Embryology and phylogeny of glucosinolate-producing plants.

已知有15个科在被子植物中产生芥子酸或硫代葡萄糖苷，包括十字花科、辣椒科、辣木科、萨尔瓦多科、大戟科（仅含木蔷薇科）。它们的系统发育关系一直不确定。在本研究项目中，我介绍了各种植物类群的胚胎学信息，并在胚胎学证据的基础上，讨论了它们之间和内部的关系。新的胚胎学信息和/或新发现总结如下。(1)Canotia与Koebelinia的亲缘关系并不密切，Koebelinia具有黑松香细胞，经常与Canotia一起讨论，而在Celastrales中比较常见。(2)萨尔瓦多科植物的种子结构与其他产硫代葡萄糖苷科植物不同。(3)Macgregoria被认为是与硫代葡萄糖苷产生科关系密切的候选植物之一，在不含硫代葡萄糖苷的Stackhousiaceae中表现良好。(4)产硫代葡萄糖苷的Drypetes属（包括Liodendron）和不产硫代葡萄糖苷的Poranthera属在大戟科都很适合。综上所述，尽管大分子证据表明14个产硫代葡萄糖苷科（不包括大戟科）是单系的，但胚胎学特征与许多其他形态或分类特征一样，都朝着不同的方向进化。

Fifteen families are known to produce myrosin or glucosinolates in angiosperms, and they include Brassicaceae, Capparaceae, Moringaceae, Salvadoraceae, Euphorbiaceae (Drypetes only). Their phylogenetic relationships have been uncertain. In the present research project I presented embryology of various plant groups whose information are still lacking and, on the basis of embryological evidence, discuss relationships among and within them. New embryological information and/or new findings are summarized as follows. (1)Canotia is not closely related to Koebelinia which has myrosin cells and has often discussed together with Canotia, and is rather at home within Celastrales. (2)Salvadoraceae have no seed structure that fits in with that of any of the other glucosinolate-producing families. (3)Macgregoria, which has been suggested as one of candidates having close relationships to the glucosinolate-producing families, is well at home in Stackhousiaceae which has no gluosinolates. (4)A glucosinolate-producing genus Drypetes (including Liodendron) and a non-glucosinolate-producing genus Poranthera both are well suitable in Euphorbiaceae embryologically.In conclusion, even if 14 glucosinolate-producing families (excluding Euphorbiaceae) are monophyletic as indicated by macromolecular evidence, embryological characters like many other morphological or taxonomical characters have evolved towards diverse directions.