Article ID Journal Published Year Pages File Type
6491629 Journal of Biotechnology 2014 6 Pages PDF
Abstract
Near infrared spectroscopy is demonstrated as a suitable method for monitoring real time cell density and concentrations of glucose and lactate during insect cell cultivation. The utility of this approach is illustrated during the cultivation of Trichoplusia ni BTI-Tn-5B1-4 insect cells in a stirred-tank bioreactor. On-line near infrared measurements are made by passing unaltered culture medium through an autoclavable near infrared flow-through sample cell during the cultivation process. Single-beam near infrared spectra were collected over the combination spectral range (5000-4000 cm−1) through a 1.5 mm path length sample. Cell density calibration model was established by uni-variable linear regressions with measured mean absorbance values of on-line spectra collected during a cultivation run. Calibration models are generated for glucose and lactate by regression analysis of both off line and on line spectra collected during a series of pre-measurement cultivation runs. Analyte-specific calibration models are generated by using a combination of spectra from both natural, unaltered samples and samples spiked with known levels of glucose and lactate. Spiked samples are used to destroy concentration correlations between solutes, thereby enhancing the selectivity of the calibration models. Absorbance spectra are used to build partial least squares calibration models for glucose and lactate. The calibration model for cell density corresponds to a univariate linear regression calibration model based on the mean absorbance between 4750 and 4250 cm−1. The standard errors of prediction are 1.54 mM, 0.83 mM, and 0.38 × 106 cells/mL for the glucose, lactate, and cell density models, respectively.
Related Topics
Physical Sciences and Engineering Chemical Engineering Bioengineering
Authors
, , ,