No abstract

Although synthetic biology progress has made it possible to produce various biofuels in more user-friendly hosts, such as Escherichia coli, the large-scale biofuel production in these non-native systems is still challenging, mostly due to the very low tolerance of these non-native hosts to the biofuel toxicity. To address the issues, in this study we determined the metabolic responses of E. coli induced by three major biofuel products, ethanol, butanol, and isobutanol, using a gas chromatography-mass spectrometry (GC-MS) approach. A metabolomic data set of 65 metabolites identified in all samples was then subjected to principal component analysis (PCA) to compare their effects and a weighted correlation network analysis (WGCNA) to identify the metabolic modules specifically responsive to each of the biofuel stresses, respectively. The PCA analysis showed that cellular responses caused by the biofuel stress were in general similar to aging cells at stationary phase, inconsistent with early studies showing a high degree of dissimilarity between metabolite responses during growth cessation as induced through stationary phases or through various environmental stress applications. The WGCNA analysis allowed identification of 2, 4, and 2 metabolic modules specifically associated with ethanol, butanol, and isobutanol treatments, respectively. The biofuel-associated modules included amino acids and osmoprotectants, such as isoleucine, valine, glycine, glutamate, and trehalose, suggesting amino acid metabolism and osmoregulation are among the key protection mechanisms against biofuel stresses in E. coli. Interestingly, no module was found associated with all three biofuel products, suggesting differential effects of each biofuel on E. coli. The findings enhanced our understanding of E. coli responses to exogenous biofuels and also demonstrated the effectiveness of the metabolomic and network analysis in identifying key targets for biofuel tolerance.

The heterotrophic dinoflagellate alga Crypthecodinium cohnii is known to accumulate lipids with a high fraction of docosahexaenoic acid (DHA). In this study, we first evaluated two antioxidant compounds, butylated hydroxyanisole (BHA) and propyl gallate (PG), for their effects on lipid accumulation in C. cohnii. The results showed that antioxidant BHA could increase lipid accumulation in C. cohnii by 8.80% at a final concentration of 30 μM, while PG had no obvious effect on lipid accumulation at the tested concentrations. To decipher the molecular mechanism responsible for the increased lipid accumulation by BHA, we employed an integrated GC-MS and LC-MS metabolomic approach to determine the time-series metabolic profiles with or without BHA, and then subjected the metabolomic data to a principal component analysis (PCA) and a weighted gene coexpression network analysis (WGCNA) network analyses to identify the key metabolic modules and metabolites possibly relevant to the increased lipid accumulation. LC-MS analysis showed that several metabolites, including NADPH, could be important for the stimulation role of BHA on lipid accumulation. Meanwhile GC-MS and network analyses allowed identification of eight metabolic modules and nine hub metabolites possibly relevant to the stimulation role of BHA in C. cohnii. The study provided a metabolomics view of the BHA mode of action on lipid accumulation in C. cohnii, and the information could be valuable for a better understanding of antioxidant effects on lipid accumulation in other microalgae as well.