c06-05-0283 Krakowsky.indd

Correlations between concentrations of cell wall components (CWCs) in the leaf sheath and stalk and resistance to stalk tunneling by the European corn borer (ECB) [Ostrinia nubilalis (Hübner)] have been reported in some maize (Zea mays L.) populations. Evaluations of resistance to ECB stalk tunneling (ECB-ST) and concentrations of neutral detergent fi ber (NDF), acid detergent fi ber (ADF), and acid detergent lignin (ADL) have been performed on recombinant inbred lines (RILs) developed from the cross of maize inbred lines B73 (susceptible to ECB-ST, low to moderate CWC concentrations) and DE811 (resistant to ECB-ST, high CWC concentrations). The objective of this study was to estimate genotypic correlations between ECB-ST and CWC concentrations and compare locations and effects of quantitative trait loci (QTL) for those traits. Genotypic correlations between ECB-ST and CWCs were not signifi cant, but clustering of QTL for ECB-ST and CWCs was observed. Negative genotypic correlations between ECBST and CWC concentrations were observed at some loci, and resistance to ECB-ST may be associated with a subset of the QTL observed for CWCs and ADF in particular. Resistance to ECB-ST may also be associated with starch concentration in the stalk, which could explain the detection of resistance alleles contributed by B73. Examination of temporal differences in CWC and starch concentrations, and markerassisted transfer of select alleles, could provide more information on mechanisms of resistance to ECB-ST. Genotypic Correlation and Multivariate QTL Analyses for Cell Wall Components and Resistance to Stalk Tunneling by the European Corn Borer in Maize M. D. Krakowsky,* M. Lee, and J. B. Holland M.D. Krakowsky and M. Lee, Dep. of Agronomy, Iowa State Univ., Ames, IA 50011; J.B. Holland, USDA-ARS, Plant Science Research Unit, Dep. of Crop Science, North Carolina State Univ., Raleigh, NC 27695. M.D. Krakowsky, current address: USDA-ARS, Crops Genetic and Breeding Research Unit, Coastal Plain Experiment Station, Tifton, GA 31794. Research conducted in partial fulfi llment of the Ph.D. degree by M.D. Krakowsky. This journal paper of the Iowa Agriculture and Home Economics Experiment Station, Ames, Iowa, Project No. 3134, was supported by Hatch Act and State of Iowa funds and the R.F. Baker Center for Plant Breeding. Received 1 May 2006. *Corresponding author (mkrakowsky@ tifton.usda.gov). Abbreviations: ADF, acid detergent fi ber; ADL, acid detergent lignin; CWCs, cell wall components; ECB, European corn borer; ECB-ST, ECB stalk tunneling; IVDMD, in vitro dry matter digestibility; NDF, neutral detergent fi ber; QTL, quantitative trait loci; RILs, recombinant inbred lines; WFISIHI, Wisconsin fi ber silica high; WFISILO, Wisconsin fi ber silica low. The biological basis of resistance to stalk tunneling by the European corn borer (ECB) [Ostrinia nubilalis (Hübner)], a major pest of temperate maize (Zea mays L.), is not well understood. The borer normally has two generations per year in the U.S. cornbelt, with the fi rst generation feeding primarily on leaf tissue and the second generation feeding primarily on leaf sheath, shank, and stalk tissues (Mason et al., 1996). Resistance to leaf feeding by ECB in temperate maize is usually conferred by the chemical 2,4-dihidroxy-7-methoxy-1,4-benzoxazin-3-one (DIMBOA), but the chemical is not associated with resistance to ECB-ST, as concentrations of DIMBOA decrease as the plant matures (Klun et al., 1967, 1970; Klun and Robinson, 1969). One possible mechanism of resistance to leaf sheath and stalk feeding is increased concentrations of CWCs, particularly fi ber and lignin, in the leaf sheath and stalk. Published in Crop Sci 47:485–490 (2007). doi: 10.2135/cropsci2006.05.0283 © Crop Science Society of America 677 S. Segoe Rd., Madison, WI 53711 USA R e p ro d u c e d fr o m C ro p S c ie n c e . P u b lis h e d b y C ro p S c ie n c e S o c ie ty o f A m e ri c a . A ll c o p y ri g h ts re s e rv e d . 