Perfusion cultures of CHO cells producing t-PA were performed using acousti
c filter cell retention. A robust off-line glucose analysis and predictive
control protocol was developed to maintain the process within approximately
0.5 mM of the glucose set point, without the need for a more fallible on-l
ine sensor. Glucose usage (the difference between the inlet and reactor glu
cose concentrations) provided an easily measured indicator of overall mediu
m utilization for mapping acceptable ranges of operation, including the edg
e of failure. Earlier onset of perfusion with a ramping glucose set point (
1.5 mM/d) resulted in improved growth and consistency during the perfusion
culture start-up. At steady state, the t-PA concentration variability incre
ased gradually with increasing glucose usage up to approximately 22 mM, the
n up to 24 mM the variability increased threefold. Peak t-PA concentrations
of over 90 mg/L were obtained by controlling at a glucose usage of approxi
mately 24 mM, but these t-PA levels were not sustainable for more than 3 da
ys. A consistent t-PA concentration of 40 mg/L was obtained at a glucose us
age of 21.5 mM. (C) 2001 John Wiley & Sons, Inc.