| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Africa Rice Center (WARDA), 01 BP 2031, Cotonou, Benin
b Dep. of Plant and Environmental Sciences, Norwegian Univ. of Life Sciences, P.O. Box 5003, N-1432, Ås, Norway
* Corresponding author (helge.skinnes{at}umb.no)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: AFLP, amplified fragment length polymorphism • ArNK, Arina x NK93604 • CIM, composite interval mapping • cM, centimorgan • d.a.i., days after inoculation • DArT, diversity arrays technology • d.f., degrees of freedom • d°, sum degree Celsius x day • DH, double haploid • DON, deoxynivalenol • FHB, Fusarium head blight • LOD, log of odds • QTL, quantitative trait loci • R2, proportion of variance explained by QTL • RFLP, restriction fragment length polymorphism • SSR, simple sequence repeat • CV, cross validation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In FHB resistance breeding efforts both in Europe and North America, the main emphasis has long been on exploiting resistance, mainly from Chinese, Japanese, and Brazilian spring wheat germplasm. However, many backcrosses are needed to incorporate these exotic sources into adapted varieties, making this a long-term strategy. An alternative short-term approach has been to exploit moderate resistance already present in adapted material, as documented by Saur (1991) and Mielke (1980). Mapping studies based on spray inoculation in field conditions in ‘Renan’ x ‘Recital’ population, for example, revealed three stable QTLs from Renan on chromosomes 2B and 5A (Gervais et al., 2003), whereas Paillard et al. (2004) reported two major QTLs from Arina from an Arina x Forno population on chromosomes 4A and 6D. However, the Arina x Forno map for chromosome 4A and 6D (Paillard et al., 2003) had two limitations: (i) the maps consisted of very few markers; and (ii) the flanking markers on either side of both chromosomes are RFLPs, which are not breeder-friendly for high-throughput molecular breeding strategies. We recently published a linkage map in a double haploid (DH) population derived from a cross between Arina (a Swiss winter wheat; accession number 01C0104276) and NK93604 (a Norwegian spring wheat breeding line) using diversity arrays technology (DArT), amplified fragment length polymorphism (AFLP), and simple sequence repeat (SSR) markers (Semagn et al., 2006). The objectives of this study were to determine the genome location, parental contributions, and effects of QTLs for FHB resistance and low DON content in the Arina x NK93604 population.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The DH population segregated for spring and winter types and only spring types were selected in the mapping population because breeding for FHB resistant spring wheat adapted to the Norwegian climatic conditions is one of the priorities in the country. Parents and the DH lines were evaluated for FHB resistance at Vollebekk Research Farm of Norwegian University of Life Sciences, Ås, Norway, for a period of 3 yr (2001–2003). The lines were sown in spring (May) and grown in hill plots of 30–40 plants, 40 x 45 cm apart in two replicates in 2001, and three replicates of 75 x 200 cm plots in 2002 and 2003 with complete randomization of genotypes within replicates. Each year the plots were treated with a combination of propiconazole and fenprophimorph 1 wk before anthesis. This treatment does not affect FHB (Henriksen and Elen, 2005) but helps to avoid the influence of other leaf diseases on FHB development. A bundle of 10 to 15 heads per line and replicate were inoculated at full flowering by spraying a conidial suspension of 10 to 15 mL Fusarium culmorum using hand sprayers (P2 Polyspray, Hozelock Ltd., Haddenham, Buckinghamshire, UK). Inoculum was produced from single spore isolates on Potato Dextrose Aagar. In the 2001 experiment, five isolates (no. 7, 8, 9, 200–104, 33–3) were obtained from BIOFORSK Crop Research Institute, Ås, and a mixture of these isolates containing 105 spores per mL was sprayed. In both 2002 and 2003 experiments, two isolates of low (isolate no. 8) and high (isolate no. 200–104) aggressiveness, respectively, each containing 5x104 spores per mL, were sprayed separately using distilled water as a control. Aggressiveness was tested on wheat seedlings according to Mesterhazy (1985). Inoculations (bundles of about 15 heads of low and high aggressiveness plus water) were made in two parallels for each plot, resulting in a total of six treatment applications per plot. Infected heads were covered with a transparent plastic bag for 2–3 d (45 degree days) as described by Mesterhazy (1995). The proportion of infected spikelets per bundle of about 15 heads was estimated visually using a linear scale from 0 (no disease) to 4 (100% of spike area diseased). Observations were made on the basis of a constant sum of days x degree Celsius after inoculation within years: 320 resp 410 d° in 2001, 275 resp 365 d° in 2002, and 170 resp 240 d° in 2003. For the 2002 experiment, both parallels of low and high aggressiveness of each plot were harvested at full ripening, dried, hand threshed, bulked, and used for deoxynivalenol (DON) determination using the fluorometric quantitation method (Malone, 1998) at CIMMYT, Mexico.
