Planta (2010) 231:319–328

DOI 10.1007/s00425-009-1054-8

O R I G I N A L A R T I CL E

Red clover coumarate 3⬘-hydroxylase (CYP98A44) is capable

of hydroxylating p-coumaroyl-shikimate but not p-coumaroyl-
malate: implications for the biosynthesis of phaselic acid
Michael L. Sullivan · Robert Zarnowski

Received: 1 October 2009 / Accepted: 28 October 2009 / Published online: 17 November 2009
© US Government 2009

Abstract Red clover (Trifolium pratense) leaves accumu- ated protein in red clover stems and leaves and cross-reacts
late several
mol of phaselic acid [2-O-caVeoyl-L-malate] with C3⬘H proteins from other species. CYP98A44 expressed
per gram fresh weight. Post-harvest oxidation of such in Saccharomyces cerevisiae is capable of hydroxylating
o-diphenols to o-quinones by endogenous polyphenol oxi- p-coumaroyl-shikimate, but not p-coumaroyl-malate. This
dases (PPO) prevents breakdown of forage protein during Wnding indicates that in red clover, phaselic acid is likely
storage. Forages like alfalfa (Medicago sativa) lack both formed by transfer of a caVeoyl moiety to malic acid,
foliar PPO activity and o-diphenols. Consequently, break- although the existence of a second C3⬘H capable of hydrox-
down of their protein upon harvest and storage results in ylating p-coumaroyl-malate cannot be deWnitively ruled out.
economic losses and release of excess nitrogen into the
environment. Understanding how red clover synthesizes Keywords p-Coumarate 3⬘-hydroxylase · Cytochrome
o-diphenols such as phaselic acid will help in the develop- P450 · o-Diphenol · Phaselic acid · Phenylpropanoid ·
ment of forages utilizing this natural system of protein pro- Forages
tection. We have proposed biosynthetic pathways in red
clover for phaselic acid that involve a speciWc hydroxycin- Abbreviations
namoyl-CoA:malate hydroxycinnamoyl transferase. It is 4CL 4-Coumarate:CoA ligase
unclear whether the transfer reaction to malate to form C3⬘H Coumarate 3⬘-hydroxylase
phaselic acid involves caVeic acid or p-coumaric acid and C4H Cinnamate-4-hydroxylase
subsequent hydroxylation of the resulting p-coumaroyl- CYP Cytochrome P450
malate. The latter would require a coumarate 3⬘-hydroxy- HCT Hydroxycinnamoyl-CoA hydroxycinnamoyl
lase (C3⬘H) capable of hydroxylating p-coumaroyl-malate, transferase
an activity not previously described. Here, a cytochrome PAL Phenylalanine ammonia lyase
P450 C3⬘H (CYP98A44) was identiWed and its gene cloned PCR Polymerase chain reaction
from red clover. CYP98A44 shares 96 and 79% amino acid PPO Polyphenol oxidase
identity with Medicago truncatula and Arabidopsis thali- qRT-PCR Quantitative real-time polymerase chain
ana C3⬘H proteins that are capable of hydroxylating p-cou- reaction
maroyl-shikimate and have been implicated in monolignol SDS-PAGE Sodium dodecyl sulfate-polyacrylamide gel
biosynthesis. CYP98A44 mRNA is expressed in stems and electrophoresis
Xowers and to a lesser extent in leaves. Immune serum SMT Sinapoyl-glucose:malate synapoyl transferase
raised against CYP98A44 recognizes a membrane-associ-

Introduction
M. L. Sullivan (&) · R. Zarnowski
US Dairy Forage Research Center,
Agricultural Research Service, US Department of Agriculture,
Red clover (Trifolium pratense) leaves accumulate large
1925 Linden Drive, Madison, WI 53706, USA amounts (approximately 5
mol g¡1 fresh weight) of 2-O-
e-mail: michael.sullivan@ars.usda.gov caVeoyl-L-malic acid (hereafter referred to as phaselic acid

123
320 Planta (2010) 231:319–328

or caVeoyl-malate) (Sullivan 2009a). Post-harvest oxida- OH OH

OH OH
tion of phaselic acid and other o-diphenols by endogenous
HO HO
polyphenol oxidases (PPO) prevents breakdown of the red O O O O CoAS O

O O
clover protein during harvest and storage (Sullivan and
(CYP98A44) (HCT1)
HatWeld 2006). Because ruminant animals poorly utilize C3’H HCT-S

degraded protein, excessive protein breakdown during har- OH CoASH OH

vest and storage of alfalfa costs US farmers an estimated OH OH OH

p-Coumaroyl Caffeoyl Caffeoyl
$100 million annually to supplement rations with the shikimic acid shikimic acid CoA

needed true protein and results in release of excess nitrogen (HCT2) Malate
(HCT1) HCT-M
into the environment (Sullivan and HatWeld 2006). Adapta- HCT-S Shikimate
O O

HO HO
tion of red clover’s natural system of protein protection to OH OH
O O O O
alfalfa and other forage crops would therefore be expected HO O CoAS O O O

(HCT2) (?)
to have positive economic and environmental impacts. PAL C4H 4CL HCT-M C3’H
Although alfalfa lacks both foliar PPO activity and o-diphe- CoASH Malate
nol PPO substrates, we have recently shown that alfalfa OH

transformed with a red clover PPO gene experiences less OH

p-Coumaric
OH
p-Coumaroyl
OH
p-Coumaroyl
OH
Phaselic
post-harvest protein degradation than wild type alfalfa, pro- acid CoA malic acid acid