486 WWW.CROPS.ORG CROP SCIENCE, VOL. 47, MARCH–APRIL 2007 Elevated levels of fi ber and lignin have been correlated with resistance to ECB-ST in several studies. The BS9(CB) population was developed for use in a recurrent selection program for resistance to both generations of ECB (Klenke et al., 1986). Highly signifi cant linear correlated responses were observed for leaf sheath NDF, ADF, and ADL over four cycles of selection, indicating a possible genetic association between concentrations of CWCs and resistance to ECB-ST. In addition, negative genotypic correlations were observed between ECB-ST and concentrations of NDF, ADF, and ADL in F 3 lines of B73 × DE811, and heritability estimates for all cell wall components (H > 70%) in this population were within the range of the heritability for resistance to ECB-ST (H = 74%; Beeghly et al., 1997). To clarify the association between CWC concentrations and ECB-ST, selection for CWC concentrations was performed in BS9(CB) Cycle 2 (BS9(CB) C2), Wisconsin fi ber silica low (WFISILO), and Wisconsin fi ber silica high (WFISIHI), to produce BS9(CB) C2-Lo and WFISILO C0, C1, and C2 (selected for lower concentrations of CWCs) and BS9(CB) C2-Hi and WFISIHI C0, C1, and C2 (selected for higher concentrations of CWCs) (Ostrander and Coors, 1997). The populations WFISILO and WFISILH diff ered signifi cantly for CWC concentrations, but not for ECBST. Selection for decreased CWCs in BS9(CB) C2-Lo and WFISILO resulted in increased ECB stalk tunneling, but selection for higher concentrations of CWCs did not consistently result in lower ECB stalk tunneling. The BS9(CB) C2-Hi population diff ered signifi cantly from BS9(CB) C2Lo for stalk NDF, ADF and lignin concentrations, and ECB stalk tunneling, but selection in WFISIHI for increased CWC concentrations did not decrease ECB stalk tunneling. The results suggest that selection for low CWC concentrations may increase susceptibility to second-generation ECB, and, conversely, selection for ECB resistance may adversely aff ect the nutritional value of maize harvested as silage (Ostrander and Coors, 1997). Genotypic correlations and comparisons of locations of QTL for ECB-ST and CWC concentrations have been reported for RILs of B52 × B73 (Cardinal and Lee, 2005). In that population, genotypic correlations between ECBST and CWC concentrations ranged from –0.27 to –0.44, and a majority of the QTL observed for ECB-ST were linked to QTL for CWCs, indicating that CWCs may play some role in expression of resistance to ECB-ST in the RILs of B52 × B73. Quantitative trait loci for resistance to ECB-ST and CWC concentrations have been mapped in RILs of B73 × DE811 (Krakowsky et al., 2004, 2005, 2006). B73 is susceptible to ECB-ST and has low to intermediate concentrations of CWCs, while DE811 is resistant to ECBST and has high concentrations of CWCs. The primary objective of this study was to estimate genotypic correlations between, and compare locations and eff ects of QTL for, ECB-ST and CWC concentrations to provide a better understanding of the relationship between these traits. Multivariate QTL mapping was used to estimate the amount of genotypic covariance between ECB-ST and CWC associated with specifi c genome regions. A secondary objective was to compare the genetic correlations and QTL estimates from this population to those reported by Cardinal and Lee (2005) for RILs of B73 × B52, which were evaluated for the same traits. MATERIALS AND METHODS The univariate QTL analyses of the B73 × DE811 population have been reported previously (Krakowsky et al., 2004, 2005, 2006). Briefl y summarizing the materials and methods from those reports, 200 RILs of B73 × DE811, along with both parental inbreds, were planted at two locations in 1998 and one location in 1999. The ECB-ST and CWC concentrations were evaluated in all environments. Two stalk-tunneling traits were originally analyzed due to the correlation between ECBST and anthesis: ECB-ST and ECB-ST adjusted for maturity. Adjustment of ECB-ST for diff erences in maturity was performed by including anthesis as a covariate in the model for calculation of least-square means for ECB-ST (Krakowsky et al., 2004). The resulting adjusted ECB-ST least-square means were used for calculation of genotypic correlations and comparisons of QTL herein. In addition, NDF and NDF adjusted for ADF (NDF adjusted) were reported for both stalk and sheath CWCs (Krakowsky et al., 2005, 2006). To separate the eff ects of different CWCs, NDF adjusted for ADF was used for calculation of genotypic correlations and comparisons of QTL herein. The ADL was not included in the analyses herein due to the low genetic variances observed in both the sheath and stalk (Krakowsky et al., 2005, 2006). Genotypic covariances and correlations (r g ) and the standard errors of the correlations were calculated herein among traits by applying standard procedures (Mode