Phylogenetic analysis of 83 plastid genes further resolves the early diversification of eudicots Michael J. Moorea,1, Pamela S. Soltisb, Charles D. Bellc, J. Gordon Burleighd, and Douglas E. Soltisd aBiology Department, Oberlin College, Oberlin, OH 44074; bFlorida Museum of Natural History, University of Florida, Gainesville, FL 32611; cDepartment of Biological Sciences, University of New Orleans, New Orleans, LA 70148; and dDepartment of Biology, University of Florida, Gainesville, FL 32611 Edited* by Michael J. Donoghue, Yale University, New Haven, CT, and approved January 26, 2010 (received for review July 14, 2009) Although Pentapetalae (comprising all core eudicots except Gun- (BS) values regardless of the partitioning strategy (Fig. 1 and Figs. nerales) include ≈70% of all angiosperms, the origin of and rela- S1–S3). The ML topologies were also similar to the maximum tionships among the major lineages of this clade have remained parsimony (MP) topology (Fig. S4). Relationships among early- largely unresolved. Phylogenetic analyses of 83 protein-coding diverging angiosperms and among many clades of Mesangio- and rRNA genes from the plastid genome for 86 species of seed spermae agree with those from the largest previous plastid genome plants, including new sequences from 25 eudicots, indicate that analyses (12, 15). Below, we focus on results regarding relation- soon after its origin, Pentapetalae diverged into three clades: (i) ships within the eudicots, with an emphasis on the ML topologies. a “superrosid” clade consisting of Rosidae, Vitaceae, and Saxifra- Within Eudicotyledoneae, a basal grade emerged, with most gales; (ii)a“superasterid” clade consisting of Berberidopsidales, nodes receiving 100% ML BS support. In all ML analyses, Santalales, Caryophyllales, and Asteridae; and (iii) Dilleniaceae. Ranunculales were sister to all remaining eudicots (BS = 100%), Maximum-likelihood analyses support the position of Dilleniaceae followed by a clade of Sabiaceae + Proteales (BS = 70–80%, as sister to superrosids, but topology tests did not reject alterna- depending on partitioning strategy) as sister to a strongly supported tive positions of Dilleniaceae as sister to Asteridae or all remaining clade (BS = 100%) of Trochodendraceae + Buxaceae + Gunner- Pentapetalae. Molecular dating analyses suggest that the major idae (Gunneridae = Gunnerales + Pentapetalae; Fig. 1 and Figs. lineages within both superrosids and superasterids arose in as S1–S3). However, the approximately unbiased (AU) test did not little as 5 million years. This phylogenetic hypothesis provides a reject alternative relationships among Sabiaceae, Proteales, and EVOLUTION crucial historical framework for future studies aimed at elucidating remaining eudicots (Table S1). In all but one ML analysis, Buxaceae the underlying causes of the morphological and species diversity were recovered as sister to Gunneridae with BS = 52–70%, in Pentapetalae. depending on partitioning strategy (Fig. 1 and Figs. S1 and S2). In the CodonPartBL partition, Trochodendraceae were sister to Angiosperm Tree of Life | Pentapetalae | plastid genome Gunneridae, but with less than 50% bootstrap support (Fig. S2). Within the strongly supported (BS = 100%) Gunneridae, he Eudicotyledoneae (sensu) (1), or eudicots, comprise Gunnerales were sister to Pentapetalae (BS = 100%). Pentape- “ ≈75% of all angiosperm species (2) and encompass enor- talae comprised three major clades in the ML trees: (i)a super- T ” – mous morphological, biochemical, and ecological diversity. More rosid clade (BS = 96 100%) of Saxifragales, Vitaceae, and “ ” – than 90% of eudicot species diversity is found within the clade Rosidae; (ii)a superasterid clade (BS = 92 100%) of Santa- lales, Berberidopsidales, Caryophyllales, and Asteridae; and (iii) Pentapetalae (1), which includes major clades such as Rosidae, – Caryophyllales, Saxifragales, Asteridae, and Santalales, as well as Dilleniaceae (Fig. 1 and Figs. S1 S3). The superrosid and super- smaller lineages such as Berberidopsidales and Dilleniaceae (3– asterid clades were also recovered in the MP trees (Fig. S4), with 8). Previous analyses of multigene data sets have failed to resolve one major difference in composition. Whereas ML placed Dillenia relationships among the major clades of Pentapetalae (6, 9). The (the single representative of Dilleniaceae in the data set) sister to superrosids (Fig. 1 and Figs. S1–S3), MP placed Dillenia sister to inability to resolve these relationships suggests that the major Caryophyllales, with 97% BS support (Fig. S4). Still, the MP lineages of Pentapetalae diverged rapidly, a hypothesis sup- analysis also supported a superrosid clade of Vitaceae + Saxi- ported by the fossil record (10, 11). However, our understanding