Metabolic Regulation, Optimization And Soft Sensor Modeling Of Streptomyces Pristinaespiralis Fermentation For Pristinamycin Production

Pristinamycin is an antibiotic of the streptogramin family produced by Streptomyces pristinaespiralis and consists of two components, i.e., pristinamycin Ⅰ (PⅠ) and pristinamycin Ⅱ (PⅡ). Due to its strong, prolonged activity against most gram-positive bacteria including drug-resistant pathogens, pristinamycin is considered as a preferred antibiotic to fight against stubborn gram-positive bacterial infections. To increase the production of pristinamycin during fermentation, this study sought to ellucidate the effects of medium components on the biosynthesis of pristinamycin with an emphasis placed upon the screening of appropriate oxygen vectors and fatty acid or amino acid precursors for inclusion, and to establish an optimized strategy that enables to enhance pristinamycin yield by feeding pristinamycin precursor. The results are summarized below.Effects of key medium components on pristinamycin biosynthesis. The effects of glucose, ammonium ions and phosphate on pristinamycin biosynthesis during the submerged culture of Streptomyces pristinaespiralis F213in flasks were assesed. Pristinamycin biosynthesis was suppressed by high concentration of glucose, ammonium ions, and phosphate in basal fermentation medium. The concentrations of glucose, phosphate, and ammonium ions exceeding30g/L,1.5mM and15mM in the basal medium respectively resulted in decreased pristinamycin yields. The high glucose and phosphate concentrations were found to increase glucose-6-phosphate accumulation for the suppression of glucose catabolism and thus pristinamycin biosynthesis, while concentrated NH4+could inactivate some enzymes involved in Embden-Meyerhof-Parnas (EMP) pathway and Hexose Monophophate pathway (HMP) but activate enzymes involved in tricarboxylic acid (TCA) cycle.Effects of oxygen vectors on pristinamycin biosynthesis. Cyclohexane, N-heptane, N-hexane and hexadecane as added oxygen vectors were compared with blank control for their effects on pristinamycin biosynthesis in flask cultures. Of those, hexadecane was the best oxygen vector to improve oxygen transfer, resulting in a pristinamycin yield97.4%higher than control. After10%(w/v) emulsified hexadecane (with0.01%dimethyl silicone) was added to the culture at the time of36h fermentation, a maximal pristinamycin yield of180.6mg/L was achieved after the fermentation continued for36h. This yield was3.1fold of the control counterpart. In addition, the emulsified hexadecane was superior to its un-emulsified form in the improvement of oxygen and nutrient transfer, thus favoring the biosynthesis of pristinamycin.Fermentation optimization based on the biosynthesis pathway of pristinamycin. Based on the known biosynthesis pathway and the metabolic regulation of pristinamycin, the fermentation strategy was further optimized to increase the yield of pristinamycin in flasks and3L bioreactor. Five fatty acids (or alcohol) and seven aminoacids were compared as pristinamycin precursors for their effects on pristinamycin biosynthesis. Of those,0.2%propanol added to the fermentation medium resulted in135.2%increase in pristinamycin yield while inclusion of1g/L glycine in the medium enhanced the yield by1.9fold compared with the control yield. The feeding concentration and time of glycine was furher optimized as0.75g/L at the time of36h fermentation. This optimized glycine feeding strategy was applied to enlarged3L bioreactor fermentation with8%resin added at the time of20h fermentation for in situ separation. The combination of resin and glycine feeding resulted in the maximal pristinamycin yield of616mg/L at the time of12h after glycine feeding. This yield was1.71,2.77and4.32fold of those from the mere glycine and resin treatments and the control, respectively. The results indicated that glycine feeding is an effective approach to enhancing pristinamycin production in the culture of S. pristinaespiralis F213with supplemented resin for in situ separation.Soft sensor modeling in fermentation process based on support vector machine. Support vector machine (S VM) theory was adopted to establish S VM model for fitting the mycelium biomass and pristinamycin yield measurements of S. pristinaespiralis F213from the3L bioreactor. The model provided better fitness to the pristinamycin yield (R2=0.987) than the biomass (R2=0.863).