CIP) (Sigma-Aldrich Co., Lisbon Portugal) and imipenem (IPM) (Cayman Chemical, Ann Arbor, MI, USA) were assessed by the microdilution method outlined by CLSI (2014). Overnight cell cultures (18 h incubation) were adjusted to a cell density of 1 × 10⁶ cells/ml and added to sterile 96-well polystyrene microtiter plates (Orange Scientific, USA) containing different concentrations (CIP-6, 8, 10, 14, 16, 24, 32, 36, 40, 44, 52, 56 and 60 µg/ml; IPM-2, 4, 6, 8, 10, 12 µg/ml) of each antibiotic in a final volume of 200 µl. A bacterial suspension without antibiotic was used as negative control. Microtiter plates were then incubated at 30 °C for 24 h. MIC matched to the lowest concentration in which the final optical density (OD) at 620 nm was equal or lower than the initial OD.

Screening of resistance genetic determinants: β-lactamases

Detection of genetic determinants was done for TEM-type, SHV-type, CTX-M-type, MOX and FOX variants of extended-spectrum-β-lactamases (ESBLs) and for the CphA, IMP-type, VIM-type and OXA-type carbapenemases. The mixture of PCR was performed in a final volume of 50 µl containing 5 µl of Biotools PCR buffer (10×), 200 µM of each nucleotide, 10 pmol of each primer (Table 1), 1 µl (1U) of Taq DNA polymerase (Biotools) and 1 µl of template genomic DNA (100–200 ng). The conservative oligonucleotide primers used were developed by Fontes (2009). The amplification conditions were: initial denaturation for 5 min at 95 °C, 35 amplification cycles composed of a denaturation step for 15 s at 94 °C, an annealing step at 55 °C for 30 s, except for bla_FOX, in which 60 °C were used, and an extension period at 72 °C for 45 s. The reaction was completed with an extension step at 72 °C for 3 min.

Bacterial surface hydrophobicity evaluation

Surface hydrophobicity was evaluated after contact angle measurement according to the procedure described by Simões, Simões & Vieira (2010). Lawns of each bacterium were prepared as described by Busscher et al. (1984) and their contact angles were determined by the sessile drop contact angle measurements using a model OCA 15Plus (DATAPHYSICS, Germany), with three pure liquids (water, formamide, and α-bromonaphthalene). The surface tension components of the three liquids were obtained from the literature (Janczuk et al., 1993). Afterwards, the hydrophobicity ($\Delta G_{bwb}^{TOT}$) and surface tension parameters (ϒ^AB—Lewis acid–base component, ϒ⁺—electron acceptor parameter, ϒ⁻—electron donor parameter) of the strains were evaluated according to the method of Van Oss et al. (1989).

Initial adhesion assay

In vitro initial adhesion was determined according to the method of Simões et al. (2007) and Simões, Simões & Vieira (2010). Sterile microtiter plates (12-well) with 2 ml of cell suspension (1 × 10⁸ cells/ml in MH) and coupons (dimensions of 1 × 1 cm) of polystyrene (PS) horizontally placed were incubated at 30 °C for 2 h in an orbital shaker at 150 rpm. PS coupons were prepared by successive immersions in a solution (5% v/v) of commercial detergent (Sonasol Pril; Henkel Ibérica S.A, Barcelona, Spain) and ultrapure water with gentle shaking (30 min). All experiments were performed in triplicate, with three independent repeats. After 2 h incubation the coupons were removed and washed in 2 ml of sterile saline solution (0.85% w/v) and the biomass was quantified by crystal violet (CV) staining (Borges et al., 2017).

Biofilm formation

Biofilm formation was performed in 96-well polystyrene microplates following the method of Stepanović et al. (2000). Briefly, overnight cultures were adjusted to an initial OD (620 nm) of 1 × 10⁸ cells/ml in MH and 200 µl aliquots were added to the microplate. The microtiter plate was incubated for 24, 48 and 72 h at 30 °C and agitated at 150 rpm. The medium (MH) was carefully and aseptically replaced daily by fresh one, after a washing step with saline solution (0.85% v/v). Negative controls contained MH broth without bacteria.

Biofilm control

To determine whether antibiotics had capability to remove 24 h old biofilms, CIP and IPM at MIC, 5 × MIC and 10 × MIC were used according to Borges et al. (2017). After 24 h exposure at 30 °C the antibiotic solutions were discarded and the biofilms were analyzed in terms of biomass by CV staining. The results are presented as percentage of biofilm mass removal when compared to biofilms non-exposed to antibiotics. The assays were performed in triplicate with three independent repeats.

