Tomato phenylacetaldehyde reductases catalyze the last step in the synthesis of the aroma volatile 2-phenylethanol

Article abstract of DOI:10.1016/j.phytochem.2007.06.005

The volatile compounds, 2-phenylacetaldehyde and 2-phenylethanol, are important for the aroma and flavor of many foods, such as ripe tomato fruits, and are also major constituents of scent of many flowers, most notably roses. While much work has gone into elucidating the pathway for 2-phenylethanol synthesis in bacteria and yeast, the pathways for synthesis in plants are not well characterized. We have identified two tomato enzymes (LePAR1 and LePAR2) that catalyze the conversion of 2-phenylacetaldehyde to 2-phenylethanol: LePAR1, a member of the large and diverse short-chain dehydrogenase/reductase family, strongly prefers 2-phenylacetaldehyde to its shorter and longer homologues (benzaldehyde and cinnamaldehyde, respectively) and does not catalyze the reverse reaction at a measurable rate; LePAR2, however, has similar affinity for 2-phenylacetaldehyde, benzaldehyde and cinnamaldehyde. To confirm the activity of these enzymes in vivo, LePAR1 and LePAR2 cDNAs were individually expressed constitutively in petunia. While wild type petunia flowers emit relatively high levels of 2-phenylacetaldehyde and lower levels of 2-phenylethanol, flowers from the transgenic plants expressing LePAR1 or LePAR2 had significantly higher levels of 2-phenylethanol and lower levels of 2-phenylacetaldehyde. The in vivo alteration of volatile emissions is an important step toward altering aroma volatiles in plants.

Full text of DOI:10.1016/j.phytochem.2007.06.005

Available online at www.sciencedirect.com
PHYTOCHEMISTRY
Phytochemistry 68 (2007) 2660–2669
www.elsevier.com/locate/phytochem
Tomato phenylacetaldehyde reductases catalyze the last step
in the synthesis of the aroma volatile 2-phenylethanol
a
b
b
Denise M. Tieman , Holly M. Loucas , Joo Young Kim ,
b
a,
*
David G. Clark , Harry J. Klee
a
Horticultural Sciences Department, University of Florida, Gainesville, FL 32611, United States
Department of Environmental Horticulture, University of Florida, Gainesville, FL 32611, United States
b
Received 31 March 2007; received in revised form 25 May 2007
Available online 17 July 2007
Abstract
The volatile compounds, 2-phenylacetaldehyde and 2-phenylethanol, are important for the aroma and flavor of many foods, such as
ripe tomato fruits, and are also major constituents of scent of many flowers, most notably roses. While much work has gone into elu-
cidating the pathway for 2-phenylethanol synthesis in bacteria and yeast, the pathways for synthesis in plants are not well characterized.
We have identified two tomato enzymes (LePAR1 and LePAR2) that catalyze the conversion of 2-phenylacetaldehyde to 2-phenyleth-
anol: LePAR1, a member of the large and diverse short-chain dehydrogenase/reductase family, strongly prefers 2-phenylacetaldehyde to
its shorter and longer homologues (benzaldehyde and cinnamaldehyde, respectively) and does not catalyze the reverse reaction at a mea-
surable rate; LePAR2, however, has similar affinity for 2-phenylacetaldehyde, benzaldehyde and cinnamaldehyde. To confirm the activ-
ity of these enzymes in vivo, LePAR1 and LePAR2 cDNAs were individually expressed constitutively in petunia. While wild type petunia
flowers emit relatively high levels of 2-phenylacetaldehyde and lower levels of 2-phenylethanol, flowers from the transgenic plants
expressing LePAR1 or LePAR2 had significantly higher levels of 2-phenylethanol and lower levels of 2-phenylacetaldehyde. The
in vivo alteration of volatile emissions is an important step toward altering aroma volatiles in plants.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Keywords: Solanum lycopersicum; Solanaceae; 2-Phenylethanol; Tomato flavor; Reductases; Flower scent
1. Introduction
are derived from phenylalanine (1) (Fig. 1) including
2-phenylacetaldehyde (3) and 2-phenylethanol (4) (Buttery,
1993; Baldwin et al., 2000). Both of these volatiles have fru-
ity/floral properties that are considered desirable, although
elevated levels of 2-phenylethanol (4) and 2-phenylacetal-
dehyde (3) have also been associated with undesirable fla-
vor in tomato fruit (Tadmor et al., 2002). Additionally
2-phenylethanol (4) and 2-phenylacetaldehyde (3) are major
constituents of scent in many flowers, with the isomers
being the major aroma volatile contributing to the scent
of rose, while the latter is the major volatile associated with
hyacinth and lilac (Knudsen et al., 1993). Indeed, 2-pheny-
lethanol (4) is the most used fragrance chemical in cosmetic
products (Clark, 1990), as well as being a major contribu-
tor to the overall flavor of diverse products including
The way in which humans perceive flavor is little under-
stood at the molecular level. For example, what we con-
sider to be the flavor of a tomato is the sum of
interactions between sugars, acids and over 400 volatile
compounds (Buttery, 1993; Baldwin et al., 2000). Of these
many volatiles, about 30 are believed to significantly
impact flavor (Buttery et al., 1971; Baldwin et al., 2004).
Moreover, these volatiles are derived from a diverse set of
precursors that include lipids, carotenoids and amino acids.
Several of the most important tomato aroma volatiles
*
Corresponding author. Tel.: +1 352 392 8249; fax: +1 352 846 2063.
E-mail address: hjklee@ifas.ufl.edu (H.J. Klee).
0031-9422/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.phytochem.2007.06.005

D.M. Tieman et al. / Phytochemistry 68 (2007) 2660–2669
2661
Fig. 1. The biosynthetic pathway for 2-phenylethanol (4) from phenylalanine (1) via phenethylamine (2) and 2-phenylacetaldehyde (3) in tomato.
cheese, bread, wine and olive oil. As a consequence, there is
much interest in natural sources of 2-phenylethanol (4) for
the flavor and fragrance industry.
