Translational Genomics for Bioenergy Production from Fuelstock Grasses: Maize as the Model Species

There is increasing worldwide economic interest and scientific focus on developing biofuel crops. Historically, biofuels were plant materials, such as grasses and wood, that could be burned to generate heat, which can be used directly or converted into electricity. Today, biofuel connotes conversion of plant biomass through fermentation into liquid fuels. Most proposed strategies aimed at meeting future U.S. energy needs include the use of such biofuels. To begin to address this need, ways to improve plants for sustainable production must be devised, especially for the high-yield grasses. Sugarcane (Saccharum officinalis) and maize (Zea mays ssp mays, commonly referred to as corn in the U.S.) are already used for ethanol production from sugar and starch, respectively, and research is expanding rapidly to recover cellulosic biofuels from temperate-zone maize, sweet sorghum (Sorghum bicolor), Miscanthus × giganteus (the perennial sterile hybrid between M. sinensis and M. sacchariflorus), and perennial switchgrass (Panicum virgatum). Here, we discuss some of the relevant characteristics of these species as fuelstocks and reasons for considering maize as a primary model for biofuel research.

TARGETING C₄ GRASSES

All the above-mentioned energy fuelstocks are members of the Panicoid subfamily of the grass family (Grass Phylogeny Working Group, 2001; Figure 1A) and use the physiologically unique and highly efficient C₄ photosynthetic pathway (Osborne and Beerling, 2006). The C₄ pathway is intrinsically more efficient than the C₃ classic Calvin cycle because the first step of C₄ photosynthesis is a carbon concentrating mechanism. In addition, C₃ plants keep their stomata open for gas exchange during photosynthesis, which causes water loss. C₄ plants can photosynthesize with stomata nearly closed, thus reducing water loss to the environment and increasing water use efficiency. Also, plants using the C₄ photosynthetic pathway are better equipped to handle high temperatures, drought, and nitrogen limitations than closely related C₃ plants (Makino et al., 2003; Raines 2006). Most C₃ grasses are closely related and belong to the BEP (Bambusoid, Ehrhartoid, and Pooid) clade (shown in Figure 1B; Edwards et al., 2007).

Figure 1.

PACCAD and BEP Clades of Poaceae, the Grass Family.

(A) Phylogenetic relationships among the fuelstock grasses: subfamily Panicoideae. Cladogram represents the robust and generally accepted taxonomy of these subfamily members. For more information, visit the Grass Genera of the World website at http://delta-intkey.com/grass/.

(B) Fuelstock plants are in the subfamily Panicoideae of the PACCAD clade, whereas rice is in the subfamily Ehrhartoideae and Brachypodium is in the subfamily Pooideae of the BEP clade. The phylogenetic tree reproduced here is from Figure 2 of the Grass Phylogeny Working Group's 2001 publication. As noted in the original publication, this represents the single most parsimonious tree for the grasses and relatives, based on eight sets of data. Length = 8752 steps, consistency index = 0.375, and retention index = 0.557. Numbers above branches are percentages of bootstrap replicates and below are Bremer support. Brackets indicate the classification for the Poaceae as reported by the Grass Phylogeny Working Group (2001).

SPECIES THAT MAKE SENSE FOR BIOFUEL PRODUCTION IN THE U.S.

Sugarcane is the basis for Brazil's independence from fossil fuels, but this tropical species is an appropriate fuelstock in only a few regions of the U.S. Maize is grown on more than 80 million acres in the US, but there is already evidence that its use as a fuelstock can drive up prices for its use as a commodity for feed, food, and industrial applications. Sorghum is another possible biofuel suitable for the temperate zone. Although sweet sorghum has many advantages for use as a fuel crop (e.g., directly fermentable sugars are present in the stalk, the grain can be used for starch fermentation, and fertilizer inputs are low), a serious seasonality problem exists: because sweet sorghum's stem sugars are unstable, processing plants would need to be equipped to deal with an entire crop within a matter of days, an unworkable model (reviewed in Gnansounou et al., 2005). The quick polymerization of free sugars into starch in sweet maize was overcome by selecting mutations in the genes responsible for polymer synthesis (Coe et al., 1988), and it is possible that sorghum could be improved using insights from sweet maize stabilization.

Like maize and sweet sorghum, Miscanthus and switchgrass can be grown in the Midwest. Importantly, these nondomesticated grasses grow on land where maize fails to thrive (e.g., Naidu et al., 2003), which would avert displacing current maize acreage. For these reasons and others, Miscanthus and switchgrass are the current front-runners as viable U.S. biofuel candidates.

MODELS FOR BIOFUEL RESEARCH AND DEVELOPMENT

Leveraging insights from model organisms for applied purposes uses basic genetic findings to solve practical problems (Varshney and Koebner, 2006), a practice which has evolved into the field of translational genomics. For the purposes of developing a fuelstock from Miscanthus, switchgrass, and future C₄ fuel production candidates, translational genomics can speed discovery and focus applied research efforts. Using the genetic and genomic information we have for model grass species, proven tactics for improving fuelstock cultivars to maximize efficiency could be developed more rapidly. Four model species for studying fuel production by grasses are currently recognized: rice, Brachypodium distachyon, sorghum, and maize.

The best feature of rice as a model is its completely sequenced genome (Goff et al., 2002). Despite this asset, research on rice is unlikely to strike oil for improving Miscanthus or switchgrass for many reasons. First and foremost, rice uses C₃ photosynthesis. Furthermore, access to rice germplasm is impeded by strictly enforced import restrictions; rice genetics efforts are laborious and, hence, gene functions are not as well characterized as in other model plants; and genetic diversity in available accessions of rice, a self-fertilizer, is low relative to outcrossing species. Brachypodium is touted as a good model for fuelstock development because (1) it has a small genome (∼355 Mbp), (2) diploid, tetraploid, and hexaploid cultivars exist, and (3) it has a small physical stature and is self-fertile during a short lifecycle with simple growth requirements (reviewed in Draper et al., 2001; Huo et al., 2006). But, like rice, Brachypodium is not closely related to Miscanthus or switchgrass and lacks C₄ photosynthesis.