vided o-diphenol PPO substrates are exogenously applied
(Sullivan and HatWeld 2006). We have been using red
clover as a model for biosynthesis of phaselic acid and
other o-diphenols in forage legumes. Understanding these
pathways in red clover should give insights into how simi-
lar biosynthetic pathways could be introduced into alfalfa
and other forages.
In the Brassicaceae, hydroxycinnamoyl-malate esters are
biosynthesized by the transfer of hydroxycinnamoyl moie-
ties from their corresponding glucose esters to malic acid
via the action of sinapoyl-glucose:malate synapoyl transfer-
ase (SMT, in Arabidopsis thaliana the product of the SNG1
gene) (Grawe et al. 1992; Lehfeldt et al. 2000). Because
there is no genetic evidence for a similar SMT enzyme in
red clover, we have proposed alternative pathways (Fig. 1)
for phaselic acid biosynthesis in red clover whereby
hydroxycinnamoyl moieties are transferred from CoA thio-
lesters to the malic acid acceptor (Sullivan 2009a). In these
proposed pathways, phenylalanine would be converted to
p-coumaroyl-CoA by the sequential action of phenylala-
nine ammonia lyase (PAL), cinnamate-4-hydroxylase
(C4H), and 4-coumarate:CoA ligase (4CL). The action of
speciWc hydroxycinnamoyl transferases (HCT) and one or
more p-coumarate 3⬘-hydroxylases (C3⬘H) would result in
the formation of phaselic acid. Formation of phaselic acid
could be accomplished either by transfer of a caVeoyl moi-
ety to malic acid (upper pathway), or transfer of a p-couma-
royl moiety to malic acid followed by 3⬘-hydroxylation of
the resulting p-coumaroyl-malate by C3⬘H (lower path-
way). Both pathways predict a hydroxycinnamoyl-
CoA:malate hydroxycinnamoyl transferase, which we have
recently identiWed in red clover (Sullivan 2009a). The
lower pathway would predict red clover has a C3⬘H
enzyme capable of 3⬘-hydroxylation of p-coumaroyl-
malate to form phaselic acid, while the upper pathway
requires an enzymatic activity capable of forming a caVeic
Fig. 1 Possible pathways for phaselic acid biosynthesis in red clover.
Proposed pathway enzymes for production of phaselic acid include
phenylalanine ammonia lyase (PAL), cinnamate 4-hydroxylase (C4H),
4-coumarate:CoA ligase (4CL), hydroxycinnamoyl-CoA:shikimate
hydroxycinnamoyl transferase (HCT-S), hydroxycinnamoyl-CoA:
malate hydroxycinnamoyl transferase (HCT-M), and p-coumarate 3⬘-
hydroxylase (C3⬘H). Red and blue arrows represent alternative path-
ways. Red clover gene products described here and elsewhere (Sulli-
van 2009a) corresponding to the enzymatic activities are indicated in
green. For simplicity, not all reactants and products are shown
acid moiety (e.g. p-coumaroyl-shikimate 3⬘-hydroxylase as
shown in Fig. 1 and described in more detail below).
Existing literature suggests that C3⬘H enzymes, which
are membrane-associated cytochrome P450 enzymes
(CYP98A subfamily) involved in the biosynthesis of mono-
lignols and other 3,4-hydroxylated phenylpropanoid sec-
ondary metabolites, do not directly hydroxylate p-coumaric
acid to caVeic acid but rather act on p-coumaroyl deriva-
tives (Schoch et al. 2006). Most commonly, the preferred
substrate for CYP98A enzymes appears to be p-coumaroyl-
shikimate, although some can hydroxylate other p-couma-
royl derivatives as well. For example, the enzyme from A.
thaliana hydroxylates shikimic and to a lesser extent quinic
acid esters of p-coumaric acid, but only poorly or not at all
p-coumaric acid or its glucose or CoA esters (Schoch et al.
2001; Franke et al. 2002). C3⬘H activities capable of
hydroxylating shikimic or quinic and 4-hydroxyphenyllac-
tic esters of p-coumaric acid are present in peltate glands of
sweet basil (Ocimum basilicum) (Gang et al. 2002). The
p-coumaroyl-shikimate/quinate hydroxylating versus p-cou-
maroyl hydroxyphenyllactate hydroxylating activities in
basil may represent diVerent enzymes because, unlike the
hydroxylating activities derived from leaves, when a cloned
basil C3⬘H gene (CYP98A13) was expressed in yeast (Sac-
charomyces cerevisiae), the recombinant protein eYciently

