Statistical design of experiment was applied to simultaneous determination of the effect of two nutritional factors, namely yeast extract and corn steep liquor, on the fermentation production of adenosine by a mutant strain of Bacillus subtilis ATCC 21616. A 22 full-factorial central composite design followed by canonical analysis of the response surfaces was employed for experimental design and analysis of the results. A second-order polynomial model was fitted to predict adenosine production given specific levels of nutrients and limited within the tested range Optimum conditions for adenosine accumulation were estimated by means of contour plot surfaces and found to be 7.27 g/1 for yeast extract and 51.84 ml/1 for corn steep liquor. Under optimum conditions,9.84 mg of adenosine per ml was predicted by the empirical. A verification experiment conducted under optimal conditions resulted in an average accumulation of 9.62 mg of adenosine per ml (97.8% of the predicted production) after 72h cultivation in a 250 ml conical flask at 32℃. This represented an improvement over 58.22% of the production achieved in the initial experiments. Canonical analysis of the response surface revealed a point of maximum located within the design space. Supplementation of antibiotics and surface-active agents, as stimulators, resulted in considerable improvement in adenosine accumulation. Among them, ampicillin and Tween 80 had the more significant stimulating effect on adenosine accumulation. In the medium supplemented with 1.0% Tween 80,12.04 mg of adenosine per ml was accumulated. Their addition time during the culture period appeared to be critical for the improvement. Furthermore, it was found that the concentrations of sodium citrate and magnesium sulfate were important. Under optimal conditions,10.65 mg and 11.96 mg of adenosine per ml were accumulated in the presence of sodium citrate (0.1%, w/v) and magnesium sulfate (0.12%, w/v), respectively. Magnesium phosphate and EDTA did not improve adenosine productivity.
|