Crop Science Illumina
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 2 December 2005
Published in Crop Sci 46:61-66 (2006)
© 2005 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Shen, X.
Right arrow Articles by Zhang, X.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Shen, X.
Right arrow Articles by Zhang, X.
Agricola
Right arrow Articles by Shen, X.
Right arrow Articles by Zhang, X.
Related Collections
Right arrow Cotton
Right arrow Crop Genetics

CROP BREEDING, GENETICS & CYTOLOGY

Mapping Fiber and Yield QTLs with Main, Epistatic, and QTL x Environment Interaction Effects in Recombinant Inbred Lines of Upland Cotton

Xinlian Shen, Tianzhen Zhang*, Wangzhen Guo, Xiefei Zhu and Xiaoyang Zhang

National Key Laboratory of Crop Genetics & Germplasm Enhancement, Cotton Research Institute, Nanjing Agricultural University, Nanjing 210095, China

* Correspondence author (cotton{at}njau.edu.cn)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Most agronomic traits of cotton (Gossypium hirsutum L.) are quantitatively inherited and affected by environment. The importance of epistasis as the genetic basis for complex traits has been reported in many crops. In this study, a linkage map was constructed by means of a recombinant inbred line (RIL) population derived from 7235xTM-1. Main effects, epistatic effects, and environmental interaction effects of quantitative trait loci (QTLs) controlling fiber and yield component traits were determined by mixed linear model based QTL mapping in two to four environments. Sixteen main effect QTLs for 12 traits and 19 digenic interaction QTLs for 12 traits were detected. The amount of variation explained by the QTLs of main effect was larger than the QTLs involved in epistatic interactions. Among seven QTLs having QTL x Environment (QE) interaction effects, five produced significant additive effects and two did not, but only one interaction showed significant additive x additive x environment (AAE) interaction effects. Some QTLs with large effects detected in specific environments could be associated with environmental response elements.

Abbreviations: QTL, quantitative trait loci • RIL, recombinant inbred line • SSR, simple sequence repeats


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
COTTON is an important cash crop in the world. Cotton fiber yield and quality traits of economic importance are typically quantitatively inherited and show continuous variation in segregating populations. A long-term challenge facing the cotton breeder is the simultaneous improvement of yield and fiber quality to meet the demands of the cotton producer as well as the textile industry. An understanding of the genetic control of fiber yield and quality traits is a prerequisite in a breeding program. In addition to additive effects, classical quantitative genetic studies have suggested the importance of epistasis and genotype x environment (GE) interaction (Miller et al., 1958; Meredith and Bridge, 1972; May and Green, 1994; May, 1999). McCarty et al. (2004) studied crosses of 14 high quality fiber germplasms and five cultivars and found that additive (A) and additive x additive (AA) epistatic effects for all agronomic and most fiber traits. Additive x environment (AE) and dominance x environments (DE) effects were detected for most traits, and significant AA x environments (AAE) interaction effects were detected for agronomic traits but not for fiber traits.

The advent of molecular marker techniques has resulted in the construction of molecular maps to facilitate the study of quantitatively inherited traits. Many studies have been performed to map QTL for fiber yield and quality traits in interspecific populations of G. hirsutum and G. barbadense or in intraspecific populations (Jiang et al., 1998; Shappley et al., 1998; Ulloa and Meredith, 2000; Kohel et al., 2001; Zhang et al., 2003; Paterson et al., 2003; Mei et al., 2004). Paterson et al. (2003) described the impact of environmental conditions on the genetic control of fiber quality. Seventeen QTLs were detected in a water-limited treatment while only two were specific to the well-watered treatment, suggesting that QTL for fiber quality were markedly affected both by general differences between growing seasons and by specific differences in water regimes. Few studies on QTL mapping in cotton have involved analyses of epistasis and QTL x environment (QE) interaction effect. Therefore, the estimated effects of QTLs may be biased. Zhu (1999) and Wang et al. (1999) proposed a new methodology for directly mapping QTLs with additive and digenic epistasis effects as well as their QE interaction based on mixed linear model approaches. Using this approach, a large number of studies have been conducted to identify epistasis effect and QE interaction effects for many important agronomic traits in other crops (Cao et al., 2001; Xing et al., 2002; Zhuang et al., 2002; Li et al., 2003). The results from these studies suggest epistatic effects and QE interaction effects play an important role in the genetic basis of quantitative traits.

