Structural components of the cell wall of maize. Study of its role in crop protection, animal digestibility and bioenergetic production

Resum

In terms of agricultural land use and production, maize (Zea mays L.) is one of the most important crops worldwide. In addition to the general uses of maize grain as food, feed, or raw material for generating industrial derivatives, maize stover could be a profitableby product for ethanol production, whereas the whole plant can also be used to produce silage for feeding cattle. Forage maize is the greatest source of energy for ruminants. In forage species, fibre comprises 300–800 mg/g dry matter contents, the same amount than grain. Unfortunately, less than 50 % of this fibre content is actually digested and used by the animal. This recalcitrance is mainly due the cell wall, its composition and to the limited digestion of the fiber by the microorganisms of rumen. To improve forage varieties, it is necessary to improve the enzymatic digestibility and forage yield. On the other hand, the two key parts of the maize plant that can be converted to bioethanol are the grain, which is mainly made up of starch, and the stover (stem and leaves) which predominantly contains lignin and cellulosic components. Bioethanol derived from maize stover presents the added advantage of not diverting basic goods to industrial uses, making possible a double exploitation of the crop as source of food or feed and energy. The main obstacle for the biomass fermentation process is again the cell wall recalcitrance, defined in this case as cell wall resistance to degradation by microbes and their enzymes, which subsequently increases the energy requirements, the cost and complexity of bio refinery operations and/or reduces the recovery of biomass carbon into desired products. Overall, the main problems that crops have to face are the different pests. Stem corn borers are diverse lepidopteran insects that feed on the stem pith of maize producing tunnels in the stems and causing important yield losses. The damage produced by second and subsequent generation larvae causes the most important impact; forage, stover and grain yields are reduced as consequence of tunnels produced in the stalk pith that increase lodging rate ad interfere with assimilate movement of nutrients toward the developing ear. Accessibility, extensibility, and digestibility of maize tissues would determine important characteristics of maize such as resistance to stem borer pests (R), suitability for ethanol production (E) and feedstuff quality and digestibility (D). From now on, those three aspects will be referred as R.E.D. These characteristics depend greatly on cell wall functionality and structure, which are controlled by the composition and organization of individual cell wall components. The secondary cell wall of plants consists in microfibrils of cellulose embedded in a matrix of hemicelluloses, pectins, hydroxycinnamic acids and lignin. In grasses, arabinoxylans (hemicellulose) play a key role in cell wall structure. Arabinoxylans chains mediate in cell wall structure by joining cellulose microfibrils and by crosslinking with lignin through hydroxycinnamic acids. The main hydroxycinnamic acids are p-coumaric acid and ferulic acid. In maize, lignins are acylated (primarily syringyl units) at the γ-position by p-coumarates. Cell-wall-bound p-coumaric acid is considered an indicator of lignification and secondary cell wall deposition; while ferulic acid participates increasing stiffness by binding to arabinoxylan chains or lignin. Ferulic acid has the ability to form dimers that contribute to the reinforcement of crosslinking within cell walls, increasing their mechanical resistance. In studies of corn borer resistance (R), several biochemical and physical characteristics (general plant traits, antibiotic compounds, toxins, and repellent or attractant metabolites) have been studied as possible defence mechanisms against borers. At the physical level, researchers have been mainly focused on the study of the cell wall as structural barrier. Cell wall fibres, represented by the main polymers of cell wall, can disrupt the development of the insects by reducing the nutritional value of the host tissue and/or by enhancing the role of the cell wall as physical barrier due to thicker or tougher cell walls. At the same time, the role that cell-wall-bound hydroxycinnamic acids and lignin play on strengthening and stiffening of the cell wall has been described as a defence mechanism against corn borer attack making the cell wall less accessible to the pest. The conversion of lignocellulosic biomass to ethanol (E) is a three step process: (i) a pre-treatment stage, followed by the (ii) hydrolytic degradation of carbohydrates to the constituent sugar monomers (saccharification), and the (iii) final fermentation of the free sugars to ethanol. The biological conversion of plant cell wall carbohydrates into fermentable sugars would depend on polysaccharide contents. Degradation of those polysaccharides is hindered by the embedding and/or cross-linkages of carbohydrates with lignin or p-hydroxycinnamic acids that impede the biomass deconstruction. Also, the degree of polymerization of the cellulose and its crystallinity index are presupposing to limit its enzymatic hydrolysis. Lignin arises as the main cell wall factor affecting digestibility (D) of the cell wall impeding the action of hydrolytic enzymes and