Swainsonine, a novel fungal metabolite: optimization of fermentative production and bioreactor operations using evolutionary programming

The optimization of bioreactor operations towards swainsonine production was performed using an artificial neural network coupled evolutionary program (EP)-based optimization algorithm fitted with experimental one-factor-at-a-time (OFAT) results. The effects of varying agitation (300-500 rpm) and aeration (0.5-2.0 vvm) rates for different incubation hours (72-108 h) were evaluated in bench top bioreactor. Prominent scale-up parameters, gassed power per unit volume (P (g)/V (L), W/m(3)) and volumetric oxygen mass transfer coefficient (K (L) a, s(-1)) were correlated with optimized conditions. A maximum of 6.59 +/- A 0.10 mu g/mL of swainsonine production was observed at 400 rpm-1.5 vvm at 84 h in OFAT experiments with corresponding P (g)/V-L and K (L) a values of 91.66 W/m(3) and 341.48 x 10(-4) s(-1), respectively. The EP optimization algorithm predicted a maximum of 10.08 mu g/mL of swainsonine at 325.47 rpm, 1.99 vvm and 80.75 h against the experimental production of 7.93 +/- A 0.52 mu g/mL at constant K (L) a (349.25 x 10(-4) s(-1)) and significantly reduced P (g)/V (L) (33.33 W/m(3)) drawn by the impellers.