Mass quantification of adhered cells and biofilms

The mass of adhered cells and biofilm was quantified by CV staining according to Borges et al. (2017). The OD was measured at 570 nm using a Microplate Reader (BIO-TEK, Model Synergy HT). Biofilm mass removal percentage (%BR) was given by the following equation: ${\displaystyle \text{\%}\mathrm{BR}=\frac{\mathrm{ODn}-\mathrm{ODw}}{\mathrm{ODn}}\times 100.}$

Where the ODn are the absorbance values of the biofilms non-exposed to the antibiotics (CIP and IPM) and ODw are the absorbance values of the biofilms exposed to antibiotics (CIP and IPM).

Adherent/biofilm bacteria classification

Based on the cutoff of the ODc quantification (value defined as three standard deviation values above the mean OD of the negative control) the bacteria were classified according to Stepanović et al. (2000) as: non-adherent/non-biofilm producer (0)-OD ≤ ODc; weakly adherent/weak biofilm producer (+)-ODc < OD ≤ 2 × ODc; moderately adherent/moderate biofilm producer (++)-2 × ODc < OD ≤ 4 × ODc; strongly adherent/strong biofilm producer (+++)-4 × ODc < OD.

Statistical analysis

Data were analyzed by analysis of variance (ANOVA) using the statistical program SPSS version 22.0 (SPSS, Inc., Chicago, IL, USA) and statistical calculations were based on a confidence level ≥ 95% (P < 0.05 was considered statistically significant).

Bacteria and antibiotic susceptibility profile

The selected bacteria belong to different genera: Acinetobacter, Klebsiella, Pseudomonas and Shewanella. S. putrefaciens was obtained from a sample of red deer (Cervus elaphus), P. fluorescens from a red fox (Vulpes vulpes) and K. pneumoniae and Acinetobacter spp. were isolated from the same sample of a greater rhea (Rhea americana). These four strains were identified through 16S rDNA sequencing and were selected due to their antibiotic resistance profiles. Acinetobacter spp. was only identified at the genus level. According to the analysis of the phylogenetic tree generated (See Fig. S1), this strain can represent a new species.

The susceptibility profile against 27 antibiotics belonging to different classes was assessed. All bacteria were multidrug-resistant, i.e., resistant to at least two antibiotics belonging to different chemical classes (Godebo, Kibru & Tassew, 2013; Exner et al., 2017) (Table 2). It was possible to observe that Acinetobacter spp. and K. pneumoniae, showed identical resistance profiles. This could be related to the fact that these bacteria were isolated from the same sample, indicating an exchange of resistance genes. Among all the isolates, S. putrefaciens was the most susceptible to the antibiotics tested. All bacteria were resistant to amoxicillin, amoxicillin/clavulanic acid, cephalothin, fosfomycin and erythromycin. Of particular concern was the resistance of all the selected bacteria to the carbapenem imipenem (IPM), an intravenous broad-spectrum antibiotic of the β-lactams group that is of exclusive hospital use (Papp-Wallace et al., 2011).

Genetic bases of antibiotic resistance

The presence of 12 β-lactamase genes (bla_CphA; bla_OXA-aer; bla_OXA-B; bla_OXA-C; bla_MOX; bla_TEM; bla_FOX; bla_SHV; bla_SHV; bla_CTX-M; bla_IMP and bla_VIM) in the selected bacteria was analyzed by PCR. The genotype results are summarized in Table 2. The bla_OXA-aer was detected in P. fluorescens and K. pneumoniae, isolated from faecal samples of a fox and a rhea, respectively. Three β-lactamases genes (bla_OXA-aer; bla_TEM; bla_VIM) were detected in the same bacterium, K. pneumoniae. S. putrefaciens and Acinetobacter spp. were resistant to IPM but apparently did not display any of the genetic determinants of IPM resistance studied in this work. No PCR specific bla_CphA; bla_OXA-B; bla_OXA-C; bla_MOX; bla_FOX; bla_SHV; bla_CTX-M; bla_IMP and bla_VIM encoding sequences were detected.

Minimal inhibitory concentration of ciprofloxacin and imipenem

K. pneumoniae was the bacterium with lowest susceptibility to the action of CIP with MIC of 60 µg/ml (Table 3). S. putrefaciens was the most resistant to IPM with MIC of 12 µg/ml. In general, IPM was the most efficient antibiotic against the four bacteria tested (MICs: 12 µg/ml for S. putrefaciens; 6 µg/ml for P. fluorescens, 6 µg/ml for K. pneumonia; 2 µg/ml for Acinetobacter spp.). According to the CLSI (2014) guidelines all isolates are considered resistant to CIP (MICs > 4 µg/ml). K. pneumonia , P. fluorescens and S. putrefaciens are resistant to IPM (MICs > 4 µg/ml). Acinetobacter spp. is classified as intermediate (MIC = 2 µg/ml).

