Quantitative description of interactions among a meat bacterial spoilage community

Spoilage bacteria are a primary factor affecting the shelf-life of vacuum-packaged (VP) beef. Common bacterial species that dominate microbial communities of VP beef include Carnobacterium maltaromaticum, Brochothrix thermosphacta, and Serratia liquefaciens. However, a quantitative description of how these species establish and dynamically interact within a beef spoilage community, under commercial VP conditions, has yet to be reported. The pH, lactic acid, and glucose concentrations of beef can markedly impact bacterial growth. The effects of these combined factors on individual growth kinetics of C. maltaromaticum, B. thermosphacta, and S. liquefaciens were studied in a simulated beef matrix (modified brain heart infusion broth; mBHI broth) within a commercial VP heatshrunk film. pH (5.5, 6.5), lactic acid (50 mM), and undissociated lactic acid (UDLA; 0.11, 1.12 mM) significantly impacted bacterial growth rate (GR) and maximum population density (MPD). Of the three species, S. liquefaciens displayed high growth and tolerance to low pH and high concentrations of lactic acid and UDLA, which was followed by C. maltaromaticum and B. thermosphacta, the latter being the most sensitive species. Current culture-based methods are not adequate to measure bacterial growth kinetics in 2- and 3-species experimental systems. Therefore, a SYBR green-based qPCR method targeting the 16S rRNA gene of C. maltaromaticum, B. thermosphacta and S. liquefaciens was developed. Primers were designed based on 16S rRNA gene sequences. DNA extraction was optimized and a comparatively high annealing temperature (65°C) used. The reaction efficiency of standard curves was high (R\(^2\) = 0.98-0.99) over a linear quantification range of >5 log CFU/ml. Coefficient of variation did not exceed 14% within or between runs. The calculated GR and MPD were not significantly different between plate count and qPCR methods. Validation in mixed culture showed a variance of less than 0.3 log CFU/ml (x̄= 0.10 log CFU/ml, R\(^2\) = 0.98). The qPCR method was then implemented to study the effect of UDLA on bacterial growth kinetics in 1-, 2- and 3-species experimental systems. mBHI broth was formulated with a range of lactic acid concentration (0, 25, 50, 75, 100 mM) and pH (5.5, 6.5) and packaged within a commercial VP film, which was then heat-shrunk following a commercial protocol to reduce oxygen permeability. The highest species sensitivity to 0 to 2.24 mM UDLA occurred in the order: B. thermosphacta > C. maltaromaticum > S. liquefaciens. No growth was observed for any species at 2.24 mM UDLA. Interaction among the species was evaluated by subtracting the GR and MPD observed in the 3-species mixed culture from that of the pure culture. For B. thermosphacta, GR and MPD were significantly inhibited in the 3-species culture at 0.56 and 1.12 mM, and from 0.06 to 1.12 mM UDLA, respectively. A trend in increase of MPD difference between individual and 3-species mixed culture was observed up to 0.22 mM UDLA. Evaluation of 2- species interactions revealed C. maltaromaticum had the greatest inhibitory effect on B. thermosphacta growth and reducing GR and MPD at 1.12 mM UDLA to 0.001 CFU/h and 3.20 CFU/ml, respectively. Unlike B. thermosphacta, GR of C. maltaromaticum was increasingly greater in the 3-species versus pure culture from 0.56 to 1.68 mM UDLA; MPD was increasingly lower in 3-species culture from 0.06 to 0.22 mM UDLA, but not at 0.56 and 1.12 mM UDLA. Unexpectedly, C. maltaromaticum grew at 1.68 mM UDLA in the 3-species mixture but not in pure culture, an effect not attributed to a change in pH during culture. The effector specie(s) causing C. maltaromaticum growth inhibition or promotion was not observed in 2-species culture, indicating the effect required a 3-species interaction. UDLA had no effect on S. liquefaciens GR or MPD in 3-species mixed culture. The ‘Jameson effect’ may have caused the reduction in B. thermosphacta MPD in 2-species co-culture with C. maltaromaticum, however the inhibitory effect produced by S. liquefaciens was likely caused by a different mechanism. Individual bacterial growth kinetics in mBHI broth were validated in irradiated VP beef, demonstrating an acceptable accuracy factor of 1.12 for all three species. In conclusion, this thesis provides the beef industry with science-based evidence to more effectively design intervention strategies to control the evolution of beef spoilage microbiomes