Crop Science Grow Your Career with CSSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
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 HighWire
Right arrow Citing Articles via ISI Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bushman, B. S.
Right arrow Articles by McMullen, M. D.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Bushman, B. S.
Right arrow Articles by McMullen, M. D.
Agricola
Right arrow Articles by Bushman, B. S.
Right arrow Articles by McMullen, M. D.
Crop Science 42:1669-1678 (2002)
© 2002 Crop Science Society of America

CELL BIOLOGY & MOLECULAR GENETICS

Two Loci Exert Major Effects on Chlorogenic Acid Synthesis in Maize Silks

Bradley S. Bushmana, Maurice E. Snookc, Justin P. Gerkeb, Stephen J. Szalmaa, Mark A. Berhowd, Katherine E. Houchinse and Michael D. McMullen*,e

a Genetics Area Program, Univ. of Missouri-Columbia, Columbia, MO 65211
b Biological Sciences Unit, Univ. of Missouri-Columbia, Columbia, MO 65211
c Richard B. Russell Research Center, Univ. of Georgia, Athens, GA 31793
d USDA-ARS National Center for Agricultural Utilization Research, Peoria, IL 61604
e USDA-ARS, Plant Genetics Research Unit, and Plant Science Unit, Univ. of Missouri, Columbia, MO 65211

* Corresponding author (mcmullenm{at}missouri.edu)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Chlorogenic acid (CGA) in maize (Zea mays L.) silks has been implicated in resistance to corn earworm (Helicoverpa zea Boddie) with a mechanism similar to the flavone, maysin. However, the genetic basis of CGA synthesis is poorly understood. Our goal was to identify candidate loci affecting the biosynthesis of CGA using quantitative trait locus (QTL) analyses on three F2 populations: (A619 x Mp708)F2, (A619 x Mo6)F2, and (Mo6 x Mp708)F2. Chlorogenic acid and flavone contents in silks were measured, linkage maps generated, and significant loci identified with composite interval mapping (CIM) and ANOVA multiple-effects models for the three populations. Of the QTLs detected, two exerted major effects; one corresponding to the p1 locus, and the other a novel locus we named qtl2. The main effect of the functional allele at the p1 locus was to increase both CGA and flavones, while the positive allele at the qtl2 locus only increased CGA. An epistatic interaction between p1 and qtl2 alleles in the (A619 x Mo6)F2 population caused an increase in CGA with a concomitant decrease in flavones. The rare ability of Mo6 and Mp708 to synthesize CGA in excess of flavones is primarily a result from the effects at the qtl2 locus.

Abbreviations: 4CL, 4-coumaroylCoA ligase • C3H, coumarate-3-hydroxylase • C4H, cinnamate-4-hydroxylase • CGA, chlorogenic acid • CHS, chalcone synthase • CIM, composite interval mapping • CQT, coumaroylCoA quinate hydroxycinnamoyl transferase • LOD, log-odds ratio • PAL, phenylalanine ammonia lyase • QTL, quantitative trait locus • SSR, simple sequence repeat.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
CHLOROGENIC ACID is a secondary metabolite synthesized in many plants (Clifford, 1999; Herrmann, 1989). Chlorogenic acid is implicated in free radical scavenging (Chen and Ho, 1997), inhibition of lipid peroxidation (Ohnishi et al., 1994), enzymatic browning of fruits and vegetables (Walker, 1995), antifungal activity (Maher et al., 1994), and host-plant resistance against insects (Dowd and Vega, 1996; Isman and Duffey, 1982). Of particular interest to our group are the activities of CGA on corn earworm. The corn earworm is a polyphageous lepidopteran pest that reduces fertility of maize and provides entry points for ear rotting fungi and other pathogens by feeding on silks and kernels. Host-plant resistance to corn earworm was detected in maize germplasm with high levels of maysin, a C-glycosylflavone (Elliger et al., 1980a), and further analysis of maysin and other flavones suggested a requirement for adjacent phenolic hydroxyl groups in the molecule to confer resistance (Elliger et al., 1980b). Upon ingestion by corn earworm larvae, these adjacent hydroxyls are oxidized to quinones that sequester available amino acids from digestion, cause oxidative stress, reduce larval weight, and lengthen days to pupation (Hurrell et al., 1982; Summers and Felton, 1994; Wiseman and Carpenter, 1995). Chlorogenic acid has a similar adjacent hydroxyl ring structure as maysin, and has also been implicated in growth-reduction of lepidopterans (Duffey and Stout, 1996; Isman and Duffey, 1982).

