Phytochemistry 72 (2011) 365–371
Contents lists available at ScienceDirect
Phytochemistry
journal homepage: www.elsevier.com/locate/phytochem
Cell wall composition as a maize defense mechanism against corn borers ⇑ Jaime Barros-Rios a, , Rosa A. Malvar a, Hans-Joachim G. Jung b, Rogelio Santiago a a Misión Biológica de Galicia (CSIC), Apartado 28, E-36080 Pontevedra, Spain b USDA-Agricultural Research Service, Plant Science Research Unit and Univ. of Minnesota, Dept. of Agronomy and Plant Genetics, 411 Borlaug Hall, 1991 Upper Buford Circle, St. Paul, MN 55108, USA article info abstract
Article history: European and Mediterranean corn borers are two of the most economically important insect pests of Received 30 April 2010 maize (Zea mays L.) in North America and southern Europe, respectively. Cell wall structure and compo- Received in revised form 13 December 2010 sition were evaluated in pith and rind tissues of resistant and susceptible inbred lines as possible corn Available online 1 February 2011 borer resistance traits. Composition of cell wall polysaccharides, lignin concentration and composition, and cell wall bound forms of hydroxycinnamic acids were measured. As expected, most of the cell wall Keywords: components were found at higher concentrations in the rind than in the pith tissues, with the exception Zea mays of galactose and total diferulate esters. Pith of resistant inbred lines had significantly higher concentra- Ostrinia nubilalis tions of total cell wall material than susceptible inbred lines, indicating that the thickness of cell walls Sesamia nonagrioides Resistance could be the initial barrier against corn borer larvae attack. Higher concentrations of cell wall xylose Hydroxycinnamates and 8-O-4-coupled diferulate were found in resistant inbreds. Stem tunneling by corn borers was nega- Polysaccharides tively correlated with concentrations of total diferulates, 8-5-diferulate and p-coumarate esters. Higher Lignin total cell wall, xylose, and 8-coupled diferulates concentrations appear to be possible mechanisms of corn Pith borer resistance. Rind Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction binoxylan, the majority of p-CA is esterified to syringyl units of lig- nin (Ralph et al., 1994). Formation of diferulates (DFA) and higher Maize (Zea mays L.) resistance to European corn borer (ECB), oligomers of FA has been shown to cross link arabinoxylan chains Ostrinia nubilalis Hübner (Lepidoptera: Crambidae) and Mediterra- (Bunzel, 2010). The deposition of DFAs in various tissues (kernel, nean corn borer (MCB), Sesamia nonagrioides Lefèbvre (Lepidop- leaf, pith, rind and nodes) has been shown to be associated with tera: Noctuidae) has been extensively evaluated because these resistance to pests such as ECB (Bergvinson et al., 1997), South- are two of the most economically important insect pests of maize western corn borer (Diatraea grandiosella Dyar) and sugarcane production in North America and southern Europe, respectively borer (Diatraea saccharalis Fabricius) (Ramputh, 2002), maize wee- (Gianessi et al., 2003; Velasco et al., 2007). Several physical and vil (Sitophilus zeamais Motschulsky) (García-Lara et al., 2004), MCB biochemical characteristics (general plant traits, antibiotic com- (Santiago et al., 2006a,b, 2008) and diseases such as Gibberella stalk pounds, repellent or attractant metabolites, etc.) have been studied and ear rot (Fusarium graminearum Schwabe) (Bily et al., 2003; as constitutive resistance mechanisms to corn borers (Malvar et al., Santiago et al., 2007). Several quantitative trait loci (QTL) for con- 2008). There has also been significant research directed toward cell centrations of cell wall esterified p-CA, esterified and etherified FA, wall concentration, composition, and structure as possible resis- and esterified 5-5 DFA and 8-O-4 DFA that were identified by Bar- tance traits to corn borers (Buendgen et al., 1990; Santiago et al., rière et al. (2008) also co-localized with QTLs identified in other 2006a). Resistance to ECB has been related to cell wall polysaccha- studies for ECB damage (Cardinal et al., 2001). ride and lignin content of maize stalks (Ostrander and Coors, 1997; Previous work of our group identified sources of maize resis- Martin et al., 2004). tance to corn borer (Butrón et al., 1999; Ordás et al., 2002). We Cell wall bound forms of hydroxycinnamic acids in cereals con- have shown that resistant inbred lines contained significantly sist largely of p-coumaric (p-CA) and ferulic (FA) acids (Hartley and higher concentrations of DFA than susceptible lines (Santiago Jones, 1978). All FA is ester linked to arabinoxylan and some of et al., 2006a,b), however, it is unknown if other changes in cell wall these FA molecules form additional covalent linkages to lignin concentration or composition are associated with corn borer resis- (Ralph et al., 1992). While some p-CA is similarly esterified to ara- tance of these maize lines. Our current objectives were (i) to deter- mine the concentration of the cell wall polysaccharide ⇑ Corresponding author. Tel.: +34 986 854 800; fax: +34 986 841 362. components, lignin concentration and composition, and hydroxy- E-mail address: [email protected] (J. Barros-Rios). cinnamates (p-CA, FA, and DFA) in pith and rind tissues of resistant
