Optimisation and kinetic modelling for the production of 5- aminolevulinic acid by Rhodopseudomonas palustris in the solid state fermentation

Abstract

The approach of the bioprocess system engineering (BPSE) serves as a systematic methodology to better understand the overall performance of complex biological system through optimisation process and development of a compatible macroscopic kinetic model. Based on the BPSE approach, the production of 5- aminolevulinic acids (ALA) by Rhodopseudomonas palustris (Rp) via solid state fermentation (SSF), using the palm empty fruit bunch as solid state medium was studied. Optimisation studies were carried out using a full-factorial design and the response surface methodology approach. A maximum ALA yield of 43.72 mg/kg was achieved under the following optimum conditions: moisture content of 63.13 %, incubation temperature of 30.3 °C, pH 7, inoculums density of 40 % (v/w), 3.64 mM glycine and 23.03 mM succinic acid for 48 hours via SSF. Three mathematical models including the Logistic, Gompertz and Luedeking-Piret models were proposed and compared based on their goodness of curve-fitting to the SSF experimental data. The Logistic model incorporated with Luedeking-Piret model was developed and best represented (R2 >0.95) the underlying kinetic behaviour of the growth of Rp, the formation of ALA and the consumption of substrates for the production of ALA by Rp in SSF at the optimum condition. The computed kinetic parameters including the maximum specific growth rate (μm= 0.232 h-1) with the maximum Rp biomass concentration (Xmax= 316.4 x 10-9 CFU.g-1) for the modelling of Rp growth; the growth-associated (α= 8.249 mg.kg-1.h-1) and non-growth associated (β = -1.660 mg.kg-1.h-1) coefficients for the modelling of ALA formation, and the Rp growth associated and the ALA formation associated on substrate consumption coefficient (YX/S = 0.132 and YP/S = 0.141) for the modelling of substrate consumption were evaluated. These values were then validated between the predicted data and the experimental data using the least square curve fitting analysis and the ordinary differential equation solver (ODE45) using the Matlab software.