2. Results and discussion
2.1. Identification of a tomato 2-phenylacetaldehyde
reductase
2-Phenylacetaldehyde (3) and 2-phenylethanol (4) have
important biological functions in plants. The latter has
long been known to possess antimicrobial properties
(Berrah and Konetzka, 1962) and its synthesis by plant
reproductive structures may indicate a protective role
for flowers and fruits. Both 2-phenylacetaldehyde (3)
and 2-phenylethanol (4) are also potent insect attractants
(see http://www.pherobase.com/; Pichersky and Gershen-
zon, 2002), with each attracting different sets of pollinat-
ing and predatory insects (Raguso et al., 2003; Zhu et al.,
2005). All of the tomato flavor-associated volatiles are
also likely to have major roles in attracting seed-dispers-
ing organisms (Baldwin et al., 1998; Goff and Klee,
2006). These multiple roles, in both defense and repro-
duction, make regulation of their synthesis critical to
plant survival. We are, therefore, interested in elucidating
the contributions of 2-phenylacetaldehyde (3) and 2-
phenylethanol (4) to these processes in plants. Therefore,
we initiated efforts to identify genes involved in the bio-
synthesis of these and related phenylalanine-derived
volatiles.
Despite the importance of 2-phenylacetaldehyde (3)
and 2-phenylethanol (4) to tomato flavor/aroma and
flower scent, the pathway for biosynthesis is not well
understood in plants. The yeast, Saccharomyces cerevi-
siae, produces 2-phenylethanol (4) from phenylalanine
(1) via phenylpyruvate and 2-phenylacetaldehyde (3)
(Vuralhan et al., 2003). Deuterium feeding studies in rose
(Rosa damascena Mill.), however, indicated the potential
for as many as four pathways in their biosynthesis
(Watanabe et al., 2002). In addition to the yeast path-
way, Watanabe et al. (2002) reported that synthesis can
occur via a phenethylamine (2)/2-phenylacetaldehyde (3)
route as well as a trans-cinnamic acid/phenyllactate path-
way. In earlier work, we had also reported that a tomato
enzyme catalyzes the first step in the pathway to 2-pheny-
lethanol (4), namely the conversion of phenylalanine (1)
to phenethylamine (2) (Tieman et al., 2006). In the pro-
posed pathway to 2-phenylethanol (4), shown in Fig. 1,
the final step is the reduction of 2-phenylacetaldehyde
(3). Herein we show that two tomato (Solanum lycopersi-
cum) aldehyde reductases catalyze the final step in the
pathway, i.e. the conversion of 2-phenylacetaldehyde (3)
to 2-phenylethanol (4).
2-Phenylacetaldehyde (3) and 2-phenylethanol (4) are
major contributors to flavor in many fresh and processed
food products. They also have important and distinct bio-
logical functions in plants as antimicrobial compounds and
as insect attractants/repellants. Thus, the balance between
the concentrations of these two volatile compounds is crit-
ical for attracting appropriate pollinating insects. Although
some information on the pathway(s) for synthesis of 2-phe-
nylacetaldehyde (3) and 2-phenylethanol (4) in plants has
recently become available, proteins catalyzing this reduc-
tive conversion have not been identified (Kaminaga et al.,
2006; Tieman et al., 2006). Knowing that both are major
contributors to tomato aroma, we searched for genes
encoding enzymes that could potentially catalyze the
reduction of the aldehyde (3) to the alcohol (4). In database
searches (http://compbio.dfci.harvard.edu/tgi/cgi-bin/tgi/
gimain.pl?gudb=tomato) for tomato genes with homology
to known alcohol dehydrogenases, we identified two ESTs
whose translated protein sequence indicated significant
similarity to putative Eucalyptus gunnii and Vigna unguicu-
lata cinnamyl alcohol dehydrogenases (Fig. 2, EgCAD1
and VuCPRD14, respectively) as determined by the
tBLASTn procedure of NCBI. However, since the cDNA
for the first clone (LePAR1) was not full-length, its full-
length cDNA was next cloned by 50-RACE. The full-length
sequence was then obtained by PCR, and confirmed by
DNA sequence analysis (GenBank Accession EF613490).
The deduced LePAR1 protein sequence has a calculated
molecular mass of 35,908 Da, and has high homology to
aldehyde reductases from many plant species (Fig. 2).
