A molecular approach to the phylogeny of bryophytes : cladistic analysis of chloroplast-encoded 16S and 23S ribosomal RNA genes

As the most primitive living lineages of embryophytes, bryophytes are critical to an understanding of the evolution of the land flora. A relatively robust phylogenetic hypothesis exists, based on morphology, ultrastructure, and chemistry, in which the classical group "bryophytes" is not monophyletic. Instead, the mosses seem to be more closely related to the tracheophytes than to the hornworts or liverworts. However, details of these relationships remain unclear. In this paper we explore the usefulness of comparative molecular studies as a potential source of independent data to test and refine this cladogram. We have generated preliminary nucleotide sequence data from portions of the 16S and 23S ribosomal RNA genes in the chloroplasts of 11 bryophytes selected from diverse groups. These data, even though they show considerable homoplasy, appear to be historically informative at this deep phylogenetic level since they support a cladogram that is identical to the morphological one. Future studies will expand the molecular comparisons by sequencing the rest of these two rRNA genes, as well as the protein-coding gene rbcL, for many more species. The bryophytes occupy a critical phylogenetic position in our understanding of the origin of the land flora. Their phylogeny has attracted interest over a long period of time (Bower 1935; Campbell 1905; Campbell 1971; Haskell 1949; Khanna 1965; Miller 1974, 1982; Schofield 1985). This interest stems from the fact that they appear to be the most basal lineages among extant land plants, and because of the fundamental diversity among the major groups of bryophytes, which has called into question the monophyly of the bryophytes or even of the land plants (i.e., embryophytes; see Crandall-Stotler 1980, 1984; Duckett & Renzaglia 1988). Much research has focused on bryophyte phylogeny; a number of different character systems have been studied in morphological, anatomical, and developmental detail (reviewed by Berthier 1972; Crandall-Stotler 1981; Frey 1971, 1981; H6bant 1977; Ligrone & Gambardella 1988; Lorch 1931; Renzaglia 1978; Schuster 1984a,b,c; Vitt 1984). Early light microscopic research concentrated on morphology, apical meristem segmentation patterns, and sexuality. More recently, ultrastructural investigations have also provided significant data (Brown & Lemmon 1988, 1990; Carothers & Rushing 1988; Duckett et al. 1983; Duckett & Renzaglia 1988; Scheirer 1980). Critical evolutionary issues in the origin and radiation of the land plants include the nature of reproduction, source of the alternation of generations, and water relationships (Graham 1985; Mishler & Churchill 1984, 1985). Such evolutionary questions can best be addressed through comparative methods that are based on reconstructions of character evolution on cladograms (Brandon 1990; Brooks & McLennan 1991; Coddington 1988; Donoghue 1989; Harvey & Pagel 1991; Maddison 1990; Mishler 1988; Ridley 1983). Such methods obviously depend on the quality of the reference cladistic topology, that should be based on many characters in addition to (or ideally with the exclusion of) the specific characters of evolutionary interest. Therefore, the time seems ripe to add molecular characters to the data base. The status of available phylogenetic characters was first analyzed cladistically by Mishler and Churchill (1984, 1985; see also Bremer et al. 1987; Graham et al. 1991; and Mishler 1986). The current hypothesis of land plant relationships based on these studies can be outlined as follows. The embryophytes (land plants, which include the bryophytes and tracheophytes), are well supported as a monophyletic group by several synapomorphies. Closely related sister groups of the embryophytes are a number of lineages of green algae that have been lumped together in the paraphyletic group "charophytes" (Mattox & Stewart 1984; Stewart & Mattox 1975). Among these, the immediate living sister group of the embryophytes appears to be the genus Coleochaete alone. Within the embryophytes, the classical inclusive group "bryophytes" also appears to be paraphyletic. The mosses alone are the sister group of the tracheophytes (the so-called "vascular plants"), 0007-2745/92/172-180$1.05/0 This content downloaded from 207.46.13.124 on Wed, 22 Jun 2016 06:07:18 UTC All use subject to http://about.jstor.org/terms 1992] MISHLER ET AL.: MOLECULAR PHYLOGENY OF BRYOPHYTES 173 as evidenced by a number of synapomorphies, putatively including xylem and phloem (Mishler & Churchill 1984). Currently, the hornworts are best placed as the sister group of the moss-tracheophyte clade, but this placement is problematical because of the number of homoplasies involving the hornworts. Finally, the liverworts are the sister group of the other extant land plants. Considerable interest has recently been focused on tapping the historical information that resides in DNA. Large portions of the genome are quite conservative over long evolutionary times, and have proved useful in resolving relationships in a number of different cases, including bacteria, fungi, plants, birds, and primates (e.g., see volumes edited by Fernholm et al. 1989; Hillis & Moritz 1990; Soltis et al. 1992). Chloroplasts of green plants have their own DNA, quite distinct from that of the nucleus (due to the ancient origin of the chloroplast from an endosymbiotic, photosynthetic bacterium). Chloroplast DNA has proven to be especially conservative over evolutionary time, and has been useful for phylogenetic studies in the flowering plants (Jansen & Palmer 1988; Palmer 1985; Palmer et al. 1988; Sytsma & Gottlieb 1986). Current understanding of chloroplast DNA evolution was reviewed by Birky (1988) and Zurawski and Clegg (1987). The work of a number of researchers on structural aspects of the chloroplast genome (reviewed by Palmer 1985) has led to the elucidation of a consensus gene order for the angiosperms that is highly conserved evolutionarily. Two inverted repeat regions (containing the ribosomal RNA genes) are separated by a large and a small single copy region. With only slight modification, this consensus gene order has also been found in a liverwort, Marchantia polymorpha (Ohyama et al. 1983, 1986), in a fern, Osmunda cinnamomea (Palmer & Stein 1982) and in a moss, Physcomitrella patens (Calie & Hughes 1987). The consensus structure of the land plant chloroplast genome provides additional strong evidence for monophyly of the group, given that the structure of the chloroplast genome in closely related "charophyte" green algal outgroups is quite different (J. R. Manhart, pers. comm.). Chloroplast studies have also confirmed the close relationship of Coleochaete to the land plants (Baldauf& Palmer 1990; Baldauf et al. 1990; Manhart & Palmer 1990). Calie and Hughes (1987) showed that the chloroplast genome of the moss Physcomitrella has two interesting exceptions to the general order and arrangement of the consensus land plant genome. First, the rpl2 and rps 19 genes are shifted from a position adjacent to the petB/petD gene cluster to a position next to the atpH gene. Second (and potentially more target DNA SPCR primer targetDNA _____________, new DNA denature and synthesize