Evolutionary Implications of Matk Indels in Poaceae1
Total Page:16
File Type:pdf, Size:1020Kb
American Journal of Botany 86(12): 1735±1741. 1999. EVOLUTIONARY IMPLICATIONS OF MATK INDELS IN POACEAE1 KHIDIR W. H ILU2 AND LAWRENCE A. ALICE Department of Biology, Derring Hall, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061-0406 Insertion/deletion events (indels) and nucleotide substitutions at the extreme 39 end of the chloroplast gene matK have been identi®ed that distinguish certain major lineages of grasses. A 1-bp (base pair) deletion creating a shift in the open reading frame (ORF) and a point mutation support the positions of Streptochaeta and Anomochloa as the two most basal lineages in Poaceae. Another 1-bp deletion resulting in early termination of the ORF is unique to Ehrharta, a member of the taxonomically disputable tribe Ehrharteae. A 6-bp insertion supports monophyly of subfamilies Panicoideae, Arundi- noideae, Centothecoideae, and Chloridoideae (PACC). This marker appears useful in de®ning PACC clade members and may have potential in providing insight into the sister-group relationship between PACC and other lineages. Alignment of deduced amino acid sequences from bryophytes, gymnosperms, and angiosperms shows that this region is relatively con- served, but variation is notably higher in Poaceae. The evolutionary implications of these changes in grasses and other plant families are addressed. Key words: grasses; indels; markers; matK; phylogeny; plants; Poaceae. Molecular characters used in systematic studies are ba- and CalderoÂn, 1979). However, Soderstrom (1981) indi- sically of three types: nucleotide substitutions, rearrange- cated that the Bambusoideae should not be regarded as ments, and insertions/deletions (indels). Indels are usually ancestral to other grasses due to certain derived anatom- more frequent in noncoding than coding regions and ical and vegetative morphological characters. Anomoch- could be useful in circumscribing lineages and de®ning loa, Streptochaeta, and Pharus, formerly classi®ed as evolutionary trends. In coding regions, all insertions and herbaceous bamboos, appear basal or distinct when in- deletions translate into nonsynonymous substitutions ei- cluded in morphological-anatomical and molecular ana- ther at the immediate site of the mutation or further lyses (Hilu and Wright, 1982; Clark, Zhang, and Wendel, downstream if a frameshift is generated. Because natural 1995; Duvall and Morton, 1996; Soreng and Davis, 1998; selection acts largely at the protein level, these mutations Hilu, Alice, and Liang, in press). When these genera are could be subject to greater selective constraints due to not included in phylogenies, either oryzoids and woody their impact on the function and/or structure of the en- bambusoids are basal (Cummings, King, and Kellogg, coded products. Conversely, synonymous substitutions 1994; Barker, Linder, and Harley, 1995; Liang and Hilu, maintain protein composition and are considered effec- 1996) or there is a basal split in the family into Pooideae tively neutral (Clegg et al., 1994; Nielsen, 1997). In sys- and the rest of the Poaceae (Davis and Soreng, 1993; tematic and evolutionary studies, informative indels not Nadot, Bajon, and Lejeune, 1994). only can be combined with other molecular characters, De®nitions of major clades in the Poaceae have also but also can work as qualitative molecular markers de- been disputed. The appearance of subfamilies Panicoi- pending on their size, position, and in¯uence on the open deae, Arundinoideae, Centothecoideae, and Chloridoi- reading frame (ORF). Such molecular markers are par- deae in one major group (named PACC; Davis and So- ticularly useful in phylogenies where resolution is low or reng, 1993) is supported by morphological, anatomical, disagreements prevail among data sets. The phylogeny of and molecular data (Hilu and Wright, 1982; Hilu and the Poaceae presents a case where such molecular mark- Esen, 1988; Esen and Hilu, 1989; Davis and Soreng, ers can be useful (Clark, Zhang, and Wendel, 1995). 1993; Barker, Linder, and Harley, 1995; Clark, Zhang, Several hypotheses have been proposed concerning the and Wendel, 1995; Duvall and Morton, 1996; Liang and most basal lineages in Poaceae. Subfamily Bambusoideae Hilu, 1996; Mathews and Sharrock, 1996; Soreng and s.l. (sensu lato) was traditionally considered the most an- Davis, 1998). In addition, using ndhF sequences, Clark, cestral group based on the presence of presumably prim- Zhang, and Wendel (1995) recognized a lineage (named itive reproductive characters (Stebbins, 1956; Soderstrom BOP) that comprises most bambusoids plus oryzoid and pooid taxa. An alternative relationship identi®ed in So- 1 Manuscript received 8 December 1998; revision accepted 22 April reng and Davis's (1998) study includes the Pooideae plus 1999. The authors thank the U.S. Department of Agriculture-Agricultural PACC as a major clade. Research Service, National Plant Germplasm System for seeds; N. In a systematic study of the Poaceae using sequences Barker, L. Clark, J. Davis, G. Fleming, S. Jacobs, O. Mistretta, and the of the plastid gene matK, we detected molecular markers Missouri and University of Bonn Botanical Gardens for providing ma- near the 39 end of the coding region that appear phylo- terial; T. Borsch and T. Bodo Slotta for unpublished sequences; Patty genetically informative. In this study, we investigate the Singer of the University of Maine DNA sequencing facility; and J. Bond, G. Davidse, and C. Neinhuis for their assistance. This work was use of these markers in evaluating current hypotheses on supported by NSF grant number DEB 9634231 to KWH. grass evolution and assess the phylogenetic position of 2 Author for correspondence. taxonomically disputable taxa. The evolutionary hypoth- 1735 1736 AMERICAN JOURNAL OF BOTANY [Vol. 86 TABLE 1. Accessions used along with their classi®cation and sources of material. BBG 5 University of Bonn Botanical Garden, KH 5 KWH collections located at Virginia Tech, MBG 5 Missouri Botanical Garden, NTPG 5 U.S. National Tropical Botanic Garden, PI 5 U.S. Department of Agriculture numbers, VTHG 5 Virginia Tech Horticulture Garden. Herbarium codes for vouchered accessions are ISC 5 Iowa State University, MO 5 Missouri Botanical Garden, NSW 5 New South Wales (Australia), and RSA 5 Rancho Santa Ana Botanical Garden. Genus/Species Subfamily Tribe Voucher Restionaceae Restio tetraphyllus Labill. BBG 00700 Joinvilleaceae Joinvillea ascendens Gaud. ex Brongn. et Gris. NTBG 800379 Poaceae Anomochloa marantoidea Brongn. Anomochloideae Anomochloeae L. Clark 1299, ISC Streptochaeta angustifolia Soderstr. Streptochaeteae L. Clark 1304, ISC Aristida adscensionis L. Arundinoideae Aristideae PI 269867 Arundo donax L. Arundineae VTHG Danthonia spicata (L.) P. Beauv. ex Roem. & Schult. KH 5505 Molinia caerulea (L.) Moench L. Clark 1165, ISC Phragmites australis (Cav.) Trin ex Steud. G. Fleming s. n. Chusquea coronalis Soderstr. & C. E. CalderoÂn Bambusoideae Bambuseae MBG Phyllostachys aurea RivieÁre & C. RivieÁre KH 9418 Sasa kurilensis (Rupr.) Makino and Shibata A. Nishiwaki s. n. Brachyelytrum erectum (Schreb.) P. Beauv. Brachyelytreae T. Wieboldt s. n. Lithachne pauci¯ora (Sw.) P. Beauv. Olyreae L. Clark 1297, ISC Olyra latifolia L. L. Clark 1191, ISC Pariana radici¯ora Sagot ex DoÈll Parianeae MBG Chasmanthium laxum (L.) H. O. Yates Centothecoideae Centotheceae KH 5716 Zeugites pittieri Hack. L. Clark 1171, ISC Chloris cucullata Bisch. Chloridoideae Chlorideae PI 315683 Cynodon dactylon L. PI 224149 Perotis rara R. Br. PI 238348 Fingerhuthia sesleriformis Nees Eragrostideae PI 299968 Leptochloa dubia (Kunth) Nees PI 216459 Plectrachne pungens (R. Br.) C. E. Hubb. S. Jacobs 8031, NSW Sporobolus indicus (L.) R. Br. PI 310309 Tridens brasiliensis Nees ex Steud. PI 310319 Lepturus repens (G. Forst.) R. Br. Leptureae Latz 10843, MO Orcuttia californica Vasey Orcuttieae Mistretta s. n., RSA Pappophorum bicolor E. Fourn. Pappophoreae PI 216526 Ehrharta longifolia Schrad. Oryzoideae Ehrharteae PI 270493 Oryza sativa L. Oryzeae GBAN-X15901 Zizania aquatica L. Zizanieae KH 9423 Andropogon gerardii Vitman Panicoideae Andropogoneae PI 215661 Hyparrhenia hirta (L.) Stapf N. Barker 1134 Sorghum bicolor (L.) Moench KH 9408 Zea mays L. GBAN-X86563 Loudetiopsis chrysothrix (Nees) Conert Arundinelleae KH 5748 Tristachya biseriate Stapf N. Barker 1126 Panicum capillare L. Paniceae PI 220025 Paspalum scrobiculatum L. PI 365511 Pharus latifolius L. Pharoideae Phareae L. Clark 1302, ISC Agrostis gigantea Roth. Pooideae Aveneae T. Wieboldt s. n. Avena sativa L. KH 1 Phalaris arundinacea L. PI 578797 Phleum pratense L. PI 303130 Bromus inermis Leyss Bromeae PI 315375 Briza erecta Lam. Poeae PI 282880 Poa pratensis L. PI 620724 Vulpia myuros (L.) C. C. Gmel. PI 217618 Nassella tenuis (Phil.) Barkworth Stipeae PI 282897 Stipa offneri Breistr. PI 287941 Hordeum vulgare L. Triticeae GenBank X64129 Leymus angustus (Trin.) Pilg. PI 547357 Triticum aestivum L. ``Massey'' Starling s. n. eses to be tested are outlined in the next section. Also MATERIALS AND METHODS addressed are patterns of nucleotide and amino acid sub- stitutions resulting from these mutations and their evo- We examined matK sequences representing 52 grass genera, 26 tribes, lutionary implications in Poaceae and other plant groups. and nine subfamilies (Table 1). In addition, we obtained seven sequenc- December 1999] HILU AND ALICEÐMATK INDELS IN POACEAE 1737 Fig. 1. Amino acid alignment of taxa representing bryophytes, gymnosperms, and angiosperms. Consensus sequences for each