fragales + Rosidae with 99% BS. When Dillenia was constrained of the origins and evolution of Pentapetalae diversity, and con- to be sister to the superrosid clade, the best MP tree was 44 steps sequently much of angiosperm diversity, is hindered by the lack longer than the best unconstrained tree. A Templeton test (16) of a well-supported phylogenetic hypothesis. failed to reject this alternative topology (n = 712; P = 0.099). For Phylogenetic analyses based on complete plastid genome ML, AU tests strongly rejected topologies in which Dillenia was sequences have resolved several enigmatic relationships within sister to Caryophyllales (Table S1); however, they did not reject angiosperms (12, 13). However, these analyses have not included the placement of Dillenia as sister to all remaining Pentapetalae or data for many crucial eudicot clades. To resolve relationships as sister to the superasterids. among the major clades of Eudicotyledoneae (with a focus on Parametric bootstrapping provided evidence of bias in the MP Pentapetalae), we performed phylogenetic analyses using a data analyses against placing Dillenia as sister to the superrosid clade. set composed of 83 genes derived from 86 complete plastid genome sequences, 25 of which were eudicot sequences generated for this study. To date, this is the largest plastid genome data set Author contributions: P.S.S. and D.E.S. designed research; M.J.M. performed research; M.J. used for phylogenetic inference and includes representatives of M., C.D.B., J.G.B., and D.E.S. analyzed data; and M.J.M., P.S.S., C.D.B., J.G.B., and D.E.S. nearly all (37 of 42) orders of eudicots sensu Angiosperm Phy- wrote the paper. logeny Group (APG) III (14). The resulting phylogenetic The authors declare no conflict of interest. hypothesis helps to clarify the diversification of Pentapetalae and *This Direct Submission article had a prearranged editor. provides an improved framework for investigating evolutionary Freely available online through the PNAS open access option. processes that accompanied this radiation. Data Deposition: The sequences reported in this paper have been deposited in the Gen- Bank database (accession nos. GQ996966–GQ998871). Results 1To whom correspondence should be addressed. E-mail: [email protected]. Phylogenetic Analyses. Maximum-likelihood (ML) analyses of the This article contains supporting information online at www.pnas.org/cgi/content/full/ 83-gene alignment yielded similar trees and bootstrap support 0907801107/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.0907801107 PNAS Early Edition | 1of6 Downloaded by guest on October 1, 2021 Acorus Dioscorea Elaeis Musa * * * 95 Oryza * Triticum 98 * Zea Monocotyledoneae * * Typha Phalaenopsis * Yucca Lemna Anethum * Daucus 96 * Panax Lonicera * Helianthus Campanulidae * * Lactuca * * Scaevola Trachelium Ilex Antirrhinum * * Epifagus Jasminum Asteridae * Atropa * Solanum 46 * * Nicotiana * Cuscuta Lamiidae 99 * Ipomoea * Coffea 43 * Nerium Eudicotyledoneae * Ehretia Pentapetalae Aucuba 88 Rhododendron Cornus * Plumbago * Spinacia Caryophyllales SUPER- * Berberidopsis 85 Phoradendron * Ximenia Santalales ASTERIDS Arabidopsis * * Brassica * Gossypium * Citrus Staphylea Malvidae 97 * Eucalyptus * Oenothera Pelargonium * Bulnesia * 98 Cucumis Quercus 55 Ficus Rosidae * Morus * * Glycine * Phaseolus * Lotus Fabidae 53 88 * Medicago Euonymus * * Oxalis * Manihot * Passiflora * Populus SUPER- 70 Heuchera * Liquidambar Saxifragales Vitaceae Vitis ROSIDS * Dillenia DILLENIACEAE Gunnera Buxus 80 * Trochodendron * Meliosma * Nelumbo Platanus basal eudicots 64 * Nandina * Ranunculus Ceratophyllum * Calycanthus Ceratophyllum 99 Liriodendron super asterids * Drimys Magnoliidae 72 Piper Rosidae * Chloranthus Illicium Chloranthaceae * 82 Saxifragales Nuphar early-diverging 64 * Nymphaea Vitis Amborella angiosperms Dillenia 91 Cycas Ginkgo Pinus 0.01 substitutions/site INSET Fig. 1. Phylogram of the best ML tree as determined by RAxML (−ln L = 1101960.962) for the 83-gene data set. Numbers associated with branches are ML bootstrap support values. Asterisks indicate ML BS = 100%; the inset box in the lower right gives ML BS values for the basalmost branches of Pentapetalae, which are too short to visualize in the overall phylogram. The tree was rooted with Cycas, Ginkgo, and Pinus. Underlined genera names indicate plastid genomes newly sequenced for this study. When data sets were simulated using the ML topology, ML Within the superrosids, the relationships among Vitaceae, analyses of all 200 simulated data sets placed Dillenia sister to Saxifragales, and rosids were not strongly supported (Fig. 1 and superrosids; however, MP analyses of only 19 of the 200 data sets Figs. S1–S4). With MP, Saxifragales were sister to a clade