(36c) Quantifying the Effect of Minimal Processing on the Kinetics and Antimicrobial Resistance of Listeria in Structured Food Model Systems Enriched with Natural Microflora

Nowadays, there is an increasing consumer demand for minimally processed foods which retain their nutritional content, natural colouring, taste and texture.^[1] To achieve such food quality, minimal food processing techniques for microbial decontamination, such as treatment with natural antimicrobial compounds, are of great interest.^{[2 ]} One such compound is nisin, which is produced by lactic acid bacteria and is approved by the Food and Drug Administration (FDA) for use in the food industry as a preservative.^[3] Additionally, minimal processing treatments can be combined with non-thermal processing, such as ultrasounds, which can have a synergistic effect and act as a hurdle for the inactivation of bacteria, thereby increasing the processing efficiency.^[4] However, the efficacy of such antimicrobial compounds, especially in combination with non-thermal processing methods for food decontamination, is still unclear.^[5] Furthermore, these techniques are less harsh than classical (thermal) processing methods (sterilisation, pasteurisation), so there is the potential for conditions to present a mild, sublethal stress which could induce an adaptive response in bacteria, allowing post-treatment survival. This is of concern for pathogenic bacteria such as Listeria, which is often associated with ready-to-eat foods and with high mortality rates. An increase in antimicrobial resistance (AMR) in Listeria species has also been reported in recent years, which raises the concern that effective treatment of listeriosis infections may become compromised in the future.^[6]

Most available studies on the efficacy of minimal processing with natural antimicrobials and/or with non-thermal technologies on inactivating food-related pathogens (including Listeria) are conducted in liquid broth systems, even though most food products are solid or solid(like) e.g. soft cheeses, meats. Cells grown as colonies in a solid system experience a significantly different environment in comparison to planktonic growth in a liquid system. More specifically, diffusional limitations of oxygen and nutrients and accumulation of (acidic) metabolic products around the colony can cause a self-induced (acid) stress that may affect the microbial kinetics and the microbial environmental response.^[7,8,9] Furthermore, microorganisms could display a different level of AMR development due to the general environmental stress adaptation and cross protection.^[10,11,12] Bacteria found in foods are also likely to grow in co-culture with other bacteria which can result in stress due to competition and/or stress due to 'foreign' extracellular metabolic products. For example, lactic acid bacteria such as L. lactis are found in similar foods to Listeria, and are natural producers of nisin, which, as mentioned above, is an antimicrobial compound of interest for use in microbial decontamination techniques.^[13] Therefore, in order to achieve a fundamental understanding of the stress response and AMR development of Listeria in food systems in response to novel food processing (i.e. nisin combined with ultrasounds), in the presence of other bacteria, it is important to conduct kinetic experiments in model systems that mimic as accurately as possible the microstructure and microflora of solid/solid(like) foods.

The present work is a systematic comparative study of the microbial dynamics of Listeria, as influenced by the presence of natural microflora such as nisin-producing lactic acid bacteria, on solid(like)/structured systems of various complexities at a range of optimal and suboptimal growth temperatures.

The model systems are created using Xanthan gum (XG) and/or Whey protein (WPI): both are used in the food industry and are stable at a wide range of temperatures, so they are good surrogates to mimic a range of real food products.

Materials and Methods:

Food model systems were prepared using Tryptic Soy Broth supplemented with 0.6% w/v Yeast Extract (TSBYE), with (1) no added gelling agent (planktonic growth), (2) 5% w/v XG (monophasic system), (3) 10% w/v WPI (monophasic system), or (4) a combination of 5% XG and 10% WPI, thus producing a biphasic complex food model system.

Growth kinetics at 10^oC, 30^oC and 37^oC were quantitatively monitored for L. innocua ATCC 33090 in monoculture, and in co-culture with L. lactis NZ9700 (nisin-producing) or L. lactis NZ9800 (non-nisin-producing) grown as planktonic cells in TSBYE, or as surface colonies on the monophasic/biphasic food model systems described. The Baranyi and Roberts mathematical model^[14] was fitted to the experimental data, and the values of the model parameters for all systems under study were analysed and compared.

