Poaceae Genomes: Going from Unattainable to Becoming a Model Clade for Comparative Plant Genomics

Genomics has an immense potential for improving our understanding of critical issues in plant growth and development, some of which can be applied to improvement of crop production. Midway into the second decade of genomics, genome and transcriptome sequencing efforts with the Poaceae are impressive given the technical and fiscal challenges presented by the typically large, repetitive genomes found within the Poaceae (Smith and Flavell, 1975; Arumuganathan and Earle, 1991; SanMiguel et al., 1996). Indeed, as of October 30, 2008, there are 10,847,522 Poaceae sequences representing 11,142 Mb (11.1 Gb) in GenBank, confirming the fast pace of sequence generation for the Poaceae. With respect to representation of genera and species within the Poaceae, 2,740 of the approximately 10,000 species reported within the Poaceae (http://www.kew.org/scihort/poaceae.html) have at least one sequence in GenBank. Genome-scale datasets represent a more narrow phylogenetic base and of the 47 Poaceae species with genome or transcriptome sequences (Table I; discussed below), 32 are derived from the BEP clade (Bambusoideae, Ehrhartoideae, Pooideae) and 15 are derived from the PACCMAD clade (Panicoideae, Arundinoideae, Chloridoideae, Centothecoideae, Micrairoideae, Aristidoideae, and Danthonioideae) and represent five subfamilies, nine tribes, and 24 genera within the Poaceae (Fig. 1). This article is intended to provide a short introduction to genome sequencing efforts in the Poaceae and an abbreviated report of completed and ongoing genome and transcriptome sequencing efforts for Poaceae species to not only demonstrate the potential of species within the Poaceae for understanding plant biological processes but also to highlight the Poaceae as a model family for comparative genomics.

Table I.

List of Poaceae species with transcriptome and genome sequence data and/or initiatives