Chlorogenic acid is a product of the phenylpropanoid biochemical pathway, and shares substrates with lignin and flavonoid branch pathways (Fig. 1) . The first step in the phenylpropanoid pathway involves the deamination of phenylalanine to cinnamate via the enzyme phenylalanine ammonia lyase (PAL). Phenylalanine ammonia lyase RNA levels respond to several regulatory signals (Costa e Silva et al., 1993; Christie et al., 1994), and were reported to be a main factor affecting CGA levels in transgenic systems (Bate et al., 1994). Synthesis of CGA from cinnamate requires four enzymes to complete two hydroxylations on the aromatic ring and a conjugation of the hydroxycinnamate and quinate moieties: cinnamate-4-hydroxylase (C4H), 4-coumaroyl-CoA ligase (4CL), coumaroyl-CoA:quinate hydroxycinnamoyltransferase (CQT), and coumarate/coumaroylquinate-3-hydroxylase (C3H). The order of events can vary among plant systems, but previous studies have suggested the pathways shown in Fig. 1 (Kuhnl et al., 1987; Rhodes and Wooltorton, 1976; Ulbrich and Zenk, 1979).



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 1. The general phenylpropanoid biochemical pathway and proposed steps leading to CGA. Enzymes are in black caps and genes in italics. PAL, phenylalanine ammonia lyase, C4H, cinnamate-4-hydroxylase; C3H, coumarate-3-hydroxylase; 4CL, 4-coumaroylCoA ligase; CQT, coumaroylCoA:quinate hydroxycinnamoyl transferase; CHS, chalcone synthase. This sequence of events is based on preferred substrates and time-course studies in other plants, and branchpoints into lignin and flavonoid biosynthesis are shown.

 
Our objective was to identify loci affecting the biosynthesis of CGA using QTL analyses, and identify candidate genes from the phenylpropanoid biosynthetic pathway that could be responsible for those loci. Previous QTL studies have detected loci and genes with major effects for maysin (McMullen et al., 1998), but populations specifically designed to address the synthesis of CGA have not been analyzed. We utilized three F2 populations from three maize inbred lines allowing a circular comparison of QTLs. The inbred parents of two populations differ in their ability to synthesize CGA in the silk tissue, and the third population was constructed from two high CGA lines. Linkage disequilibrium and practical limitations on population sizes generate QTL confidence intervals of approximately 10 centimorgans (cM) (Darvasi et al., 1993). Chromosomal regions of this map length in maize contain approximately 300 genes complicating identification of specific genes controlling quantitative traits. However, information available about the enzyme steps in the phenylpropanoid pathway necessary for CGA synthesis, and map locations of several genes involved, provide plausible and testable candidate genes for some QTLs.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Mapping Populations
The F2 populations were generated from three maize inbred lines: A619, Mo6, and Mp708. Mp708 and Mo6 synthesize high amounts of CGA (Fig. 2) , in excess of maysin, a C-glycosyl flavone typically the most abundant, extractable phenylpropanoid product in silks of most field maize varieties. Mo6 was developed from the Tennessee golden prolific line T8 at the Missouri Agricultural Experiment Station, Columbia, MO. Mo6 synthesizes high levels of CGA and moderate amounts of isoorientin, a C-glucosylflavone. Mp708 was derived from selfed selections for resistance to southwestern corn borer (Diatraea grandiosella Dyar) and fall armyworm (Spodoptera frugiperda Smith) from the cross of Mp464 x Tx601 at the Mississippi Agriculture and Forestry Experiment Station, Starkville, MS. Mp708 produces high amounts of CGA and moderate amounts of maysin, a C-glycosyl flavone. A619 was developed from (A171 x Oh43) x Oh43 at the Minnesota Agricultural Experiment Station, St. Paul, MN. A619 accumulates negligible levels of CGA and flavones in the silk tissue (Fig. 2). The populations (A619 x Mp708)F2, (A619 x Mo6)F2, (Mo6 x Mp708)F2 included 427, 348, and 190 individuals, respectively. Each F2 population was derived from two F1 ears, and was grown at the University of Missouri Agronomy Research Center, Columbia, MO, along with parent and F1 plants. The population (A619 x Mp708)F2 was grown during the summer of 1998 while the other two populations were grown in 1999. All three populations were grown under standard agronomic practices for our region.



