• Open Access

Sugarcane for bioenergy production: an assessment of yield and regulation of sucrose content

Authors

  • Alessandro J. Waclawovsky,

    1. Departamento de Bioquímica, Instituto de Química, Av. Prof. Lineu Prestes, São Paulo, Brazil
    Search for more papers by this author
    • Present address: Alessandro J. Waclawo-vsky, Universidade Tecnológica Federal do Paraná (UTFPR), Campus Dois Vizinhos, Estrada para Boa Esperança, km 04, Dois Vizinhos, PR, Brazil, 85660-000.

    • The first two authors contributed equally to this article.

  • Paloma M. Sato,

    1. Departamento de Bioquímica, Instituto de Química, Av. Prof. Lineu Prestes, São Paulo, Brazil
    Search for more papers by this author
    • The first two authors contributed equally to this article.

  • Carolina G. Lembke,

    1. Departamento de Bioquímica, Instituto de Química, Av. Prof. Lineu Prestes, São Paulo, Brazil
    Search for more papers by this author
  • Paul H. Moore,

    1. Hawaii Agriculture Research Center, Kunia, HI, USA
    Search for more papers by this author
  • Glaucia M. Souza

    Corresponding author
    1. Departamento de Bioquímica, Instituto de Química, Av. Prof. Lineu Prestes, São Paulo, Brazil
    Search for more papers by this author

* (tel +55 11 30918511; e-mail glmsouza@iq.usp.br

Summary

An increasing number of plant scientists, including breeders, agronomists, physiologists and molecular biologists, are working towards the development of new and improved energy crops. Research is increasingly focused on how to design crops specifically for bioenergy production and increased biomass generation for biofuel purposes. The most important biofuel to date is bioethanol produced from sugars (sucrose and starch). Second generation bioethanol is also being targeted for studies to allow the use of the cell wall (lignocellulose) as a source of carbon. If a crop is to be used for bioenergy production, the crop should be high yielding, fast growing, low lignin content and requiring relatively small energy inputs for its growth and harvest. Obtaining high yields in nonprime agricultural land is a key for energy crop development to allow sustainability and avoid competition with food production. Sugarcane is the most efficient bioenergy crop of tropical and subtropical regions, and biotechnological tools for the improvement of this crop are advancing rapidly. We focus this review on the studies of sugarcane genes associated with sucrose content, biomass and cell wall metabolism and the preliminary physiological characterization of cultivars that contrast for sugar and biomass yield.

Energy crops: trait improvement and yield

There has been a research surge in recent years aimed at developing alternative sources of energy that can decrease or replace the use of fossil fuel. The fluctuating prices of petroleum, its dwindling worldwide stocks and the adverse environmental effects of fossil fuel usage have collectively renewed interest in the search for alternative sources of energy. Wood, livestock manure, microbial biomass, agricultural waste, agricultural by-products and crops are a few examples of biological materials that can be used to produce bioenergy (FAO, 2000).

Several crops are being tested for bioethanol production. Ethanol is produced through the fermentation of sugars or starch from sugarcane, maize, wheat, sugar beet, cassava and others. Most of the ethanol produced in the world derives from plant juice containing sucrose from sugarcane in Brazil and starch from corn in the United States (EIA, 2008). Ethanol can also be produced from sugars derived from lignocellulosic material by hydrolysis of the cell wall using enzymes, physical and chemical treatments (Ragauskas et al., 2006). The efficiency of the lignocellulose conversion process has not yet proven to be economical, but second generation ethanol is a highly desired goal because it could significantly broaden the choice of feedstocks. It would then be possible to use noncrop plants or crop biomass fractions, that are not animal feeds or food for humans, for biofuel generation (Tollefson, 2008).

Potential energy crops include sugarcane, maize, sugar beet (Beta vulgaris), grains, elephant grass (Pennisetum purpureum), switch grass (Panicum virgatum), Miscanthus (Miscanthus giganteus) and others. For commercial markets to develop, many of these crops are being evaluated for production under short growing seasons, periodic drought, low temperatures and low input of nutrients. In an ideal situation, if the goal is to produce energy from the C-bonds of plant lignocellulose breakdown, the crop should be a high yielding, fast growing, with a cell wall that is easy to break down and requiring relatively small energy inputs for its growth and harvest. To achieve sustainability, energy crops should not require extensive use of prime agricultural lands, and they should have low cost of energy production from biomass. Basically, the crop energy output must be more than the fossil fuel energy equivalent used for its production. At present, the crop that has most successfully met the energy crop attributes above is sugarcane.

Sugarcane as an energy crop

The output to input ratio of sugarcane first generation ethanol production is around 8–10, compared to 1.6 in maize (Goldemberg, 2008). Currently, sugarcane stalks crushed and extracted for juice are subsequently burned in the sugarcane factories for the production of steam and electrical energy. Potentially, with the industry of cellulosic ethanol it is expected that the ethanol output might increase from the current 7500 to 13 000 L/ha, i.e. an increase of 40%–50%. Sugarcane second generation ethanol has not yet been used commercially, but many sugarcane breeding programmes are improving germplasm not only for sucrose yield but also for biomass yield in anticipation of upcoming technologies that may allow for efficient energy production from cellulosic residues. Sugarcane annual production per hectare (39 t/ha of dry stalks and trash; discussed below) compares favourably to other high-yield bioenergy crops such as Miscanthus (29.6 t/ha), switchgrass (10.4 t/ha) and maize (total grain plus stover, 17.6 t/ha) (Heaton et al., 2008).

