| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
a Bavarian State Research Center for Agriculture, Institute for Crop Science and Plant Breeding, Am Gereuth 8, 85354 Freising, Germany
b Technical Univ. of Munich, Chair of Plant Breeding, Am Hochanger 2, 85350 Freising-Weihenstephan, Germany
c Lochow-Petkus GmbH, Bollersener Weg 5, 29303 Bergen, Germany
* Corresponding author (Lorenz.Hartl{at}LfL.bayern.de).
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: AFLP, amplified fragment length polymorphism • BC, backcross • cM, centimorgan • DON, deoxynivalenol • EST, expressed sequence tags • FHB, Fusarium head blight • LOD, logarithm of odds • QTL, quantitative trait loci • RIL, recombinant inbred line • SSR, simple sequence repeat • STS, sequence tagged site
a Bavarian State Research Center for Agriculture, Institute for Crop Science and Plant Breeding, Am Gereuth 8, 85354 Freising, Germany
b Technical Univ. of Munich, Chair of Plant Breeding, Am Hochanger 2, 85350 Freising-Weihenstephan, Germany
c Lochow-Petkus GmbH, Bollersener Weg 5, 29303 Bergen, Germany
* Corresponding author (Lorenz.Hartl{at}LfL.bayern.de).
Many Fusarium head blight (FHB) [caused by Fusarium graminearum Schw. and F. culmorum (W.G. Sm.) Sacc.] quantitative trait loci (QTL) have been detected in both spring and winter wheat (Triticum aestivum L.). However, a QTL validation has not been performed in winter wheat. We report on the verification of resistance QTL in winter wheat, which had been previously mapped in a Dream/Lynx population. The QTL intervals of chromosomes 7BS, 6AL, and 2BL were enriched with amplified fragment length polymorphism (AFLP) markers. For a more precise estimation of the QTL effects and their influence on plant height and heading date, BC2F4 lines were created by marker-assisted selection and examined for FHB resistance. The phenotypic effects of the two main QTL on chromosomes 6AL and 7BS confirmed the previous results of the original mapping population. In the backcross population, QTL6AL and QTL7BS reduced FHB severity individually by 27% each relative to lines without resistance alleles. Both QTL had an effect on plant height, resulting in taller plants being more resistant. The combination of both QTL decreased disease severity most effectively (36%). The successfully validated QTL on chromosomes 6AL and 7BS are designated Qfhs.lfl-6AL and Qfhs.lfl-7BS, respectively. The best genotypes carried one or both of the major QTL, displaying their importance for disease resistance. In contrast to spring wheat, where QTL have been identified in an exotic genetic background, the resistance QTL validated in this study already reside in an adapted genetic background with excellent agronomic performance.
Abbreviations: AFLP, amplified fragment length polymorphism • BC, backcross • cM, centimorgan • DON, deoxynivalenol • EST, expressed sequence tags • FHB, Fusarium head blight • LOD, logarithm of odds • QTL, quantitative trait loci • RIL, recombinant inbred line • SSR, simple sequence repeat • STS, sequence tagged site
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The cultivation of resistant wheat cultivars is the most promising strategy in FHB control. However, FHB resistance breeding is a difficult task due to its quantitative inheritance (Snijders, 1990) and the high genotype x environment interactions (Miedaner et al., 2001). Therefore, marker-assisted selection could be an important tool to facilitate the selection of resistant cultivars and to enhance breeding efficiency. Schroeder and Christensen (1963) discriminated two types of FHB resistance, namely type I resistance against initial infection and type II resistance against fungal spread within the spike. By the use of spray inoculation to cause FHB in field trials, the combined effects of type I and type II resistance can be evaluated.
In recent years, quantitative trait loci (QTL) have been detected in several mapping populations of both spring and winter wheat. In spring wheat, Sumai 3 and its derivatives are the main resistance sources for the development of segregating populations. Together with several minor QTL, a major FHB resistance QTL was identified on chromosome 3BS (Qfhs.ndsu-3BS), which explained up to 60% of the phenotypic variation (R2) for type II resistance after single spikelet inoculation (Waldron et al., 1999; Bai et al., 1999; Anderson et al., 2001; Buerstmayr et al., 2002; Zhou et al., 2002), and about 20% of the phenotypic variation after spray inoculation in the field (Buerstmayr et al., 2003). Recently, this QTL has been characterized as the Fhb1 gene locus (Cuthbert et al., 2006) and fine mapped in a 1.2-cM marker interval flanked by two sequence tagged site (STS) markers (Liu et al., 2006).
