Usefulness of RNA polymerase II coding sequences for estimation of green plant phylogeny.

Phylogenetic reconstruction based on chloroplast (cp) and nuclear ribosomal (r)DNA sequences have contributed greatly to our understanding of green plant evolution. Many relationships remain unresolved, however, especially among basal angiosperm taxa. Here, both plastid and ribosomal gene analysis have generated inconclusive, and occasionally contradictory, hypotheses of deep branching topology (Chase et al. 1993; Soltis et al. 1997). Such incongruities underscore the need for sequence data from additional independent loci. The pilot study we report here was motivated by an interest in studying the early diversification of land plants. Toward this end, we PCR-amplified and sequenced the genes encoding the second largest subunit of RNA polymerase II (RPB2) from nine taxa representing tricolpate, monosulcate, and eudicot angiosperm lineages, one gymnosperm, and one liverwort species to determine the feasibility of recovering the RPB2 gene from any green plant using a single set of PCR primers. We also performed a preliminary evaluation of the potential phylogenetic utility of RPB2 sequences. Other than the widely studied largeand small-subunit rDNA genes, protein-encoding genes are the main potential source of comparative nuclear DNA sequence. In plants, many of these genes are members of repeated families, complicating phylogenetic comparisons with issues of paralogy and within-family concerted evolution. These problems were noted in phylogenetic studies with the genes for phytochrome (Mathews, Lavin, and Sharrock 1995), small heat shock proteins (Waters 1995), and cytosolic phosphoglucose isomerase (Gottlieb and Ford 1996). Currently, the nuclear gene with the broadest phylogenetic applicability in plants is alcohol dehydrogenase (Adh), but separate gene duplication events within angiosperm lineages may limit the utility of universal angiosperm Adh primers (Morton, Gaut, and Clegg 1996; Sang, Donoghue, and Zhang 1997). The identification of nonrepeated nuclear genes that evolve at rates suitable for interfamilial phylogenetic assessment is made difficult by the paucity of comparative sequence data currently available for plant nuclear genomes. In principle, promising nuclear genes for such studies are those that encode components of the multiprotein assemblies involved in the apparatuses for replication and transcription in the cell nucleus. These pro-