The cDNA for the second gene (LePAR2, GenBank Acces-
sion EF613491) was full-length, and the encoded protein
had a calculated molecular mass of 35,469 Da. The pro-
teins encoded by the LePAR1 and LePAR2 cDNAs were
most closely related to a putative cinnamyl alcohol dehy-
drogenase from E. gunnii (Goffner et al., 1998) (81% simi-
lar, 72% identical for LePAR1 and 85% similar, 73%
identical for LePAR2). They were also closely related to
a Vigna radiata aldehyde reductase (Guillen et al., 1998)
(77% similar, 67% identical for LePAR1 and 84% simi-
lar/69% identical for LePAR2) (Fig. 2). Overall, LePAR1
and LePAR2 are more closely related to reductases than

2662
D.M. Tieman et al. / Phytochemistry 68 (2007) 2660–2669
NADPH-ferrihemoprotein reductase
CrCPR
PsCPR
TaCPR
Alcohol dehydrogenase
ScAdh4p
StFDH
AtFDH
LeADH2
PhADH1
EgCAD1
LePAR1
LePAR2
VrERE
VuCPRD14
ScAdh5p
ScAdh3p
ScAdh2p
ScAdh1p
LeADH
EgCCR
ZmCCR
S. cerevisiae alcohol dehydrogenase
Cinnamoyl-CoA reductase
ZmCAD
SoCAD
MsCAD
ZmDFR
SbDFR
RhDFR
VvDFR
PhDFR
EgCAD2
NtCAD
DcDFR
AtLELI3-2
AtLCADa
Cinnamyl alcohol dehydrogenase
Dihydroflavonol-4-reductase
Fig. 2. Relationship between LePAR1, LePAR2 and other related alcohol dehydrogenases and reductases of known function. Sequences were aligned
using the ClustalW multiple sequence alignment tool and the phylogenetic tree was generated by the neighbor-joining method. The identifiers and
GenBank accession numbers for the proteins included in the tree are as follows: LePAR1and LePAR2, this paper (black arrows); LeADH, unknown
function, this paper (gray arrow); EgCAD1, Eucalyptus gunnii cinnamyl alcohol dehydrogenase, T10736; VrERE, Vigna radiata aldehyde reductase,
AAD53967; VuCPRD14, V. unguiculata cinnamyl alcohol dehydrogenase, T11610; EgCCR, E. gunnii cinnamoyl CoA reductase, T10733; ZmCCR, Zea
mays cinnamoyl-CoA reductase, CAA66707; TaCPR, Triticum aestivum cytochrome P450 reductase, CAC83301; CrCPR, Catharanthus roseus
cytochrome P450 reductase, Q05001; PsCPR, Papaver somniferum cytochrome P450 reductase, O24424; AtLCADa, Arabidopsis thaliana cinnamyl alcohol
dehydrogenase, T05624; AtELI3-2, A. thaliana cinnamyl alcohol dehydrogenase, S28043; NtCAD, Nicotiana tabacum cinnamyl alcohol dehydrogenase,
S23525; EgCAD2, E. gunnii cinnamyl alcohol dehydrogenase 2, P31655; MsCAD, Medicago sativa cinnamyl alcohol dehydrogenase, P31656; SoCAD,
Saccharum officinarum cinnamyl alcohol dehydrogenase, O82056; ZmCAD, Z. mays cinnamyl alcohol dehydrogenase, T02990; ScAdh1p, Saccharomyces
cerevisiae alcohol dehydrogenase, NP_014555; ScAdh2p, S. cerevisiae alcohol dehydrogenase, NP_014032; ScAdh3p, S. cerevisiae alcohol dehydrogenase,
NP_013800; ScAdh5p, S. cerevisiae alcohol dehydrogenase, NP_009703; PhADH1, Petunia hybrida alcohol dehydrogenase, P25141; LeADH2, Solanum
lycopersicum alcohol dehydrogenase 2, S51826; AtFDH, A. thaliana formaldehyde dehydrogenase, Q96533; StADH, Solanum tuberosum alcohol
dehydrogenase, T07179; ScAdh4p, S. cerevisiae alcohol dehydrogenase, NP_011258; ZmDFR, Z. mays dihydroflavonol-4-reductase, P51108; SbDFR,
Sorghum bicolor NADPH-dependent reductase, T03447; RhDFR, Rosa hybrida dihydroflavonol 4-reductase, BAA12723; VvDFR, Vitis vinifera
dihydroflavonol-4-reductase, P51110; DcDFR, Dianthus caryophyllus dihydroflavonol 4-reductase, T10716; PhDFR, P. hybrida dihydroflavonol-4-
reductase, S07463.
to dehydrogenases. However, their closest relative
EgCAD1 has both reductase and dehydrogenase activities,
having higher affinity for aldehydes than the corresponding
alcohols (Goffner et al., 1998).
several related substrates determined (Table 1). The highest
level of LePAR1 activity was observed with 2-phenylacetal-
dehyde (3) as a substrate (Km = 32 lM), and it also had the
highest turnover number (Kcat) with this substrate. The
Kcat/Km values show that LePAR1 was catalytically the most
active with 2-phenylacetaldehyde (3) as substrate (Table 1).
Lower activities were observed with cinnamaldehyde (6)
(Km = 1179 lM) and benzaldehyde (5) (Km = 572 lM),
whereas activity was not detected with salicylaldehyde.
Similar LePAR2 activities were observed when the sub-
strate was 2-phenylacetaldehyde (3), benzaldehyde (5) or
cinnamaldehyde (6) (Table 1). The LePAR2 turnover
numbers (Kcat) for each substrate were also similar, but
lower than that of LePAR1. Analysis of the reaction
products by gas chromatography (GC) confirmed that
LePAR1 and LePAR2 converted 2-phenylacetaldehyde
(3) to 2-phenylethanol (4) (Fig. 3), with product identi-
fication confirmed by GC–MS. The conversion of one
2.2. Activity of the tomato 2-phenylacetaldehyde reductase
The putative 2-phenylacetaldehyde reductases (LePAR1
and LePAR2) were expressed in Escherichia coli to deter-
mine activity and substrate specificity. To do this, the
LePAR1 or LePAR2 coding regions were cloned into vec-
tor pDEST15 containing a GST-tag and transformed into
E. coli BL21-AI cells for inducible expression. Enzyme
activities of the purified recombinant proteins were deter-
mined on several substrates by measuring the loss of sub-
strate over time. Maximal activity was observed between
pH 6.0 and 7.5 (data not shown), with reductase activities
of the purified proteins on 2-phenylacetaldehyde (3) and

D.M. Tieman et al. / Phytochemistry 68 (2007) 2660–2669
2663
Table 1
Kinetic parameters of LePAR1 and LePAR2
Substrate
Km (lM)
Vmax (nkatal mgꢀ1
)
Kcat (sꢀ1
)
Kcat/Km (lMꢀ1 sꢀ1
)
2-Phenylacetaldehyde (3)
LePAR1
LePAR2
32 10
503
33
31
2
0.96
0.033
O
62
4
Benzaldehyde (5)
LePAR1
LePAR2
572 256
42
176
17
11
1
0.019
0.025
O
6
Cinnamaldehyde (6)
LePAR1
LePAR2
1179 84
44 24
167
21
10
1
0.0088
0.029
O
Activities of purified recombinant LePAR1 and LePAR2 on 2-phenylacetaldehyde and related substrates were determined by measuring the conversion of
NADPH to NADP+ at pH 6.5 at 25 ꢁC. Km is presented as average SE.
molecule of 2-phenylacetaldehyde (3) to 2-phenylethanol
(4) required the conversion of one molecule of NADPH
to NADP+ (data not shown).
dehydrogenase (Fig. 2, EgCAD1), since phylogenetic anal-
ysis groups it most closely with the tomato 2-phenylacetal-
dehyde reductases (Fig. 2, LePAR1 and LePAR2).