0031-9422/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2011.01.004 366 J. Barros-Rios et al. / Phytochemistry 72 (2011) 365–371 and susceptible inbred lines of maize, and (ii) to examine the role mental Table S2). Overall, DFAs only accounted for 0.1% and of these cell wall components as maize constitutive defense mech- 0.0002% of mature maize pith and rind tissue cell walls, anisms against corn borers. respectively. A higher lignin concentration was observed in rind than pith tis- sue (70% higher) (Supplemental Table S1). This fact may partially 2. Results and discussion account for lower concentrations of esterified DFAs in the rind than pith (92% less) because some DFA molecules become cross liked to 2.1. Pith vs. rind cell-wall composition lignin through ether and other covalent linkages, such as occurs for FA (Ralph et al., 1992). Any DFAs that were ether linked to lignin Pith and rind tissues were both analyzed in order to gather could not be determined by the alkaline hydrolysis method em- comprehensive data regarding maize stalk cell walls, however, ployed in the current study. the following data presentation is focused primarily on pith tissue because that is the tissue where corn borer larvae tunnel and feed. 2.2. Environmental effects It has been assumed that the cell wall polysaccharides are indigest- ible to Lepidopterans larvae, which utilize mainly soluble carbohy- Concentrations on a DM basis of most pith cell wall components drates and proteins as nutrients (Terra et al., 1987). As expected, were significantly different between locations, except for p-CA, FA most of the cell wall components were found at higher concentra- ethers, and total DFA (data not shown). These differences in cell tion when calculated on a dry matter (DM) basis in rind than pith wall component abundance were probably due to the different tissues with the exception of galactose, uronic acids, and total DFAs growth environments at the two locations. Usually Zaragoza is (Supplemental Table S1). Compositional analysis showed that cell warmer and drier than Pontevedra during the growing season. 1 wall components accounted for 320 g kg DM in pith tissue and Although temperature profiles in 2008 were similar for the loca- 1 580 g kg DM in rind tissue (Supplemental Table S1), indicating tions, Pontevedra had greater precipitation overall (data not higher concentrations of cytoplasmic components (proteins, lipids, shown). Pith tissue at Zaragoza had higher total cell wall and indi- ash, organic acids, etc.) in pith tissue. This greater concentration of vidual cell wall component concentrations, except Klason lignin total cell wall material in rind may explain why corn borer larvae, and 5-5-DFA which were not different between locations (data particularly MCB larvae, enter the stem through the base of the not shown). As total cell wall concentration differed between loca- internode where the intercalary meristem is located and the cells tions, we examined the data on a cell wall basis to determine if are the least developed (Santiago et al., 2003; Barros et al., 2010). composition of the walls varied independently of total wall accu- While total cell wall concentration was numerically higher in resis- mulation. Zaragoza still had higher concentrations of uronic acids, tant lines for both tissues, the statistical contrast of resistant vs. arabinose, galactose, and glucose on a cell wall basis, however, Kla- susceptible lines was only significant for pith tissues. All inbred son lignin was greatest at Pontevedra (data not shown). Lower lines had higher rind cell wall concentrations than the cell wall stem internodes should have completed cell wall development concentration in pith tissue of resistant lines. by 30-d post-flowering (Jung, 2003), therefore, differences be- In pith tissues glucose was the predominant constituent of the tween locations suggest that growing conditions affected cell wall cell wall polysaccharides (58%), followed by xylose (27%), uronic development. acids (6%), arabinose (5%), mannose (2%), and galactose (2%) (Sup- Based on biomass productivity, which did not differ based on plemental Table S2). Arabinoxylan is the major hemicellulose com- measured of plant height, neither location was more stressful than ponent in most cereal cell walls, although there are large the other (data not