QTL Analysis of Flavonoid Synthesis

1
QTL Analysis of Flavonoid Synthesis Illuminates
Genetic Control of Metabolic Pathways
P.F. Byrne, Colorado State Univ. H. Kross,
USDA-ARS, Columbia, MO M.D. McMullen, USDA-ARS
and Univ. of Missouri N.W. Widstrom, USDA-ARS,
Tifton, GA M.E. Snook, Univ. Of Georgia M.A.
Berhow, USDA-ARS, Peoria, IL B.R. Wiseman,
USDA-ARS, Tifton, GA E.H. Coe, USDA-ARS and
Univ. of Missouri E.A. Lee, Univ. of Guelph
We gratefully acknowledge research support from
USDA National Research Initiative - Competitive
Grants Program - Plant Genome Awards No. 9400774,
9500636, and 97-35300-4391.
Fig. 1. The corn earworm, Helicoverpa zea
(Boddie) feeding on a maize ear (courtesy of
CIMMYT).
INTRODUCTION Maysin is a flavone glycoside found
in maize silks that retards growth of corn
earworm (CEW) larvae (Fig. 1). This insect
damages maize crops by directly consuming kernels
and by introducing ear-rotting fungi. Better
understanding of the genetic control of maysin
synthesis will facilitate development of maize
lines and hybrids with improved levels of CEW
resistance. We conducted quantitative trait
locus (QTL) analysis for maysin concentration in
multiple maize populations, not only to determine
the genetics of maysin synthesis, but also to
gain insights into quantitative flux through a
metabolic pathway. Five of our most notable
results to date are described in this poster.
4
Apparently independent synthesis of maysin and
apimaysin. A single hydroxyl group at the 3-
position of the flavonoid B-ring differentiates
maysin from apimaysin (Fig. 2). We assumed that
synthesis of both compounds occurred via the same
pathway and hypothesized that pr1, which encodes
a flavonoid-3'-hydroxylase, was the major locus
explaining the difference between maysin and
apimaysin concentrations in Population 3. We
found that pr1 did affect apimaysin concentration
(explaining 65 of the phenotypic variance), but
had no affect on maysin concentration.
Conversely, rem1 was the major QTL for maysin
(accounting for 55 of the phenotypic variance),
but was not significant for apimaysin. Our
surprising results suggest that synthesis of
these closely related compounds occurs at least
partially independently.
STRATEGY We developed mapping populations (Table
1) based on crosses between parental lines having
contrasting flavone concentrations and/or
different alleles at flavonoid pathway loci (Fig.
2). Individual plants or families were evaluated
both for phenotypic traits and for molecular
markers, which included known flavonoid loci and
closely linked markers. Thus, we were able to
determine the effect of allelic variation at
specific points in the pathway on concentrations
of biochemical end-products. We used a variety
of QTL methods (ANOVA, interval mapping, and
composite interval mapping) to detect and
characterize genomic regions affecting the
phenotypic traits. Table 1. Mapping populations
used in these studies.
5
Detection of a second locus that determines the
salmon silk phenotype. Maysin is the major
flavone present in the majority of maize lines
with normal (green) silk phenotype. Chemical
analyses of methanol extracts of silks of various
salmon silk (sm) lines or stocks revealed two
novel patterns of flavones associated with the sm
phenotype (Fig. 5). In silks of the sm1-ref line,
representing the first class, the predominant
flavone is rhamnosyl-isoorientin. A second
distinct class of sm lines, typified by the
inbred line T218, contains the flavones
isoorientin and de-rhamnosyl-maysin. Crosses
between plants of these two classes complement,
restoring both the green silk phenotype and the
presence of maysin. The genetic basis of the sm
phenotype in T218 was determined using Population
5. We identified a gene on chromosome 2, bin
2.07, required for the expression of the sm
phenotype and presence of isoorientin and
de-rhamnosylmaysin. We have designated this gene
salmon silk2 (sm2). Both sm1 and sm2 require a
functional p1 gene for expression of the sm
phenotype. The chemical structure of the flavones
associated with sm1 and sm2 suggests that both
genes are involved with addition or modification
of the sugar moieties attached to the flavone.
Fig. 5. Left ear, T218 (sm2/sm2) middle ear,
sm1-reference (sm1/sm1) right ear, F1 hybrid
(Sm1/sm1, Sm2/sm2) showing complementation.