In winter wheat, several chromosome regions could be associated with FHB resistance. In a population derived from a cross between Sincron (resistant) and F1054W (susceptible), a major resistance QTL could be detected on chromosome 1B (P < 0.001) (Ittu et al., 2000). In a Renan (resistant)/Recital (susceptible) population, main QTL were identified on chromosome 2BS and two on chromosome 5A (Gervais et al., 2003) each explaining between 9 and 14% of the phenotypic variation. Shen et al. (2003) analyzed a population derived from the cross between Fundulea 201R (resistant) and Patterson (susceptible), in which two main QTL could be found on chromosomes 1B and 3A with R2 of 19 and 13%, respectively. In another mapping population derived from the resistant cultivar Arina crossed with the susceptible cultivar Forno, chromosomes 4AL, 5BL, and 6DL were associated with FHB resistance each explaining between 10 and 22% of the phenotypic variation (Paillard et al., 2004). In addition to the above mentioned main QTL, several other chromosome regions with minor influence on FHB resistance were reported in the cited studies.
Recently, Schmolke et al. (2005) identified combined resistance QTL against initial infection and fungal spread within the spike in a Dream (resistant)/Lynx (susceptible) mapping population consisting of 145 recombinant inbred lines. Main QTL were located on chromosome 6AL explaining 19% and on chromosome 7BS explaining 21% of the phenotypic variation and are designated Qfhs.lfl-6AL and Qfhs.lfl-7BS throughout the paper. Qfhs.lfl-6AL overlapped with a QTL for plant height and Qfhs.lfl-7BS was associated with a minor effect on heading date. Quantitative trait loci of minor importance for FHB resistance could be detected on chromosomes 2BL and 1B. Because of the limited number of molecular markers in some QTL regions, the entire QTL intervals could not be characterized with molecular markers hampering the exact localization of the mentioned QTL.
Due to the limited population size and the segregating genetic background, there is an uncertainty in QTL estimation concerning localization and genetic effects, which might lead to an overestimation of the proportion of the explained phenotypic variation (Utz et al., 2000). Therefore, validation of QTL is generally recommended before using them in marker-assisted breeding programs. Although many QTL for FHB resistance have been detected in both spring and winter wheat, almost all validation and fine mapping studies refer to the major resistance QTL in spring wheat (Qfhs.ndsu-3BS) providing type II resistance. So far, no QTL validation study has been performed in winter wheat using adapted resistance donor lines.
In this study, we report on an experiment conducted to verify the presence and effects of QTL that had been previously mapped by Schmolke et al. (2005) in the winter wheat Dream/Lynx mapping population. The genetic map of the original Dream/Lynx mapping population had to be enriched with molecular markers in important linkage groups lacking flanking markers at one side of the QTL intervals. To get a better estimation of the QTL effects and their influence on plant height and heading date, BC2F4 lines were created by marker-assisted selection for the QTL of interest.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
FHB Field Trials and Inoculation
In 2005 field trials were performed at three locations: Freising, Seligenstadt, and Wohlde. The first two locations are in the south of Germany and the latter is situated in the north of Germany. The field experiment was arranged as a lattice design with two replications in each environment. Together with the selected genotypes, the parents Dream (5x) and Lynx (12x), as well as the FHB resistant cultivars Solitaer (2x) and Petrus (2x) were tested within each experiment. Each genotype was grown in two rows with a plot size of 0.6 m2.
The inoculation of plants was performed during flowering time using two highly aggressive F. culmorum isolates (FC33 and FC46), which were kindly provided by Th. Miedaner (University of Hohenheim). The spore suspension had a concentration of 5 x 105 spores mL–1 and was sprayed onto the plants with an automatic small plot sprayer (100 mL m–2). Due to differences in flowering date, all genotypes were spray-inoculated two to three times to inoculate all plants at least once at midanthesis. The evaluation of FHB severity was scored visually as percentage of infected spikelets per plot and started with the first visible symptoms 9 to 10 d after the last inoculation and was repeated two to three times at intervals of 3 to 5 d. In addition, heading date in days from January first was determined, as well as the mean plant height in centimeters per plot.