123
Planta (2010) 231:319–328
hydroxylated p-coumaroyl-shikimate and -quinate but only
poorly the 4-hydroxyphenyllactate ester (Gang et al. 2002).
In Lithospermum erythrorhizon a gene encoding a C3⬘H
capable of hydroxylating p-coumaroyl-4-hydroxyphenyl-
lactic acid has been isolated, although it has not been deter-
mined if shikimate or quinate p-coumaroyl esters are
substrates for the enzyme (Matsuno et al. 2002). In coVee,
two C3⬘H enzymes have been identiWed, one of which uti-
lizes both p-coumaroyl-shikimate and -quinate with similar
eYciency and one of which can utilize p-coumaroyl-
shikimate only (Mahesh et al. 2007). In wheat (Triticum
aestivum), activities of some C3⬘H isoforms also appear to
be plastic, with the ability to hydroxylate both p-coumaroyl
esters (e.g. p-coumaroyl-shikimate) and amides (e.g. p-cou-
maroyl-tyramine) (Morant et al. 2007). To our knowledge,
there has been no assessment of whether any characterized
C3⬘H enzymes are capable of hydroxylating p-coumaroyl-
malate, the predicted precursor of phaselic acid in the lower
pathway of Fig. 1. It seems plausible that red clover could
have such an activity, either as a highly speciWc (i.e. capable
of hydroxylating p-coumaroyl-malate only) or multispeciWc
(e.g. capable of hydroxylating both p-coumaroyl-malate
and -shikimate) enzyme. To address this possibility, we
identiWed genes corresponding to red clover C3⬘H and
characterized the encoded enzymes with respect to sub-
strate speciWcity.
Materials and methods
Reagents
All purchased reagents were of molecular biology or higher
grade. 5-O-(p-coumaroyl)-shikimate, 2-O-(p-coumaroyl)-
L-malate, and 5-O-caVeoyl-shikimate standards were pre-
pared by the method described by Hemmerle et al. (1997).
The NMR spectra of the standards were consistent with
the assigned structures. 2-O-caVeoyl-L-malate (phaselic
acid) was prepared enzymatically from caVeoyl-CoA and
malic acid using recombinant red clover hydroxycinna-
moyl-CoA:malate hydroxycinnamoyl transferase (Sullivan
2009a).
Plant materials
A red clover (T. pratense L.) genotype (lab designation
“PPO”) (Sullivan et al. 2004) selected from a population
of WI-2 germplasm (Smith and Maxwell 1980) was the
source material for isolation of CYP98A genes. This same
plant, and two additional genotypes derived from the WI-2
germplasm were used for expression and other analyses.
The plants were clonally propagated from crown pieces
and grown in a greenhouse with temperatures maintained
321
between 20 and 30°C and light intensities between 400
and 1,000
mol m¡2 s¡1. Supplemental lighting was used
when day length was <13 h day¡1. Plants were fertilized
weekly with Peter’s soluble 20-10-20 (Scott’s, Marysville,
OH).
Nucleic acid isolation and analysis
Total RNA, cDNA, and plasmid DNA were prepared and
sequencing was carried out as previously described
(Sullivan 2009a). Sequence analyses were carried out using
the Wisconsin Package Version 10 (Accelrys, San Diego,
CA), Lasergene Version 8 (DNAStar, Madison, WI), and
BLAST programs using the National Center for Biotech-
nology Information (NCBI, http://www.ncbi.nlm.nih.gov)
and The Gene Index Project (TGI, http://compbio.dfci.
Harvard.edu/tgi) web sites. Analysis of proteins and protein
domains was accomplished using the PROSITE database
(Hulo et al. 2008) and InterProScan (Zdobnov and Apweiler
2001).
Generation of a red clover CYP98A gene fragment
by polymerase chain reaction and isolation of a full-length
cDNA by library screening
Oligo dT-primed cDNA from unexpanded red clover leaves
corresponding to 25 ng total RNA was used as template
with degenerate primers 5⬘-AACCTCTACGACRTMA
AACCNGTCCG-3⬘ (forward) and 5⬘-CCYTTGGGWA
TATCRTAGCCNCC-3⬘ (reverse) in a polymerase chain
reaction (PCR) with JumpStart KlenTaq LA polymerase
(Sigma, St. Louis, MO) using 40 cycles of 30 s at 94°C,
30 s at 48°C, and 120 s at 68°C. The resulting »1,050-bp
DNA fragment was puriWed from an agarose gel using
QIAEX II resin (Qiagen, Valencia, CA) and cloned into
pGEM-T Easy (Promega, Madison, WI) according to man-
ufactures’ instructions. Approximately 106 recombinant
phage from a
ZAP II (Stratagene, La Jolla, CA) red clover
young leaf cDNA library were screened essentially as pre-
viously described (Sullivan et al. 2004) using the 1,050-bp
PCR-derived red clover CYP98A gene fragment (described
above) as the hybridization probe. Phage clones of interest
were rescued to pBluescript II SK(-) phagemids according
to the manufacturer’s protocol.
Quantitative real-time PCR analysis of gene expression
Quantitative real-time PCR (qRT-PCR) analysis of
CYP98A44 expression in various red clover tissues was
carried out as previously described (Sullivan 2009b) using
the CYP98A44-speciWc primer pair 5⬘-CCAAAGACT
AAGATTCAAGCTTCCA-3⬘ (forward) and 5⬘-CGGTTT
TATGTCGTAAAGGTTTCC-3⬘ (reverse). CYP98A44
123
322
expression levels were normalized to the expression of the
housekeeping gene EIF4A (Taylor et al. 1993) determined
using the primer pair 5⬘-GACATGGACCAAAATACTCG
TGATA-3⬘ (forward) and 5⬘-CAGTTGTTATGAGCACTC
GTGAAGA-3⬘ (reverse) (Sullivan 2009a). The average and
standard error of three biological replicates is reported with
the normalized transcript abundance in each tissue expressed
relative to Xowers, the tissue with the highest level expres-
sion.
Preparation of gene constructs for expression
of CYP98A44 in Escherichia coli and yeast
For all gene constructions standard molecular biology tech-
niques were used (Sambrook et al. 1989; Ausubel et al.
1998). When fragments for cloning were generated via
PCR, the cloned insert was sequenced to insure that no
mutations were introduced that would alter the sequence of
the translated protein.
For expression in E. coli, a plasmid containing the full-
length CYP98A44v1 cDNA was used as template in PCR
with the primers 5⬘-GCATATGGCTCTGTTTCTCACAA
TAC-3⬘ and 5⬘-GGCGGCCGCTTAGATATCAGCTGG
CACACG-3⬘ to introduce NdeI and NotI restrictions sites
(underlined) immediately before the start and stop codons,
respectively. The resulting PCR product was digested with
NdeI and NotI and inserted into pET28a (Novagen,
Madison, WI) digested with NdeI and NotI. For expression
in yeast, a similar PCR was carried out using the primers
5⬘-GGAGATCTATGGCTCTGTTTCTCACAATAC-3⬘ and
5⬘-GCGGTACCTTAGATATCAGCTGGCACACG-3⬘ to
introduce BglII and KpnI restriction sites (underlined)
