Introduction

Campylobacter spp., particularly C. jejuni and C. coli, are leading causes of foodborne illnesses worldwide, such that the European Food Safety Authority and the European Centre for Disease Prevention and Control reported campylobacteriosis as the most prevalent zoonosis in 2016 (EFSA and ECDC, 2018). Similarly, 1.3 million cases of Campylobacter-associated illnesses have been reported to occur annually in the United States (Geissler et al., 2017). The consumption and/or handling of raw or inappropriately cooked poultry products are considered major sources of campylobacteriosis (Edwards et al., 2011).

Currently, the presence or absence of Campylobacter spp. in poultry meat samples has predominantly been determined using modified charcoal-cefoperazone-deoxycholate agar (mCCDA) (Chon et al., 2016a). This medium contains third-generation antibiotics, including cephalosporin and cefoperazone, which selectively isolate Campylobacter spp. However, the routine use of antibiotics in poultry farming has encouraged bacteria to develop resistance to broad-spectrum antibiotics (such as cephalosporins) by producing extended-spectrum β-lactamase (ESBL). This enzyme hydrolyzes essential molecular structures within antibiotics (Kim et al., 2016), thus inactivating them. As a result, ESBL-producing bacteria overgrow on mCCDA plates, and thereby mask the growth, and confound the detection, of Campylobacter colonies derived from poultry samples (Chon et al., 2016a; 2016b).

To date, various ESBL inhibitors have been added to mCCDA to improve its performance by suppressing the growth of ESBL-producing competing flora such as Escherichia coli (Chon et al., 2013; Chon et al., 2016a; 2016b). While some of these inhibitor-supplemented media enable Campylobacter spp. to be detected with increased selectivity and sensitivity, they are not suitable for commercialization and/or application in either research or industrial contexts due to their high cost and/or complex production requirements. Moreover, bacteria resistant to previously developed ESBL-inhibitor/antibiotic combinations have now been identified; thus, novel ESBL inhibitors are urgently needed (Fernández-Martínez et al., 2015; Sharma et al., 2016; Stapleton et al., 1995).

Avibactam is an ESBL inhibitor that was developed in 2011, and has been shown in pure culture models and clinical trials to be capable of neutralizing a wider spectrum of β-lactamase-producing bacteria, even at low concentrations, compared to other ESBL inhibitors (e.g., clavulanic acid, tazobactam, and sulbactam) (Abboud et al., 2016; Ehmann et al., 2012). To the best of our knowledge, it has not yet been evaluated as a selective agent for the isolation of Campylobacter using culture media. Thus, the present study supplemented mCCDA with avibactam (A-mCCDA), and compared the performance of the supplemented media with that of standard (non-supplemented) mCCDA in detecting Campylobacter spp., and inhibiting the growth of competing flora derived from whole-chicken carcass-rinse samples.

Materials and methods

Bacterial strains

In total, 25 Campylobacter strains, comprising 11 C. jejuni (NCTC 11168, ATCC 33560; two human, and seven food isolates), and 14 C. coli (ATCC 33559; three human, and 10 food isolates), were used in this study. All human isolates were kindly provided by the Korea Centers for Disease Control and Prevention (KCDC; Cheongju, South Korea). Food isolates were collected in our laboratory (between 2014 and 2016) from meat products. Bacteria were grown at 42 °C for 48 h under microaerobic conditions (10% CO2, 5% O2, and 85% N2) on Columbia blood agar (Oxoid, Hampshire, UK) that was supplemented with 5% (v/v) laked horse blood (Oxoid). Microaerobic conditions were achieved using an AnaeroPack jar (Mitsubishi Gas Chemical, Tokyo, Japan), with a GENbox microaer gas generator (bioMérieux, Marcy-l’Étoile, France). The 25 utilized ESBL-producing E. coli strains were isolated from chicken carcass samples, and incubated at 37 °C on tryptic soy agar (Oxoid) for 24 h prior to use.