DNA Isolation, Bulked Segregant Analysis and Marker Analysis
The plant DNA was isolated from young green leaves using the CTAB method (Saghai Maroof et al., 1984). To enrich the QTL regions of interest with amplified fragment length polymorphism (AFLP) markers, several screenings were performed based on genotypic pools according to Giovannoni et al. (1991). Bulks of DNA from 6 to 10 genotypes of the original Dream/Lynx mapping population homozygous for opposing alleles for a desired chromosomal region were assorted and screened with around 100 AFLP primer combinations using the standard list for the AFLP primer nomenclature (http://wheat.pw.usda.gov/ggpages/keygeneAFLPs.html). The name of the AFLP markers consisted of the applied primer combination followed by the estimated fragment size in base pairs. Subsequently, AFLP markers showing expected polymorphisms between the genotypic pools were tested on the entire mapping population, which consisted of 145 RILs (Schmolke et al., 2005). In addition, 13 expressed sequence tags (EST)-derived simple sequence repeat (SSR) markers from Cornell University and Kansas State University were analyzed for the respective chromosome regions: CNL70, CNL76, CNL152, KSUM23, KSUM45, KSUM52, KSUM61, KSUM67, KSUM104, KSUM128, KSUM173, KSUM220, and KSUM247 (http://wheat.pw.usda.gov/ggpages/ITMI/EST-SSR/Cornell). For the QTL region on chromosome 7BS, seven additional SSR markers were analyzed: WMC426, WMC546, WMC606, WMC323, WMC182, WMC335, and BARC72 (http://wheat.pw.usda.gov).
Molecular markers used for selection of the QTL intervals in the final generation (BC2F4) were GWM82, GWM1011, and P69M51-175 for the QTL interval on chromosome 6AL, GWM46, BARC72, and P70M56-237 for the QTL interval on chromosome 7BS, P69M51-245, P69M62-295, and P67M52-120 for the QTL interval on chromosome 2BL (Table 1 ). The proprietary SSR GWM1011 was kindly provided by M. Röder (Institute of Plant Genetics and Crop Plant Research, Gatersleben).
|
Statistical Analysis
The existing genetic map of the Dream/Lynx mapping population was extended by additional AFLP markers resulting from the bulked segregant analyses using Mapmaker version 3.0B (Lander et al., 1987). The subsequent QTL analysis for the marker-enriched QTL intervals on chromosome 6AL, 7BS, and 2BL was recalculated with MultiQTL version 2.5 (Korol et al., 2005) using the multiple environment option.
The phenotypic data of the field trials was analyzed with Plabstat version 2N (Utz, 1995) to calculate the adjusted means for each environment and the broad-sense heritability according to Fehr (1987). The efficiency of the lattice design relative to a randomized block design was low (between 102 and 115%) indicating good reproducible phenotypic data. Thus, further analysis was performed with the SAS program version 9.1 (SAS Institute, 2004) using the individual values of each environment. Pearson correlation coefficients of FHB severity between the individual environments were calculated. Further statistical analysis showed that the residuals of the dependent variable did not follow a normal distribution, thus, a log transformation was performed to achieve normality of the data. The log transformed data was used for the subsequent analysis of variance. Analysis of variance was performed with the PROC GLM procedure to estimate sources of variance associated with FHB severity. The effects in the statistical model were locations, replication within location, BC2F1–derived families (considering the different pedigree of the tested lines), Qfhs.lfl-6AL, and Qfhs.lfl-7BS and interactions. Location, families, and their interactions were defined as random effects. Subsequently, a SCHEFFE test was conducted for multiple comparisons between the different marker classes with a probability value of P < 0.05 using means averaged across all three locations.
|
RESULTS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
The distributions of FHB infection phenotype within the four marker classes possessing susceptible (S) and/or resistant (R) alleles at Qfhs.lfl-6AL and Qfhs.lfl-7BS were relatively high (Fig. 3). Disease severity of lines with the Dream alleles alone at Qfhs.lfl-6AL were distributed over the whole range between the parents. The resistance alleles at the two QTL regions coming from the parent Dream were significantly associated with lower disease levels in comparison to the marker class with the alleles from the susceptible parent Lynx.
