Submicron infrared spectroscopy assessment of single-cell phenotypic diversity in microbial lipid production

Microbial lipid production offers a sustainable method for creating biofuels, lubricants, and high-value oils, utilizing the metabolic uniqueness of diverse organisms like bacteria, yeasts, and microalgae. However, minor physicochemical variations in bioreactors, along with subtle biochemical differences in organism's life stages, can lead to phenotypic diversity and impact the production. Therefore, monitoring, understanding and managing this diversity within bioreactors is essential in microbial biotechnology. Optical photothermal infrared (O-PTIR) spectroscopy can provide label-free chemical characterization of individual cells at sub-micron level. Here, we demonstrate the use of O-PTIR to evaluate metabolic heterogeneity within a population of oleaginous yeast Rhodotorula graminis in the production of free fatty acids (FFAs) and triacylglycerols (TAGs).

Forty yeast cells were measured by acquiring six single-point O-PTIR spectra per cell. Cell sizes were estimated from the corresponding microscopy images, while reference bulk infrared measurements of yeast biomass and pure compounds were obtained by Fourier transform infrared spectroscopies. Within the population, most of the cells have similar chemical composition, though several cells have quite different composition from the population average. Moreover, a number of cells have relatively large intra-cell chemical variability. The main chemical differences between the cells are correlated with cell sizes, and there are statistically significant size-dependent differences in cellular chemistry. Specifically, small cells have higher content of proteins than mid-size and large cells, and large cells have higher TAG-to-FFA ratio compared to mid-size cells. Characteristic wavenumbers for TAGs, FFAs and proteins can be used to estimate content of these compounds, namely 1748, 1714 and 1659 cm^- 1 respectively.

The O-PTIR method allows estimation of chemical composition of individual yeast cells and differentiation of two types of lipids (TAGs and FFAs). We have demonstrated that measurement at only four wavenumbers (the aforementioned wavenumbers for TAGs, FFAs and proteins plus one reference wavenumber at 1800 cm^- 1) provides the assessment of major chemical constituents of high importance for optimization of SCO production. We foresee that rapid data acquisition through O-PTIR imaging will significantly aid in understanding and managing phenotypic diversity in microbial cells by providing a detailed representation of individual cells for population statistics.