immediately before the start and stop codons, respectively.
The resulting PCR product was digested with BglII and
KpnI and inserted into pYeDP60 (Pompon et al. 1996)
digested with BamHI and KpnI.
Expression of CYP98A44 in E. coli for immune serum
production
The CYP98A44-pET plasmid was transformed into
BL21(DE3)RIL Codon Plus E. coli (Stratagene). A 250-mL
culture of the transformed E. coli was grown at 37°C with
shaking (225 rpm) in Luria–Bertani medium supplemented
with 50
g mL¡1 kanamycin to an optical density at
600 nm of 0.5, at which time isopropyl--D-thiogalactoside
was added to 1 mM. The culture was grown at 37°C with
shaking for an additional 3 h. The culture was harvested
and lysed using Bugbuster reagent (Novagen), then frac-
tionated into soluble and insoluble fractions according to
the manufacturer’s suggested procedures. Protein from the
insoluble inclusions was used without further puriWcation
123
Planta (2010) 231:319–328
for immune serum production in a rabbit at a commercial
facility (Harlan Bioproducts for Science, Indianapolis, IN)
using standard protocols.
Preparation of red clover tissues samples for
immunoblotting
Red clover stems and young fully expanded leaves were
powdered in liquid nitrogen using a mortar and pestle and
the powdered tissue was suspended in 3 mL g¡1 fresh
weight of 50 mM Tris, 50 mM ascorbic acid pH 7.5, 1%
(v/v) protease inhibitor cocktail (Sigma P8849, added
immediately before use). Using a wide bore pipette tip,
500
L of the tissue slurry was removed, mixed with
500
L SDS-PAGE (sodium dodecyl sulfate-polyacryl-
amide gel electrophoresis) sample buVer (Harlow and
Lane 1988), heated at 80°C for 20 min with periodic mix-
ing, and centrifuged at 22,000£g for 10 min. The resulting
supernatant represented the total, unfractionated sample.
The remaining slurry was centrifuged at 22,000£g for
10 min at 4°C. The supernatant was removed to a fresh
tube, a small amount was reserved for determination of
protein concentration, and the remainder was mixed
with an equal volume of SDS-PAGE sample buVer and
heat-treated as described above. The resulting sample rep-
resented the soluble sample. The pelleted material was
washed once by resuspending in the original volume of
extraction buVer, centrifuging at 22,000£g for 10 min at
4°C, and removing the supernatant. The pellet was resus-
pended in the original volume of extraction buVer and an
equal volume of SDS-PAGE sample buVer and heat-treated
and centrifuged as described above for the total sample.
The resulting supernatant represented the membrane-asso-
ciated sample. Protein concentration was determined for the
soluble samples using Bio-Rad Protein Assay (Bio-Rad,
Hercules, CA), and SDS-PAGE gels were loaded such that
total and membrane-associated samples corresponded to
the same amount of starting material as the soluble
samples. Because preliminary experiments indicated immu-
noblot signals for leaves were low, those SDS-PAGE
samples were concentrated by adding four volumes 1:1
(v/v) methanol:acetone (acidiWed with 1 mM HCl) and the
resulting precipitate was resuspended in one-tenth the orig-
inal volume of SDS-PAGE sample buVer. SDS-PAGE
and transfer to PVDF membranes was carried out using
standard methodologies (Harlow and Lane 1988). Immuno-
logical detection was carried out according to the PVDF
manufacturer’s suggested protocol using the prepared
CYP98A44 immune serum (diluted 1:3,000), an alkaline
phosphatase-conjugated goat anti-rabbit secondary anti-
body (Bio-Rad), and 1-Step NBT/BCIP (Thermo ScientiWc,
Rockford, IL).
Planta (2010) 231:319–328
Expression of CYP98A44 in yeast and preparation of yeast
microsomes
pYeDP60 containing the red clover CYP98A44 open read-
ing frame (described above) or pYeDP60 lacking an insert
(as a negative control) was transformed into the WAT11
yeast strain (Pompon et al. 1996) using the lithium acetate
method of Gietz and Woods (2002). Microsomes were pre-
pared from the transformed yeast using high-density cultur-
ing and mechanical disruption procedures described by
Pompon et al. (1996). These microsome preparations were
used for activity assays and in immunoblotting experiments
along with microsome preparations from yeast expressing
CYP proteins from A. thaliana (the generous gifts Clint
Chapple and Jing-Ke Weng) and sweet basil (the generous
gifts of David Gang and Anna Berim).
Preparation of microsomes from red clover tissues
Microsomes were prepared from red clover tissues using an
adaptation of previously described methods (Schaller and
Dewitt 1995; Stukkens et al. 2005). Approximately 5 g red
clover stems or young leaves were ground in a mortar and
pestle in liquid nitrogen with 0.5 g polyvinylpolypyrroli-
done and 5 g sea sand. After grinding, 25 mL of cold
microsome isolation buVer [250 mM tricine, 250 mM
sucrose, 100 mM ascorbic acid, 50 mM NaH2PO4, 2 mM
dithiothreitol, 2 mM EDTA, 0.4% (v/v) protease inhibitor
cocktail (Sigma P8849), 5 mg mL¡1 bovine serum albumin,
pH 8.2] were added. The homogenate was gently stirred on
ice and then Wltered through Miracloth (Calbiochem, La
Jolla, CA). The Wltrate was centrifuged at 10,000£g for
15 min at 4°C; then the supernatant was collected, Wltered
again, and centrifuged at 245,000£g for 90 min at 4°C. The
resulting pellet was washed by resuspending in 10 mL of
cold microsome isolation buVer and recentrifuging at
245,000£g for 60 min. The washed membrane pellet was
resuspended in 500
L of cold microsome storage buVer
(20% [v/v] glycerol in 100 mM sodium phosphate buVer,
pH 7.4). Isolated microsomes were used immediately for
enzymatic assays.
Assay for 3⬘-hydroxylase activity
Reactions for 3⬘-hydroxylase activity contained 0.2 mM
p-coumaroyl-shikimate or p-coumaroyl-malate, 1 mM
NADPH and microsomal protein from yeast (40
g mL¡1)
or red clover tissue (400–600
g mL¡1) in 100 mM sodium
phosphate buVer, pH 7.4. Reactions were incubated at 30°C
for 1 h (yeast microsomes) or up to 18 h (for red clover
microsomes), then stopped by the addition of one-tenth vol-
ume 10% (v/v) acetic acid. Reaction products were ana-
lyzed by HPLC with detection using a photodiode array
323
detector (250–500 nm) as previously described for
hydroxycinnamoyl transferase reaction products (Sullivan
2009a).
Results
IdentiWcation of a red clover C3⬘H (CYP98A44) gene
Degenerate oligonucleotide primers based on conserved