We also identified a third protein (Fig. 2, LeADH) clo-
sely related to the E. gunnii and V. unguiculata cinnamyl
alcohol dehydrogenases (Fig. 2). However, no reductase
activity on either 2-phenylacetaldehyde (3), cinnamalde-
hyde (6) or benzaldehyde (5) was observed.
2.3. Expression of LePAR1 and LePAR2 in tomato plants
Expression of LePAR1 and LePAR2 at the mRNA level
was determined for different tissues using quantitative real-
time RT-PCR. LePAR2 mRNA levels were higher thanLe-
PAR1 levels in all tissues examined. LePAR1 and LePAR2
mRNA could be detected in all tissues examined with the
highest expression occurring in flower buds (Fig. 4).
Although expression of the genes was observed throughout
fruit development, LePAR1 and LePAR2 mRNA levels
decreased somewhat during ripening. This highest expres-
sion in flowers was consistent with 2-phenylethanol (4)
emissions, which were highest on a per gram tissue basis
from tomato flowers (data not shown).
LePAR1 and LePAR2 were first identified in the TIGR
tomato EST database as having homology to cinnamyl
alcohol dehydrogenases. However, it is clear that such clas-
sifications are frequently incorrect and must be validated
experimentally. For example, Kim et al. (2004) showed that
many Arabidopsis genes annotated as putative cinnamyl
alcohol dehydrogenases actually encode proteins with
highly varied substrate specificities. The characterization
of the recombinant LePAR1 enzyme showed that it has
the highest activity upon 2-phenylacetaldehyde (3) with
much less affinity for benzaldehyde (5) and cinnamalde-
hyde (6). As has been observed with other cinnamyl alcohol
dehydrogenase/reductase enzymes, the enzymes required
NADPH as a cofactor and no activity was detected with
NADH. Activity for the reverse reaction with 2-phenyleth-
anol (4) as a substrate was not detected (<1% of forward
reaction), thus establishing the enzymes as reductases
rather than dehydrogenases.
Many other plants, such as Nicotiana species and Euca-
lyptus species also produce 2-phenylethanol (4) in their
flowers, leaves or fruits (Hellyer et al., 1966; Oka et al.,
1999; Raguso et al., 2003). Cinnamyl alcohol dehydrogen-
ases have been identified in both tobacco and Eucalyptus;
however, the activity of these enzymes on 2-phenylacetal-
dehyde (3) has not been determined (Knight et al., 1992;
Goffner et al., 1998). It would be particularly interesting
to examine the activity of the E. gunnii cinnamyl alcohol
2.4. Expression of LePARs in transgenic petunia
Tomato does not accumulate significant pools of 2-phe-
nylacetaldehyde (3). Although we did identify transgenic
tomato lines with elevated 2-phenylethanol (4) emissions,
only one was significantly elevated relative to the control
(data not shown), suggesting that synthesis of the sub-
strate, 2-phenylacetaldehyde (3), is a limiting step in 2-
phenylethanol (4) synthesis. In addition, transgenic tomato
lines with reduced LePAR expression did not have reduced
2-phenylethanol (4) levels. These results likely are the con-
sequence of additional proteins with 2-phenylacetaldehyde
reductase activity in tomato. In EST database searches,
over 50 tomato putative alcohol dehydrogenases were
found, suggesting the potential for redundancy in this gene
family.

2664
D.M. Tieman et al. / Phytochemistry 68 (2007) 2660–2669
0.006
0.005
0.004
0.003
0.002
0.001
0
0.2
0.15
0.1
0.05
0
Fig. 4. LePAR1 and LePAR2 expression levels in wild type tomato (M82)
plants. Tomato fruit RNA was prepared from immature green (IMG),
mature green (MG), breaker (Br), turning (Tu) and red ripe (Ri) stages of
development. Values are indicated SD.
flowers collected at night produce approximately 1 and
0.2 lg gfwꢀ1 hꢀ1 of 2-phenylacetaldehyde (3) and 2-
phenylethanol (4), respectively, while ripe tomato fruit
only produce 0.01 and 0.1 ng gfwꢀ1 hꢀ1 of 2-phenylacet-
aldehyde (3) and 2-phenylethanol (4), respectively. The
full-length LePAR cDNAs were introduced separately
into petunia ‘Mitchell Diploid’ under control of the con-
stitutively expressed figwort mosaic virus 35S promoter.
The levels of LePAR expression were measured in the
transgenic petunia flowers by quantitative real-time RT-
PCR (Fig. 5A and B). As expected, wild type flowers
had no detectable RNA, indicating that the real-time
PCR assay is specific for each LePAR, and is not recog-
nizing any petunia LePAR homologs. Wild type petunia
flowers emit high levels of 2-phenylacetaldehyde (3) and
relatively lower levels of 2-phenylethanol (4). A range
of 2-phenylacetaldehyde (3) and 2-phenylethanol (4) lev-
els was observed in the transgenic lines. The lines with
detectable LePAR1 or LePAR2 RNA had higher levels
of 2-phenylethanol (4) and lower levels of 2-phenylacetal-
dehyde (3) emissions than wild type flowers, whereas
lines with low expression had lower levels of 2-phenyleth-
anol (4) (Fig. 5c and d). This major shift in the ratio of
2-phenylacetaldehyde (3) to 2-phenylethanol (4) indicates
that the LePAR enzymes are recognizing 2-phenylacetal-
dehyde (3) in vivo and converting most of it to 2-pheny-
lethanol (4). These data suggest that, in contrast to
tomato, petunia 2-phenylacetaldehyde reductase activity
is normally limiting, causing 2-phenylacetaldehyde (3)
to predominate over 2-phenylethanol (4) in floral volatile
emissions.