Maize and sorghum are better models for improving C₄ biofuel grasses than rice or Brachypodium for many reasons. (1) Maize and sorghum use the C₄ photosynthetic pathway. (2) Unlike Brachypodium and rice (members of the BEP clade; see Figure 1B), maize and sorghum are in the same subfamily of the PACCAD clade (consisting of the Panicoid, Arundinoid, Chloridoid, Centothecoid, Aristidoid, and Danthonioid lineages) as Miscanthus and switchgrass, and information derived from these model species likely will translate more readily to useful information for developing fuelstock grasses as a crop as well as for modeling the technologies necessary for processing plants for lignocellulosic energy production. For example, gene and regulatory sequences of fuelstock grasses will be similar to maize and sorghum, as will regulatory circuits to manage the time, site, and amount of molecules produced. Furthermore, the cell wall composition and genes determining this key property are likely to be highly conserved within the PACCAD clade, facilitating transfer of knowledge from maize, for which mutants of cell wall synthesis are already studied, to biofuel crops. (3) Sorghum genome sequencing has been completed, and the maize genome is currently being sequenced (anticipated completion by the end of 2008).

Because many new biofuel candidates are polyploid and perennialism is a key feature of sustainable biofuel production in most current discussions, maize is particularly interesting to consider given that subspecies of Z. mays include both diploids and tetraploids that use both the annual and perennial lifestyles (Takahashi et al., 1999; Westerbergh and Doebley, 2004). In addition, maize inbred lines and maize relatives can be crossed easily to produce hybrids, which would mimic the method used to produce Miscanthus from an interspecific cross. Other reasons that maize can be considered the forerunner as a model for biofuel research include maize's incomparable body of knowledge; the large community of maize genetics, molecular biology, and physiology researchers as well as breeders, agronomists, and other field specialists who know how to translate laboratory findings into valuable traits, including continuous yield improvement in the U.S.; plus, the maize research community knows how to talk to and work with farmers.

RISKS ASSOCIATED WITH MONOCULTURE: A KEY LESSON FROM MAIZE PRODUCTION AGRICULTURE

A prime example of how discoveries from maize can and should be considered for developing a fuelstock comes from the lessons learned in the early 1970s when the U.S. maize crop failed during a pervasive Southern corn leaf blight infestation (reviewed in Levings, 1993). Commercial maize seed derives from crossing two genetically distinct parents to produce superior hybrid progeny. To save the cost of manually detasseling the male flowers from hundreds of millions of plants to produce hybrid seed, the industry adopted cytoplasmic male sterility (cms-Texas); this genetic trick ensured that plants serving as maternal parents produced no pollen and all seeds on their ears (progeny) would derive from cross-pollination by nearby male plants. The male parent transmitted dominant restorers of fertility; consequently, the progeny carried the cms trait but were fully fertile. This system not only made economic sense but ensured that all progeny were hybrid—growers received a uniform, high-quality product. Unfortunately, in 1969, the pathogen Helminthosporium maydis mutated to generate race T, a strain highly pathogenic on cms-Texas. Nearly all commercial U.S. acreage was susceptible, and massive crop losses occurred in 1969 and 1970, until the old-fashioned hand-detasseling was employed once again to generate resistant hybrid seed. One of the clear lessons from this disaster is that a monoculture of a superior strain confers tremendous risk.

Using diverse sources of any crop species will be important to the biofuel industry, particularly as acreage increases. To achieve replacement of slightly more than half of U.S. liquid fuel consumption using cellulose-derived ethanol, the U.S. Departments of Energy and Agriculture project that a combination of existing agriculture sources and new biofuel crops will be required. Current projections are that 50% of the maize stover (leaves and stalks) and wheat straw in the U.S. plus the yield from biofuel grasses, such as Miscanthus and switchgrass, grown on 40 million acres would be required (Somerville, 2007). The 40 million acres of U.S. farmland previously in production that is currently in set-aside acreage is the proposed biofuel land; much of this reservoir is contiguous to current maize production. Movement of viral, bacterial, fungal, or insect pests that are benign on maize to biofuel crops and vice versa should be considered a question of when and not if. Key components of biofuel crop development must include a genetically diverse plant germplasm base, use of crop rotation of biotypes, good management practices over time and space, and monitoring pest movement between fuelstocks and traditional crops to be ready to respond to inevitable crop production challenges. Devastating losses of one crop would lead to reduced biofuels production and potentially to energy shortages. This risk can be minimized by having a diverse crop portfolio in production and by planning how to manage use of foods other than corn for humans and animals. As the much-studied Miscanthus × giganteus is the result of a single cross producing an individual sterile triploid plant from which virtually all production acreage is derived, it has little to no genetic diversity. In addition, the National Plant Germplasm System has only 12 accessions of M. sinensis currently available and one for M. saccharifloris. Not only should the crossing program be expanded to develop more high production lines, but the wild germplasm (centered in Asia) should be collected as a reservoir for plant breeding efforts to diversify the crop's genetic base, now and in the future. In the 1990s, several European efforts to collect Miscanthus germplasm were funded but were then abandoned. For the U.S. to make strides in developing sustainable fuelstocks, the wisdom derived from both science and historical lessons must be taken into account, with particular emphasis on increasing genetic diversity from which advances in crop research and development can be drawn.