Bacterial surface hydrophobicity

The four bacteria had hydrophilic surfaces ($\Delta G_{bwb}^{TOT}$), being their values significantly different ( P < 0.05) (Table 4). Also, all the bacteria were predominantly electron donors (ϒ > ϒ⁺) and K. pneumonia and Acinetobacter sp. cell surfaces had similar electron donor abilities (P > 0.05). For P. fluorescens no electron accepting character (ϒ⁺ = 0 mJ/m²) was observed. S. putrefaciens had the highest value of polar component (ϒ^AB) (P > 0.05).

Initial monolayer adhesion

The degree of bacterial attachment was found to follow the sequence: P. fluorescens > S. putrefaciens > K. pneumoniae > Acinetobacter spp. (Fig. 1). S. putrefaciens, K. pneumoniae and Acinetobacter spp. were classified as moderately adherent while P. fluorescens was classified as strongly adherent (Table 5). P. fluorescens and Acinetobacter spp. had the highest and lowest adhesion abilities, respectively. K. pneumoniae and Acinetobacter spp. adhered with comparable extents (P > 0.05).

Biofilm formation

The isolates were studied for their biofilm-forming ability in 96-well polystyrene microtiter plates during 24, 48 and 72 h. Figure 2 shows that all bacteria were able to form biofilms in all the sampling times. P. fluorescens produced the smallest biomass amount for the three sampling times, while on the other hand K. pneumoniae produced the highest biomass amount. Biofilm production was directly proportional to the biofilm age for Acinetobacter spp. and S. putrefaciens. Regarding P. fluorescens, the biomass decreased from 24 to 48 h and remained constant for the 72 h old biofilms.

K. pneumoniae developed weak biofilms during the first 24 h. The biofilm formation ability increased at 48 h and decreased at 72 h. It was found that the degree of biofilm formation followed the sequence: 24 h old biofilms—K. pneumoniae > Acinetobacter spp. > S. putrefaciens = P. fluorescens; 48 h old biofilms K. pneumoniae > Acinetobacter spp. = S. putrefaciens > P. fluorescens and 72 h old biofilms—K. pneumoniae > Acinetobacter spp. = S. putrefaciens = P. fluorescens. According to the classification proposed by Stepanović et al. (2000) concerning the bacterial biofilm production ability (Table 5), P. fluorescens and Acinetobacter spp. showed weak and moderate biofilm production for the various sampling times, respectively. S. putrefaciens presented weak biofilm formation ability at 24 h and moderate biofilm formation ability at 48 and 72 h. K. pneumoniae demonstrated moderate ability to form biofilms at 24 h and strong biofilm ability at 48 and 72 h.

Effect of ciprofloxacin and imipenem on biofilm removal

The effects of the IPM and CIP were tested on the removal of 24 h old biofilms of S. putrefaciens, K. pneumoniae, P. fluorescens and Acinetobacter spp. (Fig. 3). In general, the increase of CIP or IPM concentrations increased biofilm removal. A significant increase of the P. fluorescens biomass reduction occurred with the increase of CIP from MIC to 5 × MIC (P < 0.05). However, the increase to 10 × MIC did not produce significant improvement in biofilm removal (P > 0.05). Similar results were observed for K. pneumoniae biofilms when exposed to IPM. For S. putrefaciens the increase of CIP concentration from MIC to 5 × MIC did not increase significantly its efficiency (Fig. 3). Moreover, for this bacterium 5 × MIC and 10 × MIC caused similar biofilm mass reductions (36%). No biofilm mass removal was obtained with IPM at MIC against S. putrefaciens, P. fluorescens and Acinetobacter spp., and with CIP at MIC and 5 × MIC against Acinetobacter spp. IPM at 10 × MIC caused the highest reductions of S. putrefaciens (36% removal) and K. pneumoniae (54% removal) biofilms. CIP at 10 × MIC caused the highest biomass reductions of P. fluorescens biofilms (40%). Significant reduction was also observed for K. pneumoniae biofilms upon the application of CIP at 10 × MIC. Acinetobacter spp. biofilms were highly difficult to remove using CIP and IPM (a maximum removal of 10% was obtained with IPM at 10 × MIC). In fact, Acinetobacter spp. biofilms were the most resistant to removal by CIP while P. fluorescens biofilms were the most resistant to IPM (P < 0.05).