Optimization Of Avermectin Fermentation Process And Its Scale-up On Industrial Scale By Using Multi-Scale Parameter Association Analysis

Avermectins, a broad-spectrum and highly effective bio-pesticide, is a prominent member of the macrolide antibiotics produced by Streptomyces avermitilis. As an effective and low toxicity pesticidal antibiotic, avermectins exceeds 50% in the pesticide market in China. But avermectin B1a industry by fermentation biotechnology in China lags far behind international advanced level. Aimed at this situation, the fermentation process physiological characteristics, glucose feeding strategy and fluid dynamics of avermectin production by S. avermitilis as well as its industrial scale-up methodology were investigated based on multi scale parameter association analysis.1. Rational breeding of high avermectin B1a yield industrial strain and its physiological characteristic study under different inoculum typesStrain Biok Av-023 used as the control was employed on screening of high-avermectin B1a-yield mutants by rational screening using He as screening pressure and resistance of high reducing sugar concentration. The result showed that the maximum avermectin B1a production of a mutant stain AV60s-32 reached 4520 IU/ml, which was 23.4% higher than the control. Also, mutant av60s-32 can withstand a higher reducing sugar concentration (increased from 2% to 5%). Oxygen uptake rate (OUR) during fermentation process was significantly different when two inoculation types were employed,i.e. spores inoculum and vegetative inoculum. Although lower OUR (15～20 mmol/1/h) occurred with spore inoculum during the early stage of fermentation, a stable OUR (9～11 mmol/1/h) was obtained in the late stage of fermentation, which corresponded to 9.8% greater avermectin B1a production than that obtained with vegetative inoculum (4493 IU/ml) in 2 M3 fermentor.2. Optimization of nutritional requirements for avermectin producion and OUR control strategy in the early stage of fermentation processIndustrial strain requires very complex nutritent for avermectin production. Experiments showed that adding 0.01% Mg2+ in the slant medium increased spore formation and avermectin production of S. avermitilis. Using statistical methods (CCD) to optimize the fermentation medium, a new avermectin production medium was obtained as follows:corn starch 14.2%, cottonseed meal 1.2%, soya bean meal 2.7%. Avermectin B1a titer of 5235 IU/ml was obtained by the optimal medium in shake flask, which was 17.6% higher than the control. It was found that OUR was strongly influenced by cell growth and antibiotics production based on multi-parameters association analysis in fermentation process. Avermectin B1a biosynthesis could be effectively enhanced when OUR was stably regulated at an appropriate level in batch fermentation of S. avermitilis. Avermectin. B1a yield was improved by controlling maximal OUR between 15 mmol/l/h and 20 mmol/l/h during cell growth phase. This stimulation effect on avermectin B1a production could be attributed to the improved supply of propionic acid and acetic acid, the precursors of Avermectin B1a, in the cells. Hence, this OUR control method during cell growth phase should be applicable to avermectin industry.3. A novel glucose feeding strategy for avermectin production based on OUR regulation.The results showed that, avermectin B1a 6430 IU/ml was achieved based on OUR control feeding strategy, which was 22.5% higher than the control (5250 IU/ml), when controlling OUR in the range of 10-12 mmol/l/h during the fed-batch phase. OUR change was found to have close relationship with the trends of pH, organic acids, amino acids and some key enzyme activities of avermectin biosynthetic pathway. After 150 h of fermentation, the accumulation of intracellular pyruvate, propionic acid and acetic acid, which were precusor organic acids for avermectin biosynthesis, was significantly higher than that of the control. The analysis of the amino acids in the broth revealed that threonine and methionine achieved the maximum accumulation at 200 h, and alanine, valine, leucine and isoleucine were higher than that of the control. These data indicated that the precursor amino acids pool was abundant when OUR control was used to regulate the glucose feeding. Further study on the key enzyme activity in avermectin biosynthesis pathway showed that citrate synthase activity was significantly increased, suggesting that the TCA metabolic flux was increased, which was consistent to the results of organic acid analysis. Methyl malonyl coenzyme A activity greatly increased at 200 h, indicating that the node of the precursors for avermectin biosynthesis was strengthened. It was beneficial to the high rate of precursor supply for avermectin biosynthesis using the OUR feeding strategy.4. Scale up of avermectin B1a production by integrating fluid dynamics and cell physiology.A relationship between the broth fluid properties and the mycelial morphology was established through morphology quantitative calculation and fluid dynamics analysis. In addition, computational fluid dynamics (CFD) was employed for the analysis of fluid dynamics in 150 m3 fermentor. It was found that fluid characteristics of avermectin production on industrial scale was improved when novel glucose feeding strategy was adopted by analyzing gas holdup, shear stress, mixing intensity, turbulent character in the bioreactor using CFD simulation. Based on flow dynamic analysis using CFD and cell physiology study of avermectin B1a production, the novel glucose feeding strategy by OUR regulation was successfully scaled up in a 150 m3 fermenter, which would provide a scientific basis for other microbial fermentation process optimization and scale up.