Statistical Analyses
Phenotypic Data
Analysis of variance was performed using the PROC GLM procedure of the SAS software package (SAS Institute Inc., Version 9.1). The segregation of the DH lines for the different traits was tested for normality using the PROC UNIVARIATE procedure of SAS. Components of variance were computed considering the effects of the environment (year) and genotype as random using PROC VARCOMP of SAS. Broad-sense heritability for FHB severity across 3 yr was calculated with the formula H2 = 
2g/(2g + 2gxy/r + 2
/ry) where 2g is the genetic variance among the DH lines, 2gxy is the genotypes-by-years interaction, 2 is the error variance, y is the number of years (the experiments) and r is the number of replications for the experiments. For DON content, repeatability within a year was estimated using the formula H2 = 2g/(2g + 2/r). The Spearman correlation coefficients were calculated using the PROC CORR procedure of SAS.
Marker Data and QTL Analyses
All 93 DH lines of the ArNK population were genotyped with AFLP, DArT and SSR markers. AFLP markers were named according to the standard list for AFLP primer nomenclature provided by KeyGene (http://wheat.pw.usda.gov/ggpages/keygeneAFLPs.html; verified 5 Oct. 2006). DArT markers were named as described by Akbari et al. (2006) with a prefix "wPt" followed by numbers corresponding to a particular clone in the genomic representation, where w stands for wheat, P for PstI (primary restriction enzyme used) and t for TaqI (secondary restriction enzyme). The genotypic data was used to construct a genetic linkage map, which spans 2596 cM with a total of 624 markers (165 AFLPs, 189 DArTs, and 270 SSRs) distributed into 21 linkage groups (Semagn et al., 2006). QTL analyses were performed on yearly averages, grand average FHB severity, and DON content in the 2002 experiment. Composite interval mapping (CIM) was performed on untransformed FHB data using the PLABQTL software, version 1.2 (Utz and Melchinger, 2003). Because of the non-normal distribution of the data for DON content (Fig. 1
), QTL analysis for the latter was performed using the log transformation option of the PLABQTL. Thirty one markers that were linked more closely than 0.11 cM were excluded from the QTL analysis to prevent ill-conditioned equation systems and the generation of synthetic new markers by the program. A whole-genome scan with CIM was conducted using the following options: automatic cofactor selection, model to determine additive effects at individual QTL and additive x additive epistatic interactions, and F-to-Enter value of ten. The dependence of QTL estimation on sampling effects was estimated by a five-fold cross validation (CV), dividing the genotypes into five subsets (using four for calibration, the last one for validation in each independent run). The LOD threshold for declaring a putative QTL was set to 3.0 as determined after 1000 permutation tests (type I error level 
= 10%). In this study, QTLs that explained <10% and 
10% of the total phenotypic variation were arbitrarily classified into minor and major QTLs, respectively. Genetic maps and QTL graphs were drawn using the MapChart program, version 2.1 (Voorrips 2002, Wageningen, The Netherlands).
