The purpose of this study was to develop non‐invasive techniques to monitor the composition of cell culture media in insect cell bioreactors. Such a monitor could be used in conjunction with a fed‐batch feeding scheme to ensure that cells are maintained in an optimal environment for growth and protein production. Glucose and glutamine concentrations in an insect cell culture bioreactor were determined off‐line with near‐infrared (NIR) absorption spectroscopy. Spectra were collected from 5000 to 4000 cm−1 with a 1.5‐mm optical path length. Partial least squares (PLS) regression was applied to correlate the collected spectra with the concentration of the desired analytes. Under the culture conditions evaluated here, glucose and glutamine concentrations ranged from 38 to 55 mM and from 3 to 13 mM, respectively. Accurate measurements of glucose and glutamine in insect cell culture samples were possible over these entire ranges. The standard error of prediction (SEP) and mean percent error (MPE) for glutamine were 0.52 mM and 5.3%, respectively. Glucose could be measured with an SEP of 1.30 mM and an MPE of 2.3%. These levels of error are quite low considering the changing complexity of the growth media due to the shifting levels of amino acids, carbohydrates, yeastolate, proteins, and cell debris. This study represents an important step in the development of noninvasive on‐line monitoring devices for cell culture bioreactors. © 1997 John Wiley & Sons, Inc. Biotechnol Bioeng 55: 11–15, 1997.

No abstract

A novel fiber-optic sensor for nitric oxide is constructed by holding a small amount of an internal reagent solution at the tip of a fiber-optic bundle with a piece of gaspermeable membrane. Nitric oxide diffuses across the membrane into this internal solution, where a chemiluminescent reaction between nitric oxide, hydrogen peroxide, and luminol takes place. The resulting light intensity is related to the concentration of nitric oxide in the sample. Results are presented from experiments to optimize the magnitude and rate of the sensor response. The resulting sensor possesses a limit of detection of 1.3 µM, a response time of 8-17 s, and a dynamic range from 5 to 40 µM. A mathematical model is derived to explain the sensor response as a function of time. Oxidation of nitric oxide by ambient oxygen and mass transport of nitric oxide through the gas-permeable membrane are considered in this model. The excellent agreement between experimental data and model predictions indicates that the oxidation of nitric oxide by oxygen is a major factor in this measurement, regardless of transducer element. The model further indicates a reaction rate constant for the oxygen oxidation of nitric oxide of 9.5((0.5) × 10 6 M -2 s -1 and a diffusion coefficient for nitric oxide in the silicone membrane of 5.0((0.2) × 10 -11 m 2 s -1 at 30 °C.

No abstract

No abstract