limiting animal energy intake. However, all variations in cell wall digestibility should not be attributed solely to the degree of lignification of the cell walls. Relationships between lignin content and digestibility are greatly influenced by the hydrophobicity of lignin, its polymerization or by cross-linkages between cell wall components, in addition to lignin monomeric composition. The structural reinforcement of cell walls by hydroxycinnamates also affects digestibility. Both the crosslinking of feruloylated arabinoxylans to G units of lignins via ether bonds mediated by ferulic acid, as well as the crosslinking of arabinoxylans, impact the organization of the lignin and cell wall networks, decreasing cell wall deconstruction by microorganisms and enzymes. Conversely, the role of p- coumaric acid in cell wall degradability is not that clear. Some author postulate that, because most p-coumaric acid is esterified to the γ -position of phenylpropanoid sidechains of syringyl units in lignin, they do not act as bridges to polysaccharide and should not directly affect cell wall degradability. Other authors propose that p-coumaric acid can induce recalcitrance because it is associated to lignification, cell wall maturity and higher proportion of secondary cell wall. As previously discussed, the structural and functional properties of the plant cell wall are controlled by the composition and organization of each of its individual components. This composition is defined by four main fractions: cellulosic, hemicellulosic, and hrydroxicinnamates fraction and lignin. Those fractions have been either positively or negatively correlated with economically important uses and aspects of the crop such as resistance to insects (R), ethanolic production (E) and feedstock digestibility (D). In this context, the general objective of this proposal includes the integrated study of the different components of the cell wall, and its influence in the characteristics that make a variety agronomically profitable for each of the purposes that have been discussed. The specific objective 1 has two particular aims, (i) the study of the interrelation between different cell wall components in diverse maize genotypes and determine the degree to which each component intervene in pest resistance, ethanolic production and forage digestibility; and (ii) determine which components of the cell wall can be used as indirect selection traits for the improvement of the economical important aspects of maize. To fulfil this objective, we carried out an exhaustive agronomical and biochemical characterization of 20 inbred lines that showed variability for corn borer resistance, ethanolic production and digestibility. The 20 genotypes were evaluated trough two years following a randomized block design with three repetitions under two infestation conditions: artificial infestation with S. nonagrioides eggs and protected with insecticide. The composition of the cell wall (cellulose, hemicellulose, lignin and hydroxycinnamates) was determined in all inbred lines and inbred suitability for different uses (forage digestibility, ethanol production) was estimated not only in internodes, but also in whole plant, in order to investigate correlations between internode and whole plant parameters. As results, we found genetic variation among the inbred lines for saccharification efficiency and all the cell wall fractions except for hemicellulose. In view of this, the germplasm represented by this subset of inbred lines could to implemented in breeding programs. However, we did not find differences among the inbred lines for corn borer resistance measured as length of the galleries produced by the larvae. For the case of saccharification efficiency, cell walls presenting higher concentration of cellulose mean greater content of potential sugars to be converted in ethanol hence enhancing saccharification efficiency. As well, cell walls presenting a high proportion of cross-linkings mediated by diferulates respond better to alkaline pre-treatment enhancing saccharification efficiency. This results are contrary to the idea that a highly crosslinked cell wall, rich in diferulates, limits cell wall deconstruction. Finally, increased digestibility of organic matter is associated to decreased concentration of fibre, ferulic acid and glucuronic acid that limit the energy intake enhancing cell wall recalcitrance; and to higher proportion of G subunits in the lignin polymer, that correspond implies a decrease of resistant bonds. Contrast analysis showed that, cell walls richer in hemicellulose and p-coumaric acid contents are more susceptible to corn borer attack, presenting greater tunnel length. Our results in the second internode evidenced the impossibility to find a maize ideotype presenting a particular cell wall composition to fit the three R.E.D. aspects, as one combination of traits would favour one aspect and negatively affect the other. However, we could still aim to obtain a putative genotype from this subset of lines with the best combination of cell wall compositional traits to be used in a particular usage; always taking into account the plasticity of the cell wall and the co-lateral effects that the modification of one trait may have into the total cell wall composition or to the total fitness of the plant. In adittion, determinations of digestibility and saccharification of the cell walls in the internode and the whole plant did