mCCDA and A-mCCDA preparation and inoculation

Control mCCDA (Oxoid) plates were prepared according to the manufacturer’s instructions as described in our previous study (Chon et al., 2016a). To determine the optimal concentration of avibactam (Avention, Incheon, South Korea) required for mCCDA supplementation to both inhibit the growth of ESBL-producing E. coli, and support the recovery of Campylobacter spp., avibactam solutions prepared with various concentrations were added into a 20 ml of mCCDA agar plate to final concentrations of 0.0625, 0.125, 0.25, 0.5, 1, 2, 4, 8, and 16 ug/mL as described in our previous study (Chon et al., 2016b). Once prepared, mCCDA and A-mCCDA plates were inoculated with approximately 1 × 104 cells each of the Campylobacter and ESBL-producing E. coli strains (suspended in phosphate-buffered saline), and incubated at 42 °C for 48 h, as described above.

Isolation of Campylobacter spp. from chicken-carcass rinses inoculated on mCCDA and A-mCCDA plates

Eighty whole chicken carcasses were purchased between May and September 2017 from four different retail stores (20 samples/store) located in Seoul, South Korea. Campylobacter spp. were detected in chicken samples as previously described (Chon et al., 2016b). Briefly, chicken carcasses were rinsed with 400 mL of buffered peptone water (Oxoid), and shaken gently for 1 min. A 25-mL aliquot of each rinse was supplemented with 2× blood-free Bolton enrichment broth (Oxoid) in a 50-mL cell-culture flask with a filter cap (SPL Life Sciences, Gyeonggi-do, South Korea), and incubated at 42 °C for 48 h under microaerobic conditions. A loopful of the rinse-inoculated enrichment broth was then inoculated onto mCCDA or A-mCCDA (0.0625 mg/L) plates, which were then incubated (42 °C for 48 h) under microaerobic conditions. Suspected Campylobacter colonies were selected and sub-cultured on Columbia blood agar, before their identities were confirmed via a previously described polymerase chain reaction (PCR) method (Denis et al., 1999).

Identification of competing flora

To identify the predominant competing flora on each medium, 40 suspected non-Campylobacter colonies (i.e., 20 colonies each isolated from mCCDA and A-mCCDA plates) were randomly selected, and identified via biochemical-identification and ESBL-production analyses that were conducted using a VITEK® 2 GN ID Card and VITEK® antimicrobial susceptibility testing (AST)-N224 Card (bioMérieux), respectively, according to the manufacturers’ instructions.

Statistical data analysis

The number of plates found to be positive for Campylobacter and non-Campylobacter colonies was compared using Fisher’s exact test, which was performed using GraphPad InStat 3 software (GraphPad Software, San Diego, CA, USA). As previously described (Chon et al., 2016b), a growth index was determined and analyzed only for media that were contaminated with non-Campylobacter colonies. A score of 1 was used to indicate the growth of a small number of colonies, while scores of 2 and 3 were used to indicate the growth of colonies on approximately half, and most of the plate, respectively. A P value < 0.05 was considered to indicate statistical significance.

Results and discussion

Optimal avibactam concentration required to produce effective A-mCCDA

Growth data for the 25 Campylobacter (11 C. jejuni, and 14 C. coli) and 25 ESBL-producing E. coli isolates that were grown on the mCCDA and A-mCCDA plates (at various concentrations of avibactam) are provided in Table 1.

Table 1 Comparison of growth of 25 Campylobacter and 25 ESBL-producing Escherichia coli strains on A-mCCDA plates containing various concentrations of avibactam
Full size table

None of the tested Campylobacter strains were inhibited on A-mCCDA containing 0.0625–4 mg/L of avibactam; however, they were inhibited on A-mCCDA containing ≥ 8 mg/L of avibactam, indicating that high concentrations of avibactam prevent Campylobacter growth (Table 1). Notably, the 14 C. coli strains analyzed survived on A-mCCDA supplemented with avibactam at all concentrations, while two and one C. jejuni strains survived on A-mCCDA supplemented with 8 mg/L and 16 mg/L, respectively (data not shown). The observed higher level of avibactam resistance exhibited by the C. coli compared to that exhibited by the C. jejuni strains is consistent with the results of previous studies, in which C. coli strains generally exhibited low levels of antibiotic-binding activity (Pezzotti et al., 2003; Tajada et al., 1996).

