Advancing Raman model calibration for perfusion bioprocesses using spiked harvest libraries

Abstract

Background Raman spectroscopy has gained popularity to monitor multiple process indicators simultaneously in biopharmaceutical processes. However, robust and specific model calibration remains a challenge due to insufficient analyte variability to train the models and high cross‐correlation of various media components and artifacts throughout the process. Main Methods A systematic Raman calibration workflow for perfusion processes enabling highly specific and fast model calibration was developed. Harvest libraries consisting of frozen harvest samples from multiple CHO cell culture bioreactors collected at different process times were established. Model calibration was subsequently performed in an offline setup using a flow cell by spiking process harvest with glucose, raffinose, galactose, mannose, and fructose. Major Results In a screening phase, Raman spectroscopy was proven capable not only to distinguish sugars with similar chemical structures in perfusion harvest but also to quantify them independently in process‐relevant concentrations. In a second phase, a robust and highly specific calibration model for simultaneous glucose (root mean square error prediction [RMSEP] = 0.32 g L−1) and raffinose (RMSEP = 0.17 g L−1) real‐time monitoring was generated and verified in a third phase during a perfusion process. Implication The proposed novel offline calibration workflow allowed proper Raman peak decoupling, reduced calibration time from months down to days, and can be applied to other analytes of interest including lactate, ammonia, amino acids, or product titer.

Raman‐controlled pyruvate feeding to control metabolic activity and product quality in continuous biomanufacturing

“…Although only one Raman model is required for feedback control, proper decoupling of other compounds such as glucose and pyruvate by harvest library spiking is of crucial importance to obtain independent and accurate Raman prediction models. [ 34 ] Furthermore, spike validation is required to distinguish models purely relying on correlation to other process compounds from models based on compound‐specific Raman bands rendering them suitable for real‐time monitoring and feedback control.…”

“…Raman model calibration was performed based on a design of experiment (DoE) approach by harvest library spiking. [ 34 ] As the final product quality control strategy was not defined at this stage, Raman prediction models for lactate, glucose, pyruvate, and ammonium were generated. These compounds were all considered interesting metabolic markers for perfusion cell culture, therefore representing potential targets to be monitored and controlled by a Raman feedback control strategy.…”

“…Raman model calibration was performed based on a design of experiment (DoE) approach by harvest library spiking. [34] As the final product quality control strategy was not defined at this stage, Raman prediction models for lactate, glucose, pyruvate, and ammonium were generated.…”

“…Model calibration was performed by spiking a harvest library as described previously. [34] A peristaltic pump (Ismatec Reglo ICC, Ismatec SA, Barrington, USA) connected to the Raman flow cell was used as an offline calibration setup for rapid spectral acquisition.…”

Process intensification for CHO cell biomanufacturing

“…These specific Raman models need to be trained and calibrated upfront as they depend on the surrounding sample matrix. A respective model calibration workflow for perfusion processes was recently reported using spiked harvest libraries 204 . This allows for certain control actions to be applied, e. g. by controlling a separate glucose feed 205 next to the usual perfusion medium which could optimize media usage beyond CSPR control.…”