View larger version (32K):
[in this window]
[in a new window]
  		Fig. 2. Distribution of CGA and flavone values from individuals of A: (A619 x Mp708)F2, B: (A619 x Mo6)F2, and C: (Mo6 x Mp708)F2 populations. The percent fresh silk weight of each chemical (CGA or flavones) was measured with reverse-phase HPLC. Parental and F1 values are shown for each population, and represent the means of 5-15 individuals.

 
Phenotypic Analysis
For all populations, primary ear shoots were covered before silk emergence. Silks were collected two days after emergence, weighed, and lyophilized. Lyophilized silks were analyzed for flavone and CGA contents by reverse-phase HPLC (Snook et al., 1989, 1993). Parental and F1 phenotype values represent a mean of 5 to 15 individuals. Phenotype analyses of (A619 x Mo6)F2 and (Mo6 x Mp708)F2 populations were conducted at the Richard B. Russell Research Center, University of Georgia, Athens, GA, and from (A619 x Mp708)F2 at the National Center for Agricultural Utilization Research, USDA-ARS, Peoria, IL. The amount of CGA and flavones were determined and reported as the percent fresh weight of the silk.

Genotype Analysis
Leaf tissue was harvested from parental, F1, and F2 plants at the mid-whorl stage, lyophilized, and DNA was extracted from lyophilized tissue. Genotypes were determined with simple sequence repeat (SSR) markers. The DNA extraction procedures and SSR reaction conditions are available at MaizeDB, www.agron.missouri.edu/ssr.html, verified April 3, 2002. Amplification products were resolved on 4 to 5% (w/v) agarose gels at 115 V, and viewed with the Eagle-Eye uv detection system (Stratagene, La Jolla, CA). In some genome areas where polymorphism was limited and agarose gels could not resolve differences, fluorescently-labelled SSR primers were used to generate polymorphic PCR products that were analyzed by the ABI3700 (PE Biosystems, Foster City, CA) DNA sequencer system. Amplification products were diluted and resolved through capillary electrophoresis and the size of the SSR product determined relative to an internal molecular weight marker, as analyzed with GENESCAN software (PE Biosystems, Foster City, CA).