C3⬘H nucleotide sequences from Medicago truncatula
(TGI tentative consensus sequence TC113827), A. thaliana
(Genbank NM_180006), and sweet basil (Genbank
AY082612) were used in PCR reactions with young red
clover leaf cDNA to generate a 1,050-bp gene fragment.
Sequence analysis of the gene fragment showed a high
degree of sequence similarity to the C3⬘H sequences on
which the PCR primer design was based (data not shown).
This fragment was used to screen a red clover young leaf
cDNA library using moderate stringency (0.4 M NaCl,
55°C) conditions for the hybridization. Using this approach,
seven hybridizing cDNAs were isolated. Sequence analysis
of these indicated they correspond to two highly similar
(>99% identity) sequences (Genbank accessions GQ497816
and GQ919201) with 1,527 bp open reading frames pre-
dicted to encode 509 amino acid proteins diVering in only
one position (Fig. 2). The encoded proteins have been des-
ignated CYP98A44v1 and CYP98A44v2 (http://drnelson.
utmem.edu/CytochromeP450.html) and contain isoleucine
and valine, respectively, at position 75. The red clover
CYP98A44 proteins are 96, 84 and 79% identical to dicot
CYP98A proteins from M. truncatula, sweet basil, and
A. thaliana and 71% identical to CYP98A proteins from
wheat, a monocot. A tBLASTn search of a collection of
approximately 22,000 EST sequences derived from red
clover leaves and 3-week-old whole plants (Sato et al.
2005) identiWed only one CYP98A-like EST (BB926702),
and this is >99% identical to CYP98A44 at the amino acid
level. Analysis of the remaining top 50 tBLASTn matches
using InterProScan (Zdobnov and Apweiler 2001)
Fig. 2 Predicted amino acid sequence of CYP98A44v1 (red clover
C3⬘H). The highlighted amino acid shows the location of the V for I
substitution present in CYP98A44v2. Sequences for the corresponding
cDNAs have been deposited in Genbank under the accessions
GQ497816 and GQ919201
123
324
indicates they all likely encode non-CYP98 cytochrome
P450 proteins.
Accumulation of CYP98A44 transcripts in red clover
tissues
Accumulation of CYP98A44 transcripts in various red clo-
ver tissues was analyzed by qRT-PCR using a primer pair
that would recognize both CYP98A44v1 and v2. Transcript
abundance was analyzed using total RNA derived from
young leaves, mature leaves, stems, and Xowers from three
red clover genotypes and signals from each sample were
normalized to the expression of the housekeeping gene
eIF4A as previously described (Sullivan 2009a). As shown
in Fig. 3, of the tissues examined, CYP98A44 transcript
accumulation was highest in Xowers and stems. Transcripts
were also detected in young and mature leaves, but at Wve-
and tenfold lower levels than seen in stems and Xowers,
respectively.
Detection of CYP98A proteins by immunoblotting
To facilitate analyses of the CYP98A44 protein, immune
serum against CYP98A44 expressed in E. coli was pro-
duced in rabbits. To analyze red clover tissues, three pro-
tein fractions were made from leaves and stems: total, in
which tissue was extracted directly with SDS-containing
buVer; soluble, in which tissue was extracted with non-
SDS-containing buVer; and membrane associated, in which
the insoluble residue from the soluble extraction was solu-
1.00
Relative Transcript Abundance
0.80
0.60
0.40
0.20
Young Mature Stems Flowers
Leaves Leaves
Fig. 3 CYP98A44 transcript accumulation in various red clover tissues.
Quantitative real-time PCR was used to measure transcript abundance
in young leaves, mature leaves, stems and Xowers of red clover. The
CYP98A44 signal from each tissue was normalized to that of the
housekeeping gene eIF4A and expressed relative to that of Xowers.
Results are the average of three biological replicates § standard error
123
Planta (2010) 231:319–328
rCYP98A44
ObCYP71A
ObCYP82A
a b
Red Clover CYP98A
SmF5H
WAT11
AtF5H
Leaves Stems (C3H)
T S M T S M Tp At Ob
Molecular Mass (kDa)
104 104
97 97
50 50
37 37
29 29
20 20
Fig. 4 Detection of C3⬘H and other CYP proteins by immunoblotting
with anti-CYP98A44 immune serum. a Immunoblot of total (T), solu-
ble (S), or membrane-associated (M) protein fractions from red clover
leaves and stems. Sample loads for each tissue type correspond to the
amount of fresh weight required to give 50 or 5
g protein for the sol-
uble fraction from leaves and stems, respectively. Microsome protein
(0.5
g) from yeast expressing recombinant CYP98A44 (rCYP98A44)
served as a positive control. b Immunoblot of microsome protein
(1
g) from yeast expressing CYP98A (C3⬘H) proteins from red clover
(Tp), A. thaliana (At), and basil (Ob); ferulic acid 5⬘-hydroxylase
(F5H) from A. thaliana (AtF5H) or Selaginella moellendorYi
(SmF5H); or basil (Ob) CYP71A and CYP82A family proteins. Micro-
some protein from the host yeast strain (WAT11) served as a negative
control
bilized with SDS-containing buVer. As shown in Fig. 4a, in
stems the immune serum recognizes a protein that comi-
grates with CYP98A44 expressed in yeast. The protein
fractionates predominantly with the SDS-soluble, mem-
brane-associated fraction as would be expected for a
CYP98A family protein. The protein is readily detected in
the membrane-associated fraction of red clover leaves as
well, although larger sample loads were required to obtain
signals similar to those detected in stems, consistent with
the observed diVerence in transcript accumulation between
stems and leaves. The large amount of protein loaded on
the SDS-PAGE gel likely resulted in the broader, more
diVuse band detected in the leaf extract.
To determine whether the CYP98A44 immune serum
recognizes CYP98A proteins from other plant species,
microsome preparations from yeast expressing CYP98A44
from red clover (Tp), CYP98A3 from A. thaliana (At), or
CYP98A13 from sweet basil (Ob) were subjected to SDS-
PAGE and immunoblotting. As shown in Fig. 4b, the
CYP98A44 immune serum detected the yeast-expressed
proteins from both A. thaliana and sweet basil. No cross-
reacting proteins were detected from the WAT11 host yeast
transformed with empty expression vector. To determine
the extent of cross-reactivity with other CYP proteins,
microsomal proteins from yeast expressing ferulic acid
5-hydroxylases from A. thaliana (CYP84A1) or Selaginella
moellendorYi (CYP788A1) or uncharacterized CYP71A and
CYP82A proteins from basil were subjected to immunoblot-
ting with the CYP98A44 immune serum. Faint cross-reacting