A recombinant inbred line (RIL) population was developed from (7235 x TM-1) F2 (Zhang et al., 2003) individuals in Upland cotton in our laboratory. QTLs conditioning yield and fiber quality traits were analyzed by composite interval mapping (Shen, Guo, Zhu, Yuan, and Zhang. 2004, unpublished). The objective of the present study is to resolve further the genetic basis of yield and fiber quality traits into components of main-effect QTLs, epistatic QTLs, and QEs, and evaluate the relative magnitudes of these components.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Mapping population
A RIL population from the cross 7235xTM-1 (Zhang et al., 2003) was developed by bulk-selfing technique. Line 7235 is one of the G. anomalum Wawr. & Peyr. introgression germplasm lines developed by crossing G. anomalum and G. hirsutum and then backcrossing to cultivars and strains with high fiber strength such as Acala 3080 and PD4381 (Qian et al., 1992), which was kindly made available from Jiangsu Academy of Agricultural Sciences. The mean fiber strength of 7235 was 29.06 cN tex–1, fiber length 34.77 mm, and micronaire 3.97 when grown at Nanjing in 1998 and 1999. The fiber strength of TM-1 was 21.04 cN tex–1, length 28.81 mm, and micronaire 4.96. TM-1, the genetic standard for upland cotton (Kohel et al., 1970), was obtained from the USDA-ARS, Southern Plains Agriculture Research Center, College Station, TX. All of the traits evaluated in this study were based on F6:8 and F6:9 generations.

The parental lines and 258 one-row RILs were evaluated at Nanjing and Guanyun Counties, Jiangsu province, China, in 2002 and 2003. There are three main cotton growing areas presently in China. Nanjing is located in Yangtze River cotton-growing valley and Guanyun is located in the Yellow River cotton-growing valley. A randomized incomplete block design with two replications and single-row plots 5 m long and 0.8 m apart was used in the field trials. Fiber samples were tested at the Test Center of Cotton Quality, Henan province (HVI SPECTRUM), in 2002, and in the Supervision, Inspection and Test Center of Cotton Quality, Ministry of Agriculture in China (HVI900) in 2003. There are no apparent differences in the two machines used to measure fiber traits except for short fiber index (SFI). Fiber, which length is shorter than 16 and 12.8 mm, is regarded as short fiber in HVI SPECTRUN and HVI900, respectively. Fiber quality traits tested in 2002 included fiber length (FL), fiber strength (FS), micronaire (FMIC), fiber elongation (FE), fiber uniformity ratio (FUR), short fiber index (SFI), fiber maturity (FMAT), yellowness (FB), reflectance (FR), and spinning consistency index (FSCI). Agronomic traits evaluated included boll size (BS), lint percentage (LP), seed index (SI), bolls per plant (BN), seed cotton yield (SY), and lint yield (LY). Fiber quality traits tested in 2003 only included FL, FS, FMIC, FE, FUR, and SFI. Agronomic traits evaluated in 2003 were BS and LP. Experiments conducted at Nanjing and Guanyun in 2002 and 2003 were designated as Env1, Env2, Env3, and Env4, respectively.

Data Analysis
The additive and AA epistatic effects QTLs as well as QTL x environment interaction effects were analyzed by the mixed-model QTL mapping approach (Zhu, 1999; Wang et al., 1999) using QTL Mapper version 1.0. Background genetic variation (BGV) due to main and epistatic effects of important markers was controlled (Wang et al., 1999). A likelihood ratio value of 11.5, which is equal to a LOD score of 2.5 (Zeng and Weir, 1996), was used as a threshold for the detection of QTL or epistasis.

Assignment of Linkage Groups to Chromosomes
Assignment of linkage groups to subgenomes and chromosomes was made with monosomic and mono-telosomic stocks in a manner described by Stelly (1993). When no chromosome inference was available, the "A" or "D" designation of the linkage group was described from that of Lacape et al. (2003) and Rong et al. (2004) after aligning groups with common SSR loci. Three linkage groups being unable to be assigned to subgenomes and chromosomes were designated LG01, LG02, and LG03.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Trait performance and linkage map
The phenotypic values for fiber quality and yield traits of the RIL population and its parents are presented in Table 1. To determine if traits were normally distributed, skewness values were calculated for all traits. All traits except SFI in 2003 fit normal distributions in every environment, suggesting suitability for QTL analysis. The reason that SFI in 2003 was not normally distributed could be due to a different machine being used to test this fiber trait, which has a different setting criterion for short fiber. So SFI was not included in QTL analysis.


View this table:
[in this window]
[in a new window]
  		Table 1. Phenotypic values for fiber quality and yield traits of RIL and their parents across two or four environments.

