Search

Persistence of antibiotic-resistant Salmonella in the food chain may depend on strain tolerance to other antimicrobials and also on biofilm formation capacity. Yet, there is limited information on sensitivity of antibiotic-resistant Salmonella to other antimicrobials, such as phenolic compounds, chemical preservatives, or antimicrobial peptides. This study aimed at correlating antimicrobial resistance and biofilm formation capacity in antibiotic-resistant, biocide-tolerant Salmonella strains from hen eggshells. A collection of 21 strains previously selected according to their antibiotic resistance and biocide tolerance phenotypes were used for the present study. Strains were inspected for their biofilm formation capacity and for their sensitivity to (i) phenolic compounds (carvacrol, thymol), (ii) chemical preservatives (sodium lactate, trisodium phosphate), and (iii) cationic antimicrobials (polymyxin B, lysozyme-EDTA). Biofilm formation capacity was not correlated with antimicrobial resistances of the planktonic Salmonella. Polymyxin B and the lysozyme-EDTA combinations showed significant ( P < 0.05) positive correlations to each other and to sodium lactate. Significant ( P < 0.05) positive correlations were also observed for benzalkonium chloride and cetrimide with carvacrol, thymol, and trisodium phosphate, or between hexadecylpyridinium chloride and carvacrol. Antibiotic resistance also correlated positively with a higher tolerance to other antimicrobials (cefotaxime, ceftazidime, and ciprofloxacin with carvacrol, thymol, and trisodium phosphate; netilmicin with thymol and trisodium phosphate; tetracycline with carvacrol and thymol). These results must be taken into consideration to ensure a proper use of antimicrobials in the poultry industry, at concentrations that do not allow coselection of biocide-tolerant, antibiotic-resistant Salmonella.

The continued agricultural exposure of bacteria to metals such as copper and zinc may result in an increased copper tolerance through the food chain. The aim of this study was to determine the Cu and Zn tolerance of bacteria from fresh produce (cucumber, zucchini, green pepper, tomato, lettuce, vegetable salad, broccoli, cabbage, carrot, green onion, onion, and mango). Isolates (506 aerobic mesophiles) from 12 different food produce products were tested for growth in a range of Cu and Zn concentrations. Selected isolates were identified using 16S rDNA sequencing, and the presence of metal resistance genes was studied using PCR amplification. More than 50% of the isolates had MICs for copper sulfate greater than 16 mM, and more than 40% had MICs greater than 4 mM for zinc chloride. Isolates with high levels of tolerance to Cu and Zn were detected in all the produce products investigated. A selection of 51 isolates with high MICs for both Cu and Zn were identified as belonging to the genera Pseudomonas (28), Enterobacter (7), Serratia (4), Leclercia (1), Bacillus (10), and Paenibacillus (1). A study of the genetic determinants of resistance in the selected gram-negative isolates revealed a high incidence of genes from the pco multicopper oxidase cluster, from the sil cluster involved in Cu and silver resistance, and from the chromate resistance gene chrB. A high percentage carried both pco and sil. The results suggest that Cu and Zn tolerance, as well as metal resistance genes, is widespread in bacteria from fresh produce.

The cyclic peptide bacteriocin enterocin AS-48 was tested (at final concentrations of 0.175, 0.613, and 1.05 AU/ml) against the exopolysaccharide-producing cider spoilage strain Pediococcus parvulus 48 in apple juice in combination with high-intensity pulsed electric field (HIPEF) treatment (35 kV/cm and 150 Hz for 4 mus and bipolar mode). The effect of the combined treatments was studied by surface response methodology, with AS-48 concentration and HIPEF treatment time as process variables. A bacteriocin concentration of 0.613 AU/ml in combination with HIPEF treatment time of 1,000 micros reduced the population of pediococci by 6.6 log cycles in apple juice and yielded an apple juice that was free from pediococci during a 30-day storage period at 4 and 22 degrees Celsius. In contrast, application of HIPEF treatment alone had no effect on the surviving pediococci during storage of juice at 22 degrees Celsius. The combined treatment significantly improved the stability of the juice against spoilage by exopolysaccharide-producing P. parvulus.

Enterocin AS-48 was tested against Staphylococcus aureus, Bacillus cereus, and Listeria monocytogenes in different kinds of desserts. The highest activity against S. aureus was detected in baker cream. However, in yogurt-type soy-based desserts and in gelatin pudding, AS-48 (175 arbitrary units [AU]/g) reduced viable cell counts of S. aureus by only 1.5 to 1.8 log units at most. The efficacy of AS-48 in puddings greatly depended on inoculum size, and viable S. aureus counts decreased below detection levels within 24 h for inocula lower than 4 to 5.5 log CFU/g. For L. monocytogenes, bacteriocin concentrations of 52.5 to 87.5 AU/g reduced viable counts below detection levels and avoided regrowth of survivors. The lowest activity was detected in yogurt-type desserts. For B. cereus, viable cell counts were reduced below detection levels for bacteriocin concentrations of 52.5 AU/g in instant pudding without soy or by 175 AU/g in the soy pudding. In gelatin pudding, AS-48 (175 AU/g) reduced viable cell counts of B. cereus below detection levels after 8 h at 10 degrees C or after 48 h at 22 degrees C. Bacteriocin addition also inhibited gelatin liquefaction caused by the proteolytic activity of B. cereus.

