Research On The Resistance Mechanism Of Yeast Based On Raman Tweezers And Multivariate Curve Resolution

Saccharomyces cerevisiae is a small bioreactor for ethanol fermentation,but during fermentation it encounters a variety of stressors that inhibit cell growth and metabolism,which affect cell growth and thus ethanol yield and production efficiency.Monitoring changes in yeast cell metabolism during ethanol fermentation can provide insight into the effects of stress on cells and how they respond to adversity stress,leading to the development of new strategies to improve yeast fermentation performance.However,the use of traditional mass spectrometry-based metabolomic analysis techniques to study metabolic changes in yeast cells is not only time-consuming and labour-intensive,but also usually yields information averaged over a large number of cells,obscuring fundamental differences between individual cells.It is well known that microbial cells in a population may exhibit heterogeneity under adverse conditions,i.e.the responses of individual cells are different.It is therefore necessary to use single cell techniques to analyse metabolic changes in yeast cells in order to further understand the mechanisms of resistance in brewer’s yeast.Raman spectroscopy is a powerful analytical technique that can obtain information about the molecular composition and chemical structure of substances and is widely used in metabolomics studies.Raman tweezers,also known as Laser Tweezer Raman Spectroscopy（LTRS）,is a combination of laser tweezers and microscopic Raman that combines the advantages of contactless optical manipulation and Raman analysis to capture individual microbial cells suspended in aqueous solution and perform fast and non-destructive real-time detection.It is more suitable for single cell analysis than conventional Raman spectroscopy.However,it remains a challenging task to identify and quantify specific biological macromolecules from single-cell Raman spectroscopy due to the numerous molecules（e.g.proteins,nucleic acids,lipids,polysaccharides,etc.）from biological samples that generate multiple overlapping peaks.The application of multivariate curve-resolving algorithms to decompose the Raman spectra of single cells is an effective method.allowing to provide pure spectra of the components of interest and their relative concentrations in individual cells without any a priori information.Therefore,in this paper,Raman tweezers were applied to collect Raman spectra of a large number of individual yeast cells at different time periods,and in combination with the multivariate curve resolution（MCR）algorithm,the metabolic changes of yeast single cells during different ethanol fermentations were investigated with a view to further understanding the resistance mechanisms of yeast cells.（1）Multiple curve resolution-alternating least squares（MCR-ALS）was applied to resolve the spectral dataset of yeast single cells during ethanol fermentation at high concentrations from the perspective of data mining.In order to extract the spectra associated with specific biomolecules and their concentration profiles,and further understand the metabolic processes and adaptation mechanisms during ethanol fermentation in brewer’s yeast.As a result,MCR-ALS successfully extracted the spectra and their concentration profiles associated with biomolecules such as phospholipids,triacylglycerols and proteins from the spectral data matrix without any a priori information.It was found that with the continuous accumulation of ethanol,the content of phospholipids in the cells decreases towards the later stages of fermentation.High concentrations of ethanol were shown to affect the structure of yeast cell membranes and increase membrane permeability.In response to ethanol stress,yeast cells increase the accumulation of triacylglycerols and synthesise a structurally specific protein in large quantities.However,there is heterogeneity in the mechanisms by which different cells respond to ethanol stress.It is shown that Raman tweezers,combined with MCR-ALS,can be used as a powerful tool for rapid analysis of metabolic changes in yeast single cells during ethanol fermentation,contributing to the understanding of metabolic heterogeneity and resistance mechanisms during yeast adversity fermentation.（2）To further understand the molecular basis behind the ethanol tolerance of ethanol-tolerant strains,Raman tweezers combined with the MCR-ALS algorithm were applied to analyse the metabolic changes during ethanol fermentation of three strains,industrial strain Bp1,laboratory strain INVSc1 and W303 a.The results revealed that Bp1 strains with better ethanol fermentation performance and higher tolerance increased the content of ergosterol and the accumulation of triacylglycerol to confer higher ethanol tolerance to the cells;meanwhile,the content of different biomacromolecules was relatively homogeneous among Bp1 cells,while the intercellular heterogeneity of INVSc1 and W303 a strains was relatively large,showing that cell heterogeneity had（3）Using non-negative matrix decomposition（3）A single-cell Raman spectroscopy dataset of yeast ethanol fermentation process under hyperosmotic stress was data mined using the non-negative matrix decomposition（NMF）algorithm.The results revealed that yeast cells increase the content of phospholipids,proteins and triacylglycerols in response to hyperosmotic stress conditions created by sorbitol.In addition,intercellular heterogeneity,in the pre-fermentation period,increased with increasing sorbitol concentration.Showing that increasing intercellular phenotypic heterogeneity may also be a cytoprotective strategy for yeast adaptation to adversity.These results demonstrate that Raman tweezers combined with the NMF algorithm can also be used to investigate the resistance mechanisms and metabolic heterogeneity in yeast fermentation processes under adversity,and have promising applications.（4）The metabolic differences between industrial strain Bp1 and experimental strain INVSc1 in ethanol fermentation under hypertonic stress created by sodium salt were analysed by applying Raman tweezers combined with the NMF algorithm.The results revealed that under sodium stress,strain INVSc1 increased the synthesis of phospholipids,proteins and polysaccharides.In contrast,the Bp1 strain with better ethanol fermentation performance and higher stress tolerance increased the synthesis of protein and polysaccharide substances in the face of the same adversity.It shows that strains with different ethanol tolerance have different mechanisms of adaptation to sodium stress.Therefore,a Raman tweezer combined with the NMF algorithm can be used to compare the metabolic differences of different strains under sodium stress for ethanol fermentation to achieve a rapid screening of yeast strains with higher ethanol tolerance.