not show any significant correlation, indicating that results obtained for the internode cannot be extrapolated to the whole plant. Stover determinations are the more commonly used in breeding programs so, from a breeding standpoint, if the aim is to improve stover it seems more appropriate to evaluate stover. The specific objective 2 aims to deepen into genomic regions and genes associated to (i) characteristics relevant for the different crop uses and, (ii) maize cell wall composition. A subset of 408 RILs of the MAGIC population together with the eight founders included as checks were tested in a single augmented design with 10 blocks through two years. Samples of maize stover without the ear were characterized for traits related to final use of maize stover which would include saccharification efficiency (SACC), digestibility of the organic matter (DOM) and stover yield; and for cell wall components (cell wall fibres and cell wall bound hydroxycinnamates). A genome-wide association analysis was completed in order to identify single nucleotide polymorphisms (SNPs) associated with cell wall components and final-use-related traits. We found variation for cell wall structural components and R.E.D related traits among the checks and the RILs with the exception of saccharification efficiency for checks that did not significantly differ. A low percentage of the variation observed for SACC, DOM and, specially, for tunnel length could be explained by variation for cell wall bound hydroxycinnamates and cell wall fibres. Moderate to high heritability values were obtained for the traits under study with the exception of SACC that stood out for presenting a low value. This result may be derived by a large variance error caused by the experimental design and/or by the sampling procedure. In the association analysis, we identified four SNPs associated with stover yield that corresponded to four different QTL, 16 SNPs associated with saccharification that could be clustered in five QTL and two SNPs corresponding with two QTL associated with digestibility of the organic matter. A total of 24 SNPs, corresponding with 15 QTL, were found significantly associated with cell wall bound hydroxycinnamates; three SNPs distributed in two QTL we associated with fibres, one associated with NDF and two with ADF. Within the intervals of the QTL previously mentioned, and in view of the annotated functions of genes within that interval, we considered as candidate genes for improving stover yield, genes that are involved in detoxification of reactive oxygen species, nitrogen assimilation, plant and organ growth and development, and yield. In the case of saccharification efficiency, we propose genes involved in the synthesis of lignin subunits and in charge of the control of the genes involved in monolignol biosynthesis and polymerization. Besides we found, genes involved in cell wall degradation and biosynthesis and related with reduced cell wall recalcitrance trough the deactivation of the gibberellic acid. Throughout the genes that we propose as candidates for cell-wall bound hydroxycinnamates, we found genes coding for peroxidases probably involved in oxidative coupling of ferulic acid to form dimers, transcription factors of the phenylpropanoid pathway, genes implicated in xyloglucan and arabinoxylan synthesis, involved in polysaccharide synthesis and modification. By last we found a gene involved in gibberellin and another one involved in esterified suberin biosynthesis. Among the candidate genes for fibre QTL, we found genes that encode transcription factors of the phenylpropanoid pathway. Besides, we found genes induced in oxidative stress scenarios and a gene involved in lignin modifications. The associations between these putative candidate genes and the R.E.D need to be validated. In this second objective we highlight genomic regions directly linked to traits that influence final use of maize. Assisted-selection using markers located in those regions would help to develop materials for higher biofuel yield per hectare and more amenable forages for cattle. For improving stover yield and DOM, the phenotypic selection should be adequate; however, due to the lower percentage of variance individually explained by the markers significantly associated with saccharification traits we suggest to perform a genomic selection approach markers distributed through the whole genome. At the same time, some improvements in the phenotyping method and experimental design for saccharification efficiency need to be further considered in the future in order to improve the estimation of breeding values. Overall, for the MAGIC population tested, we suggest a breeding strategy for direct and simultaneous improvement of pest resistance, digestibility and ethanolic production (R.E.D). This strategy is supported by the lack of co-localizations between the markers associated to R.E.D traits and cell wall traits, and the lack of correlations among R.E.D. traits evaluated. In a global vision of the results obtained from both experimental blocks, we observed that the association between the cell wall composition and the final use of maize is strong when we studied the internode of the 20 inbreds set. However, that correlation is weak/lost in the MAGIC population using the whole plant sample; cell wall composition explains a low proportion of R.E.D variance and, the QTL for cell wall components and R.E.D traits did not co-localize.