shown). Corn borer tem tunneling was different differences in the degree of arabinose substitution among tissues between the two locations, with MCB showing larger tunnels at (Hazen et al., 2003). Therefore, the high values for xylose and arab- Pontevedra than ECB tunneling at Zaragoza (Table 2). However, inose probably represent a high content of arabinose-substituted this difference in tunneling could reflect differences between the xylan (arabinoxylan), although some arabinose may also stem from damage potential of these two corn borer species rather than envi- pectic arabinan side-chains. Lignin content represented 17.8% of ronmental impacts on cell wall development. the total cell wall mass (Supplemental Table S2). Syringyl units Although differences for locations were confirmed for most cell (S) were more common than guaiacyl units (G) in lignin. Pith and wall components, a genotype location interaction was only sig- rind tissues had approximately the same S/G ratio (1.48 and 1.51, nificant for uronic acid concentration. The interaction was due to respectively), which was similar to previous data for mature maize the EP42 inbred line which had higher uronic acid concentration stalks (Lapierre, 1993). at Pontevedra than at Zaragoza (data not shown). Also, the A509 Hydroxycinnamates such as FA and p-CA are minor components inbred line had the second highest uronic acid concentration at in plant cell walls (Bunzel, 2010). Our analyses found that ester Zaragoza but the lowest concentration at Pontevedra. bound p-CA was the most abundant hydroxycinnamic acid de- tected in maize pith and rind tissues (2.1% and 2.7% of total cell 2.3. Genotypic effects wall, respectively), with FA (ester plus ether bound) also present in significant quantities (1.4% and 1.5% of total cell wall). Together Analysis by location for stem tunneling data for corn borer dam- these monomeric phenolics accounted less than the 5% of the total age showed that resistant and susceptible lines differed for length cell wall in both tissues (Supplemental Table S2). Three different DFAs were identified and quantified: 8-5-DFA, 8-O-4-DFA, and 5- Table 1 5-DFA. The 8-5-DFA was calculated as the sum of 8-5-open or Pedigree and resistance classification of four maize inbred lines used in the current study. non cyclic and 8-5-benzofuran or cyclic forms, because it has been reported that 8-5-non-cyclic form may be a product after alkaline Genotypes Pedigree Groupa hydrolysis of the cyclic form, the only naturally occurring in plant A509 A78 A109 R cell walls (Ralph et al., 1994), The dimers in order of abundance for EP39 Fino R pith tissue were 8-5-DFA (52% of total DFAs measured), 8-O-4-DFA EP42 Tomiño S 2 (35%), and 5-5-DFA (13%). Rind tissue differed with the most abun- EP47 (EP4 A239) EP4 S dant dimer being 5-5-DFA (41%), followed closely by 8-5-DFA a R, resistant; S, susceptible. Resistance classification based on tunneling by ECB (39%), and 8-O-4-DFA (20%) present in lower proportion (Supple- and MCB across multiple evaluations (Butrón et al., 1999; Ordás et al., 2002). J. Barros-Rios et al. / Phytochemistry 72 (2011) 365–371 367
Table 2 Means by location for stem tunneling, and combined means over locations for concentration of total cell wall, cell wall polysaccharide components, and lignin concentration and composition in the pith tissue of four inbred lines grown at Pontevedra and Zaragoza, Spain in 2008.
Genotypes Stem tunneld (cm) Total (g kg 1 dry Polysaccharide componentse (g kg 1 dry matter basis) Ligninf (g kg 1 dry matter basis) matter basis) Pontevedra Zaragoza Cell wall Glc Xyl Ara Man Gal UA KL S/G ratio A509 (R) 17.53 1.40ab 333b 136b 81ab 13b 4.0b 4.3b 16b 65a 1.46b EP39 (R) 7.67 0.00b 440a 207a 100a 18a 5.0a 6.6a 21a 69a 1.38c EP42 (S) 24.50 8.27ab 335b 160b 66bc 12b 4.6ab 4.3b 17b 60ab 1.99a EP47 (S) 13.30 13.67a 299b 142b 59c 13b 4.1b 4.7b 16b 50b 1.39bc LSD (P 6 0.05) 17.49 12.75 65 29 19 4 0.9 1.6 3 10 0.36 Mean resistant group 12.60 0.70 386 172 90 16 4.5 5.4 19 67 1.37 Mean susceptible group 18.90 10.97 317 151 63 12 4.3 4.5 17 55 1.64 Contrast R vs. Sg NS ** NS *** * NS NS NS ** NS
NS, non significant (P > 0.05). abc Inbred line means in the same column not sharing a superscript differ (P < 0.05). d Means for damage data, measured as stem tunnel length, were not combined over locations because each location was infested with a different species of corn borer. Pontevedra location was infested with Mediterranean corn borer and Zaragoza with European corn borer. e Glc, glucose; Xyl, xylose; Ara, arabinose; Man, mannose; Gal, galactose; UA, uronic acids. f KL, Klason lignin; S/G, syringyl-to-guaiacyl ratio as determined by pyrolysis–GC–MS. g Orthogonal contrast of resistant (R) and susceptible (S) lines. * Significant at P < 0.05. ** Significant at P < 0.01. *** Significant at P < 0.001.