Sugarcane is an important food and bioenergy source and a significant component of the economy in many countries in the tropics and subtropics. Nearly 100 countries produce sugarcane over an area of 22 million hectares—approximately 0.5% of the total world area used for agriculture (FAOSTAT, 2008). Sugarcane produces the world’s greatest crop tonnage (FAOSTAT, 2008), even though each of the major cereals—rice, wheat and maize—occupies a several-fold larger fraction of the world’s arable land.

In Brazil, the sugarcane agribusiness accounts for more than US$ 20 billion/year and is one of the main direct and indirect job generation sectors. The country produces 25% of the world’s cane sugar and is the largest producer (31 million tons/year) and exporter (19.5 million tons/year) (UNICA, 2009). It is estimated that 558 million tons of sugarcane will be produced in 2008/2009 (Conab, 2007). Half of the cane produced will be destined for bioethanol to feed an increasing number of flex-fuel cars. In 2008, 2 254 553 flex-fuel car units were sold while gasoline powered cars amounted to 639 199 (ANFAVEA, 2008). It has been calculated that to meet demand (internal and external) Brazil needs to significantly increase its ethanol production (double in 5–7 years). Bioethanol has been produced in integrated bioethanol and sugar production units from the depleted syrup after sugar manufacture (in this case it is postulated to be the ‘one and half’ generation process). In this way there is very little competition of food vs. fuel. In addition, sugar production has been in the uprise as a result of the increasing international prices An increase in ethanol production without decline of sugar production should be achieved by expanding the planted area or increasing yield on existing sugarcane lands.

Sugarcane yields

Sugarcane yield statistics are reported on an area basis as the mass of the sugarcane stalks delivered to the processing mill and the mass of the sugar produced from the harvested sugarcane material. As commercial yield statistics are averages of all genotypes, across all environments and production systems, they give no indication of yields that might be achieved if identifiable yield-limiting constraints were ameliorated. For example, in a given year there might have been a prolonged drought or a pest or disease outbreak that reduced that year’s yield below the long-term average yield. Therefore, it is informative to compare yields under a variety of production situations from commercial average (the lowest level of production), to commercial maximum, to experimental maximum (Rabbinage, 1993).

Commercial average yields are those reached industry wide under various agronomic yield-limiting constraints of weeds, pests, diseases and soil nutrient deficiencies for which loss prevention measures are generally available. Commercial yields are the average of reported yields of all varieties over all environments and from a full range of farming practices from the poorest to the best. Under poor farming conditions, a relatively small input of herbicides, pesticides or soil amendments can raise the average commercial yields to commercial maximum yields attainable under the prevailing environment. Commercial maximum yields are obtained under good farming practices including the use of good plant nutrition and irrigation when necessary. Still, commercial maximum yields can be constrained by cultivars not selected as optimum for the prevailing environment. Attempting to reduce all environmental constraints in an experimental system to increase commercial maximum yields to the yield potential is more difficult and costly than raising the average commercial yield to commercial maximum yield. Potential yield is the yield that is achieved when a cultivar is grown in the environment to which it is adapted, with water and nutrients nonlimiting, and with pests, diseases, weeds and other stresses effectively controlled (Evans, 1994; Evans and Fisher, 1999). One can consider maximum experimental yields as approaching or equivalent to potential yields.

Sugarcane experimental stations frequently communicate to the growers they serve yield data, including commercial average, commercial record yields and experimental maximum yields, as a way to compare local production to other sugarcane growing areas of the world and to monitor their own historical progress in crop improvement. Irvine, (1983) compared the commercial average and maximum sugarcane yields and then calculated the equivalent total dry matter production for six sugarcane producing countries. The yields of the three highest sunlight countries (analysed by Irvine (1983) -Australia, Colombia and South Africa- reveal extremely high cane fresh weight yields that averaged 84 t/(ha yr) (Table 1). The commercial maximum cane yield for these three countries averaged 148 t/(ha yr) and the experimental maximum averaged 212 t/(ha yr). Breeding programmes have been successfully increasing yield at around 1% a year. Recently, in Brazil, a commercial maximum yield of 260 t/ha in 13 months (Fazenda Agrovale, Bahia) and an experimental maximum of 299 t/ha were recorded (Fazenda Busato, Bom Jesus da Lapa, Bahia, RIDESA, personal communication) which exceed the average maxima reported by Irvine (1983). These exceptionally high-yield figures were obtained under irrigation in an area having low precipitation and low cloudiness hence higher solar radiation than in most sugarcane producing areas of Brazil. While the commercial averages and the commercial maxima are from large land areas and are reasonably reliable, the experimental maxima are from individual trials on smaller land areas and thus may be over estimates of yield over several hectares.

Table 1.   Average, maximum and theoretical sugarcane yields (Australia, Colombia, and South Africa) and total dry matter production
Type of yieldCane yieldBiomass*
t/(ha yr)t/(ha yr)g/(m2 d)
  1. *Cane yield was converted to biomass dry matter first by calculating stalk dry wt (t cane ha−1 yr−1 × 0.30) then adding the proportion of trash dry wt [0.65 (stalk dry wt)] as calculated from Thompson (1978). Except for the theoretical maximum, table is partial summary of Table 3 of Irvine (1983). Supplemental Table 1 lists yield in the countries considered.

