Design of Pathway-Level Bioprocess Monitoring and Control Strategies Supported by Metabolic Networks - Measurement, Monitoring, Modelling, and Control of Bioprocesses - page 202

Fig. 3 1 H-NMR spectrum for a P. pastoris supernatant sample. The black line is the acquired
spectrum, whereas the red line is the estimated sum of the individual spectra for identified
metabolites. Looking more closely into specific regions of the spectrum it is possible to identify
key metabolites
Current analytical techniques for exo- or endometabolome analysis include
nuclear magnetic resonance (NMR) spectrometry [ 37 , 38 ] and mass spectrometry
(MS) [ 39 ]. Either of them can be coupled to separation methods for higher reso-
lution. These hyphenated methods include, for instance, capillary electrophoresis
coupled to mass spectrometry (CE-MS) [ 40 ], gas chromatography mass spec-
trometry (GC-MS) [ 41 ], and liquid chromatography coupled to nuclear magnetic
resonance spectrometry (LC-NMR) [ 42 ]. For a detailed review on the application
of such methods to metabolomics refer to [ 5 , 43 , 44 ].
A fast and low-cost technique is 1 H-NMR. The time of spectral acquisition
ranges from 2 to 10 min per sample, and automatic samplers can be used. Figure 3
shows a 1 H-NMR spectrum of the supernatant of P. pastoris culture samples. In
preliminary offline tests with this technique we detected over 20 metabolites in the
extracellular phase. Bundy and coworkers [ 45 ] have detected over 80 metabolites
in the extracellular medium of P. pastoris cultures using 1 H-NMR and GC-MS as
complementary techniques.
Knowledge of the metabolome is useful since it is very closely related to
cellular phenotype. Because changes upstream accumulate downstream, changes
in the transcriptome and proteome are found amplified in the metabolome. As a
result, the metabolome allows the detection of changes that have a very small
effect on metabolic fluxes [ 37 , 46 ]. Metabolic fluxes, which can be regarded as the
phenotype of a cell, are regulated not only at transcription and translation levels,
but also by means of posttranslational events, and as such the metabolome is
considered closer to the phenotype than the transcriptome or proteome [ 34 , 47 ].