In contrast, all tested ESBL-producing E. coli strains were inhibited on A-mCCDA containing even the lowest concentration of avibactam (0.0625 mg/L). Thus, to minimize the potential inhibitory effect of avibactam against Campylobacter, this concentration of avibactam was determined to be optimal for Campylobacter screening.

In terms of cost-effectiveness, the supplementation of mCCDA with avibactam was estimated to be superior or comparable to supplementation with other ESBL inhibitors (Table 2). In fact, the calculated cost of supplementing mCCDA with a sufficient concentration of an alternative ESBL inhibitor to suppress the growth of ESBL-producing E. coli (Chon et al., 2013; 2016a; 2016b) was shown to be much higher than that of supplementing mCCDA with avibactam, with the exception of clavulanic acid. In addition, A-mCCDA displayed the lowest growth index of 1.6 (Table 2) of the selective agars analyzed, indicating that it exhibited the lowest level of contamination by competing flora. Together, these findings suggest that A-mCCDA is more appropriate for use in commercial settings than previously reported ESBL inhibitors.

Table 2 Comparison between cost and performance of supplementing mCCDA with ESBL inhibitors to suppress the growth of ESBL-producing Escherichia coli
Full size table

Capacity of mCCDA and A-mCCDA to enable the selective isolation of Campylobacter from chicken-carcass rinses

The number of A-mCCDA plates that were positive for Campylobacter growth (10/80; 12.5%) was significantly higher than that of mCCDA plates (0/80; 0%)(P < 0.05), and furthermore, significantly fewer A-mCCDA (32/80; 40%) than mCCDA (80/80; 100%) plates were contaminated with non-Campylobacter colonies (P < 0.05), indicating that the selectivity of A-mCCDA was superior to that of mCCDA. In addition, the growth index (1.62) of the non-Campylobacter spp. that were observed on A-mCCDA was approximately two times lower than that observed on mCCDA (2.98). As a result, suspected Campylobacter colonies were covered by overgrown ESBL-producing E. coli colonies on the mCCDA agar plates, whereas only Campylobacter colonies were detected on the A-mCCDA plates, allowing single Campylobacter colonies to be easily identified and isolated (Table 3, Fig. 1).

Table 3 Comparison of Campylobacter and non-Campylobacter colony growth on mCCDA and A-mCCDA plates, as indicated by the number of positive plates and observed growth indices
Full size table
Fig. 1
figure 1

Representative images showing bacterial growth on modified-Charcoal-Cefoperazone-Deoxycholate-Agar (mCCDA) and avibactam-supplemented mCCDA (A-mCCDA) plates inoculated with the same whole-chicken carcass rinse sample. (A) Both non-Campylobacter colonies (white arrows) and suspected Campylobacter colonies (gray arrows) were observed on the mCCDA plates. (B) No non-Campylobacter colonies were detected on the A-mCCDA plates; thus, suspected Campylobacter colonies were easily identified

Full size image

A limitation of this study is that the performance of A-mCCDA was not compared to those of other previously developed ESBL inhibitor (e.g., clavulanic acid or tazobactam)-supplemented mCCDAs. Moreover, given that the experiments were not performed simultaneously, direct comparisons between A-mCCDA and other previously developed and investigated ESBL-supplemented mCCDAs may be inaccurate. Nevertheless, this is the first study to evaluate avibactam as a selective agent for the isolation of Campylobacter. Further studies are required to simultaneously compare the performance of additionally modified Campylobacter-selection media in assessing various clinical, environmental, and food samples.

In conclusion, A-mCCDA enabled the highly selective identification and isolation of Campylobacter spp. compared to mCCDA, as demonstrated by the fact that the A-mCCDA plates analyzed exhibited an approximately 2.0- and 2.5-fold lower non-Campylobacter contamination rate and growth index, respectively, than the mCCDA plates. Thus, by eliminating masking colonies, the application of A-mCCDA may enable the prevalence of Campylobacter in raw poultry to be more accurately determined, especially in samples contaminated with ESBL-producing E. coli. Furthermore, supplementing mCCDA with avibactam rather than clavulanic acid or tazobactam has the advantage of achieving a wide antibiotic spectrum at an optimal concentration, while incurring lower or comparable costs.