For the production of coenzyme Q10 (CoQ10), an electron carrier in the respiration chain with antioxidant activity, we investigated effects of various nitrogen and carbon sources, aeration and agitation rates, dissolved oxygen levels, an electron flux inhibitor - azide, a proton gradient releaser - 2,4-dinitrophenol (DNP), and an oxidative stressor - hydrogen peroxide (H₂O₂) on the intracellular CoQ10 contents and cell growth in Agrobacterium tumefaciens ATCC4452. Classical mutagenesis was introduced to improve CoQ10 content and reduce exopolysaccharides, symbiotic material for plant-microbe interaction. The highest CoQ10 content was obtained at 450 rpm, 1 vvm and 30oC when CSL and sucrose were used as major nitrogen and carbon sources. The optimized fermentation condition yielded 2.01 mg/g-DCW CoQ10 contents and 21 g/L DCW, respectively. An increase of agitation speed and aeration rate decreased CoQ10 content. A decrease in dissolved oxygen level from 20 to 5%, the intracellular CoQ10 content increased about 4-fold, yielding 2 mg per g-dry cell weight at 5% dissolved oxygen level. Azide significantly increased the intracellular CoQ10 content, with the highest value of 5.3 mg per g dry cell weight at 0.45 mM of sodium azide. However, DNP (up to 200 μM) and H₂O₂ (up to 10 μM) did not affect the intracellular CoQ10 content, indicating proton gradient release and oxidative stress do not affect the biosynthesis of CoQ10. These results showed that the restricted electron flux by limited oxygen supply and the addition of azide increase the intracellular CoQ10 content, suggesting feedback regulation of CoQ10 biosynthesis by its physiological function of an electron carrier in the respiration chain. The Agrobacterium mutant, which was induced by p-fluorophenylalanine as selective marker, showed an increase in CoQ10 content slightly (approximately 20% higher than the parent). Exopolysaccharide(EPS) deficient mutants were screened based on the fluorescence of colonies on calcofluor agar plates. In flask culture, DCW (dry cell weight) of the EPS-deficient mutant was increased 2 times compared with the wild type strain but the CoQ10 content decreased to 75%. High cell-density culture was carried out by a pH-controlled fed-batch strategy which was the best suitable method in acidic material producing bacteria such as A. tumefaciens. In a 2.5 L fermenter, exopolysaccharide was produced below 5 g/L and the maximal cell mass for the exopolysaccharide deficient mutant reached 61 g/L in a shorter cultivation time, while the wild type took 86 hrs to reach 41 g/L DCW.
To diminish the exopolysaccharide from A. tumefaciens ATCC4452 and to enhance CoQ10 content through carbon flux redistribution, a putative celA gene was disrupted by homologous gene recombination. The partial region of the cellulose synthsis operon was cloned and the chloramphenicol acetyl transferase of approximately 1 kb long gene was inserted into the celAB gene for disrupting this operon. Plasmid containing defective celAB lesion was introduced to the wild-type strain. The chloramphenicol resistant colony was screened on LB with calcofluor white and chloramphenicol. Unfortunately, the selected strain did not show the property of EPS deficient but the fluorescent intensity could be distinguishable from that of the parent strain. It was considered that a single gene knock out might not be able to completely block the EPS production because there were various metabolic fluxes to synthesize other materials.
In conclusion, we demonstrated that the control of cellular respiration by respiration inhibitors or DO adjustment allowed an increase in CoQ10 contents on A. tumefaciens ATCC4452 and the higher cell density was attributed to the less EPS production.