Dereplication of Bioactive Agave Saponin Fractions: The Hidden Saponins

The good phytotoxicity and selectivity against weeds versus tomato or cress make saponin-rich fractions from Agave macroacantha, A. colorata, A. parryi, and A. parrasana attractive candidates as bioherbicides. The saponin contents have only previously been reported for A. macroacantha, and as a consequence, simultaneous dereplication has been performed on saponin-rich fractions from the other plants by mass spectrometry (MS) and nuclear magnetic resonance (NMR) spectroscopy. This strategy enables the identification of a total of 26 saponins, 14 of which have been described previously and 12 of which are proposed as new saponins. They include isomers and a new sugar chain with a β-d-apiofuranose unit. The method is corroborated by the isolation of eight dereplicated saponins from A. colorata.

Electrospray Ionization in the negative polarity mode (ESI -) was used with the following settings: sample probe capillary voltage 2800 V, sampling cone voltage 30 V, source temperature 120 °C and desolvation temperature 450 °C.Desolvation and cone gas with flow rates of 850 and 10 L/h were used, respectively.The data were acquired in the centroid mode using MS E (low collision energy, 6 eV; high collision energy ramp, 20−80 eV) over a mass range of m/z 100−2000 and a retention time range of 0-10.0 min with a 0.5 s scan time.Data acquisition and processing were carried out with MassLynx version 4.1 (Waters Inc.Milford, MA, USA, 2013).The stock solutions (1000 ppm) of the saponin-rich fractions were prepared in water:acetonitrile (6:4).All the samples were injected as a dilution 1:15 (66.7 ppm) and filtered through a PTFE syringe filter (0.22 µm) prior to analysis.

Procedure of HMAI method.
Two flowcharts (below) are tools for the identification of aglycones of saponins from the Agave species by 1 H-NMR and HMBC experiments.The decisions (inside diamonds) are named with D or S and an integer number, and they use both ranges of chemical shifts and absolute values of HMBC correlations in the flowchart.In this last case, values within the error range ±0.04 ppm and ±0.4 ppm for 1 H-and 13 C-NMR signals should be considered.Spectra should be referenced to deuterated pyridine (7.55 ppm and 135.6 ppm for 1 H-and 13 C-NMR, respectively).
Prior assignment is not needed and only proton signals (for three equivalent protons) should be distinguished between 1.6 ppm and 0.5 ppm.These signals are readily recognisable and correspond to secondary methyl groups at C-21 and C-27, which are doublets and singlets for angular methyl groups C-18 and C-19.The flowchart should be started with the methyl doublets that provide information on rings C-F.The doublet that is more shielded would be analysed first (usually C-27).
Secondly, methyl groups that give rise to singlets should be investigated and the most deshielded position will be applied in the flowchart to assign each methyl group.
In some cases, the flowchart indicates that HMBC values for a specific methyl should be revised.In this situation, taking into consideration the structural features of the remaining methyls, HMAI Tables (below) should be used.
In a case where the HMBC signals are different to those indicated in the tables, the saponin should have other structural characteristics that will require elucidation.