Fig. 3. Conversion of 2-phenylacetaldehyde (3) to 2-phenylethanol (4) by
LePAR1 and LePAR2. GC analysis of volatile components of the Control
(no enzyme), LePAR1 or LePAR2 reaction mixtures. 2-Phenylacetalde-
hyde (3) was converted to 2-phenylethanol (4) upon inclusion of
LePAR1or LePAR2.
To validate the role of the LePARs in production of
2-phenylethanol (4) in vivo, the genes were introduced
into Petunia hybrida plants. We chose petunia as an
in vivo model because they accumulate high levels of 2-
phenylacetaldehyde (3), the preferred substrate of
LePAR1 in vitro. Wild type ‘Mitchell Diploid’ petunia

D.M. Tieman et al. / Phytochemistry 68 (2007) 2660–2669
2665
Fig. 5. LePAR1 andLePAR2 expression and volatile emissions in transgenic petunia flowers. (a) Expression of LePAR1 RNA in flowers of wild type
(MD) and six independent transgenic petunia lines expressing LePAR1 (1, 5, 6, 8, 13, 17). (b) Expression of LePAR2 RNA in flowers of wild type (MD)
and six independent transgenic petunia lines expressing LePAR2 (7, 8, 10, 12, 32). RNA was quantified using Taqman quantitative RT-PCR. (c) 2-
Phenylacetaldehyde (3) and 2-phenylethanol (4) emissions from wild type and transgenic flowers expressing LePAR1 ( SE). (d) 2-Phenylacetaldehyde (3)
and 2-phenylethanol (4) emissions from wild type and transgenic flowers expressing LePAR2 ( SE).
Three of the petunia lines with the highest levels of
LePAR1 transgene expression and 2-phenylethanol (4) lev-
els were self-pollinated and homozygous progeny were
identified. Volatile emissions from flowers of these lines
had significantly decreased levels of 2-phenylacetaldehyde
(3) and increased levels of 2-phenylethanol (4) compared
to controls (Fig. 6). Although LePAR1 is active on benzal-
dehyde (5) as a substrate in vitro and benzaldehyde (5) is
present at higher levels than 2-phenylacetaldehyde (3),
the transgenic petunia plants did not, in general, have
Fig. 6. Petunia floral volatile emissions. Levels of volatiles emitted by flowers from wild type Mitchell Diploid (MD) and three independent homozygous
lines (6, 8 and 39) expressing the tomato LePAR1 cDNA ( SE, n = 6). Asterisks ( , or ) indicate results significantly different from wild type
* ** ***
(P < 0.05, 0.01 or 0.001, respectively).

2666
D.M. Tieman et al. / Phytochemistry 68 (2007) 2660–2669
significantly altered emissions of benzaldehyde (5) or ben-
zyl alcohol. Benzaldehyde (5) emissions were somewhat
lower in one transgenic line than in wild type flowers,
although benzyl alcohol levels were not significantly differ-
ent (Fig. 6, line 39). Thus, the in vivo data are consistent
with a specific function for LePAR1 in the reduction of
2-phenylacetaldehyde (3) to its alcohol (4). Overall, these
data indicate that the introduction of the LePAR1 trans-
gene results in the conversion of 2-phenylacetaldehyde (3)
to 2-phenylethanol (4) in petunia flowers. Although
LePAR1 has the highest affinity for 2-phenylacetaldehyde
(3) as a substrate, it does have activity on benzaldehyde
present in petunia with preference for either benzaldehyde
(5) or 2-phenylacetaldehyde (3). It is possible than increases
in benzyl alcohol are not seen because it is further con-
verted to other compounds. It has been predicted by meta-
bolic flux analysis that benzyl alcohol is further converted
to benzyl benzoate in petunia (Boatright et al., 2004).
LePAR1 and LePAR2 can convert 2-phenylacetaldehyde
(3) to 2-phenylethanol (4) in transgenic petunia flowers.
Petunia is an excellent system for the study of the action
of this enzyme in vivo, since the flowers emit both metabo-
lites. Transgenic plants expressing the tomato gene have
flowers that emit much reduced levels of 2-phenylacetalde-
hyde (3) and higher levels of 2-phenylethanol (4), providing
in vivo confirmation of the action of LePAR1. Interestingly,
although the levels of 2-phenylethanol (4) are higher in the
transgenic LePAR1-expressing petunia flowers, the
increases are not equivalent on a molar basis to the reduc-
tions in 2-phenylacetaldehyde (3). We cannot exclude a feed-
back system in which 2-phenylethanol (4) negatively
regulates flux through the pathway. Alternatively, a portion
of the 2-phenylethanol (4) produced in the transgenic petu-
nia flowers may be converted to 2-phenylethanol b-D-gluco-
pyranoside. Glucosylation occurs in rose flowers
administered deuterium-labeled phenylalanine. In rose
flowers, 2-phenylethyl-b-D-glucopyranoside accumulates
before flower opening and declines after flower opening
and 2-phenylethanol (4) evolution. An increase in b-glucosi-
dase activity was observed upon flower opening, consistent
with the emitted 2-phenylethanol (4) being released from the
2-phenylethyl-b-D-glucopyranoside (Watanabe et al., 2002;
(5) in vitro. However, as we would predict from the Kcat
/
Km (Table 1), we saw no significant effect of transgene
expression upon benzyl alcohol or benzaldehyde (5) levels
in the transgenic petunias.