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The distribution of mean FHB severity was approximately normal (P = 0.047); however, the Shapiro-Wilk test rejected the hypothesis of normality (P < 0.001) for DON content. For each experiment, significant variation for FHB severity and DON content were observed within the population (Table 1). In the combined analysis for FHB severity, both genotype and genotype-by-year interaction effects were highly significant (P < 0.0001). Broad-sense heritability was 0.91 for FHB mean values averaged over years, and repeatability for DON content was estimated to be 0.51. Highly significant (P < 0.0001) positive correlations between FHB severity in different years (0.71 
r 0.78) and between FHB severity and DON content (0.61 r 0.76) were detected (Table 2).
|
|
R2 1.2%). Five-fold cross validation performed on FHB severity values averaged over years revealed that the QTLs mapped on chromosomes 1AL and 1BL were consistently detected in each of the five cross validation splits. The QTLs on 6BS and 7AL were detected in only two and four out of five splits, respectively (data not shown).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The high correlations in FHB severity among years (0.71 r 0.78; P < 0.0001) and high broad-sense heritability estimated in the ArNK population are indications of the accuracy and reproducibility of the experimental and disease evaluation method. A positive, linear correlation between DON content and FHB severity was detected in the different years (0.61 r 0.76; P < 0.0001) following spray inoculation in field conditions. The coincident QTL identified for the two traits on chromosome 1AL may have contributed to these results. There are contradictory reports in the literature concerning correlations between FHB and DON content. Liu et al. (1997) did not find any significant correlation between FHB rating in the field and DON content. Somers et al. (2003) reported low to medium correlations (0.15 r 0.65) between FHB and DON content in Wuhan-1 x BacUp population, whereas Miedaner et al. (2004) reported a significant correlation (r = 0.77; P = 0.01) between the two traits. The latter results, together with ours, suggest that lines with low DON content can be indirectly identified by selecting for reduced FHB severity.
For both traits, we found transgressive segregates toward both resistance and susceptibility in the ArNK population (Fig. 1), indicating that both parents carry positive and negative alleles, as has been reported in other studies (e.g., Buerstmayr et al., 1999; Waldron et al., 1999; Miedaner et al., 2003; Liu et al., 2005). Transgressive segregation has been observed in wheat for head blight rating when medium to highly susceptible parents are crossed (Snijders, 1990; Singh et al., 1995; Buerstmayr et al., 1999; Ittu et al., 2000). The transgressive segregation observed toward lower values for FHB and DON content suggest that both parents contributed favorable alleles to their offspring; thereby creating an opportunity to identify superior gene combinations in the population by selection.
QTL Mapping
FHB Resistance
Interval mapping identified a total of 4 QTLs for FHB resistance (Table 3). Each of the individual FHB resistance QTLs in the ArNK population explained from 7.8 to 27.9% of the total phenotypic variance and together accounted for 49.1%, which resemble results from other QTL studies on moderate FHB resistance (Gervais et al., 2003; Somers et al., 2003; Paillard et al., 2004). The three QTLs identified on chromosome 1AL, 1BL, and 7AL were consistent across years. The major QTL on 1AL from NK93604 is located between wPt-5577 and Xbarc213 and explained up to 27.9% of the total phenotypic variation for FHB severity over 3 yr (Table 3). As far as we are aware, this major QTL for FHB resistance on chromosome 1AL is novel and has not been reported elsewhere.