Statistical Analysis
Linkage maps were generated by MAPMAKER/EXP, version 3.0 (Whitehead Institute, Cambridge, MA), on the basis of a minimum LOD score of 3.0 and a maximum linkage distance of 50 cM. Segregation distortion was tested by Chi-square analysis with P < 0.01. The normality of F2 population distributions was tested by the Shapiro-Wilk procedure of PROC UNIVARIATE, SAS system (SAS Inst., 1999). Marker-trait associations affecting CGA, isoorientin, maysin, and total flavone (maysin + isoorientin) were detected by composite interval mapping (CIM) and multiple-effects ANOVA models. Composite interval mapping was performed with QTL Cartographer, Version 1.14a (Basten et al., 1994, 2000), and markers used as parameters were selected from a forward/backward stepwise regression with thresholds for inclusion and exclusion at P <= 0.01. For each population and trait, we determined empirical threshold levels for declaring QTL significance at a genome-wise Type I error rate of 0.05 (Churchill and Doerge, 1994). Positional relationship of QTLs from different populations was based on location and genetic distance between markers shared by the populations. Multiple effect models were generated with PROC GLM in SAS from markers significant in stepwise regression, and from digenic interactions significant in EPISTACY, a two-way ANOVA procedure (Holland, 1998). Markers with main effects were initially identified by stepwise forward/backward selection (P <= 0.01), and inclusion in the final model required a P <= 0.05 for the Type III sums of squares (SAS Inst., 1999). All digenic interactions for markers were initially identified with EPISTACY (P <= 0.001), and resulting significant interactions were included in the final model if P <= 0.05 for the Type III sums of squares. Genotype class means were obtained through PROC MEANS using the CLASSES statement of SAS software.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Phenotypic Data
Both Mo6 and Mp708 synthesized large amounts of CGA, in excess of flavones (Fig. 2). Mo6 accumulated isoorientin instead of maysin; but when crossed to A619 or Mp708, isoorientin was replaced by maysin in the F1. To simplify analysis, we have summed isoorientin and maysin, and reported that sum as flavones. A619 synthesized negligible amounts of both CGA and flavones. In the (A619 x Mp708)F2 and (A619 x Mo6)F2 populations both CGA and flavone distributions exhibited transgressive segregation tails of high values (Fig. 2). Additionally, approximately one third of the (A619 x Mo6)F2 progeny synthesized no flavones, resulting in a bimodal distribution for that trait. The (Mo6 x Mp708)F2 population distribution for CGA was normal (P = 0.2954), but displayed a nonnormal, broad distribution of values for flavones. The phenotype means of CGA and flavones in this population were approximately equivalent to the Mp708 value, and transgressive segregant plants exceeded parental CGA and flavone values up to two-fold.

Linkage Maps
Framework maps of 10 linkage groups were generated for all three populations. For the (A619 x Mp708)F2 population, 91 SSR markers were anchored onto a map for a total genetic distance of 1570 cM. For the (A619 x Mo6)F2 population, 93 SSR markers were anchored onto a map representing a total genetic distance of 1486 cM. For the (Mo6 x Mp708)F2 population, 91 SSR markers were anchored onto a map representing a total genetic distance of 1353 cM. Polymorphism on the tip of the long arm of chromosome 7 in the (Mo6 x Mp708)F2 population could not be detected. In this population the marker bnlg1036 was located within 9.7 cM of phi127, but the SSR consensus map places these two markers in bins 2.05 and 2.08, respectively, approximately 30 cM apart. Except for this locus, marker order and approximate distances in all three populations was consistent with the SSR consensus maps in MaizeDB, www.agron.missouri.edu. Marker loci spanning bins 4.00-4.03 showed segregation distortion in the (A619 x Mp708)F2 population and markers from 4.00-4.07 were distorted in the (Mo6 x Mp708)F2 population. The genetic maps for all three populations will be made available on the MaizeDB.

QTL Detection
(A619 x Mp708)F2
For CGA synthesis, two QTLs were detected by CIM (Table 1) . The QTL with the largest effects was located on chromosome 1 in bin 1.03 near umc1021. The 10 maize chromosomes are subdivided into bins spaced approximately 20 cM apart, with the bin denoted on the right-hand side of a decimal and the chromosome on the left-hand side (Gardiner et al., 1993; Davis et al., 1999). Quantitative trait loci confidence intervals generally span 10 to 20 cM, so we describe QTL location by bin. We ascribed the QTL detected in bin 1.03 to the pericarp color1 (p1) gene from Mp708, which has a p1-wwb (white pericarp, white cob, browning silks) allele. The second major QTL was located near marker umc1065 in bin 2.06, and will be referred to as qtl2 in this report because no candidate gene from the pathway is known. The qtl2 locus explained 7.9% of the phenotypic variation, increasing CGA levels by approximately 50% (Fig. 3) . An ANOVA multiple-effects model identified five markers as main effects and three digenic interactions, accounting for 43.8% of the total phenotypic variation (Table 2) . Both QTLs detected from CIM were included as main effects, as were markers in bins 4.08, 5.00, and 6.01. bnlg381 and phi036 were not significant as main effects, but were included in significant interactions.