Progeny from three LePAR2 expressing lines were also
examined for altered volatile levels. As in the primary
transformants, 2-phenylacetaldehyde (3) levels were
reduced and 2-phenylethanol (4) levels were higher in the
transgenic flowers than in control flowers. One LePAR2
expressing line (Fig. 7, line 7) also had lower levels of benz-
aldehyde (5), although benzyl alcohol levels were not
affected. Although the Km values for 2-phenylacetaldehyde
(3) and benzaldehyde (5) were similar, the levels of benzyl
alcohol were not significantly affected in the transgenic
petunias. It is possible that the petunia homolog of
LePAR1 has higher affinity for benzaldehyde (5) and con-
tributes to the conversion of both substrates to their corre-
sponding alcohols. Alternatively, separate enzymes may be
Fig. 7. Petunia floral volatile emissions. Levels of volatiles emitted by flowers from wild type Mitchell Diploid (MD) and progeny from three independent
lines (7, 8 and 10) expressing the tomato LePAR2 cDNA ( SE, n = 6). Asterisks ( , or ) indicate results significantly different from wild type
* ** ***
(P < 0.05, 0.01 or 0.001), respectively.

D.M. Tieman et al. / Phytochemistry 68 (2007) 2660–2669
2667
Oka et al., 1999; Hayashi et al., 2004). We have detected sig-
nificant quantities of 2-phenylethyl-b-D-glucopyranoside in
petunia flowers (unpublished results).
were enclosed in glass tubes, air filtered through a hydro-
carbon trap (Agilent, Palo Alto, CA) flowed through the
tubes for 1 h with the aid of a vacuum pump connected
to a Super Q column. Volatiles collected on the Super Q
column were eluted with CH2Cl2, with the volatiles sepa-
rated on an Agilent (Palo Alto, CA) DB-5 column and ana-
lyzed on an Agilent 6890N gas chromatograph; retention
times were compared to known standards. Identities of vol-
atile peaks were confirmed by GC–MS as described by Sch-
melz et al. (2001). GC standards were purchased from
Sigma–Aldrich (St. Louis, MO).
LePAR1 is a member of the short-chain dehydrogenase/
reductase (SDR) family. This is a large and diverse family
of over 3000 enzymes from all kingdoms with broad sub-
strate specificities that include many aromatic compounds
(Persson et al., 2003). The Arabidopsis thaliana genome
contains 138 members of this diverse family. Phylogenetic
analysis groups LePAR1 with a E. gunnii enzyme called a
cinnamyl alcohol dehydrogenase; however, the Eucalyptus
gene does not group with other characterized cinnamyl
alcohol dehydrogenases (Fig. 2), and has a strong prefer-
ence for aldehyde substrates over their corresponding alco-
hols. The Eucalyptus enzyme has a broad substrate
specificity but 2-phenylacetaldehyde (3) was not tested as
a potential substrate (Goffner et al., 1998). Another closely
related aromatic aldehyde reductase from V. radiata was
shown to reduce the fungal toxin eutypine (Guillen et al.,
1998). Again, this enzyme was not tested with 2-phenyl-
acetaldehyde (3) as a substrate. Since many plants have
the capacity to synthesize 2-phenylethanol (4), it is likely
that a number of the genes annotated as cinnamyl alcohol
dehydrogenases actually encode enzymes that reduce
2-phenylacetaldehyde (3) to 2-phenylethanol (4).
3.3. LePAR expression in E. coli
A partial LePAR1 cDNA was obtained from the TIGR
database. The full-length 2-phenylacetaldehyde reductase
cDNA was cloned by 50-RACE from tomato fruit cDNAs
using primer 50TCCTTGGCCCCACCAAGAGAAAG-
CAAGTGCTGCGT. Following sequence analysis of the
50-RACE products, the full-length cDNA was obtained
by PCR. The full-length LePAR2 gene sequence was
obtained from the TIGR database. The LePAR2 coding
region was amplified from cDNA with forward primer
50CACCATGGCGATGAGAACAGTATGTGTAACAG
and reverse primer 50TCAGTAAAACTTCTTCTCTT-
TCAAGCTTTCAGC. The LeADH coding region was
amplified from cDNA with forward primer 50CAC-
CATGGAAAGTAAGAATATTGGAG and reverse pri-
mer 50TTAAATGTGGAGGAAGTTC. The sequences of
the LePAR1, LePAR2 and LeADH cDNAs have been
deposited in Genbank (Accession Nos. EF613490,
EF613491, and EF613492, respectively). The coding
regions of both LePARs were cloned into vector
pENTR/D-TOPO (Invitrogen, Carlsbad, CA). The coding
region was then recombined into vector pDEST15 contain-
ing a GST tag (Invitrogen, Carlsbad, CA), and trans-
formed into E. coli strain BL21-AI (Invitrogen, Carlsbad,
CA) for inducible protein expression. To purify GST-
tagged LePAR proteins, bacterial cultures were centrifuged
at 5000g for 5 min, followed by sonication for 1 min in 1·
PBS buffer and centrifugation at 10,000g for 15 min. The
GST-tagged protein was purified from the supernatant by
using Glutathione-Uniflow resin (BD Biosciences, Moun-
tain View, CA) according to the manufacturer’s instruc-
tions followed by removal of glutathione with PD-10
desalting columns (Amersham Biosciences, Piscataway,
NJ). Protein purity was determined by SDS–PAGE fol-
lowed by staining with Coomassie brilliant blue. Enzyme
activity of E. coli-expressed proteins was determined by
the method of Larroy et al. (2002) using 2-phenylacetalde-
hyde (3), cinnamaldehyde (6) or benzaldehyde (5) as a
substrate. Assays were performed in triplicate at 25 ꢁC in
a 100 ll reaction. The reaction was monitored at 365 nm,
and an extinction coefficient of 3.4 · 103 Mꢀ1 cmꢀ1 was
used to calculate the conversion of NADPH to NADP+.