Commercial Average843910.7
Commercial maximum1486918.8
Experimental maximum2129827.0
Theoretical maximum38117748.5

The mature sugarcane plant consists of three main parts: stalk, leaf and root system. The stalk is composed by joint segments made up of a node (where the leaf attaches the stalk and the bud and root primordial are found) and an internode. The leaf is divided into sheath and blade, separated by a blade joint. They are usually attached alternately to the nodes, thus forming two ranks on opposite sides depending on variety and growing conditions. There is an average total upper leaf surface of about 0.5 m2, and the number of green leaves per stalk is around ten. The root system has two kinds of roots: set roots which are thin, highly branched and shoot roots that replace the previous as they develop. Shoot roots are thick, fleshy and less branched and with a limited life. There is a tillering periodicity where the new tiller (shoots) develops its own roots to adapt for the changing environment (Miller and Gilbert, 2009).

The sugarcane crop can be harvested by hand labour, in which case the material taken to the mill consists of only the mature culm with the other plant materials left in the field, or the crop can be harvested mechanically in which case the material hauled to the mill includes in addition to the mature culm, some fraction of attached green leaves, immature culm and part of the blanket of dead leaves (referred to as trash). Commercial yields are reported on a fresh weight basis with an attempt to correct for the amount of trash taken to the mill so that yield data are comparable. However, to calculate the amount of biomass dry matter produced, the water content of the crop and the proportion of milled cane to trash have to be determined. The amounts of water and proportion of cane to trash are a function of cultivar, environment and season of the year the crop is harvested (Donaldson et al., 2008), so that calculations of biomass must be based on empirically determined data that are fairly consistent for 1 year sugarcane crop worldwide. Irvine (1983) used these averages to calculate t/(ha yr) dry biomass of 39 for three country commercial averages, of 69 for commercial maximum, and 98 for experimental maximum (Table 1).

Sugarcane theoretical yield potential

Experimental maximum yield approximates the crop potential yield limit, which in the case of sugarcane is approximately 212 t/(ha year) fresh weight or 98 t/(ha year) dry biomass. However, this level of yield remains lower than the theoretical yield maxima that have been calculated from models of physiological processes contributing to plant growth (Monteith, 1977; Loomis and Amthor, 1999; Long et al., 2006; Zhu et al., 2008). These models are based on the principles of yield potential (Yp) and primary production (Pn) at a given incident solar radiation as developed by Monteith (1977) and presented by Long et al., 2006, where:

  • 1Primary production of biomass (Pn = St·εi·εc/k) is the product of the annual integral of incident solar radiation (St), two efficiencies that describe broad physiological and architectural properties of the crop, i.e. the efficiency of light capture (εi) and the efficiency of conversion of the captured light (εc), and a constant (k) representing the energy content of the particular plant mass produced (MJ/kg).
  • 2Yield potential (Yp = η·Pn) is the product of primary production (Pn) and the harvest index (η) or the efficiency for partitioning of biomass into the harvested product.

Mean world distribution of daily irradiance recorded from 1990 through 2004 (published on line <http://www.soda-is.com/eng/map/#monde>) shows the annual mean daily irradiance of approximately 230 W/m2 = 19.872 MJ/m2 = 198 720 MJ/(ha d) in the sugarcane production areas of Australia, Colombia and South Africa, the three high sunlight, high mass production per unit area countries analysed by Irvine (1983). This level of irradiation can be used with the concepts of Monteith (1977) to calculate a theoretical yield maximum for sugarcane at similar high solar radiation locations such as the main sugarcane growing regions of Brazil. Very little of total solar irradiance is available to the plant for biomass production (Zhu et al., 2008). More than half of the energy is outside of the photosynthetically active region, and additional losses are associated with reflection and transmission of the incident light. Overall, the low efficiencies of light capture (εi) and poor conversion of the captured light (εc), results in only about 0.06 of the total irradiance being stored in the energy of the chemical bonds of C4 plants. Thus, the theoretical irradiant energy stored in biomass of sugarcane is reduced from the 198 720 MJ/(ha d) striking the earth to only six per cent of that or 11 923 MJ/(ha d) providing that there is maximum leaf canopy interception of photosynthetic active radiation (PAR), which is not the case for the entire crop year as discussed below.

The energy content of plant mass depends on its composition with higher quantities of energy stored in fats and proteins than in simple carbohydrates. Sugarcane is mainly composed of carbohydrates (sugar and lignocellulose) that have an energy content of (∼15.9 MJ/kg). Thus the 11 923 MJ/(ha d) of irradiant energy potentially stored by sugarcane in biomass having an energy content of 15.9 MJ/kg calculates to a maximum theoretical primary productivity yield of 749.87 kg/(ha d) = 0.750 t/(ha d) = 273.70 t/(ha yr) in high sunlight areas. However, a 1-year sugarcane crop requires approximately 140 days from ratooning to develop a canopy for maximum interception of PAR (Singels et al., 2005), and this period of reduced PAR capture will decrease the year’s total energy for producing plant mass by the quantity not captured over 70 of the 365 days of the year. Thus, the primary maximum yield would be reduced by 19.2 per cent or 221.2 t/(ha yr). One must keep in mind that yield potential is less than total primary productivity by the amount of productivity that does not end up in the harvested product. Sugarcane has a high harvest index because a majority of the plant organs are harvested. However, there is a fraction (∼0.2) of plant material that remains in stubble and roots and trash consisting of dead stalks and leaves. Sugarcane’s harvest index of 0.8 could reduce the crop primary productivity of 221.2 t/(ha yr) to a potential yield of above ground biomass of 177 t/(ha yr)or a fresh weight cane yield of 381 t/(ha yr) (Table 1). Areas receiving higher solar energy would have a higher potential yield under optimum growing conditions with selected cultivars as may be the case in the high yields reported from Brazil.