View this table:
[in this window]
[in a new window]
  		Table 1. Composite interval mapping location, significance, effects, and candidate genes of QTLs for CGA and flavones from three maize populations.

 


View larger version (36K):
[in this window]
[in a new window]
  		Fig. 3. Genotype class means for CGA and flavones at the marker genotype nearest the qtl2 locus in A: (A619 x Mp708)F2 and B: (A619 x Mo6)F2 populations. Genotype A, homozygous A619 alleles; H, heterozygous; and B, homozygous Mp708 or Mo6 alleles depending upon the population. Standard error bars are taken to the second standard deviation.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Multiple-effects models for CGA and flavone accumulation in maize silks from the three populations used in this study.

 
Four QTLs were detected for flavone synthesis by CIM. The p1 locus near umc1021 is a candidate for the effects of the QTL in bin 1.03, and accounted for 36.3% of the phenotypic variation. The recessive enhancer of maysin (rem1) candidate gene was located near the QTL in bin 9.02 (Byrne et al., 1996), but we have no candidates for the QTLs in bins 4.01 and 7.03. The QTL in bin 4.01 was also flanked by markers exhibiting segregation distortion. A multiple-effects model for flavone synthesis of six main effects and two digenic interactions explained 57.1% of the phenotypic variation. All four QTLs identified from CIM, and a locus in bin 6.01, were included as main effects.

(A619 x Mo6)F2
Several QTLs were detected for CGA in this population by CIM, including two analogous to those identified in (A619 x Mp708)F2. The QTL in bin 1.03 included the locus for the p1-wrb (white pericarp, red cob, browning silks) allele from Mo6, and accounted for 13% of the phenotypic variation. The QTL in bin 2.06 had the largest effect in this population, explaining 22.7% of the variation for CGA, and its confidence interval was consistent with that of the qtl2 locus in the first population. Within bin 5.05, the QTL confidence interval included a putative phenylalanine ammonia lyase1 (pal1) gene location. For the QTLs detected in bins 3.05 and 4.06, we could assign no candidate genes. Interestingly, A619 alleles increased CGA at these QTLs. A multiple-effects model for CGA from this population included 10 markers as main effects and six digenic interactions, representing 71.5% of the variation. Nearly all QTLs detected by CIM, and markers in bins 1.01, 2.08, 4.09, 5.01, and 10.01 were included in the model. The QTL in bin 4.06 was not a main effect despite its moderately large effects on the phenotype detected with CIM. Of the six digenic interactions, three included the qtl2 locus, including the interaction of p1 and qtl2.

Six QTLs were detected for flavone synthesis by CIM. The p1 locus is a candidate for the QTL in bin 1.02, and accounted for half of the phenotypic variance. Unlike Mp708, two QTLs for flavones were detected on chromosome 2 in this population. The location of the QTL in bin 2.06 was consistent with the location of the qtl2 locus, and the location and effects of the QTL in bin 2.08 was consistent with the action at the salmon silk2 (sm2) locus. A locus near umc1037 in bin 9.02 was also detected in this population in a similar location to the analogous QTL found in the first population. No plausible candidate gene was identified for the QTLs in bins 1.10 and 5.07, where A619 alleles increased flavones. The multiple-effects model for flavone synthesis included 11 markers and three digenic interactions, explaining 69% of the variation. Along with the six QTLs, markers in bins 2.02, 4.07, 7.02, 7.04, and 10.04 were also significant.

