Towards optimal substrate feeding for heterologous protein production in Pichia pastoris (Komagataella spp) fed‐batch processes under P_AOX1 control: a modeling aided approach

Abstract

BACKGROUND To improve the efficiency of a bioprocess, key parameters, such as yield, titer and productivity, must be considered. They are mainly dependent on the specific rates for product (qP) and cell growth (μ), and their correlation determines the most suitable feeding polices that should be applied. RESULTS The mathematical description of the Pichia pastoris (Komagataella spp) PAOX1‐based system (Mut+) expressing recombinant Rhizopus oryzae lipase (ROL), which includes cell growth, substrate consumption and product formation kinetics, was used to define the optimal profiles regarding methanol feeding‐rate (F), residual methanol concentration (S) and specific growth rate (μ). Optimal trajectories obtained using numerical optimization and algorithms based on S‐profile determination were first applied to manage the multiplicity of S‐states with both μ and qP. Once the S‐profile was determined, the F and μ profiles were obtained in a straightforward manner from mass balances and growth kinetics. From experimental results, the bioprocess efficiency improvement was confirmed in terms of a 2.2‐fold higher final titer and 3.4‐fold higher productivity compared with the best standard pre‐programmed exponential substrate feeding strategy. Additionally, 1.2‐fold higher productivity and 1.1‐fold higher final titer relative to the most promising feedback control of the methanol concentration strategy were obtained. CONCLUSION The whole model‐based approach was established as a versatile and simple platform, which would then be transferable to alternative modes of operation and even to produce other proteins of interest. © 2018 Society of Chemical Industry

Comparative study of μ‐stat methanol feeding control in fed‐batch fermentation of Pichia pastoris producing HBsAg: an open‐loop control versus recurrent artificial neural network‐based feedback control

“…By applying the proposed feedback feeding control system based on RNN, the efficiency of already established large‐scale production plants producing HBsAg in methylotrophic yeasts could be enhanced without considerable time and money investment. In general, this tool can be combined with other feeding strategies and other fed‐batch designs, especially novel and more complicated ones, to reduce any unexpected deviations in key parameters . The combination of mathematical models with computational models, based on machine learning, in the future, could significantly reduce the costs and time related to bioprocess development and scale‐up by decreasing the number of experimental investigations.…”

“…In general, this tool can be combined with other feeding strategies and other fed-batch designs, especially novel and more complicated ones, to reduce any unexpected deviations in key parameters. 11,21,36 The combination of mathematical models with computational models, based on machine learning, in the future, could significantly reduce the costs and time related to bioprocess development and scale-up by decreasing the number of experimental investigations.…”

“…Despite the technical and safety constraints related to methanol, P. pastoris with P AOX1 is still the most well‐characterized and widely studied strain; it has already been used as an industrial strain for commercial‐scale production of recombinant hepatitis B vaccine (HBV) for more than 20 years . In this specific industrial strain, hepatitis B surface antigens (HBsAgs), the main component of HBV, are produced intracellularly and then form virus‐like particles (VLPs) during the downstream process .…”

“…14 Apart from natural promoters, since the last decade, there has also been growing interest in using other synthetic-engineered promoters to increase the production rate without the major challenges related to P AOX1. 15,16 PFLD1, ICL1 PEX8, YPT1, PCAT1, PPMP20, PDAS2 and NPS are some of the highly regulated promoters with a high production level of heterologous proteins which were extensively described by Vogl et al [15][16][17] Despite the technical and safety constraints related to methanol, P. pastoris with P AOX1 is still the most well-characterized and widely studied strain; [18][19][20][21] it has already been used as an industrial strain for commercial-scale production of recombinant hepatitis B vaccine (HBV) for more than 20 years. [22][23][24] In this specific industrial strain, hepatitis B surface antigens (HBsAgs), the main component of HBV, are produced intracellularly and then form virus-like particles (VLPs) during the downstream process.…”

Real‐time optimization and control for polyhydroxybutyrate fed‐batch production at pilot plant scale

“…Different advances in sensors for bioprocess monitoring have contributed to the development of operational strategies in the fed-batch mode to enhance the efficiency in bioprocesses. 15 A recent work by Biechele et al 16 reviews the advances in this topic. On the other hand, definition and real-time implementation of optimal feeding profiles requires the combination of advanced control strategies and real-time optimization topics.…”