The relationships among the production situations can be used to identify where R&D resources might give the greatest return in increasing crop yields. Under poor production situations, yields will likely be increased by minimizing the effects of reducing factors such as pests and diseases, and then satisfying the limiting factors such as water and nutrients. Under advanced production systems, those external reducing and limiting factors may already be addressed so that greatest return might be from research aimed at altering the genetics of the crop plant to raise the potential yield. It is this potential that the remainder of this paper is addressing.

Breeding for energy cane

The Saccharum genus is a group of crop species particularly challenging for improvement. Cultivars are interspecific aneuploid hybrids. The crossing of large genomes (with multiple recent duplications that allow chromosome pairing and recombination) makes each progeny genotype a unique genome. To improve yield and other traits of interest for the development of an energy cane, research must unravel the complexities of the sugarcane genome, develop statistical genetics for highly polyploid genomes and identify genes associated with sucrose content, drought resistance, biomass and cell wall recalcitrance.

The reference domesticated species of sugarcane is S. officinarum, a group of canes with thick and juicy culms (Daniels and Roach, 1987) which crossed to wild relatives producing the natural hybrids (S. sinense and Sbarberi). These naturally occurring hybrids were selected and cultivated, and sugar extraction probably began from such hybrids (Daniels, 1975). At the end of the 19th century, Sspontaneum, a wild species with little sugar and thin culms was used in the earliest sugarcane breeding programmes in search of disease resistant genes to introgress into Sofficinarum to produce cultivars. The interspecific hybridization solved many disease problems, increased cane yield and sucrose content (Roach, 1972). All modern cultivars are derived from a few intercrossing of these hybrids (Price, 1965; Arceneaux, 1967). World collections of germplasm exist in Florida and India (Naidu and Sreenivasan, 1987; Schnell and Griffin, 1991; Schnell et al., 1997) that keep ancestor genotypes and cultivars, and many private collections of breeding programmes are also kept and used for crosses.

In São Paulo state, the biggest ethanol producer in Brazil, ethanol yield increase can be achieved ideally through the use of higher sucrose yielding cultivars rather than expanding the sugarcane growing area because sugarcane culture already occupies 70% of the agriculture land. In 2008, an expansion into pasture land has occurred in São Paulo State and the Cerrado, the central savannah-like region in Brazil (Goldemberg, 2008). Cultivation can potentially further expand into Cerrado and pasture in the centre-west regions but the sugarcane crop would likely be subject to pronounced drought stress. Drought tolerant cultivars are highly sought by growers expanding into the northeast as well. Brazil is indeed seeing an expansion of cane cultivation into centre-west and northeast regions despite low precipitation and into the south region despite it being too cold for optimal sugarcane production. This fact, aligned with the predicted future shortage of water resources, indicates that a sustained growth of cane cultivation will depend on the development of high-yielding, drought- and cold-tolerant cultivars, and adapted to poor soil conditions. Sugarcane growers in the United States of America would also benefit from cold-resistant cultivars adapted to poor or sandy soil, and breeders are crossing sugarcane and Miscanthus genotypes in search of high-yielding, stress-tolerant hybrids (Lam et al., 2009). In addition, Sspontaneum has been used as a source of stress resistance genes because it is well adapted to harsh climatic conditions (Roach and Daniels, 1987; Ming et al., 2006). Improving cold tolerance could give growers an opportunity to extend harvest into the cold season and to expand production into marginal lands and temperate regions. Increased cold/drought tolerance has been reported for specific hybrids, especially those with germplasm of Sspontaneum, Ssinense and Miscanthus and having higher fibre content (Irvine, 1977). Increasingly, Sspontaneum germplasm is being introgressed into breeding lines of programmes in Louisianna, Brazil, Barbados and Australia (ISSCT, 2009) aiming for higher fibre content and the breeding of energy cane. Interspecific crosses followed by backcrossing to established cultivars will, depending on the selection process, introgress alleles into commercial types and lead to improved stress tolerance and a higher fibre content. Expansion into noncrop areas will likely encounter marginal soils such as sandy, saline/sodic or waterlogged soils or those with mineral stress problems notably having aluminium and manganese toxicity. Genes for resistance to these soil stress problems are available in sugarcane cultivars and Saccharum species (Nuss, 1987). One can expect that new cultivars will be available in a decade or so to expand cultivation to new climatic conditions.

