Protease production by a marine bacterium Teredinobacter turnirae

The production of extracellular alkaline protease by a shipworm bacterium Teredinobacter turnirae was studied in both defined and complex media. The research focused on the improvement of protease level. Investigations were carried out in shake-flask as well as 1-L bioreactor. In the first part of the study, some fermentation conditions such as carbon source, inducer, agitation rate and air rate on protease production were performed with batch culture in defined medium. The maximum enzyme activity value obtained in Plackett-Burman media (PB) was higher than that obtained in Standard media (SM). The second part of the work, the protease production by T. turnirae fermentation using different fed-batch feeding strategies was studied. The effects of different fed-batch methods such as pulse fed-batch and constant feed rate fed-batch on protease production were determined. Compared with batch culture, the fed-batch culture had better protease production. Mass transfer characteristics of fermentation broth in batch bioreactor were investigated. Increases in both the agitation rate and the air feed rate increased volumetric mass transfer coefficients, kLa and oxygen uptake rates, ro values. V VI Production of protease by a shipworm bacterium T. turnirae, under solid substrate fermentation (SSF) was optimised. Corn protein, soybean flour and casein were used as a sole carbon and nitrogen sources. The appropriate particle size, substrate concentration, initial pH, inoculum level, agitation rate, supplementation of additional carbon sources and medium temperature was determined in shake-flask. Concurrently, the effects of agitation speed and air rate were investigated in a 1-L bioreactor with the complex medium. The fermentation kinetics of protease by T. turnirae were studied in a batch system. This study focuses on comparing different kinetic models in literature and the use of neural networks model. Various kinetic parameters were obtained for protease fermentation optimising with Genetic Algorithm Toolbox. The experimental results of growth, protease production and sucrose consumption are compared model results, and neural networks model appeared to provide a reasonable description. Key Words : Alkaline protease, Teredinobacter turnirae, Solid substrate fermentation, Process conditions, Batch fermentation, Fed-Batch fermentation, Kinetic modelling, Genetic Algorithm, Neural network modelling.