of ECB tunnels at Zaragoza; however, damage caused by MCB at the In an attempt to attribute differential impact of individual cell Pontevedra location during the year of evaluation did not differ be- wall components on resistance, independent of total cell wall con- tween resistant and susceptible lines (Table 2). Lower tempera- centration, we also evaluated the composition data on a cell wall tures during the year of evaluation adversely affected basis. The resistant lines had significantly more xylose and 8-O- development of MCB larvae (Velasco et al., 2007), causing shorter 4-DFA on a cell wall basis and significantly less glucose, mannose, tunneling than expected based on previous studies (Butrón et al., and uronic acids, and esterified FA (data not shown). Klason lignin 2009; Sandoya et al., 2008). These results do not alter the resis- concentration in cell wall material was similar among resistant and tance classification of the lines which is based on their resistance susceptible groups of inbreds (data not shown). Xylose concentra- or susceptibility response to corn borers across multiple evalua- tion was 19% greater in the resistant than susceptible lines tions (Butrón et al., 1999; Ordás et al., 2002). whereas 8-O-4-DFA was 46% greater for the resistant lines. These Significant differences among genotypes were observed for con- results suggest that 8-O-4-DFA could play a greater role in corn centration of all cell wall components (on DM basis) except FA borer resistance than 5-5 DFA or ferulate which did not differ be- ether and 5-5-DFA concentration (Table 3). In the contrast for tween resistant and susceptible maize inbreds. Higher concentra- resistant vs. susceptible lines, resistant lines had significantly high- tion of 8-O-4-DFA could lead to more fortified cell walls and er xylose, arabinose, Klason lignin, FA ethers, 8-O-4-DFA, p-CA, and therefore more resistant genotypes. Molecular modelling experi- total cell wall material than susceptible lines (Tables 2 and 3). The ments by Hatfield et al. (1999) showed that 8-coupled DFA cross resistant line EP39 had the highest concentrations of most wall links arabinoxylans inter-molecularly while some 5-5-linked DFA components. Clearly, the resistant group had more total cell wall could also form intra-molecular linkages when arabinose residues material, particularly those components abundant in secondary bearing ferulate units are positioned three xylose residues apart on walls (glucose, xylose, and lignin), which suggests that cell wall the same xylan chain. Although Obel et al. (2003) suggested that 8- thickness may be an important barrier that corn borers must 5-DFAs could also cross link intra-molecularly, as far as we know mechanically rupture in order to gain access to the cytoplasmic 8-O-4 DFAs can only link two separate arabinoxylan chains. How- nutrients contained within pith tissues. ever, at present there are no definitive data as to which of the DFAs
Table 3 Means for esterified and etherified ferulates, and esterified diferulates and p-coumarates in the pith tissue of four inbred lines grown at Pontevedra and Zaragoza, Spain in 2008.
Genotypes Ferulates (g kg 1 DM) Diferulatesd (g kg 1 DM) p-Coumarates (g kg 1 DM) Esters Ethers Total 8-5- 8-O-4- 5-5- Total Ester A509 (R) 2.49c 2.26 4.75ab 0.10b 0.08a 0.02 0.21b 8.82a EP39 (R) 3.41a 2.34 5.75a 0.17a 0.12a 0.04 0.33a 7.47a EP42 (S) 2.63bc 1.63 4.26b 0.09b 0.05c 0.02 0.16b 7.04a EP47 (S) 3.16ab 1.65 4.81ab 0.11b 0.07bc 0.04 0.22b 5.13b LSD (P 6 0.05) 0.57 0.88 1.02 0.03 0.02 0.02 0.07 1.90 Mean resistant group 2.95 2.29 5.25 0.13 0.09 0.03 0.26 8.15 Mean susceptible group 2.90 1.64 4.54 0.10 0.06 0.03 0.19 6.08 Contrast R vs. Se NS * NS NS ** NS NS ** abc Inbred line means in the same column not sharing a superscript differ (P < 0.05). d 8–5-DFA was calculated as the sum of 8–5-open or non cyclic and 8–5-benzofuran forms of this DFA because it has been reported that 8–5-open form may be a product after alkaline hydrolysis of the cyclic form (8–5-benzofuran), the only naturally occurring in plant cell walls (Ralph et al., 1994). e Orthogonal contrast of resistant (R) and susceptible (S) lines. NS, non significant (P > 0.05). * Significant at P < 0.05. ** Significant at P < 0.01. 368 J. Barros-Rios et al. / Phytochemistry 72 (2011) 365–371
Table 4 Stepwise multiple linear regression results for corn borer stem tunnel length as a function of cell wall component concentrations in the pith tissue of four inbred lines grown at Pontevedra and Zaragoza, Spain in 2008.
