Antibiotic susceptibility of bacterial isolates from probiotic products available in Italy

Objective: An emerging issue of probiotic products is the antibiotic resistance of the strains used. The aim of this study was to determine the susceptibility of the isolates of 10 probiotic products available in Italy. Materials and methods: The susceptibility of 15 strains of Lactobacillus spp., 5 Streptococcus salivarius ssp. thermophilus, 1 Enterococcus faecium and 8 Bifidobacterium spp. to several groups of antibacterial agents was determined by E-test using MRS agar for Lactobacillus spp. and E. faecium and MH agar+5% sheep blood for S. thermophilus, with different conditions of incubation. For Lactobacillus, S. thermophilus and E. faecium the MICs obtained by E-test were compared to the MICs by broth microdilution test obtained following CLSI M45-A (2006) and CLSI M100-S17 (2007) guidelines. The broth microdilution test resulted in MICs identical to those obtained with the E-test or in MICs with differences of 1 or 2 log dilution steps. All the strains of Lactobacillus were susceptible to ampicillin. Species-dependent antibiotic susceptibility was detected for cephalosporins; gentamicin and ciprofloxacin had variable activity. Intrinsic resistance to vancomycin was confirmed for L. paracasei, L. salivarius and L. plantarum. Atypical resistance to erythromycin was detected in one strain of L. salivarius. The strains of Bifidobacterium were susceptible to ampicillin, cefotaxime and erythromycin. The strains of E. faecium were susceptible to the tested antibiotics; the strain of S. thermophilus was resistant only to ciprofloxacin. The observed resistance in the strains used in the Italian probiotic products tested seemed to be intrinsic except for erythromycin in one L. salivarius strain.


Introduction
Routine antibiotic susceptibility testing of lactic acid bacteria (LAB) and bifidobacteria may be advisable in a number of instances, e.g. for checking the biosafety of potential probiotic isolates (1Á3).
In fact, there is concern over the possible spread of antibiotic resistance determinants from bacteria used in probiotic products. However, there is still a lack of agreement on the minimum inhibitory concentration (MIC) interpretative breakpoints, mainly for Lactobacillus spp. (1Á4). The Panel on Additives and Products or substances used in Animal Feed (FEE-DAP) (5) recently proposed microbiological breakpoints that can be relevant to identify strains with acquired and potentially transferable antibiotic resistance in microbial feed additives. However, the results provided by different methods cannot be compared because they are influenced by test media, growth conditions and test assay. Moreover, another aspect of probiotic bacteria is that under certain circumstances they can cause infections in humans. A new CLSI document, 'Methods for Antimicrobial Dilution and Disk Susceptibility Testing of Infrequently Isolated or Fastidious Bacteria', was published in 2006 as approved guidelines (M45/A) (6) and includes Lactobacillus spp., Leuconostoc spp. and Pediococcus spp. The document suggests broth microdilution MIC tests for these genera, with cationadjusted Mueller Hinton broth (CAMHB-LHB) supplemented with 2.5Á5% lysed horse blood and MIC interpretative criteria. The aim of the study was to determine the susceptibility of isolates from probiotic products available in Italy.

Materials and methods
The MICs of several groups of antimicrobial agents were determined against 15 strains of Lactobacillus spp., 5 Streptococcus thermophilus, 1 Enterococcus      (8).
For the antimicrobial susceptibility of Bifidobacterium spp. (CLSI guidelines are not currently available) the MICs were obtained by the E-test on LSM (lactic acid bacteria susceptibility test medium) (9) supplemented with cysteine (anaerobic incubation at 378C for 48 h).

Results and discussion
The broth microdilution test resulted in MICs identical to those obtained with the E-test (only one value reported) or in MICs with differences of plus 1 or 2 log dilution steps (Table I).
Intrinsic resistance to vancomycin was confirmed for L. paracasei, L. salivarius and L. plantarum (MIC ]32 mg/L). The other species had MICs 54 mg/L, that is the microbiological and clinical breakpoint.
All the strains of Lactobacillus were b-lactamasenegative and susceptible to ampicillin by both breakpoints used. Species-dependent antibiotic susceptibility was detected for the cephalosporins tested; a wide range of MICs was observed, mainly in L. acidophilus. A typical resistance to erythromycin was detected for one strain of L. salivarius according to CLSI and FEEDAP breakpoints (MIC ]8 mg/L). This strain was also resistant to other antibiotics and more studies are needed to determine if the resistance is intrinsic or transferable. According to both breakpoints, resistance to gentamicin (MIC ]16 mg/L) was present in strains of L. salivarius, L. acidophilus and L. paracasei. For Lactobacillus spp. ciprofloxacin had variable activity. The strain of E. faecium was susceptible to the tested antibiotics; the strain of S. thermophilus was resistant only to ciprofloxacin (Table I).
The strain of Bifidobacterium spp. had MICs for vancomycin, ampicillin, cefotaxime and erythromycin 52 mg/L. The MICs of gentamicin were ]8 mg/ L (Table II).
On the basis of microbiological (5) and clinical breakpoints (6), the observed resistances in the tested strains, genus-or species-dependent, seemed to be intrinsic, except for erythromycin in one L. salivarius strain.
In conclusion, MICs obtained following the CLSI guidelines (2006) (6) for Lactobacillus may give new data to better define the cut-off values for separating strains with acquired resistance from susceptible strains (3,4). Moreover, these results confirm that the E-test on LSM plus cysteine is an applicable technique for susceptibility testing of bifidobacteria and generally for testing individual strains, e.g. by a probiotic culture producer (9).

Declaration of interest:
The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.