Interactions between bacterial strains isolated from vacuum-packaged Australian beef primals

Vacuum-packaged (VP) beef produced and packaged in Australia can have an unusually long shelf-life. This observation has been attributed, in part, to superior abattoir hygiene, but there is an absence of robust scientific evidence to support this claim. While plant hygiene may be an important factor in extending VP beef shelf-life, there are likely other contributing factors. Bacteria rarely exist in isolation and occur as members of a microbial niche. Numerous published studies have described the composition of bacterial species within food, including the influence of the environment; however, limited attention has been given to understanding how bacteria interact within foods, and how this contributes to the overall formation of a microbial community. The aim of the present study was to define how specific environmental factors, relevant to Australian VP beef, influence the interactions among bacteria. Interactions among representative species of bacteria isolated from Australian VP beef primals were investigated. From a set of strains, 39 isolates inhibiting (effectors) other bacteria, and 20 isolates actively inhibited (targets) by effectors, were selected for further study. These isolates represented ten bacterial genera; including Carnobacterium, Pseudomonas, Hafnia, Serratia, Yersinia, Rahnella, Brochothrix, Bacillus, Leuconostoc, and Staphylococcus. A number of interactions were observed, with 28.6% inhibiting and 4.2% promoting target isolate growth. All lactic acid bacteria (LAB) inhibited other species, especially Carnobacterium maltaromaticum, which inhibited the growth of a wider range of target isolates, compared to other LAB. The majority of Pseudomonas isolates antagonised growth of approximately one-half of target isolates. Two Bacillus spp. each inhibited the growth of 16 target bacteria. The majority of effector isolates that enhanced target isolate growth were Gram-negative bacteria, including Pseudomonas spp. and Enterobacteriaceae. The mechanisms of interactions were partially characterised for eight effector-target isolate combinations. The inhibitory effects of two isolates of C. maltaromaticum and one isolate of Bacillus subtilis was mediated by heat-stable, pH-stable, proteinaceous substances found in cell-free supernatants (CFS). In contrast, live cells were required for the inhibitory activity of three isolates of Bacillus sp., Pseudomonas putida, and a Pseudomonas sp. against corresponding isolates of Yersinia enterocolitica, C. maltaromaticum and B. subtilis, yet this inhibitory effect did not require direct contact between effector and target cells. Compounds produced by B. subtilis and Serratia sp. that promoted the growth of Pseudomonas lundensis were non-proteinaceous and were heat- and pH-stable. The next phases of the thesis quantified the effect of simulated intrinsic VP beef factors and associated extrinsic storage conditions (i.e. pH, temperature, atmosphere, glucose, and lactic acid) on: 1) the sensitivity of target isolate C. maltaromaticum strain D8c to inhibition caused by effector strain C. maltaromaticum D0h and, 2) the production of inhibitory factor(s) by C. maltaromaticum D0h. In the former study, all five environmental factors significantly (P < 0.05) affected sensitivity of C. maltaromaticum D8c to D0h CFS inhibitory activity. Inhibition sensitivity was relatively higher at low pH (5.5), at higher concentrations of glucose (5.55 mM) and lactic acid (50 mM), and under aerobic conditions. The sensitivity of strain D8c did not correlate linearly with temperature; since inhibition was greatest at 15°C, followed by 7, -1, and 25°C. Preliminary models were produced to describe D8c sensitivity. Furthermore, the influence of pH, atmosphere, glucose, and lactic acid on production of inhibitory compounds by D0h was studied at 25°C. It was found pH produced the greatest influence on inhibitor production, compared to atmosphere, glucose, and lactic acid. The lowest amount of inhibitor was produced at an initial medium pH of 5.5. Lactic acid significantly reduced production, but only at an initial pH of 5.5. A two-factor interaction was observed between glucose and pH; relatively high concentrations of glucose (5.55 mM) enhanced the production at pH 6.5, whereas production decreased at pH 5.5. Atmosphere did not significantly affect inhibitory activity. In conclusion, numerous interactions among the bacterial community of VP beef were described, and potentially drive formation of the microbial spoilage community, as influenced by environment. Intraspecific interactions between two C. maltaromaticum isolates were significantly affected by pH, atmosphere, lactic acid, glucose, and temperature. These findings, and the resulting models, may improve the understanding of putative interactions among spoilage bacteria in meat, in particular C. maltaromaticum, one of the most dominant bacterial species on chilled VP beef.