The apparent Km and Vmax values were determined using
Lineweaver–Burk plots. Protein concentrations were
2.5. Concluding remarks
The identification of genes encoding enzymes in the
pathways to flavor and aroma volatiles should enable
researchers to manipulate fruit flavor and flower scent by
conventional breeding or metabolic engineering. Indeed,
expression of LePAR1 orLePAR2 in transgenic petunias
does significantly alter the volatile profile of flowers. It will
be interesting to determine the effects of the altered volatile
emissions on attraction of pollinating insects as well as
human preferences. This is an step toward elucidation of
the pathway(s) to the flavor and aroma volatile, 2-phenyl-
ethanol (4).
3. Experimental
3.1. Plant material
Tomato (S. lycopersicum cv. M82) plants were grown in
the greenhouse under standard conditions (Tieman et al.,
2001), whereas petunia plants (Petunia · hybrida, cv.
Mitchell Diploid) were greenhouse grown as described ear-
lier (Underwood et al., 2005).
3.2. Volatile collection
Petunia volatiles were collected from fully open flowers
with nonyl acetate as an internal standard as described by
Schmelz et al. (2003). Flowers were harvested in the even-
ing after sunset when volatile levels were high. Flowers

2668
D.M. Tieman et al. / Phytochemistry 68 (2007) 2660–2669
measured using a Bradford protein determination kit using
BSA as a standard (Bio-Rad Laboratories, Hercules, CA).
Dehydrogenase activity of the reverse reaction was mea-
sured using 2-phenylethanol (4) as a substrate, and the con-
version of NADP+ to NADPH was monitored. For GC
analysis of reaction products the reaction mixture contain-
ing nonyl acetate as a control for recovery was extracted
twice with an equal volume of CH2Cl2. Volatiles were sep-
arated on an Agilent DB-5 column and analyzed on an
Agilent 6890N gas chromatograph with a flame ionization
detector and retention times compared to known stan-
dards. GC–MS was performed on an Agilent 6890N gas
chromatograph with an Agilent 5975 mass selective detec-
tor. The reaction mixtures produced the characteristic 2-
phenylethanol (4) mass-spectral fragments of m/z 102, 91,
77, 65, 51 as well as the parent ion of m/z = 122.
supported by funding from the National Science Founda-
tion (DBI-0211875 to H.J.K.) as well as the Florida Agri-
cultural Experiment Station.
References
Baldwin, E.A., Goodner, K., Plotto, A., Pritchett, K., Einstein, M., 2004.
Effect of volatiles and their concentration on perception of tomato
descriptors. J. Food Sci. 69, S310–S318.
Baldwin, E.A., Scott, J.W., Einsein, M.A., Malundo, T.M.M., Carr, B.T.,
Shewfelt, R.L., Tandon, K.S., 1998. Relationship between sensory and
instrumental analysis of tomato flavor. J. Am. Soc. Hort. Sci. 123,
906–915.
Baldwin, E.A., Scott, J.W., Shewmaker, C.K., Schuch, W., 2000. Flavor
trivia and tomato aroma: biochemistry and possible mechanisms for
control of important aroma components. HortScience 35, 1013–1022.
Berrah, G., Konetzka, W.A., 1962. Selective and reversible inhibition of
the synthesis of bacterial deoxyribonucleic acid by phenethyl alcohol.
J. Bacteriol. 83, 738–744.
3.4. RNA expression analysis
Boatright, J., Negre, F., Chen, X., Kish, C.M., Wood, B., Peel, G., Orlova,
I., Gang, D., Rhodes, D., Dudareva, N., 2004. Understanding in vivo
benzenoid metabolism in petunia petal tissue. Plant Phys. 135, 1–19.
Buttery, R.G., 1993. Quantitative and sensory aspects of flavour of tomato
and other vegetables and fruits. In: Acree, T.E., Teranishi, R. (Eds.),
Flavor Science: Sensible Principles and Techniques. Amer. Chem. Soc.,
Washington, DC, pp. 259–286.
Buttery, R.G., Seifert, R.M., Guadagni, D.G., Ling, L.C., 1971. Charac-
terization of additional volatile components of tomato. J. Agric. Food
Chem. 19, 524–529.
Clark, G.S., 1990. Phenethyl alcohol. Perfum. Flavor 15, 37–44.
Goff, S.A., Klee, H.J., 2006. Plant volatile compounds: sensory cues for
health and nutritional value? Science 311, 815–819.
Goffner, D., Van Doorsselaere, J., Yahiaoui, N., Samaj, J., Grima-
Pettenati, J., Boudet, A.M., 1998. A novel aromatic alcohol dehydro-
genase in higher plants: molecular cloning and expression. Plant Mol.
Biol. 36, 755–765.
Total RNA was extracted using a Qiagen (Valencia,
CA) total RNA extraction kit followed by DNase treat-
ment to remove any contaminating DNA. LePAR mRNA
levels were measured by real-time quantitative RT-PCR
using Taqman One-Step RT-PCR reagents and a Gene-
Amp 5700 Sequence Detection System (Applied Biosys-
tems, Foster City, CA). Primers and probe for real-time
PCR were as follows: LePAR1 Taqman probe, 506-FAM-
CCATCAGTAAAACGAGTTGTTTTAACGTCTTCCA-
30BHQ; LePAR1 forward primer, CGGGTCATGTGC-
CAAAGC; LePAR1 reverse primer, GGCTGACCAC-
TGTAAGCAACTG; LePAR2 Taqman probe, 6-FAM-
TCCCTGGATGGAGCTAAGGAGAGACTCC-30BHQ;
LePAR2 forward primer, ATGATCAAAAAAAGGTA-
GATCACTTGAC; LePAR2 reverse primer, CCAGTA-
GGTCCGCTTTGAACA. For absolute quantification of
RNA, a standard curve was constructed from LePAR1
RNA. LePAR1 RNA was synthesized by in vitro transcrip-
tion of the coding sequence as described previously (Tie-
man et al., 2001).