Dependent variable Location-speciesa Intercept Independent variable b coefficient Pr < Fb R2
Stem tunnel length Pontevedra-MCB 38.17 Total DFAs 511 0.07 0.86 8-5-DFA 816 0.05 0.14 Stem tunnel length Zaragoza-ECB 24.52 p-CA ester 3 0.03 0.94
a Mediterranean corn borer (MCB) and European corn borer (ECB). b F, test statistic used to reject or fail to reject the null hypothesis; Pr, probability of obtaining the F test statistic, assuming that the null hypothesis is true. form intra and/or intermolecular cross-links. Grabber et al. (2002) The combination of total DFAs and 8-5-DFA explained 100% of also showed that 8-coupled DFAs may assume a greater role in cell stem tunneling variation at Pontevedra, with total DFAs alone wall stiffening than 5-5 DFA and ferulate because a substantial pro- explaining 86% of the variation (Table 4). Total DFAs had a negative portion of 5-5 DFA and ferulate mediated cross links between regression coefficient whereas 8-5-DFA had a positive regression xylans and lignin could be lost via de-esterification, if 8-b0-lactone coefficient. Such contrasting signs for regression coefficients are structures with monolignols were the major product of 8-b0-cou- common when one independent variables accounts for most of pling during lignification. However, it is not known whether these the variation in the dependent variable, resulting in an error for 8-b0-lactone structures are actually present in the plant or whether the relationship between the less important independent variable they are artifacts created during the analytical procedure. with the dependent variable. This mathematical artifact was cor- In addition, 8-5-DFA (P < 0.08) and total DFAs (P < 0.07) showed roborated by removing all DFAs from the regression analysis ex- a trend toward higher concentrations in the resistant maize in- cept 8-5-DFA and re-running the regression analysis. breds. However, 5-5-coupled DFA concentration was not different Subsequently, a negative regression coefficient was found for 8- (P = 0.82) between resistant and susceptible groups. Previous re- 5-DFA explaining 78% of the tunnel length variation. As 8-5-DFAs search by our group (Santiago et al., 2006a) also found higher con- accounted for half of all DFAs in pith tissue, our results suggest that centrations of 8-O-4- and 8-5-DFAs in the pith tissue of a larger higher concentrations of 8-coupled DFAs are related to length group of MCB resistant lines. Higher DFA ester concentrations in reductions of MCB tunnels. the resistant lines would be expected to lead to more DFA ethers At the Zaragoza location only p-CA was retained in the regres- (currently under investigation), hence the cell wall would be more sion model, with a negative regression coefficient explaining 94% fortified through these additional cross links and corn borer larvae of ECB stem tunneling variation (Table 4). This suggests that higher would have greater difficulty accessing pith tissue nutrients. A pre- concentrations of p-CA are related to shorter ECB tunneling. In vious study suggested a role for ester linked DFA in greater cell maize, p-CA is primarily esterified to syringyl units as a terminal wall adhesion, increasing the tissue mechanical strength in Chi- molecule on lignin polymers and does not function as a cross link- nese water chestnut (Eleocharis dulcis [Burm. f.] Trin. ex Henschel) ing agent between wall matrix polymers (Ralph et al., 1994). tissues (Parker et al., 2003). A similar hypothesis regarding maize Although p-CA concentration is a good indicator of lignin deposi- cell wall fortification against ECB was suggested by Bergvinson tion in grasses associated with maturity (Morrison et al., 1998)it et al. (1997). is not obvious how p-CA would affect cell wall rigidity and tough- Resistant lines had greater feruloylarabinoxylan concentration ness, and thereby corn borer resistance. These results suggest that (total FA plus total DFA plus arabinoxylan) than susceptible lines ECB damage is impacted by maize cell wall structure, but the nat- on both a dry matter (P < 0.07) and cell wall (P < 0.01) basis. How- ure of this effect is unclear. ever, the degree of xylan substitution with arabinose (arabinose- to-xylose ratio) and the degree of feruloylation of arabinose (total 3. Conclusions FA plus DFA-to-arabinose ratio) did not differ (P > 0.10) between the resistant and susceptible lines. As a result, the xylan cross link- Pith tissue cell walls may play a role as a defense mechanism of ing (total FA plus DFA-to-xylose ratio) was not different (P > 0.10) maize against corn borers in three ways: (i) thickness of the cell between the two groups of maize lines. We suggest that the higher walls, resistant lines had significantly more total cell wall material 8-coupled DFAs and total DFA concentrations (P < 0.10) in the than susceptible inbred lines; (ii) monosaccharide composition of resistant lines resulted from the higher concentration of feruloylar- cell wall polysaccharides, higher concentration of xylose in resis- abinoxylan chains in the cell wall, allowing more frequent dimer- tant inbred lines; and (iii) increased cell wall stiffening in the resis- ization between FA residues to occur, because more tant genotypes, ability of the 8-coupled DFAs to cross link feruloylarabinoxylan polymers present per unit of cell wall proba- arabinoxylan chains. A plant breeding project is currently under- bly placed FA molecules in closer physical proximity. way to develop maize lines with divergent concentrations of DFAs in order to test the role of these cross linking agents on corn borer 2.4. Regression analyses resistance.
