Splendor in the Grasses

Grasses provide over half of the world's caloric intake and are major components of many terrestrial ecosystems. They are also used in cultivated landscapes, as construction materials, and as biofuel feedstocks. The grasses form a single genetic system (Bennetzen and Freeling, 1993), in which genomic, morphological, and physiological similarity permits generalizations among species. In addition, comparisons among and between the cereals can provide novel insights into the evolutionary history of this group of 10,000 species (Kellogg and Birchler, 1993; Kellogg and Shaffer, 1993). The articles in this Focus Issue on the Grasses share two general themes. First, many aspects of grass biology, including many of the genes characterized here, appear to be unique to the grasses. Second, the breadth and depth of genomic and genetic resources in the grasses are enormously powerful. These themes are outlined below.

THE GRASSES ARE UNIQUE, WITH UNIQUE UNDERLYING BIOLOGY

Most of the articles in this issue characterize one or more genes that appear to be grass-specific. Some of these are novel proteins (e.g. Tie-dyed1, described by Ma et al. [2009]), whereas others are novel paralogs, e.g. many of the floral genes described by Thompson and Hake (2009). Paterson et al. (2009) outline mechanisms by which novelty may arise among and between the grass genomes. The studies in this issue thus build on the computational work of Campbell et al. (2007), who identified genes that occurred in grasses and in no other plant genomes. The subtitle by Colasanti and Coneva (2009) is particularly apt in this regard: “something borrowed, something new.” Many genes and gene functions are similar to those of the model eudicot Arabidopsis (“borrowed”), but others are restricted to grasses (“new”).

Morphological similarities among the cereals are particularly striking. Four articles (Colasanti and Coneva, 2009; McSteen, 2009; Thompson and Hake, 2009; Yuan et al., 2009) address the genetic basis of grass architecture and development. They show that the developmental biology of the cereals is unique in many ways compared to Arabidopsis; this is hardly surprising given the many years of evolution since monocots and eudicots last shared a common ancestor.

The distinctive architecture of the grass plant in part reflects differential regulation of hormones (McSteen, 2009; Wu et al., 2009). The grasses are a particularly good system for dissecting hormone effects because the plants produce so many different types of meristems. For example, axillary buds near the base of the stems are regulated differently from those farther up, and buds in the inflorescence can have a variety of different fates. Each responds in a different way to hormonal stimuli and is regulated differentially by different proteins. Seed dispersal is also distinctive, and of both agronomic and ecological importance; in wild species, seeds are dropped or the inflorescence breaks up in various ways. Sang (2009) summarizes recent data assessing the genetic basis of disarticulation and its involvement in the domestication process.

Much of the biochemical research reported here focuses on polysaccharides of various sorts. Fincher (2009) summarizes recent work on cell walls in the Poaceae. While many such studies are driven by the nascent biofuels industry, understanding the complexity of cell wall biosynthesis is important in its own right. The articles by Lasseur et al. (2009) and Boehlein et al. (2009) demonstrate how few changes are required to change the specificity of a biosynthetic enzyme (fructan:fructan 6G-fructosyltransferase and ADP-Glc pyrophosphorylase, respectively), which then produces a polysaccharide with appreciably different structure and biochemical properties. The biochemical change is of industrial importance, particularly in the case of starch, but also presumably affects the physiology and ultimately the fitness of the plant.

Many grasses use the C₄ photosynthetic pathway, which has evolved independently multiple times creating a replicated experiment in evolution (Christin et al., 2009). Appropriate comparisons among C₄ species and their close C₃ relatives will provide new tools for dissecting the pathway. At the single gene/species level, advances in our understanding of photorespiration in C₄ photosynthesis are reported by Zelitch et al. (2009). New data are also accumulating on how carbohydrates are transported. Ma et al. (2009) describe a novel protein (Tie-dyed1) involved in loading carbohydrates into the phloem. Braun and Slewinski (2009) place this new protein in the context of other proteins involved in phloem loading.

Investigations of plant pathogen interactions have focused profitably on specific types of resistance and interactions with single highly noxious pathogens. However, Manosalva et al. (2009) focus on proteins (germin-like proteins) that confer more generalized resistance. The high copy number of these proteins makes them challenging to analyze, but clearly resistance to a broad range of pathogens is as important as specific defense against a single pathogen. Meng et al. (2009) describe the barley Bln1 gene that is induced by the fungal pathogen Blumeria graminis; the gene product reduces the plant's baseline defense capacity, thereby permitting access by the fungus. It is not clear whether the main function of Bln1 is to regulate pathogen susceptibility and resistance, or whether it has a function in normal plant metabolism and is simply co-opted by the fungus for its own nefarious purposes. Degenhardt (2009) provides an ecological perspective by describing indirect defenses in the grasses, in which a volatile compound produced by the plant attracts an enemy of the pathogen or herbivore, while Trail (2009) focuses on the genomic resources now available for Fusarium, a major fungal pest of cereal crops.

