Predicting the performance of chain elongating microbiomes through flow cytometric fingerprinting.

Abstract

As part of the circular bio-economy paradigm shift, waste management and valorisation practices have moved away from sanitation and towards the production of added-value compounds. Recently, the development of mixed culture bioprocess for the conversion of waste(water) to platform chemicals, such as medium chain carboxylic acids, has attracted significant interest. Often, the microbiology of these novel bioprocesses is less diverse and more prone to disturbances, which can lead to process failure. This issue can be tackled by implementing an advanced monitoring strategy based on the microbiology of the process. In this study, flow cytometry was used to monitor the microbiology of lactic acid chain elongation for the production of caproic acid, and assess its performance both qualitatively and quantitatively. Two continuous stirred tank reactors for chain elongation were monitored flow cytometrically for over 336 days. Through community typing, four specific community types could be identified and correlated to both a specific functionality and genotypic diversity. Additionally, the machine-learning algorithms trained in this study demonstrated the ability to predict production rates of, amongst others, caproic acid with high accuracy in the present (R² > 0.87) and intermediate accuracy in the near future (R² > 0.63). The identification of specific community types and the development of predictive algorithms form the basis of advanced bioprocess monitoring based on flow cytometry, and have the potential to improve bioprocess control and optimization, leading to better product quality and yields.