Analyzing Mammalian Cell Culture using Raman Spectroscopy

By AZoM

Table of Contents

Introduction
Benefits of Raman Systems
Experimental
Data Analysis
Conclusions
About Kaiser Optical Systems

Introduction

Mammalian cell cultures are complicated processes by which cells are cultured in controlled conditions. In the industrial scenario, mammalian cultures are basic to the manufacture of viral vaccines and other biological products. Chinese hamster ovary cells in the 1980s began revolutionizing mass culture for medicines and continue to remain the workhorse of the industry for therapeutic protein production.

The work described here involves development and application of new bioprocess analytics according to the aims of Quality by Design (QbD) and the FDA Process Technology Initiative of the US. Present research is focused on obtaining a better understanding of mammalian cell cultures by cultivating CHO cells for therapeutic protein production. In order to ensure healthy progression of the cell culture, it is important to understand and monitor the stages of this bioprocess.

Benefits of Raman Systems

Raman spectroscopy is useful for PAT and QbD applications as it allows real-time, quick, in-situ monitoring and bioprocess control. In complicated systems such as bioprocess cell cultures, it is important to quantitatively study specific components without interference from other species in the system. Vibrational spectroscopic techniques such as near-infrared spectroscopy and Raman spectroscopy produce fingerprint-like spectra of biological and organic molecules so that specific peaks can be selected to be used in quantitative analysis. These techniques have also been deployed successfully for control of bioprocesses and real-time monitoring as they can obtain data in a few seconds, require minimal sample preparation and are agreeable to immersion or non-contact sampling optics.

Choosing between NIR and Raman comes down to the fact that water has a very weak Raman profile so there is no interference with species analysis in aqueous systems that are common in the bioprocessing environment. Raman and NIR spectra of glucose and lactic acid in water overlaid with a pure water spectrum is shown in Figure 1. Raman spectra show high chemical specificity with little interference from water, whereas, water dominates the NIR spectra.

Figure 1. Comparison of Raman and NIR spectra of glucose and lactic acid in water. Raman spectra were processed using a 3rd order baseline correction and a 1st order derivative was applied to NIR spectra.

Experimental

The spectroscopic specificity of CHO cell culture components has been studied in this research.

Te experimental steps followed are:

  • Aqueous laboratory samples of glucose, lactic acid, and glutamine were prepared to typical concentrations seen in mammalian cell cultures and analyzed by Raman spectroscopy using a RamanRxn2 analyzer equipped with a 785-nm Invictus laser and a 457 mm in situ probe.
  • Conducting of a series of four CHO cell culture batches, referred to as Batches 1-4, was done using a 2-L autoclavable glass bioreactor. Seed cultures were grown in spinner flasks, and the bioreactor was inoculated at a cell density of 2 x 105 cells/mL.
  • Bioreactor conditions were chosen to increase cell growth rate and all conditions and control parameters used were intended to simulate industrial fed-batch CHO cell culture processes.
  • Grab samples were removed from the bioreactor approximately twice per day and analyzed off-line using a BioProfile® Basic 100 automated analyzer throughout the course of each grab.
  • The Raman probe was inserted directly through the headplate into the bioreactor and Raman spectra were collected continuously every 20 minutes (30 s exposure, 40 accumulations) until the run was completed.

Data Analysis

Several analysis methods were used to analyze the in-line Raman data. The simplest Raman peak area trending was used for observing relative changes in glucose concentration over batch time by trending the intensity of a glucose peak located at 1135 cm-1. A quantitative univariate model for glucose concentration was then created by correlating off-line reference data to the baseline integrated areas under the 1135-cm-1 glucose peak; spectra that were collected just before and after samples were removed from the bioreactor were matched to off-line samples.

The univariate calibration model was built using data from Batches 1-3. Finally, partial least squares (PLS) multivariate calibration models were built for glucose, lactate, and glutamine using a multivariate data analysis package (GRAMS PLSplus/IQ™).Reference measurements collected during Batches 1-3 for each component were correlated to larger regions of the Raman spectra using PLS (one model per constituent). Univariate and PLS models were then used to predict Batch 4 (validation batch) concentrations.

Conclusions

As observed in Table 1 and Figures 2 and 3, the calibration model developed using Batches 1–3 was highly effective in predicting the component concentrations in Batch 4, with R2 values of 0.99 for each component. These results show that Raman is suited for real-time, in situ monitoring and analysis of cell cultures. Raman offers simple and precise analysis of aqueous-based systems and chemical specificity for cell culture nutrients and metabolites.

Table 1. Summary of calibration and validation results for glucose, lactate and glutamine PLS models. SEC = standard error of calibration, SEP = standard error of prediction.

Constituent Self Prediction Cross Validation Batch 4 Validation
R2 SEC (g/L) R2 SECV (g/L) SEP (g/L)
Glucose 0.99 0.16 0.97 0.33 0.34
Lactate 0.99 0.06 0.96 0.18 0.14
Glutamine 0.99 0.03 0.97 0.04 0.06

Figure 2. Calibration results for univariate and PLS glucose models. SECV = Standard error of cross validation.

Figure 3. Validation results for glucose: univariate (A) and PLS (B), lactate (C) and glutamine (D) models showing Batch 4 measured and Raman predicted concentrations over time.

About Kaiser Optical Systems

Kaiser Optical Systems, Inc. is a world leader in spectrographic instrumentation and applied holographic technology. Principal products include Raman sensors and instrumentation, advanced holographic components for spectroscopy, telecommunications, astronomy and ultra-fast sciences, and display systems for aircraft. Principal offices and the manufacturing facility are located in Ann Arbor, Michigan.

Their products and services are deployed throughout the world in such diverse applications as pharmaceutical and chemical manufacturing, nanotechnology, telecommunications, education, homeland defense, deep-sea exploration and astronomy. From particles smaller than a human hair to objects as large as planets, these products are providing our customers unique insights into both today’s as well as “age-old” questions.

Kaiser Optical Systems was founded in 1979 to meet the need for diffractive or holographic optics for the Head-Up Display (HUD) market. Kaiser entered the spectroscopy market in 1990 with the introduction of the holographic notch filter. In 1993 Kaiser released their first Raman analyzer, the HoloProbe. The company was founded in 1979 and is a subsidiary of Rockwell Collins.

To better serve the European community, in 1998 Kaiser Optical Systems opened a new subsidiary in Europe. Kaiser Optical Systems SARL is located in Lyon, France. Kaiser Optical Systems SARL supervises our distributor network within Europe.

This information has been sourced, reviewed and adapted from materials provided by Kaiser Optical Systems.

For more information on this source, please visit Kaiser Optical Systems.

Date Added: Apr 9, 2013 | Updated: Jun 11, 2013
Ask A Question

Do you have a question you'd like to ask regarding this article?

Leave your feedback
Submit