Liquids Analysis Techniques

Cell Culture Media: NMR Approaches to Evaluating Quality and Nutrient Consumption

The use of cell culture media in the production of biologics and biosimilars is a growing field in the pharmaceutical industry.  Typical media contain a defined mixture of nutrients, the majority of which being small molecules.  Precise control of the nutrient composition over time is critical to optimal cell culture performance and the quality of the resulting product.  Nuclear Magnetic Resonance (NMR) spectroscopy was used for monitoring the nutrient mixtures in this study to take advantage of its unique characteristics of being quantitative, non-destructive, and requiring only minimal sample preparation.1,2  An additional reason for selecting NMR for this analysis, was to simultaneously monitor the production of desired and un-desired components along with consumption of nutrients during the course of the reaction. Ability to monitor the components would enable adjustment of parameters or component quantity for optimal results.

Bruker’s AssureNMRTM was used to automate the identification and quantification of key mixture components in Dulbecco’s Modified Eagle Media (DMEM) AT0683 containing 4.5 g/L glucose and L-glutamine. This media additionally contains a selection of amino acids, vitamins and inorganic salts. The spectral database SBASE, BBIOREFCODE, containing 1D and 2D NMR spectra of 700 physiologically important metabolites at various pH values, was used to ‘populate’ an AssureNMR analysis ‘quant method’ with the expected DMEM components (Figure 1). SBASE entries of all organic components of the media recipe were present in BBIOREFCODE. Regions were adjusted to match actual resonances observed in the DMEM spectrum.

To test the AssureNMR method, 100 ul of pH 7.0 phosphate buffer in D2O was added to 500 ul DMEM media.  Spectra were obtained on a Bruker AVIII-500 HD spectrometer with a Prodigy CryoProbe. All concentrations were calibrated against a known external standard. Figure 2 shows the quantification results for several key components of DMEM obtained by AssureNMR.

Figure 1. SBASE entries at pH7 were used to populate the AssureNMR quant method to reflect the fermentation conditions and expected NMR sample conditions.


Figure 2. Quantification of key components in DMEM. Note, total D-glucose concentration = alpha + beta (2252.41 + 2198.25 = 4,450.66 mg/l).

Region and Resonance Selective Excitation for Enhancement of Minor Components

Monitoring the entire spectra of cell culture media was achieved utilizing well defined general water suppression sequences, in the case above, water presaturation.  In order to monitor specific nutrients/metabolites or specific classes of compounds such as aromatics, sugars, amino acids, etc. selective pulse sequences such as SELective Double Pulse Field Gradient Perfect Echo (SELDPFGPE) were used to selectively excite only the region/resonance of interest.  As a result, the ability to detect and quantify minor analytes in the presence of high concentrations of other matrix components was greatly enhanced.  Figure 3 demonstrates the use of the SELDPGPE technique.

Figure 3. Selective excitation of DMEM utilizing SELDPFGPE. A) Full spectrum with water presaturation, B) Region selective 10.0-6.0 ppm, C) Region selective 4.0-2.5 ppm, D) Resonance selective 5.30-5.00 ppm (alpha-D-glucose anomeric), E) Resonance selective 4.62-4.48 ppm (beta-D-glucose anomeric)

Monitoring of Nutrient Utilization and Product Formation.

An initial model system was designed by adding baker’s yeast to a sample of DMEM and monitoring by NMR at 3.0, 6.0, and 18.5 Hr.  As expected, the glucose was rapidly converted, mainly to ethanol.  An intermediate compound exhibiting an anomeric doublet at 5.918 ppm was observed.  Quantification of the other components monitored in Figure 2 (amino acids) showed little change.  An AssureNMR method was generated to quantitate glucose, the intermediate anomeric, ethanol, and acetic acid (Figure 4, Table 1).

Figure 4. Spectra of DMEM (black) and DMEM 3 Hr. after fermentation(red). a: alpha-D-glucose, b; beta-D-glucose, c: intermediate anomeric, d: ethanol, e: acetic acid.


Table 1. Quantities of reactants and products before fermentation and at 3.0, 6.0, and18.5 hours.

Effect of Deuterium Lock on Spectral Quality and Long Term Stability.

Conventional NMR knowledge dictates the addition of a deuterated lock solvent (such as D2O) to the media to obtain long term stability, of field and shims.  In a reaction monitoring situation, this would lead to an extra step of removing an aliquot at a given time, adding D2O, and then obtaining the spectra.  In order to determine spectrometer stability without a lock solvent, a spectrum of neat DMEM was obtained without lock utilizing the Topshim routine to shim on the residual water peak.  After 12 hours the sample was re-inserted, Topshim re-ran, and the spectrum obtained.  Note that the spectra are indistinguishable (Figure 5).


Figure 5. Spectra of neat DMEM unlocked (black) and after 12 Hr. (red).