Molecular resources for sugarcane improvement

The form or phenotype of the commercial cultivars has changed considerably since the first ancestor Saccharum genotypes. Originally it was a grass with thin stalks that accumulated little sugar but it evolved to have thicker culms with juicier and sweeter internodes. The form of a plant is frequently associated to changes in regulatory elements such as cis-regulatory elements (CREs) and transcription factors (TFs) (Doebley and Lukens, 1998; Costa et al., 2005), a fact well illustrated by the molecular events associated to maize (Doebley and Wang, 1997; White and Doebley, 1998) and rice (Li et al., 2006a) domestication. Most of the traits considered in the selection process of breeding programmes have a quantitative nature and are controlled by many loci (QTL’s), such as Brix rate (soluble solids measured during plant development), sucrose content, diameter and number of stalks, fibre content, resistance to pests and flowering, precociousness, diseases, etc. Some QTLs associated to stress tolerance code for TFs that control metabolic pathways (McMullen et al., 1998). TFs have been associated with tolerance to many stresses including drought and cold (Yamaguchi-Shinozaki and Shinozaki, 1994, 2005; Kasuga et al., 1999; Zhu, 2002) and have been studied in sugarcane by groups researching putative targets for the biotechnological improvement of this crop. Recently, QTL discovery has been aided by the identification of functional markers. A sucrose synthase-derived marker was associated with a putative QTL having a high negative effect on cane yield and also with a QTL having a positive effect on sucrose content (Pinto et al., 2009). This approach was made possible by the availability of Expression Sequence Tag (EST) collections.

Several EST collections have been developed (Carson and Botha, 2000, 2002; Casu et al., 2001, 2003; Vettore et al., 2003; Ma et al., 2004; Bower et al., 2005) that helped with gene discovery in sugarcane. The SUCEST EST sequencing project represents the largest EST collection and is represented by 43,141 putative transcripts known as the sugarcane assembled sequences (SAS) http://sucest-fun.org (Arruda, 2001; Grivet and Arruda, 2001; Vettore et al., 2003). cDNA microarrays have been used to determine transcript distribution among sugarcane tissues and to identify ubiquitous and tissue-enriched gene expression (Papini-Terzi et al., 2005). Sugarcane culms have been extensively analysed at the gene expression level. Gene expression related to sucrose metabolism has been shown to decrease during culm maturation and genes related to cellulose, lignin and cell wall metabolism to be modulated during internode development (Casu et al., 2001, 2003, 2004, 2007; Watt et al., 2005). An in silico analysis of tissue expression patterns for cell wall-related genes has been done for 459 SAS of the SUCEST database and reflected quite well the expected physiological characteristics of the tissues (Lima et al., 2001).

The SUCEST-FUN project (http://sucest-fun.org) identified many genes altered in elevated Brix (percentage of solids, in sugarcane corresponds mainly to sucrose) and in response to drought, including TFs and protein kinases (PK), in a large number of cultivars and genotypes (Rocha et al., 2007; Papini-Terzi et al., 2009). This indicates that sugarcane breeding programmes when selecting for high sucrose content have inadvertently selected for gene expression changes of certain regulatory genes. Gene expression studies may be helpful in the identification of eQTLs (expression quantitative trait loci).

Sugarcane is highly syntenic with other grasses even though there have been extensive rearrangements associated with polyploidy (D’Hont et al., 1994; Grivet et al., 1994; Dufour et al., 1997; Glaszmann et al., 1997; Gaut et al., 2000). Sorghum may present the simplest syntenic relation to sugarcane (Grivet et al., 1994; Dufour et al., 1997; Glaszmann et al., 1997; Guimaraes et al., 1997; Ming et al., 1998; Asnaghi et al., 2000) and is most likely among the grasses the one to significantly contribute to sugarcane QTL studies (Ming et al., 2002; Jordan et al., 2004). The recent release of the sorghum and maize genome sequence data (http://www.phytozome.net/sorghum; http://www.maizegdb.org/) will certainly help in sugarcane map-based isolation of genes and identification of regulatory networks, CREs and promoters. Recently, an international consortium has been formed (http://bioenfapesp.org) to sequence the sugarcane genome that may accelerate the development of markers for yield and abiotic stress tolerance. Sugarcane cultivars have a very complex genome, a veritable challenge for its sequencing. With around 760–926 Mbp (monoploid genome) (D’Hont and Glaszmann, 2001) the basic Saccharum genome is twice the size of rice (389 Mbp), similar to Sorghum (760 Mbp), and significantly smaller than maize (2500 Mbp). On average, a sugarcane locus is represented by ten alleles. The somatic cell genome (2C) of the modern cultivar R570 (2n = ∼115) was estimated to be around 10 000 Mb (D’Hont and Glaszmann, 2005). With next-generation sequencing technologies, researchers expect to have a draft of several sugarcane cultivars defined in a few years that will greatly enable studies on genome structure, comparative genomics and marker identification for agronomic traits of interest.

