Development of a FCM data acquisition and processing pipeline for evaluating pure cultures of Limosilactobacillus reuteri and co-cultures with Saccharomyces cerevisiae

Abstract

Flow cytometry (FCM) is a powerful method for analysing cell characteristics, for cell quantification and for separation of different species. An automated FCM processing pipeline for lactic acid bacteria (LAB) was developed, and an analytical pipeline for co-cultures of LAB and Saccharomyces cerevisiae was designed. Data processing pipelines for Limosilactobacillus reuteri DSM 17938 sampled over time were in this project compared based on manual gating, fixed gating, and clustering algorithm k-means in MATLAB. Automated k-means strategies were found to perform well, they were fast and flexible to changes of the cell subpopulations over time, a more biologically relevant gating. The optimized k-means strategy was in reference to the manual gating well correlated, r = 0.94. The highest accuracy of gating was for the live cell population, 0.96 ± 0.18, and the lowest accuracy was for the damaged cell population, 0.50 ± 0.22. Furthermore, a workflow for analysing co-cultures of L. reuteri and S. cerevisiae was developed focusing on the cultivation conditions, the staining with fluorescent dyes and the FCM data processing. A potential mutualism of L. reuteri and recombinant yeast over-expressing TRPV1 receptor was investigated using the assay developed.

Popular Abstract

Probiotic bacteria are live microorganisms, which can be beneficial to the human gut-flora when consumed. They can be found naturally in fermented foods or be taken as supplements. The probiotic bacterium Limosilactobacillus reuteri has been found to carry several health-promoting properties, including prevention and treatment of many gut disorders like irritable bowel syndrome, antibiotic associated diarrhoea and infant colic. L. reuteri has been found to display antagonistic activity towards TRPV1, which is an important pain receptor in mammals. Modulation of TRPV1 is believed to be involved in the beneficial effects of the bacterium for the treatment of abdominal pain.

Process conditions during production of the bacterium significantly influence the final properties of the final probiotic product. The goal of the production process is to efficiently achieve a bacterial culture with optimal robustness and probiotic properties. For this purpose, it is important to be able to rapidly monitor cell counts as well as the physiological fitness and bioactivity of the culture. Flow cytometry (FCM) is a powerful method for analysing multiple properties of individual cells in a culture at a rate of thousands of cells per second. FCM is based on light, where cells scatter the light differently depending on their properties. This makes it possible to determine in which part of its growth phase the cells are, the size of the cells and its bioactivity. It is useful both for monitoring bioprocess unit operations, as well as for bioactivity assays. A challenge with FCM is however that a large amount of data is generated and needs to be efficiently compiled and interpreted in a so called data processing pipeline. This can be complex and tedious, not least since the data is commonly processed manually. It is a time-consuming endeavour and can be quite subjective to the user’s expertise.

In this project, I have investigated experimental and computational methods to generate and analyse large FCM data sets for evaluating multiple properties of L. reuteri cultures during production. Two different applications were investigated: (1) The use of online FCM to analyse L. reuteri during batch production in stirred tank reactors (STRs); and (2) The use of FCM to analyse the bioactivity of L. reuteri in a yeast-based TRPV1 modulation assay.

To make the processing of data from a large number of samples generated by online FCM less time-consuming and less biased an alternative way to process data was developed, where the data would be analysed automatically. This type of processing can be applied to cultures of the probiotic bacteria L. reuteri, and possibly other lactic acid bacteria. The automated method for processing this data divides the whole cell population into smaller parts, based on how similar the individual cells are. It is advantageous compared to manual processing since it is quicker and less biased. It is optimal to use for batch-cultivations, where samples are taken repeatedly during some time, producing many data samples. The data processing pipeline developed in this project allows for hundreds of samples or more to be processed quickly, and accurately, as well as visualized for easy interpretation.

To investigate the ability of L. reuteri to modulate the TRPV1 pain receptor a method to enable co-cultivation with yeast and subsequent FCM analysis was first needed to be developed. This was done by investigating which environmental conditions enabled growth of both the bacterium and the yeast together, as well as enabled FCM measurement of TRPV1 modulation. The developed FCM method also made it possible to analyse fitness effects of both yeast and bacterial cells separately, which was used to quantify effects on cell fitness by cultivating them together.

In summary, the advancements of research provided in this thesis comes from the development of experimental and computational methods that simplifies flow cytometry data processing of both pure and mixed cultures of L. reuteri and S. cerevisiae, and that can be used as tool for process monitoring.

• author: Tomasson, Julia
• supervisor: Magnus Carlquist; Nikhil Seshagiri Rao
• organization: Applied Microbiology
• course: KMBM05 20211
• year: 2021
• type: H2 - Master's Degree (Two Years)
• subject: Biology and Life Sciences
• keywords: Microbiology, Flow cytometry, automation, data processing, clustering, K-means, co-culture, probiotics, L. reuteri, S. cerevisiae, Applied microbiology
• language: English
• id: 9058299
• date added to LUP: 2021-06-28 13:06:16
• date last changed: 2021-06-28 13:06:16