(Mo6 x Mp708)F2
This cross involved two high CGA producing parents and completed a three-way circle of F2 populations. For CGA synthesis, neither the p1 nor qtl2 locus, which exerted the largest effects in the two predecessor populations, was significant in this population. The location of the QTL detected on chromosome 2 in this population was consistent with the sm2 location, rather than the location of qtl2. The colorless2 (c2) gene, encoding the enzyme chalcone synthase, was located within the confidence interval of the QTL in bin 4.08, but markers in that region exhibited segregation distortion. A multiple-effects model in this population included five markers and one digenic interaction to account for 70.7% of the phenotypic variation. Quantitative trait loci detected through CIM in bins 2.08, 3.06, 4.08, and 7.05 were also significant main effects, as was a marker in bin 5.06.

For flavone synthesis, four QTLs were detected by CIM. The location of the most significant QTL, in bin 2.08, was the same position as the sm2 candidate locus for CGA. No candidate genes could be identified for the QTLs in bins 1.09, 3.06, or 6.01. A multiple-effects model for flavones included five main effect markers and two digenic interactions to explain 83.5% of the phenotypic variation. Interestingly, only QTLs in bins 2.08 and 6.01 detected by CIM were also significant in multiple-effects models, while markers in bins 1.06, 3.03, and 8.05 were specific for the latter.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
The p1 Locus
In the (A619 x Mp708)F2 and (A619 x Mo6)F2 populations, the p1 locus showed large effects on both CGA and flavones (Table 1). Mp708 and Mo6 each contained different p1 alleles that expressed in silks, p1-wwb and p1-wrb, respectively. The A619 (p1-www) allele is nonfunctional for silk expression. P1 is a Myb-domain transcription factor known to activate genes in the flavonoid biochemical pathway downstream of CGA synthesis (Grotewold et al., 1994), but a basis for its effects on CGA is unclear. Prior observations reported an increase in phenylpropanoid products and several-fold induction of expression of phenylpropanoid pathway genes in transgenic p1-expressing cultures (Grotewold et al., 1998; Bruce et al., 2000), consistent with p1's candidacy for a QTL for CGA synthesis. Additionally, a QTL for CGA in the p1 region was noticed previously in the (GT114 x GT119)F2 population (Byrne et al., 1996), and has since been detected in several other populations (Szalma et al., 2002, this issue; Guo et al., 2001).

The p1 alleles might affect CGA through induction of flavonoid specific genes that increase the flow of substrates toward flavone synthesis, and thus indirectly increase CGA. Within the (A619 x Mp708)F2 and (A619 x Mo6)F2 populations functional p1 alleles increased flavones in an additive manner, but their effects on CGA were dominant (Table 1). This dichotomous gene action might have resulted if flavonoid enzymes reacted in a steady-state manner as substrate flow increased, while the branch toward CGA was quickly saturated from the effects of one p1 allele. A question remains whether the P1 protein might also induce transcription of genes in the phenylpropanoid pathway. In a (W23 c2/whp1/in1/p1-www x Mp708)F2 population, individuals with functional p1 alleles showed increased CGA levels in the absence of flavone-producing chalcone synthase activity (Szalma et al., 2002). Since p1 still affected CGA in the absence of flavone production, its effects on CGA can be separated from flavone synthesis. This is supported by transgenic p1 induction of the phenylpropanoid gene pal (Bruce et al., 2000), which was implicated as the major enzyme affecting CGA in tobacco (Bate et al., 1994). The action of p1 on CGA accumulation in maize silk likely results from both inducing genes directly involved in CGA synthesis and increasing substrate flow indirectly by induction of flavonoid genes.