Physiology and regulation of sucrose accumulation

Under certain conditions, sugarcane partitions carbon into sucrose that accumulates in the internodes to up to 50% of its dry weight (0.7 m) (Moore, 1995a) but there is very little knowledge on the regulation of this process. Sucrose, the carbon compound fixed by photosynthesis and translocated from the leaves to various sink tissues may be stored as sucrose or partitioned between respiration, including both catabolic (glycolysis) and anabolic (gluconeogenesis) processes, and an insoluble cell wall component consisting of cellulose and lignin. Meristematic sink tissues, including the shoot and root apical meristems and the stalk intercalary meristems, metabolize the incoming sucrose into a hexose pool that is used for respiration, building the insoluble component, and depending on environmental conditions and age of the stalk internodes, can be synthesized back into sucrose for storage. During culm maturation there is a redirection of incoming carbon from the insoluble and respiratory components to sucrose storage. The cycle of degradation and synthesis of sucrose in the culm parenchyma tissue is regulated through the activity of fructose-6-phosphate 1-phosphotransferase, which appears to direct the rate of sucrose accumulation by limiting glycolytic carbon flux (Groenewald and Botha, 2007). Factors regulating the amount of sucrose stored are yet to be determined but include cell water relations and properties of the sucrose molecule such as its solubility and transport by membranes (Moore and Cosgrove, 1991). Molecules that might be determining cell water content when sucrose is accumulating are aquaporins less expressed in internodes of high Brix plants (Papini-Terzi et al., 2009) that in Arabidopsis have been shown to be involved in carbon partitioning (Ma et al., 2004).

It is important to keep in mind that the sucrose content of sugarcane is a trait that should be kept high while production of cellulosic ethanol becomes economical. In designing the ideal energy cane it will be important to keep sucrose levels high, increase biomass yield and alter the cell wall for enhanced saccharification. Accomplishing these three goals through metabolic engineering would require altering carbon fixation and partitioning in such a highly controlled manner that would be extremely complex, if it is at all possible. If bagasse hydrolysis for ethanol production becomes very efficient, it may be worth sacrificing sugar for increased fibre content. Either way one must understand carbon partitioning control mechanisms to devise strategies to alter it. Transcription factors and protein kinases are greatly important in defining signalling and gene networks by regulating key steps in signal transduction through phosphorylation cascades and promoter activation/inactivation. Knowledge of the regulatory networks controlling carbon metabolism will be critical to increase yield without deleterious effects on sucrose metabolism. Recently, the putative 1647 unique sugarcane TFs have been re-categorized (http://grassius.org) and catalogued into 47 categories (Gray et al., 2009; Yilmaz et al., 2009). Likewise, sugarcane PKs were re-annotated and catalogued into the SUCAST (Sugarcane Signal Transduction) Catalogue (http://sucest-fun.org). The SUCAST db contains 1031 PKs categorized based on BLAST, Pfam (Sonnhammer et al., 1998), SMART (Schultz et al., 1998) and a phylogenetic approach (Rocha et al., 2007).

When gene expression was compared among genotypes with high and low sucrose content (Papini-Terzi et al., 2009), several TFs were found associated with this trait. Among the differentially expressed TFs were two helix-loop-helix, two Homeobox Knotted1-homeodomain genes, one MYB and several TFs responsive to auxin, ethylene and gibberellin totalling over 20 TFs correlated to sucrose content. Auxin- and ethylene-responsive TFs such as ARFs and EILs were consistently associated with sucrose content together with the genes responsible for the biosynthesis of these hormones indicating a predominant role for them in sugar content regulation or responses.

An experiment comparing culm maturation in 30 genotypes grown in the field identified developmentally regulated genes related to hormone signalling, stress response, sugar transport, lignin biosynthesis, fibre content, PKs, PPases and TFs (Papini-Terzi et al., 2009). Protein phosphorylation appeared to have a dominant role in the process of internode development as noted by the large number of PKs, particularly from the SNF-related/SnRK kinase family of proteins which were differentially expressed between high Brix and low Brix plants or between mature and immature internodes and are responsive to ABA and drought (Papini-Terzi et al., 2009). In yeast and higher plants, these kinases are known regulators of carbohydrate metabolism (Woods et al., 1994; Barker et al., 1996; Douglas et al., 1997; Halford and Hardie, 1998; Sugden et al., 1999; Rocha et al., 2007). It is possible to make a direct parallel of a putative regulatory role for an SnRK1 and 14-3-3 proteins in sucrose accumulation because several members of this family of proteins were differentially regulated in sugarcane and together they have been shown to phosphorylate and inhibit a sucrose phosphate synthase (Toroser et al., 1998; Sugden et al., 1999). SnRK2 and SnRK3 were also identified as regulated during culm development including two osmotic stress-activated kinases (OSA-PK) and three CBL-interacting protein kinases (CIPK). OSA-PKs and CIPKs mediate drought, saline and cold stress responses (Boudsocq and Lauriere, 2005) which indicates that drought responses and sucrose content may indeed be related. A CIPK14 from Arabidopsis thaliana has been shown to contain sucrose responsive elements in its promoters (Lee et al., 2005) and several sugarcane CIPKs were shown to be responsive to sucrose when sugarcane seedlings were exposed to it (Papini-Terzi et al., 2009). SnRKs are responsive to ABA and drought (Boudsocq and Lauriere, 2005) and ABA, sucrose and drought signalling appear to be correlated in sugarcane because a comparison of drought responses, sucrose responses, high sucrose and low sucrose plants led to the finding that 30% of the genes associated with sucrose content are also modulated by drought including the cane SnRKs and PP2Cs homologous to ABI1 e ABI2 (Papini-Terzi et al., 2009). ABIs have been shown to be induced by ABA and to mediate stomatal closure (Merlot et al., 2001; Tahtiharju and Palva, 2001; Nambara and Marion-Poll, 2005) which may impinge on photosynthesis efficiency and yield. This is relevant because the drought responses are also regulated by ABA and in sugarcane water stress may be a trigger for sucrose accumulation (Inman-Bamber et al., 2008).

