Analysis Of Phylogenetic Relationships Of Brassicaceae Species

The Brassicaceae with the characters of six stamens in a tetradynamous pattern (two short and four long), cruciform corolla, distinct capsular fruit and pungency, is always recognized as a rather natural monophyletic group. However, at present, the deep phylogenetic relationships within the lineages, the delimitation of some genera, the placement of many species are still unsolved, based on a few molecular analysis of the whole family. Moreover Brassicaceae is a group with frequent interspecific hybridization, and recently, a successful natural inter-tribal hybridization event is reported in genus Pachycladon. Therefore, analysis of phylogenetic relationships of Brassicaceae would also show new insights into elucidating the origin for genus Pachycladon.Using maximum parsimony, maximum likelihood, and neighbor-joining methods, sequences of nuclear chalcone synthase gene (Chs), arginine decarboxylas gene (Add) and rbcL were analyzed to elucidate:the phylogenetic relationships of taxa within the Brassicaceae family, and the origin of allopolyploid genus Pachycladon.1. Phylogenetic relationship of Brassicaceae speciesBased on DNA fragments analyses results, except for the well-supported sister-group relationship of tirbe Aethionemeae to the rest of the family, most of the rest tribes fell into one of three larger monophyletic groups (Lineagesâ… -â…¢). The phylogenetic relationships among the larger tribal groups were either undetected or with poor support, with only Lineageâ…¡received consensus bootstrap support. The monophyletic clades not included in three major lineages were Thlaspideae, Eutremeae, Noccaeeae, Cochlearieae, Alysseae, Arabideae and Anastaticeae clade. Lineageâ… Two distinct types of sequences were identified at a Chs locus for all Pachycladon species. Each type of sequence formed a monophyletic Pachycladon clade, one monophyletic Pachycladon clade together with Boechereae clade, Halimolobeae clade and Crucihimalaya clade (including Crucihimalaya, Transberingia) nested within tribe Cardamineae, and then all together they form a moderate supported clade. As a result, tribe Camelineae was paraphyletic.Coronopus didymu and Cardaria drabas nested with Lepidium revealed that they should fall within Lepidium, so our result supported that Coronopus didymu and Cardaria drabas should be renamed as Lepidium didymu and Lepidium draba.Lineageâ…¡Conringia planisiliqua shared a sister relationship with Pachypterygium multicaule, Tauscheria lasiocarpa. and Isatis alba, thus C. planisiliqua should be a member of tribe Isatideae, not in Orychophragmus clade.Brassica was divided into two evolutionary lineages as before:the nigra lineage and the rapa/oleracea lineage. Raphanus was more closely related to B. rapa/oleracea than that to B. nigra with high bootstrap values. Raphanus should be a member of Brassica.In Adc2 phylogenetic tree Sinapis alba was close to Brassica, but in Chs phylogenetic tree S. alba and Orychophragmus violaceus formed a clade. This was probably caused by lineage sorting of nuclear gene alleles in Chs locus. S. alba should be a member of Brassica.Lineageâ…¢In Chs phylogenetic tree, four Desideria species (D. baiogoinensis, D. himalayensis, D. linearis and D. stewartii), a species of Phaeonychium (P. jafrii) and species of Solms-laubachia formed a well-supported monophyletic subclade. Both Desideria and Solms-laubachia were polyphyletic, as they appeared in more than two positions in subclade, within which P. jafrii was embedded. Therefore, our results indicated that Solms-laubachia should be expanded to include Desideria and P. jafrii.Except for Lineagesâ… -â…¢, Cithareloma vernum and Pugionium cornutum formed a well-supported clade, and then with a close relationship with Lobularia maritime in rbcL phylogenetic tree. We agreed that Pugionium is a mumber of Anastaticeae. In Adc phylogenetic trees, Christolea niyaensis and Pugionium cornutum formed a well-supported clade, This was probably because incongruence between Adc gene tree and the ture phylogenetic trees of Brassicaceae.2. Origination of allopolyploid PachycladonOur Chs phylogenetic tree included all major tribes that could potentially contain the parents for the two genomes of Pachycladon, and revealed that two distinct types of sequences were identified at a Chs locus for all Pachycladon species. Each type of sequence formed a monophyletic Pachycladon clade, one monophyletic Pachycladon clade together with Boechereae, Halimolobeae, Crucihimalaya clade (including Crucihimalaya and Transberingia) nested within tribe Cardamineae, the other one nested within tribe Smelowskiea.Because of the limit of sample, the relationship of Boechereae, Halimolobeae, Australia-New Zealand clade (including Pachycladon), Crucihimalaya clade and Camelineae was still unclear. Therefore, one parent for species of Pachycladon appeared to be most closely associated with these groups, the other parent was proved to be in the tribe Smelowskieae.Phylogenetic analysis of the chloroplast rbcL gene suggested that the monophyletic Pachycladon clade was not close to tribe Smelowskieae. So, paternal parent was species of tribe Smelowskieae, and the maternal parent for species of inter-tribal allopolyploid genus Pachycladon appeared to be most closely associated with tribes Boechereae, Halimolobeae, Australia-New Zealand clade, Crucihimalaya and Transberingia.