Stepwise multiple linear regression analysis was performed in order to better understand the relationship between corn borer 4. Experimental damage and cell wall components. Tunnel length was the depen- dent variable, while cell wall components that were significant 4.1. Plant materials and experimental design (P < 0.05) or tended towards significance (P < 0.10) in the resistant vs. susceptible contrasts were included as independent variables Four maize inbred lines were selected for their consistent resis- (total cell wall, xylose, arabinose, Klason lignin, FA ether, total tance or susceptibility to corn borers across multiple evaluations FA, 8-5-DFA, 8-O-4-DFA, total DFAs, and p-CA). As different corn (Table 1). Inbred lines were grown at two locations in Spain, Pont- borer species were used to infest the maize inbreds at the two loca- evedra and Zaragoza, in 2008. Pontevedra (42° 300 N, 8° 460 W) is a tions, the multiple regression analysis was conducted separately coastal location in northwestern Spain and is approximately at sea by location in case ECB and MCB respond differently to cell wall level, whereas Zaragoza (41° 440 N, 0° 4700 W) is inland and 250 m structure with regard to extent of tunneling. above sea level. The most abundant corn borer at Pontevedra is J. Barros-Rios et al. / Phytochemistry 72 (2011) 365–371 369
MCB whereas ECB is predominant at Zaragoza (Malvar et al., 1993). combustion in a muffle furnace for 6 h at 450 °C. Monomeric com- The field experimental design at both locations was a randomized position of maize lignin was determined by pyrolysis–gas chroma- complete block design with three replicates. Each plot had two tography–mass spectral analysis as described by Ralph and rows spaced 0.80 m apart and each row consisted of 25 two-kernel Hatfield (1991). The total ion abundance data for six syringyl com- hills spaced 0.21 m apart. After thinning to one plant per hill, plant pounds (2,6-dimethoxyphenol; 2,6-dimethoxy-4-methylphenol; density was approximately 60 000 plants ha 1. Cultural operations, 4-ethyl-2,6-dimethoxyphenol; 2,6-dimethoxyvinyilphenol; 2,6- fertilization, and weed control were carried out according to local dimethoxy-4-propenylphenol; and syringaldehyde) and five practices and crop requirements. guaiacyl compounds (guaiacol, 4-ethylguaiacol, 4-methylguaiacol, To accurately define each genotype’s silking time, plots were isoeugenol, and vanillin) were collected. The ion abundance data checked until 50% of plants were showing silks. At silking, five for each compound was normalized for the guaiacol yield from plants in each plot were artificially infested with MCB in Ponteve- each sample and then the S/G ratio was calculated by summation dra and ECB in Zaragoza by placing an egg mass between the shank of the syringyl and guaiacyl compounds (Jung and Buxton, 1994). of the main ear and the stem. This infestation treatment has been demonstrated to be sufficient to guarantee corn borer damage 4.2.3. Cell wall hydroxycinnamic acids analysis (Butrón et al., 1999). Natural corn borer infestation and damage Ester-linked FA and p-CA monomers were extracted from simi- during the year of evaluation at both locations was minor. At har- lar starch-free, alcohol insoluble residues with 2 M NaOH at 39 °C vest, the stem of infested plants was split longitudinally and corn for 24 h (Jung and Shalita-Jones, 1990). Alkaline extracts were acid- borer tunnel length was measured (Table 2). Five to eight non-in- ified to pH 1.5–1.6 with concentrated phosphoric acid. The acidi- fested plants were collected for stalk composition analysis. Based fied extract was filtered through a Whatman filter 0.45 lm pore on previous studies (Jung, 2003; Santiago et al., 2006a) samples size, loaded on a C18 solid-phase extraction column (Supelco for analysis were collected 30 d after silking when internode elon- Inc., Bellefonte, PA, USA), the column was washed with 2 ml of gation had ceased. Pith and rind material were manually separated the same NaOH/phosphoric acid solution (pH 1.6) as that of the from the second and third elongated, above-ground internodes, samples, and the hydroxycinnamic acids were eluted with two frozen ( 20 °C), lyophilized, and ground through a 0.75 mm screen 2.5 ml 50% methanol washes. The eluted samples were brought in a Pulverisette 14 rotor mill (Fritsch GmbH, Oberstein, Germany) to a final volume of 10 ml and stored at 20 °C until they were [Mention of a proprietary product does not constitute a recom- analyzed. The FA and p-CA released by the alkaline extraction were mendation or warranty of the product by Misión Biológica de Gali- analyzed with an Agilent 1100 high pressure liquid chromatogra- cia, USDA, or the University of Minnesota, and does not imply phy (HPLC) system (Agilent Technologies, Wilmington, DE, USA) approval to the exclusion of other suitable products]. fitted with a diode array detector and a Spherisorb-ODS2, C18, 5 lm column (Waters Corp., Millford, MA, USA). Samples (20 ll) 4.2. Chemical determinations were eluded with a 97.7:0.3:2.0 (vol/vol) water–glacial acetic acid–butanol solvent for 15 min, followed by a methanol wash of 4.2.1. Cell wall polysaccharide analysis the column, at a flow rate of 1.8 ml min 1 (Jung and Shalita-Jones, The Uppsala Dietary Fiber method (Theander et al., 1995) was 1990). Hydroxycinnamic acids were detected at 320 nm and quan- used to measure cell wall polysaccharide components and lignin. tified using the external calibration method. A starch free, alcohol-insoluble residue was prepared according Total (ester- and ether-linked) FA monomers in the cell wall to Theander and Westerlund (1986). Acetate buffer (5 ml, 0.1 M, were extracted with 4 M NaOH for 2 h at 170 °C from starch-free, pH 5.0) and 0.1 ml of heat-stable a-amylase Termamyl 120 L (EC alcohol-insoluble residues (Iiyama et al., 1990). Alkaline extracts 3.2.1.1, from Bacillus licheniformis, 120 KNU/g) (Sigma Chemical were treated as described above to isolate and quantify hydroxy- Co., St. Louis, MO, USA) were added to 100 mg samples and heated cinnamic acids. Ether-linked FA was calculated as the difference at 90 °C for 60 min. After the mixture cooled to 50 °C, 0.2 ml of between total and ester-linked FA concentrations of each sample amyloglucosidase AMG 300 L (EC 3.2.1.3, from Aspergillus niger, (Iiyama et al., 1990). 300 AGU/g) (Sigma) was added to the samples which were then Ester bound DFAs were extracted based on a procedure previ- heated for 3 h at 60 °C. Sufficient 95% ethanol was subsequently ously described (Santiago et al., 2006a) with minor modifications. added to achieve a final concentration of 80% ethanol and the sam- One gram of ground material was extracted with 30 ml of 80% ple was held at 4 °C overnight. The crude cell wall preparation was methanol. The suspension was homogenized for 30 s with a Hei- recovered by centrifugation, washed twice with 80% ethanol and dolph mixer (Heidolph Instruments GmbH & Co. KG, Schwabach, once with acetone, and allowed to air dry under a hood. Samples Germany) before being centrifuged at 1000g for 10 min. After cen- were suspended in 12 M sulfuric acid at 30 °C for 1 h, followed trifugation, the pellet containing ester-bound phenols incorporated by dilution with water to 0.3 M sulfuric acid and heating in an in the cell wall was shaken in 20 ml of 2 M NaOH under nitrogen autoclave for 1 h at 117 °C to hydrolyze the cell wall polysaccha- flow and darkness for 4 h. The pH of alkali-treated samples was rides. After acid hydrolysis, the neutral sugar components (glucose, lowered to 2.0 with 6 N HCl. After centrifugation, the supernatant xylose, arabinose, mannose, and galactose) in the filtrate were was collected and the pellet washed twice with distilled water quantified by gas chromatography as alditol acetate derivatives (10 ml each). Supernatants were pooled and then extracted twice (Theander et al., 1995). Inositol was used as an internal standard with ethyl acetate (40 ml each). Collected organic fractions were to correct for volume variation. Neutral sugar data were converted combined and reduced to dryness using a Speed Vac (Thermo Fisher to an anhydro-sugar basis. The acidic sugars (glucuronic, galact- Scientific Inc., MA, USA) for 5 h. The final extract was dissolved in uronic, and 4-O-methylglucuronic acids) were measured as total 3 ml of HPLC-grade methanol. All of the extracts were stored at uronic acids by the colorimetric method of Ahmed and Labavitch 20 °C prior to HPLC analysis. Samples were filtered through a (1977), in aliquots of the 0.3 M sulfuric acid solution sampled be- 2 lm pore poly (tetrafluoroethylene) filter (Chromatographic fore heating, using glucuronic acid as the reference standard. Specialties, Brockville, ON, Canada) before analysis. Analyses were performed using a 2690 Waters Separations Module (Waters, Mil- 4.2.2. Lignin analysis ford, MA, USA) equipped with a Waters 996 photodiode array Klason lignin was determined as the insoluble residue from the detector and a Waters YMC ODSAM narrow-bore column two-stage acid hydrolysis retained on a glass fibre filter mat in a (100 2 mm i.d., 3 lm particle size). Elution conditions with a mo- coarse-porosity Gooch crucible and corrected for ash content by bile phase system of acetonitrile (solvent A) and trifluoroacetic 370 J. Barros-Rios et al. / Phytochemistry 72 (2011) 365–371 acid (0.05%) in water (solvent B) were as follows: initial conditions digestibility in the maize recombinant inbred line progeny F838 F286. Plant 10:90 (A/B), changing to 30:70 in 3.5 min, then to 32:68 in 6.5 min, Sci. 175, 585–595. Barros, J., Malvar, R.A., Butrón, A., Santiago, R., 2010. Combining abilities in maize for then to 100:0 in 4 min, then isocratic elution with 100:0 for the length of the internode basal ring, the entry point of the Mediterranean corn 4.5 min, finally returning to the initial conditions in 3 min. The mo- borer larvae. Plant Breeding 129, 1–3. bile phase flow rate was 0.3 ml min 1, the total analysis time was Bergvinson, D.J., Arnason, J.T., Hamilton, R.I., 1997. Phytochemical changes during recurrent selection for resistance to the European corn borer. Crop Sci. 37, 21.5 min, and the sample injection volume was 4 ll. 1567–1572. Retention times and UV spectra were compared with freshly Bily, A.C., Reid, L.M., Taylor, J.H., Johnston, D., Malouin, C., Burt, A.J., Bakan, B., prepared standard solutions of 5-5-DFA, kindly provided by the Regnault-Roger, C., Pauls, K.P., Arnason, J.T., Philogene, B.J.R., 2003. Dehydrodimers of ferulic acid in maize grain pericarp and aleurone: group of Dr. J.T. Arnason (University of Ottawa, Ontario, Canada). resistance factors to Fusarium graminearum. Phytopathology 93, 712–719. The absorption UV spectra of other DFAs were compared with pre- Buendgen, M.R., Coors, J.G., Grombacher, A.W., Russell, W.A., 1990. European corn viously published spectra (Waldron et al., 1996) and absorbance at borer resistance and cell wall composition of three maize populations. Crop Sci. 30, 505–510. 