|
DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
To get a more precise genetic map for the QTL regions on chromosomes 6AL, 7BS, and 2BL, we have successfully enriched the existing map with additional molecular markers. This enabled us to select genotypes without recombination within the QTL regions and subsequently to get a more precise estimation of the QTL effects. In the original map, the QTL regions were located at the end of the linkage groups (Schmolke et al., 2005), which impaired the exact localization of the peak of the LOD curve. The authors concluded that the position of the resistance QTL at the end of the linkage groups, is the consequence of phenotypic selection during the breeding process of the resistant cultivar Dream resulting in low polymorphic regions next to the QTL regions. Nevertheless, all three chromosome regions mentioned above could be enriched with at least two AFLP markers using a pooling strategy. Additionally, the SSR marker BARC72 could be appended to the linkage group of chromosome 7BS. The lack of more polymorphic SSR markers for the desired genomic regions might be due to the fact that most of the ordered SSR markers were EST-derived SSRs, which generally show a lower degree of polymorphism. Furthermore, there were almost no additional published SSR markers for the regions of interest. Thus, the anonymous AFLP marker system provided a good tool to further enrich these genomic regions.
For a more precise estimation of the QTL effects and their influence on plant height and heading date, the development of lines with a homogeneous genetic background by backcross- and selfing generations combined with marker-assisted selection seems suitable to establish more homogeneous plants for field trials and to reduce influences of the genetic background. Thus, phenotypic effects of particular QTL can be estimated more precisely. For QTL validation and fine mapping studies in spring wheat, near-isogenic lines for specific QTL regions were developed (Zhou et al., 2003; Salameh and Buerstmayr, 2004; Reynolds and Anderson, 2004; Cuthbert et al., 2006). Another approach for the successful validation of Qfhs.ndsu-3BS was performed by Lemmens et al. (2005) who found a major QTL for DON resistance after application of the toxin in the ear in the same genomic region as Qfhs.ndsu-3BS.
In the current study, field trials were performed at three locations using spray inoculation, which resulted in high broad-sense heritability for FHB severity and high correlation coefficients of FHB severity between individual environments, which together indicated a good reproducibility of the phenotypic data. All field trials were performed in one year (2005), but other results from our group also showed high correlation coefficients and broad-sense heritabilities between years with the used inoculation technique. To further ensure the reproducibility of the phenotypic data in the current study, field trials were performed at different geographic locations of Germany which differ in mean annual temperature, mean annual precipitation, and height above sea level.
There is a risk in QTL analysis that effects might be overestimated or QTL localization might be inaccurate (Utz et al., 2000). However, the results of our study in the genetic background of the recurrent parent Lynx agreed well with the previous results of the original mapping population concerning the two main QTL (Table 3). Qfhs.lfl-6AL and Qfhs.lfl-7BS showed a coefficient of determination of 9% in the Dream/Lynx mapping population and 12% in the current study. The effects of Qfhs.lfl-6AL and Qfhs.lfl-7BS (absolute 7%; relative 16%) in the supplementary experiment mentioned above (data not shown), also lay in the same range as in the original mapping population. A significant QTL x QTL interaction was found showing the two resistance QTL (Qfhs.lfl-6AL and Qfhs.lfl-7BS) were not additive. A similar, but nonsignificant interaction of FHB resistance QTL, was observed in a pyramiding approach by Miedaner et al. (2006). The nonsignificant QTL x family interactions (Table 2) indicate that both QTL act independent of the background of the three BC2F1 families.
A strong influence of Qfhs.lfl-6AL on plant height was detected in the mapping population as well as in the backcross population. It might be that the improved FHB resistance is due to the morphological attribute of taller plant height. However, among the genotypes with QTL were some lines combining a good FHB resistance level with an acceptable plant height (data not shown), which indicates that concerning the mentioned QTL, FHB resistance and plant height might be inherited independently by tightly linked loci. This has been confirmed by Hilton et al. (1999) and Buerstmayr et al. (2000) who also found a negative correlation between FHB severity and plant height. Nevertheless, they were able to select genotypes combining a good resistance level with a relatively short plant height. Recently, McCartney et al. (2007) introgressed FHB resistance alleles of Nyuubai, Sumai 3, and Wuhan-1 into elite Canadian spring wheat germplasm and evaluated their effects on different traits (FHB index, DON, plant height, etc.). The most effective resistance allele on chromosome 4B coming from Wuhan-1 was associated with a 9.3-cm increase in plant height. The authors also assumed that this phenomenon was due to a tight linkage between two loci rather than to pleiotrophic effects. There was no effect of Qfhs.lfl-7BS on plant height in the Dream/Lynx mapping population, but the Dream alleles at Qfhs.lfl-7BS increased plant height 9 cm in the backcross population. In the mapping population, Qfhs.lfl-7BS had a minor influence on heading date, but this effect could not be observed in the backcross population. The minor QTL for FHB resistance on chromosome 2BL was not detected in the backcross population and may be masked by the strong effects of Qfhs.lfl-6AL and Qfhs.lfl-7BS. Our findings indicated that main QTL can be detected reliably even in small populations. Nevertheless, a verification of FHB resistance QTL is recommended prior to their use for marker-assisted selection, particularly when the individual effects of QTL are small. In doing so, QTL effects can be estimated more precisely and a good confidence is created that the QTL would be effective for FHB resistance in backcross breeding programs.
