Metabolic Engineering By Genome Shuffling Of Riboflavin-producing Bacillus.Subtilis

A series of genetic engineering Bacillus subtilis for riboflavin production wasobtained by transforming integration or autonomous rib-plasmids. The effect ofriboflavin operon dosage and its integration modes on riboflavin production wereresearched in this article. By increasing riboflavin operon dosage, riboflavinproduction of engineering strain was improved stably. The traits of riboflavin-producing B.subtilis were improved by two rounds of genome shuffling. Abiochemical reaction network was assembled with metabolites, then the analysis andcomparison of metabolic fluxes in batch cultures of three different engineering strainsabout the second round genome shuffling were conducted. The main results presentedin this work are as follows: Integration rib-plasmids were inserted in host chromosome at amyE,thrC and ribloci by two different modes, and the result indicated that integration of rib-plasmid bya double-cross event cause higher riboflavin yield than a single-cross event. As forhost B.subtilis RH13, integration of rib-plasmid at amyE locus cause more riboflavinyield than the other two loci, and as for another host B.subtilis 24R7, integration ofrib-plasmid at three loci almost cause equal increase of riboflavin production. Based on theory of gene amplification, deregulated riboflavin operon wasamplified to different copies in host chromosome. It shows that with amplification ofthe riboflavin operon in B.subtilis RH13, riboflavin production is raised from 0.43g/Lto 3.0g/L. When riboflavin operon dosage reached 6~7, amplifying more riboflavinoperons is no use for further improvement of yields of riboflavin. With amplificationof the riboflavin operon in B.subtilis 24R7, riboflavin production only had an increaseof 70% because of the instability of the reiterated sequences in the chromosome of24R7. Excessive riboflavin operon dosage in riboflavin-overproducing strains exertremarkable negative effect on cell growth, even is fatal to the host cells. Withautonomous rib-plasmid pMX45 being transformed, riboflavin production inB.subtilis 24R7 was raised from 1.1g/L to about 3g/L. The instability of anotherautonomous rib-plasmid pRB49 made it very difficult to enhance riboflavinproduction efficiently. Based on the theory of genome shuffling, a recombinant engineering strainB.subtilis RH33/pMX45 with improved phenotype was obtained by two rounds of IIABSTRACTgenome shuffling. In batch culture with sucrose as the carbon source, it could produceabout 4.2g/L riboflavin, which is 43% more than that of B.subtilis 24/pMX45, a strainfor commercial production of riboflavin. In addition, B.subtilis RH33/pMX45 couldassimilate glucose quickly for riboflavin biosynthesis, in batch culture with glucose asthe carbon source, it could produce about 4.4g/L riboflavin, which is about 100%more than that of B.subtilis 24/pMX45. A biochemical reaction network of engineering Bacillus subtilis was assembledwith metabolites, and a stoichiometric model including 34 equations and 39 variableswas used to analyze metabolic fluxes in batch cultures of three different engineeringstrains. The result of metabolic fluxes analysis indicated that flux in PP pathway hadno significant effect on riboflavin production, and the fluxes in the reactions fromRibulose-5-P to riboflavin must be enhanced for further increase of riboflavinproduction. The primary limiting factor for improvement of riboflavin production inB.subtilis RH33/pMX45 lies in the reactions between 5-P-ribulose and GTP.