and Robinson, 1959). The necessary computations were performed using the MANOVA statement in PROC GLM of the software package SAS, with entries and environments treated as random eff ects (SAS Institute, Inc., 1999). To investigate the eff ects of QTL on ECB-ST adjusted for maturity and individual CWCs simultaneously, multivariate multiple regression analysis was conducted on two traits, using marker loci as regressor variables, as described by RobertsonHoyt et al. (2006). The initial regression analysis model contained one marker locus nearest each QTL peak from the fi nal QTL models for each trait. Multivariate analysis was conducted using the MANOVA option of Proc GLM in SAS version 8.2 (SAS Institute, Inc., 1999). A single degree of freedom contrast was defi ned for each marker locus or epistatic interaction. The initial model was overspecifi ed and included several nonsignifi cant terms; therefore, backward selection was used to develop a model in which all terms had signifi cant eff ects. The marker contrast with the largest Type III p value of Wilks’ lambda statistic (which tests the null hypothesis of no marker eff ect on the two traits simultaneously) was dropped from the model. The resulting new model was tested in the same way, and this process continued iteratively until all marker loci had significant (p = 0.05) eff ects according to Wilks’ lambda statistic. R e p ro d u c e d fr o m C ro p S c ie n c e . P u b lis h e d b y C ro p S c ie n c e S o c ie ty o f A m e ri c a . A ll c o p y ri g h ts re s e rv e d . CROP SCIENCE, VOL. 47, MARCH–APRIL 2007 WWW.CROPS.ORG 487 The genotypic covariance between the two traits associated with each marker locus in the fi nal model was estimated by the method of moments, using the Type III mean cross products for each marker contrast obtained from the Proc GLM analysis. To obtain unbiased estimates of the covariances, the coeffi cients of the covariance components were based on the equations of Charcosset and Gallais (1996). These estimates were used to calculate the proportion of total genotypic covariance between the two traits due to each marker locus. RESULTS AND DISCUSSION The genotypic correlations between ECB-ST adjusted for maturity and leaf sheath NDF adjusted and ADF (–0.08 [standard error (S.E.) = 0.10] and 0.02 [S.E. = 0.07], respectively) and between ECB-ST adjusted for maturity and stalk NDF adjusted and ADF (0.25 [S.E. = 0.26] and 0.13 [S.E. = 0.13], respectively) were not signifi cantly diff erent from zero and were lower than those reported by Beeghly et al. (1997) for leaf sheath in the F 3 lines of B73 × DE811. The genotypic correlations reported by Cardinal and Lee (2005) for ECB-ST with leaf sheath and stalk ADF in the RILs of B73 × B52 were –0.44 and –0.36, respectively, which are much higher than those observed herein. The NDF adjusted for ADF was not evaluated in the RILs of B73 × B52. These results suggest that the associations between ECB-ST adjusted for maturity and concentrations of CWCs observed in other studies may not be applicable to this population. In most of the cases this may be due to genetic heterogeneity, but that would not apply to the B73 × DE811 F 3 lines. That population was evaluated in only 1 yr at two locations (Beeghly et al., 1997), potentially confounding the results with the eff ects of the environments. It is also possible that resistance to ECB-ST and increased concentrations of CWCs are correlated in the F 3 lines because both traits were predominantly associated with alleles from DE811. This association may have been maintained in the F 3 lines and dissipated during the additional meioses (and recombination) that occurred during the development of RILs. The RILs of B73 × B52 were evaluated in the same environments as the RILs of B73 × DE811 during 1998, reducing the eff ect of environment on comparisons between those populations. Ten QTL were detected for ECB-ST adjusted for maturity, 9 of which are located within 25 cM of QTL for at least one CWC (Krakowsky et al., 2004, 2005, 2006) (Fig. 1). However, the expected relationship between the signs of the QTL eff ects for the two traits was only observed in about half the instances where QTL were linked; an allele from DE811 that was associated with resistance to ECB-ST adjusted for maturity was as likely to be associated with increased concentrations of CWCs as decreased concentrations of CWCs. There was also no apparent correlation between