Only one genotype with Dream alleles at both QTL reached the resistance level of Dream indicating that additional unidentified genomic regions with moderate to small effects might be involved in FHB resistance in winter wheat. This is also supported by the findings of different research groups (Gervais et al., 2003; Shen et al., 2003; Paillard et al., 2004; Schmolke et al., 2005), suggesting a more polygenic inheritance of FHB resistance in European adapted winter wheat compared to spring wheat derived from Sumai 3 where FHB resistance is under control of a few major QTL like Qfhs.ndsu-3BS. Recently, this QTL has been fine mapped (Cuthbert et al., 2006; Liu et al., 2006) and redesignated as Fhb1, a major gene providing type II resistance against FHB. The higher number of genomic regions involved in FHB resistance in winter wheat lines reported to date, and the resulting lower effects of individual QTL compared to Fhb1 might explain that, so far, no validation and fine mapping studies have been performed in winter wheat. However, the validation and fine mapping of detected QTL is not only recommended for the application of marker-assisted selection in breeding programs, but also for a subsequent map-based cloning approach to finally identify the genes that are responsible for FHB resistance.
Phenotypic variation within marker classes was relatively high (Fig. 3). This might be due to additional unknown segregating loci for FHB resistance, since the population was founded on three different BC2F1 plants (Dream/4*Lynx), in which 12.5% of the loci were still segregating for the alleles of Dream. Recombination events between marker and QTL seem unlikely, because the QTL regions were characterized by molecular markers flanking both sides of the QTL peaks (Fig. 1). The best genotypes concerning FHB resistance carried one or both resistance alleles at the major QTL indicating their importance for good Fusarium resistance. Among the genotypes with QTL were some lines combining a good FHB resistance level with an acceptable plant height (data not shown). Possibly, these lines could be valuable as crossing parents in breeding programs.
In addition, the donor (Dream) of the resistance QTL Qfhs.lfl-6AL and Qfhs.lfl-7BS is characterized by very good baking quality and yield potential which almost reaches the best cultivars. Thus, penalties associated with linkage drag might not play an important role compared to breeding with exotic resistance sources. Interesting backcross lines have already been propagated for yield trials to further evaluate their potential as crossing parents in breeding programs. Pyramiding different FHB QTL offers the chance to further improve disease resistance. When using nonadapted material as a FHB resistance donor in backcross programs, marker-assisted selection on the QTL of interest should be accompanied by the monitoring of the fixation of the elite genetic background to minimize linkage drag (Somers et al., 2005). However, by using European adapted material as the FHB resistance source, the resistance level could be further improved without reducing the agronomic performance of lines.
Further investigations are underway to develop STS markers and to fine map the QTL region on chromosome 7BS to get linked markers, essential for the application of marker-assisted selection in breeding programs as well as for a subsequent map-based cloning approach.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Received for publication December 19, 2006.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. H. Sneller, P. Paul, and M. Guttieri Characterization of Resistance to Fusarium Head Blight in an Eastern U.S. Soft Red Winter Wheat Population Crop Sci., December 30, 2009; 50(1): 123 - 133. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
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]
|
|
|
|
|
|
T. Miedaner and H.-H. Voss Effect of Dwarfing Rht Genes on Fusarium Head Blight Resistance in Two Sets of Near-Isogenic Lines of Wheat and Check Cultivars Crop Sci., November 24, 2008; 48(6): 2115 - 2122. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
