Metabolomic analyses using NMR
Analyses were performed either in vitro from perchloric acid (PCA) cell extracts or in vivo using freshly harvested cells. Extracts were prepared from 10-g (wet wt) cells quickly filtered, washed with pure water at 0°C, and thrown into liquid nitrogen. Frozen samples with 0.7 ml of 70% (v/v) PCA were ground to a fine powder with a mortar and pestle at liquid nitrogen temperature. The frozen powder was then thawed at 0°C and the resulting thick suspension was centrifuged at 15,000 g for 10 min at 0°C to remove particulate matter. The supernatant was neutralised to pH 5.0 with 2 M KHCO3 to precipitate PCA as KClO4, centrifuged at 15,000 g for 5 min and lyophilised. The freeze-dried material was dissolved in 2.0 ml water containing 10% 2H2O for further NMR adjustment and 1 μM sodium azide to avoid fermentation when unfrozen, and it was stored at -20°C.
In vitro NMR analyses were performed on a Bruker AMX 400 wide bore spectrometer (Bruker Instruments, Inc., Billerica, MA, USA) equipped with a 10-mm multinuclear-probe. The probe was tuned at 100.6 and 162.0 MHz for 13C- and 31P-NMR, respectively. The deuterium resonance of 2H2O was used as a lock signal. Spectra were recorded at 295 K. 13C-NMR spectra were the result of 3600 transients with a 6-s repetition time (6 h) recorded with 90° pulses (11 μs), a 20 kHz spectral width, and a Waltz-16 1H decoupling sequence with 2.5 W and 0.5 W during acquisition time and delay, respectively. Free induction decays were collected as 32,000 data points, zero-filled to 64,000 and processed with a 0.2-Hz exponential line broadening. 31P-NMR spectra were the result of 1000 transients with a 3.6-s repetition time (1 h) recorded with 70° pulses (15 μs), a 8.2 kHz spectral width, and a Waltz-16 1H decoupling sequence with 1 W during acquisition and 0.5 W during delay, respectively. Free induction decays were collected as 16,000 data points, zero-filled to 32,000 and processed with a 0.2-Hz exponential line broadening.
For 13C-NMR analyses, 4 μmol 1,2-cyclohexylenedinitrilotetraacetic acid (CDTA) was added to the PCA extract to chelate Mn2+, 150 μmol of maleate was added for calibration, and the pH was adjusted to 7.4. Spectra were referenced to the -CH = CH- peak of maleate positioned at 131.4 ppm. For 31P-NMR analyses, all divalent cations were chelated by the addition of sufficient amounts of CDTA, 2 μmol methylphosphonate was added for calibration, and the samples were buffered by addition of 150 μmol Hepes at pH 7.4. Spectra were referenced to the peak of methylphosphonate positioned at 22.67 ppm. The identification of the peaks of resonance was done by comparing the spectra of standard solutions of known compounds at pH 7.4 with that of the PCA extracts. The definitive assignments were made after running a series of spectra of the extracts spiked with authentic compounds, at different pHs to separate potentially overlapping peaks . To accurately quantify compounds identified on spectra, the intensities of their different resonance peaks were referred to those of the reference compounds added to samples before grinding. Twenty seconds recycling time was used to obtain fully relaxed spectra. The integration function of the spectrometer was utilized to compare the intensity of resonance peaks.
In vivo NMR analyses were performed on the same spectrometer equipped with a 25-mm multinuclear-probe. The deuterium resonance of 2H2O was used as a lock signal. Cells (10 g FW) were placed in a 25-mm NMR tube and oxygenated as described by Aubert et al.  with a perfusion flux of 20 ml min-1 sufficient for a perfect oxygenation of all cells at 22°C. The 4 l of perfusion NM contained the macro-nutrients (sucrose, KNO3, NH4NO3, KCl, Ca[NO3]2, and MgSO4) normally present in 200 ml of Lamport's or Murashige and Skoog's media, according to cell strains, which is sufficient for the growth of 10 g of cells over several days and limits decoupling-related temperature elevation at the level of analyzed cells. To further improve the signal-to-noise ratio, micro-nutrients, and in particular Mn2+, were not added to NM. The temperature of the perfusing NM was adjusted in a thermoregulated water bath outside the magnet.
13C-NMR spectra were acquired by accumulating 900 scans recorded with 90° pulses (70 μs) at 5.6-s intervals, a 20.7 kHz spectral width, and a Waltz-16 1H decoupling sequence with 4 W and 0.5 W during acquisition time and delay, respectively. Free induction decays were collected as 16,000 data points, zero-filled to 32,000, and processed with a 2-Hz exponential line broadening. Spectra were referenced to hexamethyldisiloxane contained in a capillary inserted inside the central outlet perfusion tube at 2.7 ppm. 31P-NMR spectra were acquired as described by Pratt et al. . Spectra were referenced to methylenediphosphonate (MDP, pH 8.9) contained in the same capillary, at 17.38 ppm.
The identification of the resonance peaks was performed by comparing the in vivo spectra of perfused cells to those of PCA extracts prepared from these cells and adjusted at different pHs, as described above. In vivo quantification was performed by comparing the spectra of analyzed cells with those of the extracts prepared from the same amount of cells, and using the MDP reference. The concentration of metabolites in the cytoplasm and in the vacuole were calculated as indicated in Pratt et al. . Cytoplasmic and vacuolar pH (cyt- and vac-pH) was estimated from the chemical shift of the pools of Pi (cyt- and vac-Pi) present in these two compartments as described by Gout et al. .
Other analytical methods
The cell samples FW and the growth of cell suspensions were measured as described by Bligny and Leguay . The oxygen uptake by cells was measured at 5°C and 22°C in their respective cell culture media. O2-uptake was monitored polarographically using a Clark-type oxygen-electrode (Hansatech Ltd King's Lynn, UK). 50 mg cell (FW) was stirred in the 1 ml measurement chamber filled with NM. The O2 concentration in the air-saturated NM was taken as 250 μM at 22°C and 360 μM at 5°C according to Truesdale and Downing . Uncoupled respiration was measured after adding 2 μM cyanide p-trifluoromethoxyphenylhydrazone (FCCP). For photosynthesis measurements, Arabidopsis cells were illuminated with 500 μmol m-2 s-1 PPFD.
The sucrose present in NM was measured as described by Bergmeyer , using invertase, hexokinase, and glucose 6-P dehydrogenase, and by 13C-NMR.
Neutral red was be used as a vital stain to detect the presence of dead cells in cultures.
When means ± SD are given, the statistical Student's t-test was applied to the data with P values ≤ 0.05.