Species	Common Name	Clade (Subfamily, Tribe)	EST No.a	Statusb	Reference or URL
BEP clade
Phyllostachys edulis	Moso bamboo	Bambusoideae, Bambuseae	2,141	ESTs
Oryza alta		Ehrhartoideae, Oryzeae	2	Physical map, BES	http://www.omap.org/
Oryza australiensis		Ehrhartoideae, Oryzeae	0	Physical map, BES	http://www.omap.org/
Oryza brachyantha		Ehrhartoideae, Oryzeae	0	Physical map, BES	http://www.omap.org/
Oryza coarctata		Ehrhartoideae, Oryzeae	499	Physical map, BES	http://www.omap.org/
Oryza glaberrima	African rice, Guang fu dao	Ehrhartoideae, Oryzeae	0	Physical map, BES	http://www.omap.org/
Oryza granulata	Yo li dao	Ehrhartoideae, Oryzeae	144,859	Physical map, BES, ESTs	http://www.omap.org/
Oryza minuta		Ehrhartoideae, Oryzeae	5,760	Physical map, BES, ESTs	http://www.omap.org/
Oryza nivara		Ehrhartoideae, Oryzeae	0	Physical map, BES	http://www.omap.org/
Oryza officinalis	Yao yong dao	Ehrhartoideae, Oryzeae	1,471	Physical map, BES, ESTs	http://www.omap.org/
Oryza punctata	Red rice	Ehrhartoideae, Oryzeae	806	Physical map, BES	http://www.omap.org/
Oryza ridleyi		Ehrhartoideae, Oryzeae	0	Physical map, BES	http://www.omap.org/
Oryza rufipogon	Brownbeard rice, red rice	Ehrhartoideae, Oryzeae	1	Physical map, BES	http://www.omap.org/
Ricec	Rice, Dao	Ehrhartoideae, Oryzeae	1,220,877	Resequencing, ESTs	McNally et al. (2006); http://oryzasnp.plantbiology.msu.edu; http://irfgc.irri.org/index.php?option=com_content&task=view&id=14&Itemid=106
Rice sp. indicac	Rice, Dao	Ehrhartoideae, Oryzeae		WGS draft, ESTs	Yu et al. (2002, 2005)
Rice sp. japonicac	Rice, Dao	Ehrhartoideae, Oryzeae		WGS draft, finished, ESTs	Barry (2001); Goff et al. (2002); Yu et al. (2002, 2005); The International Rice Genome Sequencing Project (2005)
Agrostis capillaris	Waipu, Colonial bent grass	Pooideae, Aveneae	7,743	ESTs
Agrostis stolonifera	Creeping bent grass	Pooideae, Aveneae	9,114	ESTs
Avena sativa	Oat	Pooideae, Aveneae	7,633	ESTs
B. distachyon	Purple false brome	Pooideae, Brachypodieae	20,449	WGS in progress, ESTs	http://www.jgi.doe.gov/sequencing/why/51281.html
Festuca arundinacea	Tall fescue	Pooideae, Poeae	44,377	ESTs
Lolium multiflorum	Italian ryegrass	Pooideae, Poeae	5,968	ESTs
Lolium perenne	Perennial ryegrass	Pooideae, Poeae	1,816	ESTs
Lolium temulentum	Darnel	Pooideae, Poeae	6,336	ESTs
Puccinellia tenuiflora	Alkali grass	Pooideae, Poeae	4,252	ESTs
Aegilops speltoides	Goat grass	Pooideae, Triticeae	4,315	ESTs
Aegilops tauschii	Tausch's goatgrass	Pooideae, Triticeae	116	Physical map	http://wheatdb.ucdavis.edu:8080/wheatdb/
Hordeum vulgare	Barley	Pooideae, Triticeae	502,895	ESTs, Physical map	http://barleygenome.org; http://phymap.ucdavis.edu:8080/barley/
Leymus chinensis		Pooideae, Triticeae	1,692	ESTs
Leymus cinereus × Leymus triticoides	Great Basin wild rye × Creeping wild rye	Pooideae, Triticeae	28,786	ESTs
Secale cereale	Rye	Pooideae, Triticeae	9,298	ESTs
Wheat	Bread wheat	Pooideae, Triticeae	1,051,304	ESTs, Physical map	http://www.wheatgenome.org/; http://urgi.versailles.inra.fr/projects/Triticum/eng/
Triticum monococcum	Einkorn wheat	Pooideae, Triticeae	11,190	ESTs
Triticum turgidum	Rivet wheat, Poulard wheat	Pooideae, Triticeae	30,874	ESTs
PACCMAD clade
Cynodon dactylon	Bermuda grass	Chloridoideae, Cynodonteae	20,148	ESTs
Eleusine coracana	Finger millet	Chloridoideae, Cynodonteae	1,749	ESTs
Eragrostis curvula	Weeping love grass	Chloridoideae, Cynodonteae	12,295	ESTs
Eragrostis tef	Tef	Chloridoideae, Cynodonteae	2,816	ESTs
Spartina alterniflora	Smooth cordgrass	Chloridoideae, Cynodonteae	1,255	ESTs
Saccharum hybrid cultivar	Sugarcane	Panicoideae, Andropogoneae	9,415	ESTs
Saccharum officinarum	Sugarcane	Panicoideae, Andropogoneae	252,698	ESTs
Sorghum	Sorghum	Panicoideae, Andropogoneae	209,814	Gene enrichment (complete), WGS draft (complete), ESTs; Physical map	Bedell et al. (2005); http://www.jgi.doe.gov/sequencing/why/3060.html
Sorghum halepense	Johnson grass	Panicoideae, Andropogoneae	1,965	ESTs
Sorghum propinquum		Panicoideae, Andropogoneae	20,881	ESTs
Maize	Maize, Corn	Panicoideae, Andropogoneae	1,464,855	Gene enrichment (complete); Draft BAC by BAC with targeted finishing, ESTs; Physical map	Whitelaw et al. (2003), Palmer et al. (2003); http://www.maizesequence.org/index.html
Cenchrus ciliaris	Buffelgrass, African foxtail grass	Panicoideae, Paniceae	21,729	ESTs
Pennisetum glaucum	Pearl millet	Panicoideae, Paniceae	2,848	ESTs
Setaria italica	Foxtail millet	Panicoideae, Paniceae	2,124	Planned WGS, ESTs	http://www.jgi.doe.gov/sequencing/why/99178.html

^aNumbers of ESTs were obtained from dbEST, October 17, 2008.

^bBES, BAC end sequence dataset. Resequencing refers to a project using hybridization-based resequencing to identify single nucleotide polymorphisms in 20 rice accessions.

^cAll rice ESTs are combined and reported as rice rather than separating them into indica and japonica.

Figure 1.

Phylogeny of grasses for which genome sequence data has been or will be generated in the near future. Asterisk indicates one poorly supported node. Polyploid species for which genome relationships are known are shown to the right of the diploids, with lines indicating ancestry. Most are tetraploid but one (wheat [T. aestivum]) is hexaploid.

EXPRESSED SEQUENCE TAGS: THE START OF THE GENOMICS ERA IN THE POACEAE

The first genome-scale sequences generated for the Poaceae were ESTs that represent the transcribed portion of a genome and provide a rapid, economic approach to sampling the gene space of an organism. As early access sequence datasets, ESTs can be used for (1) development of genetic markers (for example, see Harushima et al., 1998), (2) electronic gene expression analyses (for example, see Ewing et al., 1999), (3) improvement of structural annotation (Haas et al., 2003), and (4) functional genomic resources for use in overexpression, in vitro expression, and gene-silencing studies. In 1997, the first set of ESTs for a Poaceae species was reported for rice (Oryza sativa; Yamamoto and Sasaki, 1997). Now, 11 years later, there are 36 Poaceae species with EST collections greater than 1,000 sequences (Table I); of these, >1 million ESTs are available for three Poaceae species (maize [Zea mays], rice, wheat [Triticum aestivum]). Collectively, as of October 30, 2008, within the dbEST division of GenBank, there were 5,491,939 Poaceae EST sequences totaling 2,764 Mb (2.76 Gb). With the high degree of repetitive sequences within the majority of Poaceae species, coupled with the availability of high-throughput next generation sequencing platforms for transcriptome sequencing (Cloonan et al., 2008; Rosenkranz et al., 2008), ESTs will continue to provide a rich source of genic sequences for grass researchers and it should be fully envisioned that within the coming years, EST collections will be available for thousands of Poaceae species.