The effect of ultrasound on the inactivation of L. innocua stationary phase colonies, either in mono- or co-culture with both L. lactis strains, was investigated using an ultrasonic bath at a frequency of 20kHz and power of 20W, for a treatment time of 5 minutes. Inactivation kinetics were monitored, and the inactivation model of Geeraerd et al. was fitted to the experimental data.^[15]

The spatial distribution and size of stationary phase colonies on the surface of each system, both in mono- and co-culture, was investigated for all conditions under study using a commercial scanning laser confocal microscope (CLSM). Phase separation in the XG/WPI biphasic system was visualised using Rhodamine B, which binds to the protein phase only allowing for definition between the two phases. L. innocua colonies were visualised using DAPI, while a Green Fluorescent Protein variant of L. lactis was used to visualise these colonies and thus differentiate between species.

Results and Discussion:

The growth kinetics, i.e. lag phase, exponential growth rate and/or stationary phase, of Listeria were affected by the structure of the solid(like) systems and by the presence of L. lactis (both strains), and were different to the liquid. Differences were also observed in the growth kinetics of the nisin- and non-nisin-producing L. lactis strains, both in mono- and co-culture. These differences may affect the production rate and/or quantity of nisin, and thus also affect its efficacy as a microbial decontamination method. Additionally, the action of nisin on Listeria was different depending on the food model system and the incubation temperature, and the combination of nisin with ultrasonic treatment resulted in differences in the inactivation kinetics, i.e., inactivation rate, and shoulder period.

The use of CLSM techniques showed differences in microbial spatial organisation. More specifically, in monoculture, L. innocua and both L. lactis strains demonstrate preferential growth on the protein phase of the biphasic gelled system. By contrast, in co-culture both L. lactis strains grow on the Xanthan gum phase, while L. innocua spreads through both phases. This is the first reported observation of such changing preferential growth in a biphasic, co-culture system.

The differences observed in growth location for L. innocua and both L. lactis strains can have a significant impact on their interactions as competitive species, and on the efficacy of nisin produced by L. lactis NZ9700, in particular its ability to reach L. innocua colonies. Additionally, the colony size and growth location can have a significant impact on the environmental stress response within a colony, i.e., self-induced acid stress, starvation/metabolic stress, oxidative stress, etc. Thus, significantly different responses to microbial decontamination processes are possible for systems of varying microstructure in the presence of natural microflora.

Significance and Impact:

Our findings give a systematic quantitative insight on the impact of co-culture with nisin-producing L. lactis on the growth of Listeria, in food model systems of varying structural complexity. They highlight the importance of accounting for bacterial stress adaptation in solid(like) systems when designing novel decontamination processes. Additionally, the microscopic differences observed using CLSM techniques are important to account for when designing such processes, as the colony size and location (both of the pathogen and the naturally present organism) directly affects the environmental stress within a colony, hence different survival rates of Listeria may be observed in different systems.

Acknowledgements:

This work was supported by the Department of Chemical and Process Engineering of the University of Surrey as well as an Impact Acceleration Grant (IAA-KN9149C) of the University of Surrey, an IAA-EPSRC Grant (RN0281J), and the Royal Society.

References:

[1] Valdramidis, V.P., Koutsoumanis, K.P., 2016. Curr. Opin. Food Sci. 12, 63–69.

[2] Troy, D.J. et al., 2016. MESC 120, 2–9.

[3] Gharsallaoui, A. et al., 2016. Crit. Rev. Food Sci. Nutr. 56, 1262–1274.

[4] Liao, H. et al., 2018. Ultrason. Sonochem. 42, 244–249.

[5] Li, W.-R. et al., 2010. Appl. Microbiol. Biotechnol. 85, 1115–1122.

[6] The European Union summary report on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks in 2014, EFSA, 2015.

[7] Aspridou, Z. et al., 2014. Food Res. Int. 64, 683–691.

[8] Velliou, E.G. et al., 2013. Food Res. Int. 54, 1746–1752.

[9] Noriega, E. et al., 2013. Food Microbiol. 36, 355–364.

[10] Velliou, E.G. et al., 2011. Food Microbiol. 28, 702–711.

[11] Velliou, E.G. et al., 2010. Lett. Appl. Microbiol. 52, 116–122.

[12] Velliou, E.G. et al., 2012. J. Therm. Biol. 37, 72–78.

[13] Scatassa, M.L. et al., 2017. Ital. J. Food Saf. 6 (6191), 13-17.

[14] Baranyi, J., Roberts, T.A., 1994. Int. J. Food Microbiol. 23, 277–294.

[15] Geeraerd, A.H. et al., 2000. Int. J. Food Microbiol. 59, 185–209.