The QTL on chromosome 1BL is flanked by P43/M62–400 and wPt-3475 markers, and explained up to 19.6% of the phenotypic variation for FHB severity across years. P43/M62–400 mapped 1.8 cM distal to P33/M59–474, whereas wPt-3475 mapped approximately 12 cM proximal to Xwmc44. Both flanking markers and P33/M59–474 appeared to be diagnostic as they were consistently absent and present in all highly resistant and highly susceptible DH lines, respectively (data not shown). QTLs for FHB resistance have been reported on chromosome 1B elsewhere in different populations (Buerstmayr et al., 2002; Shen et al., 2003; Lin et al., 2004; Paillard et al., 2004; Steiner et al., 2004; Zhang et al., 2004; Zhou et al., 2004; Schmolke et al., 2005), with an explained variance ranging from 5.5% in the Frontana/Remus population (Steiner et al., 2004) to 15.6% in the Wangshuibai/Alondra population (Zhang et al., 2004). Paillard et al. (2004) reported a minor QTL conferring FHB resistance from Forno in the interval between Xgwm268 and Xwmc44 on chromosome 1BL. Xwmc44 also has been reported to be located in the region containing Lr46, which confers resistance to leaf rust severity in wheat (Nelson et al., 1997; Singh et al., 1998; Suenaga et al., 2003). These results, together with ours, provided evidence that there is possibly a common locus on 1BL close to Xwmc44 for FHB and rust resistance.
The QTL on chromosome 7AL explained up to 14.8% of the phenotypic variation for mean FHB severity over 3 yr. Chromosome 7A was reported to confer FHB resistance using both monosomic analysis (Yu, 1991; Buerstmayr et al., 1999) and molecular markers (Zhou et al., 2002a, 2004). Based on evaluation of Type II FHB resistance of two sets of Chinese Spring x Sumai 3 chromosome substitution lines, Zhou et al. (2002a) reported that chromosome 7A from Sumai 3 had a larger effect on FHB resistance than any of the other 20 chromosomes. The same authors also detected a minor QTL on chromosome 7AL flanked by AFLP marker pAG/mCTGA149 and SSR marker Xgwm1083 (Zhou et al., 2004). The latter was located proximal to Xgwm276, which is the flanking marker for the QTL detected on 7AL in the present study, suggesting that these QTLs are identical. On chromosome 6BS, CIM identified a QTL for mean FHB severity 6 cM proximal to P46/M62–107 in the interval between P46/M62–107 and P45/M60–265 (Table 3). This QTL showed a minor effect and the LOD score was close to the threshold. As the use of CIM enhances the power for QTL detection in the background of major QTLs, we suggest that the QTL located in the interval of P46/M62–107 and P45/M60–265 is not a false-positive but that the effect of this locus on FHB resistance needs to be verified.
Given the heritability of 0.91, the QTLs detected in this study explained up to 49.1% of the total phenotypic variations for FHB severity on the basis of three year averages. The unexplained variation probably results from a combination of several factors, including undetected genes, epistasis, and experimental error. Paillard et al. (2004) hypothesized that the resistance in Arina is controlled by many QTLs with minor effects, which might have remained undetected with the currently available markers. The major QTL for FHB resistance from Arina was mapped on chromosome 1BL and the minor QTL on 6BS. Both QTLs from Arina in the present study were coincident with an unpublished reported by Draeger et al. (2004) but disagreed with Paillard et al. (2004), who reported 2 major QTLs (4AL and 6DL) derived from Arina in the Arina/Forno population. The microsatellite flanking markers associated with resistance in Arina in the Arina/Forno population were screened in the present study and the polymorphic markers on chromosomes 4A and 5A were mapped to the same position as Paillard et al. (2003) but none were associated with FHB resistance. Furthermore, the linkage map for chromosome 4A has more marker density in our study (47 markers and 2.8 cM/marker; Semagn et al., 2006) than the map for Arina/Forno (18 markers and 5.8 cM/marker; Paillard et al., 2003).