GETTING AT THE GENOME: GENOME SEQUENCES

With continued advancements in technology and concomitant reductions in costs over the last decade, whole-genome sequences have been generated for multiple species within the Poaceae (Table I). Rice was not only the first crop species but also the second plant species with a genome sequence (Barry, 2001; Goff et al., 2002; Yu et al., 2002; The International Rice Genome Sequencing Project, 2005; Yu et al., 2005). Currently, genome sequence is available for two subspecies of rice, indica and japonica. Perhaps most importantly for all current and future Poaceae genome sequences, a high-quality, near-complete genome sequence is available for the Nipponbare cultivar of japonica rice (The International Rice Genome Sequencing Project, 2005) that will most likely provide the only gold standard reference genome for the Poaceae for the near future. Indeed, the reference Nipponbare sequence was used to resequence a set of 20 rice lines using hybridization-based sequencing to identify single nucleotide polymorphisms (McNally et al., 2006; http://irfgc.irri.org/index.php?option=com_content&task=view&id=14&Itemid=106; http://oryzasnp.plantbiology.msu.edu/). Draft genome sequences are also available for sorghum (Sorghum bicolor), Brachypodium distachyon, and maize with analyses and full public release anticipated in 2009 (Table I). A genome project is planned for foxtail millet (Setaria italica) by the U.S. Department of Energy Joint Genome Institute (Table I). Physical resources in the form of bacterial artificial chromosome (BAC) map clones have been developed for a number of Poaceae species in advance of genome sequencing (Table I). Most notably, the OMAP project (Kim et al., 2008) has generated physical maps and BAC end sequences for 12 Oryza species in support of comparative genomics within this important genus.

In addition to transcriptome and whole-genome sequences, large sets of genomic sequences are available within the GSS, HTG, WGS, and PLN divisions of GenBank. Within the GSS division, which includes gene enrichment as well as BAC end sequences, 5,072,454 Poaceae sequences (3,337 Mb) are available. Although the maize and sorghum gene enrichment sequences within the GSS division have now been superceded by draft genome sequences, the gene enrichment approaches of methylation filtration and high Cot were highly successful in generating genic sequences for maize and sorghum, thereby providing early access to the gene space (Palmer et al., 2003; Whitelaw et al., 2003; Bedell et al., 2005). Other Poaceae sequences in GenBank include 179,196 sequences (1,169 Mb) within the PLN division, 18,655 sequences (3,086 Mb) within the HTG division, and 85,280 sequences (785 Mb) within the WGS division.

It should be noted that the majority of the sequence available currently for the Poaceae are derived from a few species of high agricultural importance (rice, maize, wheat, and sorghum). As shown in Figure 2, although 13 of the 47 species with genome-scale datasets, resources, or initiatives listed in Table I have >100 Mb of total sequence in GenBank, three-quarters of the sequence are from maize or Oryza species reflective of the heavy bias in Poaceae genome sequencing projects to date. However, with access to the next generation of genome sequencing technologies (Margulies et al., 2005; Holt et al., 2008; Sarin et al., 2008), it can certainly be envisaged that researchers will have access to dozens of Poaceae genomes in the near future. Furthermore, application of these next generation sequencing technologies along with techniques to enrich for subfractions of the genome (Albert et al., 2007; Hodges et al., 2007; Okou et al., 2007) will greatly enhance resequencing of additional cultivars or accessions, thereby providing an unlimited set of sequence resources to examine genome diversity at the species level.

Figure 2.

Genome sequence availability for 47 Poaceae species with genome projects. Sequence was downloaded for all 47 species from GenBank (October, 2008) and summed for all divisions. Thirteen species are represented individually in the pie chart; sequence for 34 species with less than 100 Mb of total sequence in GenBank were grouped into Other.

Certainly, this is an exciting time to be engaged in Poaceae research as even if genomics is not your research discipline, access to not just one but multiple Poaceae genome sequences provides not only a robust set of resources for biological inquiries, but also provides a perspective of gene function in a phylogenetic context. With this deluge of genomic sequence data, the storage, handling, analysis, and use of the large-scale genomic sequence and annotation data becomes problematic for most researchers. Consequently, resources, databases, and analyses tools need to be developed to ensure these genome datasets can be used in a feasible and intelligent manner, thereby maximizing the return on the investment of obtaining the genome sequence. Certainly, Poaceae researchers are not alone in forging a path through the morass of genome sequence data in the early 21st century, and the tools, resources, software, and knowledge gained from other genomic research endeavors throughout the Tree of Life will be instrumental in obtaining a full understanding of the pan-Poaceae genome.