The complete external transcribed spacer of 18S-26S rDNA: amplification and phylogenetic utility at low taxonomic levels in asteraceae and closely allied families

For molecular phylogenetic reconstruction of some intrageneric groups of plants, a DNA region is needed that evolves more rapidly than the internal transcribed spacer (ITS) of the 18S-26S nuclear ribosomal DNA (nrDNA) repeat. If the region identified is nuclear, it would also be desirable for it to undergo rapid concerted evolution to eliminate problems with coalescence. The external transcribed spacer (ETS) of the nrDNA repeat has shown promise for intrageneric phylogenetic reconstruction, but only the 3' end of the region has been utilized for phylogenetic reconstruction and "universal" primers for PCR amplification have been elusive. We present a method for reliably amplifying and sequencing the entire ETS throughout Asteraceae and some closely allied families. We also show that the ETS is more variable and phylogenetically informative than the ITS in three disparate genera of Asteraceae-Argyranthemum (tribe Anthemideae), Asteriscus (tribe Inuleae), and Helianthus (tribe Heliantheae). The full ETS was amplified using a primer (ETS1f) within the intergenic spacer in combination with a primer (18S-2L) in the 5' end of the highly conserved 18S gene. ETS1f was designed to correspond to a highly conserved region found in Helianthus and Crepis, which are in separate subfamilies of Asteraceae. ETS1f/18S-2L primed in all of the tribes of Asteraceae as well as exemplar taxa from Campanulaceae, Goodeniaceae, and Calyceraceae. For both Argyranthemum and Asteriscus, we were able to directly sequence the ETS PCR products when a single band was produced. When multiple bands were produced, we gel-purified and occasionally cloned the band of interest before sequencing. Although PCR produced single bands for Helianthus species, it was necessary to clone Helianthus amplifications prior to sequencing due to multiple intragenomic ETS repeat types. Alignment of ETS sequences for Argyranthemum and Asteriscus was straightforward and unambiguous despite some subrepeat structure in the 5' end. For Helianthus, different numbers of large tandem subrepeats in different species required analysis of the orthology of the subrepeats prior to alignment. In all three genera, the ETS provided more informative variation for phylogenetic reconstruction and allowed better resolution of relationships than the ITS. Although cloned sequences from Helianthus differed, intragenomic clones consistently formed clades. This result indicated that concerted evolution was proceeding rapidly enough in ETS that species-specific phylogenetic signal was retained. It should be now be possible to use the entire ETS for phylogenetic reconstruction of recently diverged lineages in Asteraceae and at least three other families (approximately 26,000 species or about 8% of all angiosperms).