Comparative metabolicprofiling to investigate the contribution of O. oeni MLF starter cultures to redwine composition
Sulette Malherbe • Andreas G.J. Tredoux • Hélène H. Nieuwoudt • Maret du Toit*
*
Supplementary material
HS-SPME method optimization and validation (supplementary material)
A method was developed for the quantification of diacetyl, acetoin, 2,3-pentanedione and a selection of aldehydes using a HS-SPME GC-MS method. The sample preparation procedure was optimized (using a PEG fiber) by evaluating different sample volumes (5 ml, 10 ml, 20 ml) (data not shown), sample dilutions (undiluted, 5x, 10x) (Fig. 1) and sodium chloride (NaCl) concentrations (1 g, 2 g, 3g) (Fig. 2) to obtain optimal peak shape and sensitivity for all the analytes of interest. It was found that although different combinations of dilution and salt addition provided variable sensitivity for different compounds, the combination of diluting the sample 10x together with the addition of 2g NaCl provided the best overall sensitivity and chromatographic performance by minimizing interfering matrix effects, especially that of ethanol.
Fig. 1 Different sample dilutions (undiluted, 5x, 10x) evaluated during sample preparation optimisation for the analysis of carbonyl compounds in wine.
Fig. 2 Different sodium chloride (NaCl) concentrations (1g, 2g, 3g) evaluated during sample preparation optimisation for the analysis of carbonyl compounds in wine.
In addition to the optimization of the sample preparation technique, a number of fibers with different extraction phases (CAR/DVB/PDMS), polyacrylate (PA), polydimethylsyloxane/divinylbenzene (PDMS/DVB) and carbowax/divinylbenzene (CW/DVB)], different extraction times (10 min, 30 min and 60 min), extraction temperatures (30˚C, 50˚C and 80˚C), injection modes (split, splitless and pulsed split), split ratios (1:2, 1:5 and 1:10), injection temperatures, desorption times and temperatures were evaluated during the method development stages (data not shown). The optimum conditions were selected according to selectivity in the case of the fiber selection and chromatographic peak shape and intensity obtained with regard to the other parameters tested. A chromatogram of the optimised method is shown in Fig. 3. Method calibration and a selection of the validation data is shown in Table 1.
Fig. 3 A GC-MS chromatogram (SIM mode) showing carbonyl compounds associated with metabolism of LAB in a synthetic wine that could be determined with a quantification limit of 5 μg/l for the compounds listed in Table 1. The chromatogram corresponds to a wine spiked with 15 μg/lof analytes
Table 1 Peak identification, slope of calibration graphs, linear dynamic ranges, coefficients (R2), limits of detection (LODs) and reproducibility, R.S.D. (%)
Peak / Compounds / tR (min) / m/z / Linearitya / Range(μg/l) / LOD
(mg* or μg/l) / R.S.D. (%)b
Slope / R2
1 / 2-pentanone* / 10.5 / 43, 86
2 / diacetyl (2,3-butanedione)c / 10.7 / 43, 86 / 0.0136 / 0.985 / 1-50 / 1* / 4.82
3 / 2,3-pentanedione c / 13.6 / 57, 100 / 0.0204 / 0.998 / 1-50 / 0.1* / 6.03
4 / E-2-hexenal / 19.2 / 69, 83, 98 / 0.09 / 0.999 / 1-50 / 5 / 10.6
5 / octanal / 21.3 / 69, 84, 110 / 0.2684 / 0.994 / 1-50 / 0.1 / 3.36
6 / acetoin (3-hydroxy-2-butanone) c / 21.5 / 45, 88 / 0.0059 / 0.994 / 1-50 / 1 / 13.2
7 / E-2-heptenal / 22.5 / 55, 83, 112 / 0.2857 / 0.990 / 1-50 / 1.0 / 8.38
8 / nonanal / 24.3 / 82, 98, 114 / 0.5542 / 0.997 / 1-50 / 0.1 / 5.48
9 / E-2-octenal / 25.5 / 70, 83, 97 / 0.6607 / 0.995 / 1-50 / 1.0 / 10.8
10 / decanal / 27.2 / 82, 95, 112 / 0.8031 / 0.999 / 1-50 / 0.1 / 6.65
12 / E-2-nonenal / 28.3 / 83, 70, 96 / 1.5309 / 0.974 / 1-50 / 1.0 / 9.52
15 / trans-2, cis-6-nonadienal / 29.8 / 69, 70, 81 / 0.6419 / 0.992 / 1-50 / 0.1 / 8.35
an = 5 aqueous/ethanol calibration solutions; bMean of 6 determinations on 3 different days; cmg/l