Hybridization and complex evolution of Barbarea vulgaris and related species (Brassicaceae).

Barbarea, winter-cress, is a genus of 29 species in Brassicaceae, the mustard family, which has emerged as a model for evolution of plant defence and specialised metabolites. Notably, some Barbarea species have evolved the ability to produce triterpenoid saponins as the only ones in Brassicaceae, some of which make plants resistant to important herbivores. Resistance has, however, been lost in a distinct group of plants within B. vulgaris ssp. arcuata, which is genetically strongly diverged from other B. vulgaris plants. This divergence is not reflected present in taxonomy. Thus, a phylogeny is needed to understand evolution and defence in Barbarea. Here, we analysed the nuclear ITS and the plastid matK, ndhF, rps16, and psbA-trnH DNA regions from seven out of 29 Barbarea species, 57 accessions of B. vulgaris, 10 accessions of other Barbarea species, and eight outgroup species, in addition to sequences available from GenBank. All Barbarea species formed a highly supported monophyletic group, separated from sister genera. Several clades seem to have radiated within the genus with no simple branching pattern, and discordant nuclear and plastid DNA phylogenies indicate reticulate evolution and chloroplast capture. One of the complex patterns may have resulted from chloroplast capture of a non-Nordic Barbarea species not included in the study. Two pairs of species were almost identical, B. australis and B. grayi, and B. orthoceras and B. stricta. Despite hybridization, chloroplast capture, and incongruence among the plastid and nuclear DNA data, the high level of intraspecific diversity, coupled with lineage specificity, lead us to recognize three groups of Barbarea vulgaris: G-type (glabrous) and P-type (pubescent) individuals of the current B. vulgaris ssp. arcuata as two distinct groups and the current B. vulgaris ssp. vulgaris as the third. Despite the high molecular diversity below species level, the evolutionary history of the saponin-based resistance remains unsettled due to unresolved basal branching.

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.

A library of ion trap MS2 spectra and HPLC retention times reported here allowed distinction in plants of at least 70 known glucosinolates (GSLs) and some additional proposed GSLs. We determined GSL profiles of selected members of the tribe Cardamineae (Brassicaceae) as well as Reseda (Resedaceae) used as outgroup in evolutionary studies. We included several accessions of each species and a range of organs, and paid attention to minor peaks and GSLs not detected. In this way, we obtained GSL profiles of Barbarea australis, Barbarea grayi, Planodes virginica selected for its apparent intermediacy between Barbarea and the remaining tribe and family, and Rorippa sylvestris and Nasturtium officinale, for which the presence of acyl derivatives of GSLs was previously untested. We also screened Armoracia rusticana, with a remarkably diverse GSL profile, the emerging model species Cardamine hirsuta, for which we discovered a GSL polymorphism, and Reseda luteola and Reseda odorata. The potential for aliphatic GSL biosynthesis in Barbarea vulgaris was of interest, and we subjected P-type and G-type B. vulgaris to several induction regimes in an attempt to induce aliphatic GSL. However, aliphatic GSLs were not detected in any of the B. vulgaris types. We characterized the investigated chemotypes phylogenetically, based on nuclear rDNA internal transcribed spacer (ITS) sequences, in order to understand their relation to the species B. vulgaris in general, and found them to be representative of the species as it occurs in Europe, as far as documented in available ITS-sequence repositories. In short, we provide GSL profiles of a wide variety of tribe Cardamineae plants and conclude aliphatic GSLs to be absent or below our limit of detection in two major evolutionary lines of B. vulgaris. Concerning analytical chemistry, we conclude that availability of authentic reference compounds or reference materials is critical for reliable GSL analysis and characterize two publicly available reference materials: seeds of P. virginica and N. officinale.