Microstructure and Textural Properties of Yoghurts Produced by Exopolysaccharides- Producing Starter Cultures

Research output: Book/ReportPh.D. thesisResearch

  • Lanjun Zhang
Exopolysaccharides (EPS) play a central role in contributing to desired sensory characteristics to yoghurts when they are synthesized in situ by lactic acid bacterial (LAB) during fermentation. The mechanisms of how the EPS affect the textural properties of yoghurts still are not clear. In order to gain an insight into the complex matrices of yoghurts to understand the distribution pattern of EPS and the interplay between EPS and protein gel, one of the effective solutions is to directly visualize EPS at a microstructural level.
The method, we further developed in this project based on a previous study, provided a novel approach that the distribution pattern of EPS and their interplay with protein gel can be simultaneously visualized by the two selected lectin probes of concanavalin A (ConA) and wheat germ agglutinin (WGA) as well as rhodamine B using confocal laser scanning microscopy (CLSM) when these three fluorescent dyes were combined together in a yoghurt sample. Use of phosphate-buffered saline supplemented with CaCl2 and MgCl2 (PBS-Ca-Mg buffer, pH 7.4) preserved the binding affinities of both lectins of ConA and WGA to EPS under acidic conditions.
The newly developed method of EPS visualization was further applied to characterize the microstructure in conjunction with the texture analysis of yoghurts that were produced by four different EPS-producing starter cultures and mechanically treated post-fermentation at 0, 1, 2, and 4 bar. Each starter culture was composed of one strain Streptococcus thermophilus (ST) and one strain Lactobacillus delbrueckii ssp. bulgaricus (LB). Two Lb. bulgaricus strains were used in combination with two S. thermophilus strains. The choice of ST strain was the major determinant for the rheological properties of the yoghurts, since one of the ST strains conferred a ropy texture and resulted in yoghurts with decreased water holding capacity and an open microstructure. In addition, one of the LB strains with high amounts of EPS producing capacity improved water retention, when it was combined with an ST strain that produced negligible amounts of EPS. A moderate mechanical treatment post-fermentation (i.e. 1 bar) was able to further improve the textural properties, possibly due to structural rearrangements causing improved protein-protein interactions and/or a better distribution of EPS.
A total 286 of CLSM micrographs acquired from the four types of yoghurts mechanically processed at the four levels of pressure were subjected to image analysis. Pixel counting was applied to threshold images and revealed co-localization of colors that was not immediately apparent in the original images. The gel structures of the manufactured yoghurts were evaluated using the analysis of fractal curves. The novel method of ANOVA-Simultaneous Component Analysis (ASCA) was employed to perform the analyses, and it not only statistically assessed the distribution pattern of EPS detected by both lectin probes and the variations of gel structures of the yoghurts in terms of pixel counting and fractal analysis, but also highlighted the significant differences affected by the yoghurt types and the mechanical treatment pressures. Moreover, ASCA models discovered that yoghurts highly correlated with EPS visualized by ConA exhibited heterogeneous gel structures, whereas the yoghurts positively correlated with the EPS visualized by WGA displayed uniform protein networks with higher self-similarity. Irrespective of yoghurt types, the yoghurts were shown to become more homogenous as pressure was increased. However, the yoghurt without EPS was more susceptible to whey separation compared to other yoghurt types.
We attempted to build a prediction model of Partial Least Squares (PLS) to correlate the findings of the image analysis with the specific textural parameters obtained in the Lab for predicting new product. However, we were unable to successfully construct this model. This was mainly due to the limitations of our technique, in terms of both image acquisition and image analysis.
Original languageEnglish
PublisherDepartment of Food Science, Faculty of Science, University of Copenhagen
Publication statusPublished - 2015

ID: 156371547