News

E. coli is a common inhabitant found in our intestines. Most strains are harmless, but there are pathogenic variants that can cause gastrointestinal problems. The investigation studied how non-pathogenic strains adapt to different conditions in order to predict the behaviour of more harmful strains.

The main problem with this type of bacteria is its presence in some foods. The pathogenic versions of E. coli are commonly found in cattle. From there it can enter the food chain when poor hygiene procedures exist related to premises, utensils or food handling. Poorly cooked meat, unpasteurised dairy products, contaminated and poorly washed vegetables or contaminated water may contain the toxin produced by the bacteria, which is the largest worldwide cause of gastroenteritis.

The research team used mathematical models to study the variability between pathogenic and non-pathogenic strains of the bacteria to describe their behaviour. This was done via cultivating different pathogenic and non-pathogenic strains of E. coli in three different ways: in the laboratory, milk and meat juice. After this, the growth of micro-organisms was calculated using a variety of parameters, such as the lag time (the time it takes the bacteria to adapt to the medium) and the maximum growth rate. The three mediums used in the experiment attempted to recreate ambient temperature conditions that were above refrigeration levels (15, 20 or 25ºC) or the optimum temperature level for the micro-organism (30, 35 or 40ºC).

The study demonstrated that there is very little variability between pathogenic and non-pathogenic strains, with both displaying similar behaviour under experimental conditions. The principal investigator, Emiliano Quinto, from the Faculty of Nutrition and Food Science at the University of Valladolid said, ‘The values for the lag time and maximum growth rate are comparable and therefore the data from one strain can be used to predict the behaviour of other strains.’

Two statistical estimation strategies were used to do this, both of which were developed in collaboration with a researcher at UC3M. ‘These advanced probabilistic models use a general distribution which encompasses a broad family of standard distributions,’ explained Juan Miguel Marin, a researcher at the Department of Statistics at UC3M. The first of these is parametric, based on a general distribution related to positive measurements, and the second is non-parametric which uses bootstrap resampling methods.

Knowledge Transfer

The main objective of the research, which was recently published in the scientific journal Food Research International, is to predict the behaviour of more dangerous bacteria. In order to do this, it is important to understand the time it takes for bacteria to adapt to a medium and estimate the reproductive speed. This information has practical applications for risk analysis within the food industry. With the data obtained from non-pathogenic strains of E. coli, the behaviour of pathogenic strains can be predicted and their spread within the food chain can be tackled. In other words, this means understanding when to use the technical tools available in the food industry (cooling, sterilisation or pasteurisation) in order to stop bacteria reproduction and in turn limit the production of the toxin associated with intestinal problems. ‘The food industry has a major interest in understanding when to use certain technical treatments. The ideal is to tackle the bacteria before it has time to adapt and therefore deny it the time it needs to reproduce,’ added Emiliano Quinto.

Bibliography:

E.J.Quinto, J.M.Marin, I. Caroa, J. Mateo, M.P.Redondo-del-Rio, B.de-Mateo-Silleras, D.W.Schaffner, “Bootstrap parametric GB2 and bootstrap nonparametric distributions for studying Shiga toxin-producing Escherichia coli strains growth rate variability”. Food Research International. Volume 120, June 2019, Pages 829-838.  https://doi.org/10.1016/j.foodres.2018.11.045