Bioprocess Convergence Using Sentinel Genes for Process Parameter Tuning

Abstract

In vitro cell culture is a vital part of therapeutic biomolecule, cell therapy and vaccine production. In order to assure product identity, safety, potency and efficacy, controlled manufacturing processes are essential. There is a great need for a new approach of genomic translational studies for setting measurable cellular attributes, for enhanced process understanding during scale-up. The applications of setting such measures are multitude; for process parameter tuning to optimally scale-up a bioprocess, and be able to predict process outcomes. Although bioreactor scale-up has been thoroughly studied by matching fluid mixing and mass-transfer parameters, for example kLa, P/V, and OUR, there is still a gap in knowledge of cellular response itself to environmental changes during scale-change. As a starting point we performed scale-down based on matching kLa, from a bubble-aerated, impeller-agitated bench-scale 5L bioreactor to the novel 30mL minibioreactor, which resulted in growth and product related discrepancies. For identifying sentinel genes during bioprocess, we focused on differentially expressed transcripts between the two systems having cellular functions such as oxidative stress response, DNA damage response, apoptosis, cellular metabolism, and protein folding. In this study we have 1)Provided a novel resource in the form of baseline gene expression profiles of a mouse hybridoma cell line in a typical 5L bench-scale bubble-aerated and impeller-agitated bioreactor, from early-exponential-phase to early-death-phase in batch culture. Gene expression profiles that were variable as well as invariant during the course of 5L versus 30mL-minibioreactor culture are summarized from microarray findings and verified first by semi-quantitative RT-PCR. 2)Identified and validated total of 18 cell-type specific sentinel gene transcripts that were differentially expressed as a result of mixing and mass-transfer parameter differences during bioreactor scale-change. Using these sentinel genes, we correlated the cell growth parameter differences due to insufficient scale-down, activation of free-radical induced apoptosis protection gene Superoxide dismutase, along with lower Cytochrome C apoptosis gene expression, and adaptation of cellular metabolism to the challenges imposed by its micro-environment. 3)Improved scale-down by tuning process parameters such as mixing and bubble size, guided by gene analysis. This approach resulted in achieving better growth and environmental profiles convergence, between the two model systems. 4)Demonstrated sentinel gene profile convergence as a result of bioprocess convergence, thus strengthening the sentinel gene findings, using quantitative real-time-PCR. The results of this study would facilitate setting cellular attributes for process prediction at the physiological level, for bioprocess scale-up. We envision that this study would serve as a starting point for developing quantitative-RT-PCR assays that could be performed near-at-line during the cell culture process itself.