• CONCLUSIONS
• Although our study focused specifically on the
  flavonoid pathway, we believe our results are
  also relevant
• to other metabolic pathways influencing important
  agronomic traits.
• Because many pathways are known to be controlled
  by regulatory loci, the importance of the
  regulator p1 in our study may be broadly
  applicable to the manipulation of other
  quantitative traits.
• A pathway leading to a desired agronomic trait
  may be affected by loci on intersecting pathways,
  as shown by the effect of a1 on maysin
  concentration.
• Different sets of loci may control the synthesis
  of closely related compounds, as we observed for
  maysin and apimaysin.
• Even in well known systems like the flavonoid
  pathway of maize, there are new loci, functions,
  and mechanisms awaiting discovery.
• REFERENCES
• Byrne, P.F., M.D. McMullen, M.E. Snook, T.A.
  Musket, J.M. Theuri, N.W. Widstrom, B.R. Wiseman,
  and E.H. Coe. 1996. Quantitative trait loci and
  metabolic pathways Genetic control of the
  concentration of maysin, a corn earworm
  resistance factor, in maize silks. Proc. Natl.
  Acad. Sci. USA 938820-8825.

Number of families or individuals
RESULTS The significance and additivity of the
regulatory locus p1 in controlling maysin
concentration. p1 encodes a transcription
activator that regulates flavone synthesis in
silks, as well as phlobaphene synthesis in cob
and pericarp tissue (Fig. 3). In Populations 1,
4, and 5, which segregated for null and
functional p1 alleles, we found that p1 accounted
for 58, 44, and 52, respectively, of the
phenotypic variance for maysin concentration.
Moreover, the effect of p1 was strikingly
additive as shown for Population 4 in Fig. 4.
The influence of another regulatory locus, in1,
on maysin concentration was suggested by results
for Population 4.
1
Fig. 2. Flavonoid pathway of maize. The
p1-controlled portion of the pathway is shown on
the light tan background (excluding the upper
right corner). Positions of loci and enzymes are
shown when known or hypothesized, with candidate
loci in parentheses. The position in the pathway
of 3-hydroxylation is unknown.
Maysin concentration ( fresh silk wt.)
0.7 0.6 0.5 0.4 0.3 0.2 0.1
a1 genotype A H B
2
Detection of a locus designated rem1 that
enhances maysin content when homozygous recessive
in the presence of functional p1. QTL analyses
of maysin concentration in populations 1, 2, and
3 revealed a locus on chromosome 9S that was not
previously known to affect maysin synthesis. We
named the locus rem1 (recessive enhancer of
maysin1). The locus showed a strong epistatic
interaction with p1. Based on the position of
the locus and its interaction with p1, we
hypothesized that it corresponds to the bp1 locus
(Byrne et al., 1996).
A H B
3
Evidence for the interconnectedness of pathway
branches. In Population 4, a null allele at a1,
which is not directly on the pathway leading to
maysin, caused a major increase in maysin
concentration and a decrease in
3-deoxyanthocyanin concentration. We hypothesize
that this result is due to blocking of the
3-deoxyanthocyanin pathway by nonfunctional a1
and shunting of intermediates to the flavone
pathway. As with rem1, there was a strong
interaction of a1 with p1 (Fig. 4).
Fig. 3. Effect of variation in p1 alleles on cob
and kernel color.
p1genotype
Fig. 4. Epistasis between p1 and a1 for silk
maysin concentration in the population (W23a1 x
GT119)F2. The a1 effect is expressed only with a
functional p1 allele. Ahomozygous for the W23a1
allele (functional p1, nonfunctional a1)
Hheterozygous and B homozygous for the GT119
allele (nonfunctional p1, functional a1).