A collection comprising 14 isolates of Pediococcus pentosaceus and one Pediococcus acidilactici from the fermentation of caper fruits was studied. All isolates showed very similar fermentation profiles and produced a limited number of exoenzymes. All isolates carried large plasmids of diverse sizes between 20 and 55 kb, while some also contained smaller plasmids between 10 and 16 kb. Cluster analysis of plasmid profiles revealed four main groups with various degrees of similarities. All amino acid decarboxylation tests were negative, suggesting that pediococci are not involved in generation of biogenic amines. None of the isolates showed hemolytic activity. Antimicrobial resistance tests revealed that all isolates were sensitive to 11 different antimicrobials while being resistant to ciprofloxacin (MIC > or =2 mg/liter) and intrinsically resistant to vancomycin (MIC > or =16 mg/liter) and teicoplanin (MIC > or =16 mg/liter).

Antimicrobial resistance in enterococci is a matter of concern. A collection of 272 strains (including 107 Enterococcus faecalis and 165 Enterococcus faecium strains) isolated from meat and dairy products, seafood, vegetable foods, wildflowers, animal feces (ewe, goat, horse, mule), and hospitals were tested for sensitivity to biocides of different classes (quaternary ammonium compounds, a bisphenol, and a biguanide) and copper sulfate. Most isolates were inhibited at 25 mg of benzalkonium chloride or cetrimide per liter or at 2.5 mg of hexadecylpyridinium chloride per liter. Few isolates had MICs higher than 25 mg/liter for benzalkonium chloride (2.2%), cetrimide (0.74%), or hexadecylpyridinium chloride (0.37%), although they were all inhibited at 250 mg/liter. The population response to triclosan was very homogeneous, and most isolates (98.16%) were inhibited at 250 mg of triclosan per liter. Chlorhexidine showed the greatest variability, with MICs in a range from 2.5 to 2,500 mg/liter. Remarkably, 74.57% of isolates from clinical samples required 2,500 mg of chlorhexidine per liter for inhibition, compared to much-lower concentrations required for most isolates from other sources. Enterococci were inhibited by copper sulfate in a concentration range from 4 to 16 mM, with no bimodal distribution. However, most isolates required 12 mM (41.91%) or 16 mM (47.43%) for inhibition. The highest percentages of isolates requiring 16 mM CuSO4 were from vegetable foods, seafood, and wildflowers. The results from the present study suggest intermediate levels of copper tolerance and a low incidence of biocide tolerance in the enterococci investigated, except for chlorhexidine in clinical isolates.

The broad-spectrum bacteriocin enterocin AS-48 was tested for biopreservation of ready-to-eat vegetable foods (soups and purees) against aerobic mesophilic endospore-forming bacteria. By adding AS-48 (10 microg/ml), Bacillus cereus LWL1 was completely inhibited in all six vegetable products tested (natural vegetable cream, asparagus cream, traditional soup, homemade-style traditional soup, vegetable soup, and vichyssoise) for up to 30 days at 6, 15, and 22 degrees C. A collection of strains isolated from spoiled purees showed slightly higher resistance to AS-48 in the order Paenibacillus sp. > Bacillus macroides > B. cereus, although they were also completely inhibited in natural vegetable cream by AS-48 at 10 microg/ml. However, cocktails of five or eight strains composed of B. cereus (three strains), B. macroides (two strains), and Paenibacillus sp., Paenibacillus polymyxa, and Paenibacillus amylolyticus showed higher bacteriocin resistance with AS-48 of up to 50 microg/ml required for complete inactivation in natural vegetable cream stored at 22 degrees C. Repetitive extragenic palindromic sequence-based PCR (REP-PCR) analysis showed that paenibacilli (along with some B. cereus) was the predominant survivor in the cocktails after bacteriocin treatment. To increase the effectiveness of enterocin AS-48, the bacteriocin was tested (at 20 microg/ml) against the eight-strain cocktail in natural vegetable cream in combination with other antimicrobials. The combination of AS-48 and nisin had a slight but significant additive effect. Bactericidal activity was greatly enhanced by phenolic compounds (carvacrol, eugenol, geraniol, and hydrocinnamic acid), achieving a rapid and complete inactivation of bacilli in the tested puree at 22 degrees C.

Poto poto (a maize sourdough) and degue (a pearl millet-based food) are two traditional African fermented foods. The molecular biology of toxigenic and pathogenic bacteria found in those foods is largely unknown. The purpose of this study was to study the phylogenetic relatedness and toxigenic potential of 26 Bacillus cereus group isolates from these traditional fermented foods. The relatedness of the isolates was evaluated with repetitive element sequence-based PCR (REP-PCR) and 16S rDNA sequencing analysis. A multiplex real-time PCR assay targeting the lef and capC genes of B. anthracis pXO1 and pXO2 plasmids and the sspE chromosomal gene of B. cereus and B. anthracis also was carried out. Melting curve analysis of the sspE amplification product was used to differentiate B. cereus from B. anthracis, and the presence of the B. cereus enterotoxin genes was determined with PCR amplification. Isolates had 15 different REP-PCR profiles, according to which they could be clustered into four groups. 16S rDNA sequencing analysis identified 23 isolates as B. cereus or B. anthracis and three isolates as B. cereus or Bacillus sp. Multiplex real-time PCR amplification indicated the absence of the lef and capC genes of B. anthracis pXO 1 and pXO2 plasmids, and melting curve analysis revealed amplification of the 71-bp sspE product typical of B. cereus in all isolates instead of the 188-bp amplicon of B. anthracis, confirming the identity of these isolates as B. cereus. Four isolates had amylolytic activity. All isolates had lecithinase activity and beta-hemolytic activity. Enterotoxin production was detected in two isolates. The emetic toxin gene was not detected in any isolate. The nheB toxin gene was detected in 19 isolates by PCR amplification; one of these isolates also contained the hblD (L1) gene. The cytotoxin K cytK-1 gene was not detected, but the cytK-2 gene was clearly detected in six isolates.