 
Except for FUR, SFI, yellowness, FR, and SI, phenotypic differences between the two parents were significant (Table 1). The RIL lines showed large variation for all traits within each environment and transgressive segregation was observed across all environments.

All together, 1378 pairs of SSR primers were used to survey the polymorphism between TM-1 and 7235. Out these, 131 primer pairs produced 137 polymorphic loci between two parents. The genetic map for this RIL population was constructed based on these SSR markers. It contained 110 SSR markers distributed among 22 chromosomes and/or linkage groups covering 810.07 cM and was used for QTL mapping (Shen et al., 2004, unpublished). Twenty-seven loci remained unlinked.

Analysis of QTL with Additive Effects
A total of 23 QTLs with additive main effects (A) and/or additive x environment interaction effects (AE) were detected (Table 2). Out of 23 QTLs, 16 were found to be conditioned by significant additive effects and two exhibited significant AE effects, while five were significant for both additive and AE effects. QTLs with a large effect (>10% of the phenotypic variation) were found in at least two environments. Only seven QTLs showed consistency in all environments. Stability of QTL was trait-dependent. QTLs for seed index showed the best consistency. All QTLs for this trait could be found simultaneously in two environments.


View this table:
[in this window]
[in a new window]
  		Table 2. Additive(A) and additive x environment (AE) interaction effects of QTLs across two or four environments.

 
A major QTL for fiber strength on LGD03 explained 26.01% of phenotypic variance, which had previously been located on Chr.10 in a (7235xTM-1) F2 population with bulk segregate analysis (BSA) by interval mapping in our laboratory (Zhang et al., 2003). This was corrected in recent studies after aligning groups with common SSR loci reported by Lacape et al. (2003) and Rong et al. (2004) (Shen et al., 2005).

Although only seven QTLs showed significant QE interaction effect, all three cases of QE interaction were observed in this study.

  1. QTLs detected in one environment that showed QE interaction effects, such as the QTL for FS on Chr.25 and the QTL for FL on LGD02. The QE interaction effects of these two QTLs were larger than the main effects. The data indicated that the genetic effects of these two loci were cumulative effects of genetic main effects and QE interaction effects; however large QE effects affected their expression in different environments.
  2. QTL detected in all environments that showed significant QE interaction effects, such as QTL for boll size on LGD03 and QTL for seed index on LGD03. The QE interaction effects of these two QTLs were smaller than the corresponding main effects, indicating that the genetic effects of these two loci were mainly controlled by main effect.
  3. QTLs with no significant main effect that showed QE effect, such as QTL for BN on Chr.23 and QTL for LY on LGD08.

Analysis of QTL with Epistatic Effects
A total of 19 digenic epistatic interactions (AA) and/or epistasis x environment interaction effects (AAE) were detected for 12 traits (Table 3), suggesting that two-locus interaction was widespread in the entire genome. Fourteen and five AA interactions effects in favor of recombinants and the parental genotype combination, respectively, were detected. Only one significant AAE effect was detected.


View this table:
[in this window]
[in a new window]
  		Table 3. Epistatic effects of QTLs for fiber quality and yield traits.

 
The AA interaction detected included all three types of epistasis classified on basis of whether the QTLs involved exhibited main effects or not, namely interactions between QTLs, between QTL and background loci, or between complementary loci (Li, 1998). The majority of the interactions (70%) occurred between markers not linked to QTL affecting the same traits, indicating that these loci might play the role of modifying agents that tend to activate other loci or modify the action of other loci.

Variation explained by interaction effects ranged from 2.42 to 10.14%, which was lower than that of corresponding main effects. The importance of AA interaction effects in total genetic effects may be trait-dependent. For FUR, epistasis effects may be an important component of genetic variation. Three significant AA interaction effects for this trait between complementary loci were detected, explaining 23.67% of the total variation for PV. None of the main effect QTLs were detected in any environment. For fiber strength and seed index, the relative contribution of additive effects were larger than for AA interaction effects (36.14 and 3.6% for FS, 45.72 and 2.56% for SI, respectively), indicating fiber strength and seed index are mainly controlled by additive effects. For bolls per plant and seed cotton yield, the additive effects and AA interaction effects accounted for 21.52 and 15.93% for BN, 15.64 and 17.63% for SY of phenotypic variances, respectively.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
AA interaction and heterosis
The genetic basis of heterosis has been attributed to several factors: overdominance at a single locus (East, 1936), pseudo-overdominance or dominance complementation, and multilocus epistatic interactions (Allard, 1996). With the advent of QTL mapping, more and more findings suggest that interaction among different genetic loci leads to the production of heterosis (Allard, 1996; Li et al., 1997; Yu et al., 1997). Strong evidence for heterosis related to epistasis has been clearly demonstrated using a set recombinant lines from near-isogenic lines (subNIL) in tomato (Lycopersicon hirsutum Humb. & Bonpl., Monforte and Tanksley, 2000; Lin et al., 2000). One report rejected the single-locus overdominance hypothesis after the overdominance effect at a QTL locus was studied more in depth (Graham et al., 1997).

