Continuous Production of Penicillin-G Using Immobilized Living Cells of P̀enicillium Chrysogenum'

A fluidized-bed bioreactor system using cells of P. chrysogenum immobilized via a previously developed polymer coating technique was investigated for the continuous fermentation of penicillin-G. Early results indicated that one important variable was the rate of cell growth: overgrowth of the bioreactor volume by immobilized biocatalyst significantly reduced the useful lifetime of the continuous fermentations. Current research has focused on efforts to control the rate of cell growth via nutrient limitation. Initial studies involving the use of phosphate-limitation revealed that, in the absence of com steep liquor, phosphate level had no effect on cell growth. This implied that com steep liquor (or some component thereof) was necessary for achieving measurable production-phase cell growth. The effect of com steep liquor on both cell growth and penicillin formation was quantified through a series of immobilized cell repeated-batch shake flask experiments. A com steep liquor level of 2 g/L and a carbon-source (lactose) level of 4 g/L resulted in maximal penicillin-G formation with relatively little cell growth. Semi-empirical kinetic expressions based on Monod and Monod substrate inhibition models were developed to describe the effects of both com steep liquor and lactose on cell growth and product formation. Continuous fluidized-bed fermentations were performed using the optimal levels of com steep liquor and lactose identified during the repeated-batch shake flask investigations. The resulting fermentations lasted up to 480 hours with no loss of mixing or aeration capacity, and no significant increase in the average diameter of the bioparticles. This simultaneous optimization of both product formation and cell growth is essential to the successful extension of the useful production phase during continuous operation. Productivities and yields of penicillin-G in the repeated-batch and continuous fermentations were 4 times greater than in the traditional free cell fed-batch approach. A mathematical model consisting of a series of material balances and kinetic expressions was developed which incorporated the effects of com steep liquor and lactose on both cell growth and product formation. The model predictions were compared with experimental results from the continuous fluidized-bed fermentations, and were found to accurately predict the penicillin-G and lactose concentration profiles.