THE ITS REGION OF NUCLEAR RIBOSOMAL DNA - A VALUABLE SOURCE OF EVIDENCE ON ANGIOSPERM PHYLOGENY

The internal transcribed spacer (ITS) region of 18S-26S nuclear riboso mal DNA (nrDNA) has proven to be a useful source of characters for phy logenetic studies in many angiosperm families. The two spacers of this region, ITS-1 and ITS-2 (each < 300 bp), can be readily amplified by PCR and sequenced using universal primers, even from DNAs of herbarium specimens. Despite high copy numbers of both spacers, the near unifor mity of ITS paralogues, attributed to rapid concerted evolution, allow s direct sequencing of pooled PCR products in many species. Divergent paralogues, where detected, require cloning, but may offer a means of obtaining multiple estimates of organismal relationships and of determ ining placement of the root in a phylogeny independent of outgroup con siderations. In reported studies, variation between ITS sequences is m ostly attributable to point mutations. A relatively minor proportion o f sites is affected by insertions or deletions (indels) among sequence s that are similar enough to have retained sufficient signal for phylo genetic analysis. Within these limits, sequence alignment is generally unambiguous except in small regions of apparently lower structural co nstraint. Phylogenetic analyses of combined data sets from both spacer s, where examined, yield trees with greater resolution and internal su pport than analyses based on either spacer alone. This beneficial effe ct of simultaneous analysis is not surprising based on the low number of useful characters in each spacer. This effect also suggests high co mplementarity of spacer data, in accord with similarity in size, seque nce variability, and G + C content of ITS-1 and ITS-2 in most investig ated groups of closely related angiosperms. Non-independent evolution of ITS sites involved in intraspacer RNA base-pairing may occur, given possible functional constraints, but preliminary secondary structure analyses of ITS-2 in Calycadenia (Asteraceae) show no definite evidenc e of compensatory spacer mutations. As expected, levels of ITS sequenc e variation suitable for phylogenetic analysis are found at various ta xonomic levels within families, depending on the lineage. The apparent rates of ITS molecular evolution are roughly correlated with plant li fe-form, as with chloroplast DNA (cpDNA) data, but reasons for this ob servation are unclear. ITS characters have improved our understanding of angiosperm phylogeny in several groups by (1) corroborating earlier unexpected findings, (2) resolving conflicts between other data sets, (3) improving resolution of species relationships, or (4) providing d irect evidence of reticulate evolution. Hybridization or sorting of an cestral polymorphism in a lineage can complicate interpretation of tre es based on any type of evolutionary evidence, including ITS or cpDNA sequences, particularly in the absence of at least one independent phy logenetic data set from the same organisms. The need for phylogenetic markers from the nuclear genome, to complement the rapidly growing bod y of cpDNA data, makes the ITS region a particularly valuable resource for plant systematists.