Stress arising from water deficit, defined as any water content of a tissue or cell that is below the highest water content exhibited at the most hydrated state, depends on the level of the deficit and the rate at which it developed. When the water deficit develops slowly enough to allow changes in developmental processes, it has several effects, the most sensitive of which are a reduction in leaf expansion and the closing of leaf stomata. The photosynthetic rate of a leaf is typically much more tolerant to mild water stress than is cell expansion. Photosynthate translocation is the least sensitive of these responses to water deficit because it is not reduced until the stress becomes severe. The differential sensitivity of these developmental processes is that the net effect of the onset of drought is to hasten the accumulation of carbohydrates in the leaves and in storage sinks of the sugarcane plant (Hartt, 1936). In sugarcane the accumulation of sucrose in storage parenchyma is called ripening. The central idea in sugarcane ripening by drought is to cause a gradual reduction of the tissue moisture level to compel the plant through a series of drought reactions that begin with reducing cell expansion and the formation of new internodes without much inhibition of photosynthesis. The outcome of this reduced consumption of sucrose for metabolic energy and the formation of new cells is an increased sucrose content (Gosnell and Lonsdale, 1974). Sugarcane crop managers commonly use drought and other growth-inhibiting stresses to ripen the crop just prior to harvest.

Gene expression differences in cultivars that contrast for sucrose and biomass

It is possible that some of the genotypes analysed by Papini-Terzi et al., (2009) also differ in biomass content. A continued agronomic evaluation is necessary to assess how gene expression in the selected genotypes is related to other characteristics, such as cell wall composition, growth rates, internode size and width, number of internodes and drought tolerance. With this in mind, we started an evaluation of many parameters in addition to Brix among cultivars. Sucrose accumulation dynamics, for instance, varies among cultivars. Breeding programmes have selected for early and late accumulators that will reach high Brix at different periods of the season. We illustrate this showing preliminary data for four cultivars that were compared for sucrose accumulation through the season in both plant and ratoon crops, sucrose content along internodes in different developmental stages, internode width, biomass accumulation and plant height (Figure 1). Cultivars SP91-1049 (V2) and SP89-1115 (V4) accumulate sucrose early in the season and can be harvested from March onwards while SP83-2847 (V1) and SP94-3116 (V3) correspond to varieties with low Brix accumulation rates to be harvested late in the season (Figure 1a,b). When Brix measures were taken along the culm, we observed that these varieties also differ in the internode developmental pattern. V1 and V3 show a sucrose accumulation delay in the first internodes and do not reach the same high levels of accumulation in the mature internodes as observed for V2 and V4 (Figure 1c). In addition, V2 and V4 have thicker culms (Figure 1d) and produce more culm mass (Figure 1e) without an apparent change in leaf mass (Figure 1e), plant height, number of internodes or internode length (Figure 1f).

Figure 1.

 Comparison of the total soluble solids content (Brix) (a, b and c) and developmental performance (d, e and f) in four sugarcane varieties: V1 (SP83-2847) and V3 (SP94-3116) correspond to varieties with low Brix accumulation rate and late harvest genotypes while V2 (SP91-1049) and V4 (SP89-1115) correspond to high Brix accumulation rate and early harvest genotypes. Four different plants for each variety were used in the soluble solids content and developmental performance measurements. The measures of Brix were done with cane plant (a) and cane ratoon (b) and all developmental (d, e and f) and Brix (c) analyses were done with six-month-old plants. Brix measurements were done using a portable refractometer. Plant height was measured from the base of the culm until the end of the longest leaf. Dry mass was obtained by drying all the leaves or all the culms from each of the four plants separately in an oven until constant weight was achieved. The data for plant height, culm length, internode number and dry mass were analysed by one-way analysis of variance followed by tukey test at 0.05. Means with the same letter are not significantly different, and means with different letters are significantly different.

To verify whether genes, specially protein kinases, previously seen to be associated with sucrose content in genotypes (Papini-Terzi et al., 2009) had altered expression in commercial cultivars (V1, V2, V3 and V4), gene expression was determined by quantitative real-time PCR (qRT-PCR). High Brix cultivars exhibited increased expression of two SnRK2 (CIPK-8 and CIPK-16) and a SnRK1 (SnRK1-2), confirming a role for these PKs in sucrose or biomass accumulation (Figure 2). The data implicate also the regulatory subunits of the SnRK1 kinases (AKINβλ) (Bouly et al., 1999; Lumbreras et al., 2001) in sucrose content regulation or biomass accumulation.

Figure 2.

 Comparison of the mRNA expression of cell wall metabolism enzymes (a and b) and protein kinases (c, d, e and f) in four sugarcane varieties (V1, V2, V3 and V4; the same varieties of Figure 2). Gene expression data were obtained from total RNA extracted from internode 1 of twelve-month-old plants that was used as template for cDNA synthesis. qRT-PCR was done as described in Rocha et al. (2007), with a pool from four different biological replicates for each variety. Each point represents an average (± standard error) of there technical replicates. The gene expression ratio was analysed by a log-normal model, and the probability P was calculated in relation to a reference (V1 was the reference for V2 and V3 was the reference for V4). The  0.95 denotes significant differences on expression levels of V2 and V4 from V1 and V3 (reference), respectively. The P value was also higher than 0.95 if the references were inverted (V1 as the reference of V4 and V3 as the reference for V2). Low Brix = low Brix accumulation varieties; High Brix = high Brix accumulation varieties.