the size of the additive eff ects for linked QTL for ECB-ST adjusted for maturity and CWC concentrations. These factors may explain the absence of signifi cant overall genotypic correlations between ECB-ST adjusted for maturity and CWC concentrations in this population. Only three of the nine QTL for ECB-ST in the RILs of B73 × B52 were linked (within 25 cM) to QTL for ECB-ST adjusted for maturity in the RILs of B73 × DE811 (Krakowsky et al., 2004; Cardinal and Lee, 2005). Two of those QTL are on chromosome 2, and the patterns in the RILs of B73 × B52 are similar to those observed herein; at umc8, the allele(s) from B73 is associated with increased ECB-ST and stalk ADF concentrations, while at umc4 the allele(s) from B73 is associated with increased ECB-ST and reduced stalk ADF concentrations. The third QTL is on chromosome 9 and is not associated with QTL for ADF. The lack of consistent localization of QTL across the populations may indicate genetic heterogeneity for resistance to ECB-ST (Krakowsky et al., 2004). The clustering of QTL for ECB-ST adjusted for maturity and CWC concentrations raises the question of whether the QTL for the two traits are linked by random chance or if the QTL represent pleiotropic genes aff ecting both traits. Five QTL each for sheath and stalk NDF adjusted were linked to QTL for ECB-ST adjusted for maturity, with no clear association between NDF concentrations and resistance to ECB-ST. For sheath NDF adjusted, the allele from one parent was associated with resistance to tunneling and increased CWC concentrations in only two cases, and for stalk NDF the expected relationship was observed in three cases. Three QTL for ECB-ST adjusted for maturity were linked to QTL for sheath ADF, with the allele from one parent associated with resistance to stalk tunneling and increased concentrations of sheath ADF in two cases, while six QTL for ECB-ST adjusted for maturity were linked to QTL for stalk ADF, with the expected relationship observed in three cases. The NDF adjusted may be uncorrelated or even positively correlated with ECB-ST, because it represents the relatively digestible hemicellulosic fraction of the cell wall; in the sheath, NDF adjusted is higher in B73 than in DE811 (Krakowsky et al., 2006). In the case of ADF, however, a pattern does appear to be present: eight of the nine QTL for ECB-ST adjusted for maturity were linked to QTL for sheath or stalk ADF, and the alleles from DE811 were associated with increased ADF concentrations at all of these QTL. While this pattern could indicate that ECB-ST adjusted for maturity is negatively correlated with ADF in inbred DE811 (i.e., increased concentrations of ADF are correlated with decreased levels of stalk tunneling), it would also hold that the opposite is true for inbred B73. The proportions of the genotypic covariances associated with linked QTL for ECB-ST adjusted for maturity and sheath and stalk NDF adjusted and ADF are listed in Table 1. While the genetic correlations between the traits indicate relatively small genotypic covariances, R e p ro d u c e d fr o m C ro p S c ie n c e . P u b lis h e d b y C ro p S c ie n c e S o c ie ty o f A m e ri c a . A ll c o p y ri g h ts re s e rv e d . 488 WWW.CROPS.ORG CROP SCIENCE, VOL. 47, MARCH–APRIL 2007 dissection of those covariances can provide information on the genetic basis of a relationship between two traits that would not otherwise be apparent. The total genetic variances explained by the CWC QTL in Table 1 are relatively small as compared to those for the full models reported previously (Krakowsky et al., 2005, 2006), due to the smaller number of QTL in the models herein. The proportion of the genotypic covariance between specifi c CWCs and ECB-ST explained by mapped QTL was greater than 1 or less than –1 for several stalk and sheath NDF adjusted QTL. This is possible because covariances can be negative or positive, and covariances of equal magnitude, but opposite sign, will cancel each other out when summed across loci in the total genotypic covariance. Figure 1. Linkage map of B73 × DE811 recombinant inbred lines and location of quantitative trait loci (QTL) for cell wall components and stalk tunneling. Underlined loci exhibited segregation distortion (p < 0.05). Solid shapes denote QTL for which the allele from DE811 is associated with an increase in the trait, while outlined shapes denote QTL for which the allele from B73 is associated with an increase