Comparison and implementation of different control strategies for improving production of rHSA using Pichia pastoris

Abstract

Recombinant protein production by various host organisms is strictly regulated, which is determined mainly by host cell genetics and maintenance of process variables within desirable limits in a reactor. Since the process dynamics typically exhibit nonlinear interactions amongst the critical process variables, the commonly employed controllers are suboptimal for control of most bioprocesses. In this article, we present the implementation of an augmented controller for improving the production of recombinant human serum albumin (rHSA) using Pichia pastoris. The augmented controller consists of a decoupled adaptive control (DAC) and a decoupled input-output linearizing control (DIOLC) in a combined architecture that permits simultaneous control of substrate concentration and dissolved oxygen concentration, sampled at different rates during fed-batch fermentation. The performance of DAC-DIOLC controller was compared to the performance of DAC and DIOLC implemented individually on the process. Our results show that the augmented controller better regulates the substrate and dissolved oxygen (DO), along the desired control trajectory compared to the individual DAC and DIOLC controllers. The performance of the augmented DAC-DIOLC controller showed deviations of only 0.021 g/L from substrate set point compared to 0.09 g/L in PID and 0.031 g/L in DIOLC, and a 1.12% deviation from DO set point compared to 11.86% in PID and 2.58% in DAC. Overall, stricter control achieved using the augmented controller resulted in a 1.5 fold increase in rHSA production.