3.3 Engineering ZM4 with the self-flocculating phenotype
To elucidate the functions of genes involved in the biosynthesis and degradation of c-di-GMP, we targeted genes encoding proteins with PDE activities for c-di-GMP degradation to explore the effect of their combinatory knockout on the development of the self-flocculating phenotype in ZM4.
When both ZMO1055 and ZMO0401 were deleted, the flocculation efficiency of the bacterial cells improved to 24.4%, and the flocculation efficiency was further improved to 34.5% for the double knockout mutant with both ZMO1055 and ZMO1487 deleted (Fig. 5A). No further improvement in the self-flocculating phenotype was observed when all three genes were deleted in ZM4. However, such genetic manipulation is preferred for reducing the genome of Z. mobilis to engineer this species as a more reliable method to accommodate heterogeneous genes more effectively.
Cellulose fibrils have been validated as the chemical basis for developing the self-flocculating phenotype in ZM401(Xia et al., 2018). ZM4 also contains a bacterial cellulose synthase (bcs ) operon composed of ZMO1082, ZMO1083, ZMO1084, and ZMO1085. ZMO1082 was predicted to be a putative gene encoding a short peptide composed of 67 amino acid residues only (Xia et al., 2018), which is less likely to be functional and thus can be manipulated together with ZMO1083. Therefore, we engineered ZM4 with the overexpression of ZMO1082-1083 and ZMO1082-1084 and the whole bcs operon ZMO1082-1085 to investigate their contribution to the self-flocculation of the bacterial cells. As a result, the development of the self-flocculating phenotype was observed in the mutants with their flocculating efficiencies of 28.9%, 43.1% and 66.7%, respectively (Fig. 5A).
When both strategies were employed in ZM4, enhancing its intracellular accumulation of c-di-GMP to 94.42 pg/mg protein through the deletion of ZMO1055, ZMO0401, and ZMO1487, and the biosynthesis of cellulose fibrils through the overexpression of the bcs operon, a flocculating efficiency of 97.3% was observed for the bacterial cells, which was higher than that of 92.5% detected with ZM401 (Fig. 5A). Morphologies were further shown for ZM4 strains engineered with the deletion of ZMO1055, ZMO0401, and ZMO1487, overexpression of the bcs operon, and the combination of these two strategies (Fig. 5B).
Fig. 5
When industrial strains are engineered with new phenotypes, such as the self-flocculation of microbial cells for more advantages, their production performance should not be compromised. Z. mobilis is ethanologenic, and suitable for producing cellulosic ethanol. Therefore, we compared ethanol fermentation performance between the strain engineered with the self-flocculating phenotype and its unicellular wild-type ZM4. As can be seen (Fig. S1), no difference was observed when medium supplemented with 100 g/L glucose, equivalent to total sugars in the hydrolysate of lignocellulosic biomass, was fermented to produce ethanol.