Online monitoring of the SARS-CoV-2 VLP bioreactor production process by Raman spectroscopy

Abstract

In the development of biopharmaceuticals, it is essential to optimize the cost and quality of the final product. To this end, it is essential to monitor and control critical parameters of the procedure from the bench scale to facilitate scaling up to the industrial plant while maintaining the quality of the bioproduct. Techniques for monitoring the process, such as spectroscopy, are used to guarantee such conditions. Specifically, in online monitoring, results are generated continuously and preferably without compromising the biochemical integrity of the system inside the bioreactor or throughout the purification steps. Raman spectroscopy is a non-destructive, non-invasive, and highly specific method that facilitates monitoring, whose data can be correlated using chemometric methods to generate predictive models of the biochemical parameters of interest. Therefore, the present project aims to analyze Raman spectroscopy data captured every 13 seconds in online monitoring of the production of SARS-CoV-2 virus-like particles (VLP) in a bioreactor, correlating them using multivariate analysis techniques of partial least squares (PLS) and artificial neural network (ANN). Different spectra filtering techniques will also be applied in advance to improve the models. Through statistical comparison and simulation, it will be possible to identify the best models generated, capable of estimating nutrients, metabolites, cell density, concentration of viable cells, and viral titration, allowing monitoring, optimization, and control in the next stages of bioprocess development up to commercial scale at a relatively lower analytical cost and time.