Performance testing of Zymomonas mobilis metabolically engineered for cofermentation of glucose, xylose, and arabinose

IOGEN Corporation of Ottawa, Canada, has recently built a 40t/d biomass-to-ethanol demonstration plant adjacent to its enzyme production facility. It has partnered with the University of Toronto to test the C6/C5 cofermenta-tion performance characteristics of the National Renewable Energy Labora-tory's metabolically engineered Zymomonas mobilis using various biomass hydrolysates. IOGEN's feedstocks are primarily agricultural wastes such as corn stover and wheat straw. Integrated recombinant Z. mobilis strain AX101 grows on D-xylose and/or L-arabinose as the sole carbon/energy sources and ferments these pentose sugars to ethanol in high yield. Strain AX101 lacks the tetracycline resistance gene that was a common feature of other recombinant Zm constructs. Genomic integration provides reliable cofermentation performance in the absence of antibiotics, another characteristic making strain AX101 attractive for industrial cellulosic ethanol production. In this work, IOGEN's biomass hydrolysate was simulated by a pure sugar medium containing 6% (w/v) glucose, 3% xylose, and 0.35% arabinose. At a level of 3 g/L (dry solids), corn steep liquor with inorganic nitrogen (0.8 g/L of ammonium chloride or 1.2 g/L of diammonium phosphate) was a cost-effective nutritional supplement. In the absence of acetic acid, the maximum volumetric ethanol productivity of a continuous fermentation at pH 5.0 was 3.54 g/L.h. During prolonged continuous fermentation, the efficiency of sugar-to-ethanol conversion (based on total sugar load) was maintained at >85%. At a level of 0.25% (w/v) acetic acid, the productivity decreased to 1.17 g/L.h at pH 5.5. Unlike integrated, xylose-utilizing rec Zm strain C25, strain AX101 produces less lactic acid as byproduct, owing to the fact that the Escherichia coli arabinose genes are inserted into a region of the host chromosome tentatively assigned to the gene for D-lactic acid dehydrogenase. In pH-controlled batch fermentations with sugar mixtures, the order of sugar exhaustion from the medium was glucose followed by xylose and arabinose. Both the total sugar load and the sugar ratio were shown to be important determinants for efficient cofermentation. Ethanol at a level of 3% (w/v) was implicated as both inhibitory to pentose fermentation and as a potentiator of acetic acid inhibition of pentose fermentation at pH 5.5. The effect of ethanol may have been underestimated in other assessments of acetic acid sensitivity. This work underscores the importance of employing similar assay conditions in making comparative assessments of biocatalyst fermentation performance.