Abstract
Selected bacterial cultures are used on an industrial scale for biopreservation of fresh fermented dairy products. The selected cultures provide efficient inhibition of yeasts and molds and serve as an alternative to the use of purified or synthetic antimicrobials. Despite the fact that numerous compounds have been identified as antifungal based on a strategy of bioassay guided fractionation, the factors for antifungal effect remain unexplained. Lack of understanding about the mechanism(s) responsible for the effect restricts development of new cultures with antifungal properties as well as the application into other food matrices.
The scope of the thesis was to develop and apply a chromatography mass spectrometry based metabolomic footprint workflow for the investigation of the mechanisms behind the antifungal properties of a co‐culture, consisting of Lactobacillus paracasei (LAB A) and Propionibacterium freudenreichii subsp. shermanii (PAB A).
The strategy was to investigate the effect of the composition of exometabolome of the co‐culture on mold growth. A biological model system was developed in order to have a simplified system for studying the growth of bacteria and the subsequent effect on mold growth represented by two strains of Penicillium (manuscript III). Characterization of mold growth was performed by a spectral clustering algorithm on data from multispectral imaging (manuscript VI). Untargeted analysis of the exometabolome was performed on liquid chromatography/mass spectrometry
(LC/MS) instrumentation and a combination of methods selected based on expected compound classes in the exometabolome (paper I). In order to extend the coverage of the exometabolome, low molecular weight and volatile compounds were analyzed after pre‐derivatisation or headspace sampling by gas chromatography/mass spectrometry (GC/MS). Data from the untargeted LC/MS analysis was processed using feature selection and subsequent multivariate data analysis.
The antifungal properties of the co‐culture were composed of several components when fermented in the model system. The major contribution came from the LAB A cells which produced diacetyl, either secreted or formed by oxidative decarboxylation of secreted acetolactate (manuscript IV).
Production of diacetyl occurred without any perturbation of the system by mold hyphae, but was enhanced by the presence of PAB A ferments. Presence of live LAB A cells was required to maintain a diacetyl concentration sufficient for the antifungal effect. Over time, the concentration of diacetyl decreased and mold developed similar to an acidified un‐inoculated medium. Besides diacetyl, production of lactic acid and other 2‐hydroxy acids contributed weakly to the antifungal effect.
The effect of diacetyl on mold growth in a yoghurt food system was verified. Additionally, relatively higher amounts of diacetyl were produced in yoghurt with LAB A culture than in one without, indicating that diacetyl is also a major contributor to the effect in food systems.
Comparative analysis of three Lb. paracasei strains demonstrated that the cell‐free ferments only possessed a weak inhibitory effect on mold growth. The three strains could be classified according to their exometabolome and their relative inhibitory effect towards molds. Based on the study, three known and three non‐previously reported metabolites were identified as having antifungal properties (manuscript V).
In conclusion, the research conducted in this project demonstrated the potential of untargeted analysis of the exometabolome in combination with multivariate data analysis for building new understanding of biological phenomena in food.
The scope of the thesis was to develop and apply a chromatography mass spectrometry based metabolomic footprint workflow for the investigation of the mechanisms behind the antifungal properties of a co‐culture, consisting of Lactobacillus paracasei (LAB A) and Propionibacterium freudenreichii subsp. shermanii (PAB A).
The strategy was to investigate the effect of the composition of exometabolome of the co‐culture on mold growth. A biological model system was developed in order to have a simplified system for studying the growth of bacteria and the subsequent effect on mold growth represented by two strains of Penicillium (manuscript III). Characterization of mold growth was performed by a spectral clustering algorithm on data from multispectral imaging (manuscript VI). Untargeted analysis of the exometabolome was performed on liquid chromatography/mass spectrometry
(LC/MS) instrumentation and a combination of methods selected based on expected compound classes in the exometabolome (paper I). In order to extend the coverage of the exometabolome, low molecular weight and volatile compounds were analyzed after pre‐derivatisation or headspace sampling by gas chromatography/mass spectrometry (GC/MS). Data from the untargeted LC/MS analysis was processed using feature selection and subsequent multivariate data analysis.
The antifungal properties of the co‐culture were composed of several components when fermented in the model system. The major contribution came from the LAB A cells which produced diacetyl, either secreted or formed by oxidative decarboxylation of secreted acetolactate (manuscript IV).
Production of diacetyl occurred without any perturbation of the system by mold hyphae, but was enhanced by the presence of PAB A ferments. Presence of live LAB A cells was required to maintain a diacetyl concentration sufficient for the antifungal effect. Over time, the concentration of diacetyl decreased and mold developed similar to an acidified un‐inoculated medium. Besides diacetyl, production of lactic acid and other 2‐hydroxy acids contributed weakly to the antifungal effect.
The effect of diacetyl on mold growth in a yoghurt food system was verified. Additionally, relatively higher amounts of diacetyl were produced in yoghurt with LAB A culture than in one without, indicating that diacetyl is also a major contributor to the effect in food systems.
Comparative analysis of three Lb. paracasei strains demonstrated that the cell‐free ferments only possessed a weak inhibitory effect on mold growth. The three strains could be classified according to their exometabolome and their relative inhibitory effect towards molds. Based on the study, three known and three non‐previously reported metabolites were identified as having antifungal properties (manuscript V).
In conclusion, the research conducted in this project demonstrated the potential of untargeted analysis of the exometabolome in combination with multivariate data analysis for building new understanding of biological phenomena in food.
| Originalsprog | Engelsk |
|---|
| Forlag | Department of Food Science, Faculty of Science, University of Copenhagen |
|---|---|
| Antal sider | 262 |
| Status | Udgivet - 2014 |