Optimization of culture medium and modeling of curdlan production from Paenibacillus polymyxa by RSM and ANN

Abstract

Paenibacillus polymyxa ATCC 21830 was used for the production of curdlan gum for first time. A Box-Behnken experimental design was applied to optimize six variables of batch fermentation culture each at three levels. Statistical analyses were employed to investigate the direct and interactive effects of variables on curdlan production. Optimum cultural conditions were temperature (50°C), pH (7), fermentation time (96 h), glucose (100 g/L), yeast extract (3 g/L) and agitation speed (150 rpm). The yield of curdlan production was 6.89 g/L at optimum condition medium. Response surface methodology (RSM) and artificial neural network (ANN) were used to model cultural conditions of curdlan production. The maximum yield of curdlan production were predicted to be 6.68 and 6.85 g/L by RSM and ANN at optimum condition. The prediction capabilities of RSM and ANN were then statistically compared. The results showed that the ANN model is much more accurate in prediction as compared to the RSM. The infrared (IR) and NMR spectra, the thermogram of DSC and pattern of X-ray diffraction for the curdlan of the present study were almost identical to those of the commercial curdlan sample. The average molecular weight of the purified curdlan was determined to be 170 kDa by gel permeation chromatography.

Keywords: Curdlan gum; Optimization; Paenibacillus polymyxa.