In the present study, the majority of AA interaction between alleles from different parents (recombinant types) resulted in better fiber quality and productivity (Table 3). Two AA interactions which involved four loci without detectable QTL additive main effects resulted in more BN. This finding supports the suggestion that epistasis in cotton may contribute to heterosis. In that case, one can fix the portion of heterosis owing to AA effects by use of marker-assisted selection (MAS). The MAS strategy would, therefore, be a promising approach to utilizing heterosis. Li et al. (1997) found that recombinant type interactions tend to result in reduced fitness in rice (Oryza sativa L.), suggesting that epistatic loci affecting grain yield components act in a predominantly complementary manner, which is the result of strong selection for adaptation conditioned by epistatic gene complexes. A high outcrossing rate in cotton makes it possible that different loci from different parents were pyramided to adapt to changing environments during population evolution. The fact that the interaction detected largely included loci without detectable QTL additive effects further confirmed the importance of epistasis in cotton breeding. Selection for increased trait values must also pay attention to the best multilocus gene combinations as well as major genes.

AA interaction is likely greatly affected by environments because interaction between the same loci could not be identified at more than one environment (data not shown); however, there was only one significant AAE effect. The cause might be that minor AAE effects cannot be detected because of the limited power of statistical methods or the LOD threshold we set to detect QTLs. Another cause might be that the sample of environments involved in this study was insufficient to distinguish genetic effects from nongenetic effects. Another hypothesis is that epistatic loci for the same trait are indistinguishable under different environments and need to regulate interaction between different loci to adapt to changing environments, or the AAE interaction may be relatively unimportant in this population.

About QE Interaction
Higher plants are dynamic living systems in which change occurs constantly from germination to maturity. The pattern of change is rarely the same from genotype to genotype in one environment or for a single genotype grown in different environments (Allard and Bradshaw, 1964). One of the major goals for plant breeders is to develop genotypes with high yield and superior quality potential and the ability that are stable across environments. This is particularly true for cotton breeders in China because there are three main cotton growing areas which exhibit large differences in environmental conditions including rainfall, temperature, daylength, and soil type leading to large genotype x environment interactions. There are two main ways in which a cultivar can achieve stability. First, identification of the nonenvironment-specific QTLs or QTLs with minor QE interaction effects should be particularly useful in MAS manipulation of fiber yield and quality. Second, development of widely adapted cultivars by pyramiding different QTLs each adapted to a different range of environments. Such varieties should adjust their genotypic or phenotypic state in response to changing environment.

The fundamental way to evaluate stability is to investigate the complex causes of QTL–environment interaction at the molecular level. Interest in regulation of gene action has received great stimulus recently with increased cloning of genes. Several reports revealed that environmental response elements play important roles in gene expression. Dolferus et al. (1994) working with the Arabidopsis Adh gene found that upstream sequences differed in their effects on the expression of the Adh promoter under three environmental stress conditions: hypoxia, low temperature, and drought. Deletion mapping revealed a critical region was essential for expression of the Adh promoter under all three environmental stresses. Detection of such regions will be very significant to stably regulate gene expression. In this study, some QTLs with large effects such as fiber maturity on LGD03, which could only be detected in specific environments, could be associated with some environmental response elements. Research on gene regulation appears to offer a great opportunity to understand QE interaction, and ultimately, to produce cultivars that will cope with unpredictable fluctuations in environment.


   ACKNOWLEDGMENTS
 
We thank two anonymous reviewers for their comments. This program was financially supported by the Key Project of Chinese Ministry of Education (10418), IAEA (12846), National High-tech Program (2004AA211172) and National Science Foundation in China (30070483, 30270806).

Received for publication January 18, 2005.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Shen, X.
Right arrow Articles by Zhang, X.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Shen, X.
Right arrow Articles by Zhang, X.
Agricola
Right arrow Articles by Shen, X.
Right arrow Articles by Zhang, X.
Related Collections
Right arrow Cotton
Right arrow Crop Genetics

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