PK expression differences appear to be correlated to cell wall biosynthesis and expansion as well because a cellulose synthase (from the CesA family) and a UDP-glucose dehydrogenase were observed to be more expressed in the high Brix/high biomass cultivars (Figure 2). UDP-glucose dehydrogenase is responsible for the double oxidation (four electrons) of UDP-glucose, producing UDP-glucuronic acid which is a substrate to pectin and hemicelluloses synthesis. This enzyme has been characterized and extracted from young internodes of sugarcane stems (Turner and Botha, 2002). In the SUCEST database there are 52 known cellulose synthase genes belonging to CesA and CsI family. In the study conducted by Casu et al., (2007) they identified 119 transcripts differentially expressed during culm development. Some members of the CesA gene family were found to be coregulated, and two major patterns of expression were detected: high expression in the maturing stem or high expression in both maturing and mature stems showing a relationship with the primary cell wall synthesis. The members of the CsI gene family were consistently more abundant in the young stem. The cellulose synthase shown in our study as associated with sucrose content and biomass is not one of the family members described by the Casu and colleagues’ work.

Targeting the cell wall

It is evident that if sugarcane is to be improved for bioenergy production, a significant number of cultivars and genotypes need to be further evaluated at the biochemical and physiological level. It is possible, for instance, that high Brix genotypes differ in their saccharification potential (i.e. some may be more amenable for acid and enzymatic hydrolysis and cellulosic ethanol production than others). Several genes with a putative function in cell wall metabolism were identified as associated with sucrose content (Papini-Terzi et al., 2009) such as the expansins. Expansins may act in the relaxation of the cell wall, possibly by breaking the bonds between cellulose microfibrils and matrix polysaccharides (McQueen-Mason and Cosgrove, 1994; Cosgrove et al., 2002) allowing for cell expansion. These observations are corroborated by the identification of an XTH altered at the expression level during culm maturation. XTHs can hydrolyse xyloglucans, major components of plant cell walls and transglycosylate residues into growing xyloglucan chains that may be important during tissue expansion for sucrose accumulation (Farrokhi et al., 2006). It is important to note though that the structure of the sugarcane cell wall is currently unknown. A preliminary analysis of the composition of the fibres found a relatively high proportion of arabinoxylan with cellulose along with lower amounts of beta-glucan and pectins (Silva, 2005), but studies on the sugar linkages and overall architecture of the wall have not been reported yet. Silencing of lignin biosynthesis genes has been shown to benefit sugar release for lignocellulosic biomass fermentation (Chen and Dixon, 2007); thus, it will be interesting to test whether altered biomass has been selected for during the breeding process at the level of cell wall architecture, polymer architecture or cell wall expansion. At any rate, the alteration of cell wall biosynthesis genes in association with Brix content is an interesting indication of a correlation between these processes. Silencing or over-expression of some of these genes may lead to altered cell wall or increased sucrose content.

Future prospects

Although a lot is known on sugarcane’s biology and cultivation, we are only at the beginning of the detailed biochemical and genetic characterization needed for this crop (Moore, 1995b). There is an impending need to develop biotechnology that will allow for a sustained industry of sugarcane. Solid knowledge on photosynthesis, sucrose and biomass accumulation processes can only be achieved by combining multiple experimental approaches coupled to the computational analysis of large sets of data. Research is imminently needed in sugarcane to pave the way for a systems biology approach. Ongoing efforts are aimed to use EST and whole genome sequencing data, transcriptome analysis, functional data on genes by the analysis of transgenics and a thorough characterization of sugarcane varieties including growth, development, physiology in response to stress and agronomical/industrial description to build a biotechnology platform for this crop.

To our knowledge, there are no commercial transgenic sugarcane cultivars. If we take into account that sugarcane is capable of increasing sucrose up to 25% more than what is currently available (Grof and Campbell, 2001), we predict a great benefit if we can successfully target sucrose metabolism genes for increased accumulation. In parallel, we also need to keep increasing yields for the generation of an energy cane. Traditional breeding can be improved considerably if the breeders have biotechnological tools available, such as genes that could be used as markers in the selection of genotypes. QTLs have been associated with sucrose content in sugarcane (Silva and Bressiani, 2005; Pinto et al., 2009) but not mapped yet to TFs. With similar tools as those used in Arabidopsis (Davuluri et al., 2003; Molina and Grotewold, 2005; Palaniswamy et al., 2006) and in mouse and human (Jin et al., 2006; Li et al., 2006b), we will possibly be able to identify polymorphic promoters and CREs that will be of great importance in the study of sugarcane regulatory networks and in the generation of an energy cane.

Acknowledgements

This work was funded by FAPESP (Fundação de Amparo à Pesquisa do Estado de São Paulo). G.M.S. is recepient of a CNPq fellowship. We are indebted to Marcio Barbosa, Geraldo Veríssimo, Marcelo Menossi, Eugenio Ulian, William Burniquist, Sabrina Chabregas and Maria Cristina Falco and Cane Technology Center (CTC) for their valuable help in discussing experiments and data.

Ancillary