Metabolomic and Transcriptomic Analyses of Escherichia coli for Efficient Fermentation of L-Fucose

Abstract

L-Fucose, one of the major monomeric sugars in brown algae, possesses high potential for use in the large-scale production of bio-based products. Although fucose catabolic pathways have been enzymatically evaluated, the effects of fucose as a carbon source on intracellular metabolism in industrial microorganisms such as Escherichia coli are still not identified. To elucidate the effects of fucose on cellular metabolism and to find clues for efficient conversion of fucose into bio-based products, comparative metabolomic and transcriptomic analyses were performed on E. coli on L-fucose and on D-glucose as a control. When fucose was the carbon source for E. coli, integration of the two omics analyses revealed that excess gluconeogenesis and quorum sensing led to severe depletion of ATP, resulting in accumulation and export of fucose extracellularly. Therefore, metabolic engineering and optimization are needed for E. coil to more efficiently ferment fucose. This is the first multi-omics study investigating the effects of fucose on cellular metabolism in E. coli. These omics data and their biological interpretation could be used to assist metabolic engineering of E. coli producing bio-based products using fucose-containing brown macroalgae.

Keywords: Escherichia coli; brown macroalgae; fermentation; fucose; metabolomics; transcriptomics.

Figure 1

Comparison of the growth and fermentation product profiles of E. coli cultured on fucose and glucose as the carbon source (mean ± SD): (A) cell density recorded as optical density at 600 nm (OD₆₀₀); (B) concentrations of substrate; and (C) by-products (three independent replicates).

Figure 2

PCA (A) score and (B) loading plots of 102 intracellular metabolites of E. coli in the lag, exponential, and stationary phases cultured on fucose and glucose as the carbon source (six replicates; three independent replicates × two technical replicates).

Figure 3

MetaMapp analysis of 102 intracellular metabolites of E. coli cultured on fucose and glucose as the carbon source in: (A) the lag phase; (B) exponential phase; and (C) stationary phase. Classes of metabolites are represented by shapes. Significant increases and decreases in metabolite abundance are represented by color (p < 0.05). Magnitudes of fold changes are represented by the size of symbols and labels. Biochemical and structural similarities are represented by the orange and gray edges, respectively (six replicates; three independent replicates × two technical replicates).

Figure 4

Hierarchical clustering analysis (FDR adjusted p-value < 0.01, ANOVA) of significantly increased or decreased (A) intracellular and (B) extracellular metabolites of E. coli in the lag, exponential, and stationary phases cultured on fucose as the carbon source. Clustering of the model was based on Pearson’s correlation coefficient and average linkage methods (six replicates; three independent replicates × two technical replicates).

Figure 5

Comparison of transcription levels in (A) central carbon metabolism and (B) abundance of cofactors of E. coli in the exponential phase cultured on fucose and glucose. Significant changes to transcript levels are represented by color (p-value < 0.05, and fold changes > 2.0). Significant changes in cofactors are represented by * (three independent replicates; p-value < 0.05).

Figure 6

Heat map of 284 transcripts significantly changed (p-value < 0.05) in E. coli in the exponential phase cultured on fucose and glucose (three independent replicates).