The converse of pathogenicity is symbiosis. Chen et al. (2009) describe ion channel proteins that are conserved between legumes and rice. In legumes the proteins are needed for both nodulation and establishment of mycorrhizal associations, whereas in rice they are necessary for the latter.

With respect to abiotic stress responses, Nakashima et al. (2009) summarize similarities as well as differences in transcriptional regulatory networks between Arabidopsis and the grasses. Ryan et al. (2009) describe citrate efflux in wheat as a second mechanism for aluminum resistance and characterize a gene with similarity to a citrate efflux protein in barley (HvAACT1). Independently, Yokosho et al. (2009) describe a related protein in rice, OsFRDL1, which is involved in iron translocation, is expressed within the pericycle, and does not affect citrate efflux in the roots. Constant recycling of cellular contents is part of normal growth and development, and all of the necessary genes (and therefore presumably proteins) are conserved among Arabidopsis, rice, and maize (Chung et al., 2009). However, as with most other components of the cellular machinery, the numbers of genes differ between the two cereals and the eudicot model, and one protein (ATG7) contains a large duplication in rice and maize, which may hint at a novel grass-specific function.

It is impossible to consider the biology of the cereals without also considering their economic importance. Sabelli and Larkins (2009) provide a thorough review of endosperm development, clearly one of the central agricultural features of the Poaceae. Edgerton (2009) provides a view of how maize yield must change over the coming years, and how some of the basic research described in this issue might affect commercial applications. Distelfeld et al. (2009) describe the VRN2 locus of bread wheat, durum wheat, and a variety of diploid wheat relatives, which suffices to make the plant require cold weather (vernalization) for flowering. The extensive polymorphism at this locus raises the possibility of breeding for lines that do or do not require vernalization, depending on the particular local environment.

GRASSES HAVE POWERFUL GENOMIC AND GENETIC RESOURCES

A second theme in this issue is the power of genomic approaches to provide not only data in the form of genome sequence and transcriptome profiles but also insights into genome biology (Messing 2009; Wicker et al., 2009). An astounding 47 Poaceae species have transcriptome or genome sequence datasets available (Buell, 2009). More significantly, since the first grass genome (rice) was sequenced in 2002, three others have been added (maize, sorghum, and Brachypodium distachyon). B. distachyon and foxtail millet (genome sequence anticipated in 2009; Doust et al., 2009) are so-called “sequence-driven” model species, in which a genome sequence will be completed before development of the accompanying genetic and functional genomic resources/technologies, such as germplasm collections, mutants, efficient transformation methods, knockout collections, etc. While technical issues with large repetitive genomes such as barley and wheat have been a “speed bump” in the genomics race, they are not absolute impediments as highlighted by construction of a physical map of the 1-Gb chromosome 3B of hexaploid wheat (Paux et al., 2008) and development of a consortium for sequencing the diploid (5-Gb) barley genome (Schulte et al., 2009).

Other tools developed for the grasses are mutant and germplasm collections, databases, and functional genomic methodologies that are critical to understanding gene function and, ultimately, biological processes. Germplasm collections for the cereals are well established throughout the world (Sachs, 2009). Rice has extensive mutant collections made available by the International Rice Functional Genomics Consortium (Krishnan et al., 2009). Conventional TILLING (Targeting Induced Local Lesions IN Genomes), a reverse genetics approach to help understand gene function, is well advanced in the grasses, and modifications, such as Eco-TILLING and SequeTILLING, offer even deeper surveys of genetic diversity (Weil, 2009). Coupled with these genome wide approaches, virus-induced gene silencing (VIGS) provides a rapid and targeted approach to disrupt genes in select grass species (Scofield and Nelson, 2009). Adaptation of VIGS to additional species as well as the report that VIGS is effective in adult tissues will enable high-throughput screens in the grasses. The structural (i.e. sequence) and functional resources for the Poaceae have resulted literally in a deluge of data that is best housed in databases that can be queried and archived. As highlighted by Childs (2009), multiple bioinformatic resources facilitate navigation and utilization of the voluminous datasets available for the Poaceae; many of these focus on particular taxa or biological features such as grass transcription factors as described by Yilmaz et al. (2009).

OUTLOOK

We anticipate that future Poaceae research will be highly comparative in nature, providing a powerful synergistic effect among researchers studying the cereals. Comparison of multiple Poaceae species will provide insights to supplement those obtained from comparison of multiple mutant lines within a single species. Access to genome sequences, functional genomic resources, and diverse germplasm collections from grass species will enable more efficient and facile determination of biological aspects that are unique to a species, apply to all Poaceae, or characterize all angiosperms. With respect to agricultural uses of the Poaceae, not only will the information from each major cereal species be applicable to the other major crops, but such information can also be used to translate information to minor grains with small research communities and emerging crops or wild species with limited resources.