325 nm was used for quantification. Bunzel, M., 2010. Chemistry and occurrence of hydroxycinnamate oligomers. Total cell wall concentration was calculated as the sum of glu- Phytochem. Rev. 9, 47–64. cose, xylose, arabinose, mannose, galactose, uronic acids, Klason Butrón, A., Malvar, R.A., Cartea, M.E., Ordás, A., Velasco, P., 1999. Resistance of maize lignin, total FA, and ester-linked p-CA. The DFAs were not included inbreds to pink stem borer. Crop Sci. 39, 102–107. Butrón, A., Revilla, P., Sandoya, G., Ordás, A., Malvar, R.A., 2009. Resistance to reduce in the cell wall concentration calculation because some samples corn borer damage in maize for bread, in Spain. Crop Prot. 38, 134–138. were lost and their minor contribution to total cell wall concentra- Cardinal, A.J., Lee, M., Sharopova, N., Woodman-Clikeman, W.L., Long, M.J., 2001. tion. While most compositional analyses were done in duplicate, Genetic mapping and analysis of quantitative trait loci for resistance to stalk tunneling by the European corn borer in maize. Crop. Sci. 41, 835–845. only single lab replication was possible for S/G ratio determination García-Lara, S., Bergvinson, D., Burt, A.J., Ramputh, A.I., Díaz-Pontones, D.M., and analysis of DFAs (20% of the samples were analyzed in dupli- Arnason, J.T., 2004. The role of pericarp cell wall components in maize weevil cate) because of limited sample quantity of the isolated tissues. resistance. Crop Sci. 44, 1546–1552. Gianessi, L., Sankula, S., Reigner, N., 2003. Plant Biotechnology: Potential Impact for However, the repeatability of the DFA and S/G analytical methods Improving Pest Management in European Agriculture. A Summary of Three Case is high and previous experience has shown that statistically Studies. The National Center for Food and Agricultural Policy, Washington, DC. significant differences can be detected in replicated field trials with
Ramputh, A.I., 2002. Soluble and cell wall Bound phenolic-mediated insect Santiago, R., Souto, X.C., Sotelo, J., Butrón, A., Malvar, R.A., 2003. Relationship resistance in corn and sorghum. Ph.D. Dissertation, Ottawa-Carleton Institute between maize stem structural characteristics and resistance to pink stem of Biology, Ottawa, ON, Canada. borer (Lepidoptera: Noctuidae) attack. J. Econ. Entomol. 96, 1563–1570. Sandoya, G., Butrón, A., Álvarez, A., Ordás, A., Malvar, R.A., 2008. Direct response of SAS Institute, 2007. The SAS System. SAS Online Doc. HTML Format Version Eight. maize synthetic to recurrent selection for resistance to stem borers. Crop Sci. SAS Institute Inc., Cary, North Carolina. 48, 113–118. Terra, W.R., Valentin, A., Santos, C.D., 1987. Utilization of sugars, hemicellulose, Santiago, R., Butrón, A., Arnason, J.T., Reid, L.M., Souto, X.C., Malvar, R.A., starch, protein, fat and minerals by Erimo’is ello larvae and the digestive role of 2006a. Putative role of pith cell wall phenylpropanoids in Sesamia their midgut hydrolases. Insect Biochem. 17, 1143–1147. nonagrioides (Lepidoptera: Noctuidae) resistance. J. Agric. Food Chem. 54, Theander, O., Aman, P., Westerlund, E., Andersson, R., Pettersson, D., 1995. Total 2274–2279. dietary fiber determined as neutral sugar residues, uronic acid residues, and Santiago, R., Butrón, A., Reid, L.M., Arnason, J.T., Sandoya, G., Souto, X.C., Malvar, R.A., Klason lignin (the Uppsala method): collaborative study. J. AOAC Int. 78, 1030– 2006b. Diferulate content of maize sheaths is associated with resistance to the 1044. Mediterranean corn borer Sesamia nonagrioides (Lepidoptera: Noctuidae). J. Theander, O., Westerlund, E., 1986. Studies on dietary fiber. 3. Improved procedures Agric. Food Chem. 54, 9140–9144. for analysis of dietary fiber. J. Agric. Food Chem. 34, 330–336. Santiago, R., Reid, L.M., Arnason, J.T., Zhu, X.Y., Martinez, N., Malvar, R.A., 2007. Velasco, P., Revilla, P., Monetti, L., Butrón, A., Ordás, A., Malvar, R.A., 2007. Corn Phenolics in maize genotypes differing in susceptibility to Gibberella stalk rot borers (Lepidoptera: Noctuidae; Crambidae) in northwestern Spain: population (Fusarium graminearum Schwabe). J. Agric. Food Chem. 55, 5186–5193. dynamics and distribution. Maydica 52, 195–203. Santiago, R., Sandoya, G., Butrón, A., Barros, J., Malvar, R.A., 2008. Changes in Waldron, K.W., Parr, A.J., Ng, A., Ralph, J., Williamson, G., 1996. Cell wall esterified phenolic concentrations during recurrent selection for resistance to the phenolic dimers: identification and quantification by reversed phase high Mediterranean corn borer (Sesamia nonagrioides Lef.). J. Agric. Food Chem. 56, performance liquid chromatography and diode array detection. Phytochem. 8017–8022. Anal. 7, 305–312.