Planta (2010) 231:319–328
signals were detected for the F5H preparation from
S. moellendorYi and A. thaliana, the former being stronger
than the latter. The sweet basil CYP71A and CYP82A pro-
teins were not detected with the immune serum. These
results indicate that the CYP98A44 immune serum pro-
duced will cross react with CYP98A proteins from a variety
of plant species and has some cross reactivity with at least
some other CYP proteins.
CYP98A44 can hydroxylate p-coumaroyl-shikimate
but not p-coumaroyl-malate
Most characterized CYP98A proteins are capable of
eYciently hydroxylating p-coumaroyl-shikimate, which
appears to be its preferred substrate, at the 3⬘ position to
form caVeoyl-shikimate (Schoch et al. 2006). None of the
characterized enzymes has been assessed for hydroxylation
of p-coumaroyl-malate, however. To determine whether
CYP98A44 is capable of hydroxylating p-coumaroyl-
malate, an activity that would be expected if it were func-
tioning in the lower pathway of Fig. 1, the CYP98A44v1
open reading frame was cloned into a yeast expression vec-
tor and transformed into WAT11, a yeast strain expressing
A. thaliana cytochrome P450 reductase. This system has
been widely used to characterize the enzymatic activities
of plant cytochrome P450 proteins. Microsomes were pre-
pared from the transformed yeast and used as an enzyme
source in hydroxylation reactions. Microsomes from yeast
transformed with the empty expression vector served as
negative control. As shown in Fig. 5a, when microsomes
from the CYP98A44-expressing yeast were incubated with
p-coumaroyl-shikimate, in <1 h, virtually all the substrate
was converted to caVeoyl-shikimate. The reaction required
a 20000 12.57
18000
µAbsorbance Units
16000
14000
10.62
12000
Caffeoyl-
10000 shikimate
(product)
8000
6000
4000
2000
0 0
10 11 12
Time (min)
Fig. 5 Substrate speciWcity of CYP98A44. Microsomes from yeast
expressing red clover C3⬘H (CYP98A44) were incubated with p-cou-
maroyl esters as indicated. Reaction mixtures prior to and after incuba-
tion with the recombinant enzyme (background and foreground of each
panel, respectively) were resolved by HPLC and detected using a pho-
todiode array detector (250–500 nm) as described in “Materials and
325
NADPH and the CYP98A44 protein since incubations
containing microsomes from the vector control yeast did
not result in hydroxylation of the substrate (data not
shown). When p-coumaroyl-malate was incubated with the
CYP98A44-expressing yeast microsomes, no hydroxyl-
ation of the starting material was detected even though the
reaction conditions used were the same as those that had
resulted in nearly complete hydroxylation of p-coumaroyl-
shikimate to caVeoyl-shikimate (Fig. 5b). To rule out the
possibility that the single amino acid substitution present in
CYP98A44v2 might enable the protein to utilize p-couma-
royl-malate, this was also expressed in yeast and assessed
for utilization of p-coumaroyl-malate. Like CYP98A44v1,
CYP98A44v2 is capable of hydroxylating p-coumaroyl-
shikimate, but not p-coumaroyl-malate. These results
suggest that either red clover utilizes an enzyme other than
CYP98A44 to carry out hydroxylation of p-coumaroyl-
malate to form phaselic acid or that synthesis of phaselic
acid in red clover is accomplished via transfer of caVeoyl
moieties to malic acid, with the possibility that those caVe-
oyl moieties are formed by the action of CYP98A44 on
p-coumaroyl-shikimate (upper pathway of Fig. 1).
To address the possibility that an enzyme in red
clover other than CYP98A44 is capable of carrying out a
3⬘-hydroxylation of p-coumaroyl-malate, microsomes were
isolated from red clover stems or leaves. These were incu-
bated with either p-coumaroyl-shikimate or -malate and the
reactions analyzed for formation of the corresponding
caVeoyl esters. The red clover microsome preparations
were capable of catalyzing the formation of small amounts
of caVeoyl-shikimate from p-coumaroyl-shikimate, but we
were unable to detect any formation of caVeoyl-malate
(phaselic acid) from p-coumaroyl-malate.
13.50
b
p-Coumaroyl-
p-Coumaroyl- 14000 13.42 malate
(substrate)
shikimate
µAbsorbance Units
(substrate) 12000
Caffeoyl-malate
product (phaselic acid,
10000 elution at ~11.7 min)
is not detected
8000
6000
p-Coumaric aci d
4000 (contaminant)
11.53
2000 11.45
13 14 15 10 11 12 13 14 15
Time (min)
methods”. a Incubation of CYP98A44 with p-coumaroyl-shikimate.
b Incubation of CYP98A44 with p-coumaroyl-malate. The elution
peak of p-coumaric acid (a contaminant in the p-coumaroyl-malate
preparation used as the substrate) is indicated. Elution time of a phas-
elic acid standard is approximately 11.7 min
123
326
Discussion
Red clover leaves contain large amounts of hydroxycinna-
moyl-malate esters, particularly phaselic acid [2-O-
caVeoyl-L-malate], which accumulates to approximately
5
mol g¡1 fresh weight (Sullivan 2009a). In the Brassica-
ceae, hydroxycinnamoyl-malate esters are synthesized by
transfer of hydroxycinnamoyl moieties from a glucose ester
to malic acid by the action of SMT. Because there is no
genetic evidence for SMT in red clover (Sullivan 2009a),
we have proposed alternative pathways (Fig. 1) for phaselic
acid biosynthesis in red clover whereby hydroxycinnamoyl
moieties are transferred from CoA thiolesters to malic acid
acceptors in a manner analogous to chlorogenic acid (caVe-
oyl-quinate) biosynthesis in the Solanaceae (Niggeweg
et al. 2004). In the proposed pathways, phaselic acid could
be formed either by transfer of a p-coumaroyl moiety to
malic acid followed by 3⬘-hydroxylation or by transfer of a
caVeoyl moiety to malic acid. We have recently described a
novel red clover hydroxycinnamoyl-CoA:malate hydroxy-
cinnamoyl transferase (HCT2) from red clover capable of
transferring p-coumaroyl or caVeoyl moieties from their
corresponding CoA thiolesters to malic acid as would be
predicted for either the upper or lower pathway of Fig. 1
(Sullivan 2009a). In vitro speciWc activity of HCT2 is
approximately sevenfold higher for p-coumaroyl-CoA
compared to caVeoyl-CoA, which could suggest phaselic
acid formation in vivo proceeds via the lower pathway of
Fig. 1. Such a pathway would predict a C3⬘H enzyme capa-
ble of directly hydroxylating p-coumaroyl-malate to form
phaselic acid.
Using a homology-based library screen, we identiWed