The enzyme steps involved in CGA synthesis must be coordinately regulated along with lignins, flavonoids, and other phenylpropanoid branch pathways to match the varying needs of the plant. Myb-domain transcription factors C1 and P1 each induce specific flavonoid and anthocyanin genes in maize, and Myb-domain gene products in several other plants induced transcription of phenylpropanoid and lignin genes (Martin and Paz-Ares, 1997). The strong effect of Myb-domain proteins on target genes is often reflected in QTL analyses, and was seen for p1 in the current study. Approximately half of all digenic epistatic interactions in the multiple-effects models of the (A619 x Mp708)F2 and (A619 x Mo6)F2 populations included the p1 locus as one of the interacting markers (Table 2), and the other locus invariably depended on functional p1 alleles to affect the trait (flavone or CGA). Thus, transcription factors should be the first, and most easily detectable, targets in manipulating traits involving pathways (Grotewold et al., 1998).

The qtl2 Locus
Within bin 2.06, we detected a QTL with large effects on CGA in the (A619 x Mp708)F2 and (A619 x Mo6)F2 populations that we named qtl2. This QTL was located near umc1065 in bin 2.06, and confirmed in other populations utilizing Mp708 as a parent (Szalma et al., 2002). Mo6-derived populations also segregated for a nonfunctional sm2 allele that was located in bin 2.08, linked to qtl2, and identified as a significant locus for flavone synthesis. The sm2 mutation in Mo6 associates with an accumulation of isoorientin, a C-glucosylflavone lacking the rhamnosyl moiety of maysin (McMullen, unpublished data). The maximum likelihood position of the sm2 QTL affecting maysin in the (A619 x Mo6)F2 population was approximately 20 cM distal to the qtl2 affecting CGA. Additionally, in the (A619 x Mo6)F2 population a QTL affecting flavone was detected near the putative qtl2 location, again 20 cM apart from sm2 (Table 1). In the (Mo6 x Mp708)F2 population, a QTL was located in bin 2.08 affecting both CGA and flavones. As (Mo6 x Mp708)F2 likely did not segregate for qtl2, but did segregate for sm2, the effects at that region were attributed to sm2. Thus, although the sm2 locus was linked to qtl2, and slightly affected CGA synthesis in the (Mo6 x Mp708)F2 population, it was distinct from the gene causing the qtl2 locus effects.

Mo6 and Mp708 exhibit a rare ability to synthesize more CGA than flavones. This ability is not a result of p1 function, which is present and segregating in many populations that do not synthesize appreciable amounts of CGA. Rather, it results primarily from the effects of the qtl2 locus, which to our knowledge has only been detected in lines that synthesize significant amounts of CGA. On the basis of genotype means, qtl2 had the ability to increase CGA without significant changes in flavones (Fig. 3). However, when qtl2 class means were subdivided into p1 alleles, the increase in CGA associated with qtl2 was concomitant with a decrease in flavones in the presence of a functional p1 allele classes in the (A619 x Mo6)F2 population (Fig. 4) . This trend was also seen in the (A619 x Mp708)F2 population. Interestingly, without functional p1 alleles the effects of the qtl2 locus on CGA were negligible, and without qtl2 the effects of p1 on CGA were negligible. It seems that as the p1 gene product turns on the phenylpropanoid and flavonoid pathways, qtl2 diverts substrates toward specific portions of the CGA biosynthetic branch.



View larger version (38K):
[in this window]
[in a new window]
  		Fig. 4. Genotype class means for CGA and flavones of markers nearest to p1 and qtl2 from A: (A619 x Mp708)F2 and B: (A619 x Mo6)F2 populations. Genotypes of p1 are represented by the bottom row of letters, and subdivided into qtl2 genotypes where A, homozygous A619 alleles; H, heterozygous alleles; and B, homozygous Mp708 or Mo6 alleles dependent upon the population. Standard error bars are taken to the second standard deviation.