It is possible that some QTLs could have remained undetected due to incomplete coverage of chromosomes on the D genome. However, the absence of QTLs from Arina at least on chromosome 4A is highly unlikely to be caused by an incomplete coverage of the chromosome since map density in the ArNK is at least twice (Semagn et al., 2006) that of the Arina/Forno (Paillard et al., 2003). The ArNK population segregated for both winter and spring types but only the latter was used in this study because (i) FHB is a greater problem in Norwegian spring wheat than winter wheat, and (ii) it was not possible to successfully grow both spring and winter types in the same experiment under our conditions. It is evident that such exclusion of winter types from FHB evaluation affects mapping of vernalization genes but it is unlikely to contribute to the lack of coincident QTLs between the ArNK and Arina/Forno populations. The most likely reasons may be differences in population size, disease evaluation method, Fusarium species used, genetic background effects from the recipient parent, and/or genotype by environment interactions. The main constraint in the present study was the population size which was primarily due to difficulty with generating sufficient numbers of spring types in the segregating population. It also has been reported in other studies that modest numbers of individuals (up to 100) are sufficient for detecting most major QTLs (e.g., Lynch and Walsh, 1998; Somers et al., 2003). The ArNK population was evaluated for FHB severity by spraying each hill-plot independently with a macroconidial suspension of F. culmorum using hand sprayers and the plastic bag method of Mesterhazy (1995), whereas Paillard et al. (2004) sprayed F. graminearum using a motor driven backpack sprayer. We believe that the inoculation method used in our study is more precise as it enables inoculation of each genotype separately at the same stage of flowering compared to the simultaneous mass inoculation of all genotypes by Paillard et al. (2004). Although the correlations between resistance ratings using different Fusarium species have been moderate to high (0.61 r 0.97; P < 0.01; Mesterhazy et al., 2005), use of different Fusarium species might still have contributed to differences in the two studies. Furthermore, the number of genes segregating in mapping populations may also vary depending on genetic backgrounds and the magnitude of the difference in resistance between parents (Kolb et al., 2001).
Deoxynivalenol Content
We identified two major QTLs for DON content on chromosomes 1AL and 2AS, which together explained 34% of the variation. The QTL for DON content on chromosome 1AL overlapped with the QTL for FHB resistance (Table 3, Fig. 2), which could be due to tight linkage or pleiotropy. Bai et al. (2000) have shown that a QTL-conditioning low DON content is located at the same region as a gene controlling disease spreads but it was not known if they were pleiotropic or linked. According to Kolb et al. (2001), the presence of coincident QTLs caused by tight linkage can be determined by screening large numbers of recombinants to break up the linkage, which remains to be done in the ArNK population. Other research showed that the association between the two traits is low, especially in the field conditions (e.g., Mesterhazy et al., 1999), perhaps due to environmental variation or different mechanisms underlying different FHB resistant-expression.
QTL controlling DON content independent of FHB resistance was detected in the ArNK population on chromosome 2AS at the intervals between wPt-6148 and Xbarc124.1. The marker Xbarc124.1 has been mapped 13.5 cM proximal to Xgwm636 (Fig. 2) and the latter cosegregated with Xgwm614 in the Synthetic x Opata linkage map (Song et al., 2005). Xgwm614 also was found to be significantly associated with QTL for FHB resistance in recombinant inbred lines derived from Ning7840 x Clark (Zhou et al., 2002b). These QTLs are likely to be alleles of the same locus.
Future investigations will involve verification of the QTL regions for FHB resistance and DON content on chromosomes 1AL, 1BL, 2AS, and 7AL in an independent genetic background.
|
ACKNOWLEDGMENTS |
|---|
Received for publication February 16, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Liu, M. D. Hall, C. A. Griffey, and A. L. McKendry Meta-Analysis of QTL Associated with Fusarium Head Blight Resistance in Wheat Crop Sci., October 22, 2009; 49(6): 1955 - 1968. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Bonin and F. L. Kolb Resistance to Fusarium Head Blight and Kernel Damage in a Winter Wheat Recombinant Inbred Line Population Crop Sci., June 26, 2009; 49(4): 1304 - 1312. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. A. Abate, S. Liu, and A. L. McKendry Quantitative Trait Loci Associated with Deoxynivalenol Content and Kernel Quality in the Soft Red Winter Wheat 'Ernie' Crop Sci., July 1, 2008; 48(4): 1408 - 1418. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