red clover genes that encode a C3⬘H (CYP98A44). When
expressed in yeast, the enzyme is capable catalyzing the
hydroxylation of p-coumaroyl-shikimate to caVeoyl-shi-
kimate, an activity that has been described for most C3⬘H
enzymes characterized to date (Schoch et al. 2006). Since
hydroxylation of p-coumaroyl-shikimate is a key step in the
formation of monolignol lignin precursors (Schoch et al.
2001; Franke et al. 2002; Reddy et al. 2005), this observed
enzymatic activity, along with the relatively high level of
expression of the gene in red clover stems indicates a likely
role CYP98A44 plays in ligniWcation in red clover.
Red clover CYP98A44 is unable, however, to catalyze
hydroxylation of p-coumaroyl-malate to phaselic acid. This
Wnding suggests that either there is another enzyme in red
clover capable of catalyzing the hydroxylation of p-couma-
royl-malate to phaselic acid, or that formation of phaselic
acid in red clover is accomplished by direct transfer of
caVeoyl moieties to malic acid (upper pathway of Fig. 1).
There is no evidence, however, of additional CYP98 family
C3⬘H enzymes in red clover that could potentially catalyze
the hydroxylation of p-coumaroyl-malate. The library
123
Planta (2010) 231:319–328
screen in this study identiWed only CYP98A44 but was car-
ried out under moderate stringency and should have
allowed identiWcation of related genes with >70–75%
nucleotide identity. It would be expected that such a screen
would identify even relatively divergent members of a red
clover CYP98A gene family given that members of the rel-
atively diverse family of wheat CYP98A genes share
greater than 75% sequence identity overall and have sub-
stantial (>150 bp) stretches with even higher levels of
sequence identity (Morant et al. 2007). Additionally,
tBLASTn searches of a collection of red clover ESTs using
CYP98A44 as the query identiWed only CYP98A44 and
other non-CYP98 cytochrome P450 family members.
Although it would be diYcult to rule out that a member of a
poorly characterized, non-CYP98 cytochrome P450 family
is capable of carrying out hydroxylation of p-coumaroyl-
malate, we have been unable to detect such an activity in
microsome preparations from red clover tissues. Although
many characterized PPOs have phenolase activity and are
capable of forming an o-diphenol from a monophenol
(SteVens et al. 1994), this activity has not been detected for
red clover PPO in vitro (R. HatWeld, unpublished data), and
such a reaction seems unlikely in intact plant tissues since
PPO is localized to the thylaloid lumen (Sullivan et al.
2004) and hydroxycinnamoyl-malate esters are likely local-
ized to the cytoplasm or vacuole (Sharma and Strack 1985).
Together, these data and observations strongly suggest
that red clover does not have an enzyme capable of hydrox-
ylating p-coumaroyl-malate to form phaselic acid and
instead forms phaselic acid by direct transfer of caVeoyl-
moieties to malate (top pathway of Fig. 1). Such caVeoyl
moieties could be formed via the action of red clover
HCT1, a hydroxycinnamoyl-CoA:shikimate hydroxycin-
namoyl transferase (Sullivan 2009a) and subsequent
3⬘-hydroxylation by CYP98A44. Although the in vitro
preference of HCT2 for p-coumaroyl-CoA over caVeoyl-
CoA would suggest a p-coumaroyl-malate intermediate in
phaselic acid biosynthesis (lower pathway of Fig. 1), red
clover leaf extracts contain nearly tenfold more p-couma-
royl-CoA:shikimate transferase activity than p-coumaroyl-
CoA:malate transferase activity (Sullivan 2009a). Thus, in
vivo, p-coumaroyl moieties may be far more likely to be
transferred to shikimate than malate in red clover leaves,
and consequently more Xux might be expected through the
upper, rather than lower, pathway of Fig. 1. Such diVeren-
tial Xux, along with the lack of an enzyme capable of
hydroxylating p-coumaroyl-malate is also consistent with
the observation that p-coumaroyl-malate does accumulate
in red clover leaves, but to about fourfold lower levels than
phaselic acid (Sullivan 2009a and additional unpublished
data).
Oxidation of phaselic acid and other o-diphenol com-
pounds in red clover by endogenous PPOs results in
Planta (2010) 231:319–328
reduced post-harvest proteolysis (Sullivan and HatWeld
2006). Ruminant animals poorly utilize the nitrogen in
degraded protein necessitating the supplementation of diets
with true protein and resulting in excretion of excess nitro-
gen into the environment as urea. Thus, preventing post-
harvest protein losses would have positive economic and
environmental impacts. There are also indications that
forages rich in PPO and o-diphenols help protect protein
and lipid in the rumen (Lee et al. 2004), with the potential
for additional nitrogen utilization eYciency and improved
lipid proWles of animal products, respectively. Unfortu-
nately, many important forages such as alfalfa do not
accumulate the required o-diphenols. Alfalfa is known to
have hydroxycinnamoyl-CoA:shikimate hydroxycinnamoyl
transferase and p-coumaroyl-shikimate 3⬘-hydroxylase
activities (Reddy et al. 2005; Shadle et al. 2007). These
observations, along with the results presented here, i.e. that
a special C3⬘H capable of hydroxylating p-coumaroyl-
malate is likely not required for phaselic acid biosynthesis,
indicate that reconstructing an o-diphenol biosynthetic
pathway in alfalfa might be feasible by the introduction of a
relatively small set of genes.
Acknowledgments This work was supported by the US Department
of Agriculture-Cooperative State Research, Education, and Extension
Service-National Research Initiative Competitive Grants Program
(Grant no. 2009-35318-05048). We wish to thank Sara Zerbel, Lisa
Koch, and Kendal Cooper for excellent technical assistance; Paul
Schatz and John Ralph for providing hydroxycinnamoyl ester stan-
dards; Clint Chapple, Jing-Ke Weng, Dave Gang, and Anna Berim for
providing P450 microsome samples; Philippe Urban and Denis Pom-
pon for providing pYeDP60 and WAT11; Sheryl Rakowski and Mar-
cin Filutowicz for access to ultracentrifugation equipment; and Jane
Marita and Ron HatWeld for technical advice and helpful discussions
regarding this work. Mention of trade names or commercial products
in this article is solely for the purpose of providing speciWc information
and does not imply recommendation or endorsement by the US Depart-
ment of Agriculture.