 
As CGA is implicated in host-plant resistance to corn earworm, breeders might be interested in increasing CGA in silks. In many individuals of the (Mo6 x Mp708)F2 population, CGA levels reached an excess of 0.40% of the fresh silk weight. Assuming a comparable activity against corn earworm on the basis of similar chemical structure, this amount of CGA is well above a 0.20% fresh weight threshold of maysin reported to reduce the weight of corn earworm larvae by 50% (Wiseman et al., 1992). Additionally, many individuals from this population synthesized both CGA and flavones, each in amounts well above the 0.20% threshold level. Although these individuals have alleles with equivalent effects at both the p1 and qtl2 loci, several novel QTLs were detected in this study with the potential to further increase CGA.

Minor QTLs
Quantitative trait loci for both CGA and flavones were detected that exhibited small effects, and for which candidates were assigned based on location alone. The pal1 locus was within a confidence interval of a QTL detected in the (A619 x Mo6)F2 population for CGA. A c2 candidate locus in bin 4.08 was detected by CIM and the multiple-effects model in the (Mo6 x Mp708)F2 population for CGA, in which the allele from Mp708 increased CGA. Additionally, an epistatic interaction between the c2 candidate QTL and a marker in bin 3.04 from the (A619 x Mp708)F2 population was detected. A QTL for flavones in bin 9.02 was detected in both high-low populations, but only exerted small effects. The rem1 candidate gene for that QTL was characterized by its ability to increase maysin through recessive inheritance (Byrne et al., 1996, 1998), but in both high-low populations in this study the QTL effects in that region were nearly additive.

The remaining QTLs had no candidate genes that could explain their effects. Among these, there were two to three QTLs detected on chromosome 3 affecting CGA. Indeed, in eight different F2 populations assayed for CGA accumulation in silks, effects from QTLs in bins 3.03-3.06 were detected (this paper; Szalma et al., 2002; McMullen, unpublished data). As they were detected across different inbred backgrounds and environments, their influence on CGA accumulation in silk tissue is unequivocal. The genetic basis of QTLs for CGA on chromosome 3 warrant further investigation, and specific populations must be established to effectively study their effects.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
Of the QTLs detected affecting CGA biosynthesis, the p1 and qtl2 loci had the largest effects in the (A619 x Mp708)F2 and (A619 x Mo6)F2 populations. The p1 locus exhibited additive gene effects on flavone accumulation in silks, but dominant effects on CGA accumulation. The qtl2 locus main effects increased CGA levels, but in the presence of functional p1 alleles qtl2 increased CGA and decreased flavones. In the (Mo6 x Mp708)F2 population neither locus was segregating and neither QTL was detected, as might be expected if genetic variation at those loci were abolished by similar alleles in the parents. These studies provide candidate genes on which to focus genetic tests to manipulate CGA levels.


   ACKNOWLEDGMENTS
 
We are grateful for the technical assistance of C.J. Roebke, Alberto Maurer, and Chris Browne. This research was supported by USDA NRI-CGP Plant Genome Grant #97-35300-4391 (MDM) and NIH Graduate Training Grant 2T32GM 08396 (BSB), with funds provided to USDA-ARS (MDM). Brand names are important to factually report on available data; however, the USDA neither guarantees nor warrants the standard of the product, and the use of the name by USDA implies no approval of the product to the exclusion of others that may also be suitable.

Received for publication October 8, 2001.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Crop Sci.Home page
S. J. Szalma, M. E. Snook, B. S. Bushman, K. E. Houchins, and M. D. McMullen
Duplicate Loci as QTL: The Role of Chalcone Synthase Loci in Flavone and Phenylpropanoid Biosynthesis in Maize
Crop Sci., September 1, 2002; 42(5): 1679 - 1687.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
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 HighWire
Right arrow Citing Articles via ISI Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bushman, B. S.
Right arrow Articles by McMullen, M. D.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Bushman, B. S.
Right arrow Articles by McMullen, M. D.
Agricola
Right arrow Articles by Bushman, B. S.
Right arrow Articles by McMullen, M. D.

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