Cardamineae, Dumortier, Fl. Belg

Agerbirk, Niels, Hansen, Cecilie Cetti, Olsen, Carl Erik, Kiefer, Christiane, Hauser, Thure P., Christensen, Stina, Jensen, Karen R., Ørgaard, Marian, Pattison, David I., Lange, Conny Bruun Asmussen, Cipollini, Don & Koch, Marcus A., 2021, Glucosinolate profiles and phylogeny in Barbarea compared to other tribe Cardamineae (Brassicaceae) and Reseda (Resedaceae), based on a library of ion trap HPLC-MS / MS data of reference desulfoglucosinolates, Phytochemistry (112658) 185, pp. 1-19 : 6-7

2.2. Glucosinolate profiles in the tribe Cardamineae

The selected tribe Cardamineae species were investigated for GSLs with the overall evolutionary history in mind. In particular, we searched for similarities to the deviating genus Barbarea and C. pratensis . The characteristic GSLs in Barbarea include the homoPhe-derived “glucobarbarin” (BAR, 40 S), which is (S)-2-hydroxy-2-phenylethylGSL, epiglucobarbarin (EBAR, 40 R), which is the epimer (R)-2-hydroxy-2- phenylethylGSL, the Trp-derived IM (43), and a number of derivatives. Among the derivatives are the disubstituted 1,4moIM (138) (usually only detectable in roots), phenolic derivatives of both PE, BAR and EBAR, and isoferuloyl derivatives of the same and of IM (only known from seeds) (Supplementary Figure 1 ). Interestingly, the “innovative” C. pratensis also accumulates 1,4moIM ( Olsen et al., 2016). So far unique GSLs in C. pratensis are two hydroxylated GSLs with a 3-methylpentyl skeleton, 2h3mPe (149) and 3hmPe (141), both predicted to be biosynthesized from 2homoIle ( Agerbirk et al., 2010; Olsen et al., 2016). One of them (149) was likewise confined to roots ( Agerbirk et al., 2010). These recently discovered GSLs were found in some American and commercial accessions, while Danish accessions of C. pratensis contained a variety of other BCAA-derived GSLs and Phe or Tyr-derived GSLs ( Agerbirk et al., 2010).

A variety of organs were investigated whenever possible, in order to maximize the chance of detecting critical structures. In tables in this section, we consistently distinguished tentative from conclusive identifications based on an MS2 dependent criterion: For each species, conclusive identification of each reported GSLs was concluded when two independent characteristics of a peak matched an authentic standard (Blaˇzevic´et al., 2020): (1) correct t R of the relevant m / z value was observed in extracted ion chromatograms, and (2) the correct MS2 spectrum was observed. Correctness of t R and MS2 were based on comparison with an authentic standard ( Table 2 ). When conclusive identification was obtained from one organ, minor peaks from other organs were accepted as conclusive as well. In contrast, tentative identification (Blaˇzevic´et al., 2020) was concluded when a clearly distinguishable peak was observed but the MS2 criterion was not fulfilled for any organ of the species, or when reasonable t R and MS2 was observed but an authentic standard was not available. In either case, the tentative nature was indicated by labelling the data with an asterisk (*). It follows from these criteria that identification of GSLs for which authentic standards were not available were by definition tentative, even for members of a homologous series with expectable MS2 . In those cases, the GSL number in the entire table was labelled with an asterisks, essentially 3mSp (95 *), 9mSOn (68 *), 5mSp (94 *) and 9mSn ([89] *) , S2 hBuen (24 S *) and 2RhaOBZ (109 *). For individual levels of GSLs of this small group, an additional asterisk at individual data points means that also MS2 confirmation was lacking, while individual data points without an asterisk represent data validated by expectable MS2 for the side chain type .

Documentation for botanical identity and specification of origin followed recent recommendations ( Zidorn, 2017) as far as possible (Section 4.2.).

Figures

Fig. 1. MS2 spectra of pairs of desulfoglucosinolates with and without a side chain double bond. Four short chain desulfoglucosinolates were investigated, including Na+ adducts of all (A–D) and in addition H+ adducts of the methylthio substituted (E–F), as indicated in each spectrum. The desulfo derivative of the putative 9mSn ([89]), poorly characterized in the literature, was also investigated (G).