References
Ausubel FM, Brent R, Kingston RE, Moore DD, Deidman JG, Smith
JA, Struhl K (eds) (1998) Current protocols in molecular biology.
Wiley, New York
Franke R, Humphreys JM, Hemm MR, Denault JW, Ruegger MO,
Cusumano JC, Chapple C (2002) The Arabidopsis REF8 gene en-
codes the 3-hydroxylase of phenylpropanoid metabolism. Plant J
30:33–45
Gang DR, Beuerle T, Ullmann P, Werck-Reichart D, Pichersky E
(2002) DiVerential production of meta-hydroxylated phenlyprop-
anoids in sweet basil peltate gladular trichomes and leaves is con-
trolled by the activities of speciWc acyltransferases and
hydroxylases. Plant Physiol 130:1536–1544
Gietz RD, Woods RA (2002) Transformation of yeast by lithium ace-
tate/single-stranded carrier DNA/polyethylene glycol method.
Methods Enzymol 350:87–96
Grawe W, Bachhuber P, Mock HP, Strack D (1992) PuriWcation and
characterization of sinapolyglucose—malate sinapoyltransferase
from Raphanus sativus L. Planta 187:236–241
327
Harlow E, Lane D (1988) Antibodies: a laboratory manual. Cold
Spring Harbor Laboratory, New York, p 508
Hemmerle H, Burger HJ, Below P, Schubert G, Rippel R, Schindler
PW, Paulus E, Herling AW (1997) Chlorogenic acid and syn-
thetic chlorogenic acid derivatives: novel inhibitors of hepatic
glucose-6-phosphate translocase. J Med Chem 40:137–145
Hulo N, Bairoch A, Bulliard V, Cerutti L, Cuche BA, de Castro E,
Lachaize C, Langendijk-Genevaux PS, Sigrist CJA (2008) The
20 years of PROSITE. Nucleic Acids Res 36:D245–D249
Lee MRF, Winters AL, Scollan ND, Dewhurst RJ, Theodorou MK,
Minchin FR (2004) Plant-mediated lipolysis and proteolysis in
red clover with diVerent polyphenol oxidase activities. J Sci Food
Agric 84:1639–1645
Lehfeldt C, Shirley AM, Meyer K, Ruegger MO, Cusumano JC, Viit-
anen PV, Strack D, Chapple C (2000) Cloning of the SNG1 gene
of arabidopsis reveals a role for a serine carboxypeptidase-like
protein as an acyltransferase in secondary metabolism. Plant Cell
12:1295–1306
Mahesh V, Million-Rousseau R, Ullmann P, Chabrillange N, Busta-
mante J, Mondolot L, Morant M, Noirot M, Hamon S, de Kochko
A, Werck-Reichhart D, Campa C (2007) Functional characteriza-
tion of two p-coumaroyl ester 3⬘-hydroxylase genes from coVee
tree: evidence of a candidate for chlorogenic acid biosynthesis.
Plant Mol Biol 64:145–159
Matsuno M, Nagatsu A, Ogihara Y, Ellis BE, Mizukami H (2002)
CYP98A6 from Lithospermum erythrorhizon encodes 4-couma-
royl-4⬘-hydroxyphenyllactic acid 3-hydroxylase involved in ros-
marinic acid biosynthesis. FEBS Lett 514:219–224
Morant M, Schoch GA, Ullmann P, Ertunc T, Little D, Olsen CE, Pet-
ersen M, Negrel J, Werck-Reichhart D (2007) Catalytic activity,
duplication and evolution of the CYP98 cytochrome P450 family
in wheat. Plant Mol Biol 63:1–19
Niggeweg R, Michael AJ, Martin C (2004) Engineering plants with in-
creased levels of the antioxidant chlorogenic acid. Nat Biotechnol
22:746–754
Pompon D, Louerat B, Bronine A, Urban P (1996) Yeast expression of
animal and plant P450s in optimized redox environments. Meth-
ods Enzymol 272:51–64
Reddy MSS, Chen F, Shadle G, Jackson L, Aljoe H, Dixon RA (2005)
Targeted down-regulation of cytochrome P450 enzymes for for-
age quality improvement in alfalfa (Medicago sativa L.). Proc
Natl Acad Sci USA 102:16573–16578
Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning, a labo-
ratory manual. Cold Spring Harbor Laboratory Press, Cold Spring
Harbor
Sato S, Isobe S, Asamizu E, Ohmido N, Kataoka R, Nakamura Y,
Kaneko T, Sakurai N, Okumura K, Klimenko I, Sasamoto S,
Wada T, Watanabe A, Kohara M, Fujishiro T, Tabata S (2005)
Comprehensive structural analysis of the genome of red clover
(Trifolium pratense L.). DNA Res 12:301–364
Schaller GE, Dewitt ND (1995) Analysis of the H+-ATPase and other
proteins of the arabidopsis plasma membrane. Methods Cell Biol
50:129–148
Schoch G, Goepfert S, Morant M, Hehn A, Meyer D, Ullmann P, Wer-
ck-Reichhart D (2001) CYP98A3 from Arabidopsis thaliana is a
3⬘-hydroxylase of phenolic esters, a missing link in the phenyl-
propanoid pathway. J Biol Chem 276:36566–36574
Schoch GA, Morant M, Abdulrazzak N, Asnaghi C, Goepfert S, Petersen
M, Ullmann P, Werck-Reichhart D (2006) The meta-hydroxylation
step in the phenylpropanoid pathway: a new level of complexity in
the pathway and its regulation. Environ Chem Lett 4:127–136
Shadle G, Chen F, Reddy MSS, Jackson L, Nakashima J, Dixon RA
(2007) Down-regulation of hydroxycinnamoyl CoA: shikimate
hydroxycinnamoyl transferase in transgenic alfalfa aVects ligniW-
cation, development and forage quality. Phytochemistry 68:
1521–1529
123
328
Sharma V, Strack D (1985) Vacuolar localization of 1-sinapoylglu-
cose—L-malate sinapoyltransferase in protoplasts from cotyle-
dons of Raphanus sativus. Planta 163:563–568
Smith RR, Maxwell DP (1980) Registration of WI-1 and WI-2 red
clover. Crop Sci 20:831
SteVens JC, Harel E, Hunt MD (1994) Polyphenol oxidase. In: Ellis
BE, Kuroki GW, StaVord HA (eds) Genetic engineering
of plant secondary metabolism. Plenum Press, New York,
pp 275–312
Stukkens Y, Bultreys A, Grec S, Trombik T, Vanham D, Boutry M
(2005) NpPDR1, a pleiotropic drug resistance-type ATP-bind-
ing cassette transporter from Nicotiana plumbaginifolia,
plays a major role in plant pathogen defense. Plant Physiol 139:
341–352
Sullivan ML (2009a) A novel red clover hydroxycinnamoyl transfer-
ase has enzymatic activities consistent with a role in phaselic acid
biosynthesis. Plant Physiol 150:1866–1879
123
Planta (2010) 231:319–328
Sullivan ML (2009b) Phenylalanine ammonia lyase genes in red clo-
ver: expression in whole plants and in response to yeast fungal
elicitor. Biol Plant 53:301–306
Sullivan ML, HatWeld RD (2006) Polyphenol oxidase and o-diphenols
inhibit postharvest proteolysis in red clover and alfalfa. Crop Sci
46:662–670
Sullivan ML, HatWeld RD, Thoma SL, Samac DA (2004) Cloning and
characterization of red clover polyphenol oxidase cDNAs and
expression of active protein in Escherichia coli and transgenic
alfalfa. Plant Physiol 136:3234–3244
Taylor CB, Bariola PA, Delcardayre SB, Raines RT, Green PJ (1993)
Rns2—a senescence-associated RNase of arabidopsis that
diverged from the S-RNases before speciation. Proc Natl Acad
Sci USA 90:5118–5122
Zdobnov EM, Apweiler R (2001) InterProScan—an integration plat-
form for the signature-recognition methods in InterPro. Bioinfor-
matics 17:847–848