Guillen, P., Guis, M., Martinez-Reina, G., Colrat, S., Clamyrac, S.,
Deswarte, C., Bouzayen, M., Roustan, J.-P., Fallot, J., Pech, J.C.,
Latche, A., 1998. A novel NADPH-dependent aldehyde reductase gene
from Vigna radiata confers resistance to the grapevine fungal toxin
eutypine. Plant J. 16, 335–343.
Hayashi, S., Yagi, K., Ishikawa, T., Kawasaki, M., Asai, T., Picone, J.,
Turnbull, C., Hiaratake, J., Sakata, K., Takada, M., Ogawa, K.,
Watanabe, N., 2004. Emission of 2-phenylethanol from its b-D-
glucopyranoside and the biogenesis of these compounds from [2H8]L-
phenylalanine in rose flowers. Tetrahedron 60, 7005–7013.
Hellyer, R.O., Lassak, E.V., McKern, H.H.G., 1966. The volatile oils of
the genus Eucalyptus (family Myrtaceae). V. The leaf oil of E.
aggregata Deane and Maiden. Aust. J. Chem. 19, 1765–1767.
Jorgensen, R.A., Cluster, P.D., English, J., Que, Q., Napoli, C.A., 1996.
Chalcone synthase cosuppression phenotypes in petunia flowers:
comparison of sense vs. antisense constructs and single-copy vs.
complex T-DNA sequences. Plant Mol. Biol. 31, 957–973.
Kaminaga, Y., Schnepp, J., Peel, G., Kish, C.M., Ben-Nissan, G., Weiss,
D., Orlova, I., Lavie, O., Rhodes, D., Wood, K., Porterfield, D.M.,
Cooper, A.J.L., Schloss, J.V., Pichersky, E., Vainstein, A., Dudareva,
N., 2006. Plant phenylacetaldehyde synthase is a bifunctional homo-
tetrameric enzyme that catalyzes phenylalanine decarboxylation and
oxidation. J. Biol. Chem. 281, 23357–23366.
3.5. Production of transgenic petunia plants
The full-length 2-phenylacetaldehyde reductase cDNAs
were cloned into a vector under the control of the figwort
mosaic virus 35S promoter (Richins et al., 1987) and fol-
lowed by the Agrobacterium nopaline synthase (nos) 30 ter-
minator. The transgene was introduced into P. hybrida cv.
Mitchell Diploid by the method of Jorgensen et al. (1996)
with kanamycin resistance as a selectable marker. Trans-
genic plants were grown to maturity in a greenhouse and
self-pollinated to obtain seed.
Kim, S.-J., Kim, M.-R., Bedgar, D.L., Moinuddin, S.G.A., Cardenas,
C.L., Davin, L.B., Kang, C., Lewis, N.G., 2004. Functional reclassi-
fication of the putative cinnamyl alcohol dehydrogenase multigene
family in Arabidopsis. Proc. Natl. Acad. Sci. USA 101, 1455–1460.
Knight, M.E., Halpin, C., Schuch, W., 1992. Identification and charac-
terization of cDNA clones encoding cinnamyl alcohol dehydrogenase
from tobacco. Plant Mol. Biol. 19, 739–801.
Acknowledgements
We thank Dr. Andrew Hanson for helpful discussion
and critical reading of the manuscript. This work was

D.M. Tieman et al. / Phytochemistry 68 (2007) 2660–2669
2669
Knudsen, J.T., Tollsten, L., Bergstron, L.G., 1993. Floral scents – a
checklist of volatile compounds isolated by head-space techniques.
Phytochemistry 33, 253–280.
Larroy, C., Fernandez, M.R., Gonzalez, E., Pares, X., Biosca, J.A., 2002.
Characterization of the Saccharomyces cerevisiae YMR318C (ADH6)
gene product as a broad specificity NADPH-dependent alcohol dehy-
drogenase: relevance in aldehyde reduction. Biochem. J. 361, 163–172.
Oka, N., Ohishi, H., Hatano, T., Hornberger, M., Sakata, K., Watababe,
N., 1999. Aroma evolution during flower opening in Rosa damascena
Mill. Z. Naturforsch. 54c, 889–895.
Tadmor, Y., Fridman, E., Gur, A., Larkov, O., Lastochkin, E., Ravid, U.,
Zamir, D., Lewinsohn, E., 2002. Identification of malodorous, a wild
species allele affecting tomato aroma that was selected against during
domestication. J. Agric. Food Chem. 50, 2005–2009.
Tieman, D.M., Ciardi, J.A., Taylor, M.G., Klee, H.J., 2001. Members of
the tomato LeEIL (EIN3-like) gene family are functionally redundant
and regulate ethylene responses throughout plant development. Plant
J. 26, 47–58.
Tieman, D., Taylor, M., Schauer, N., Fernie, A.R., Hanson, A.D.,
Klee, H.J., 2006. Tomato aromatic amino acid decarboxylases
participate in synthesis of the flavor volatiles 2-phenylethanol and
2-phenylacetaldehyde. Proc. Natl. Acad. Sci. USA 103, 8287–
8292.
Persson, B., Kallberg, Y., Oppermann, U., Jo¨rnvall, H., 2003. Coenzyme-
based functional assignments of short-chain dehydrogease/reductases
(SDRs). Chem. Biol. Interact. 143, 271–278.
Pichersky, E., Gershenzon, J., 2002. The formation and function of plant
volatiles: perfumes for pollinator attraction and defense. Curr. Opin.
Plant Biol. 5, 237–243.
Raguso, R.A., Levin, R.A., Foose, S.E., Holmberg, M.W., McDade,
L.A., 2003. Fragrance chemistry, nocturnal rhythms and pollination
‘‘syndromes’’ in Nicotiana. Phytochem. 63, 265–284.
Richins, R.D., Scholthof, H.B., Shepard, R.J., 1987. Sequence of figwort
mosaic virus DNA (caulimovirus group). Nucl. Acids Res. 15, 8451–
8466.
Schmelz, E.A., Alborn, H.T., Banchio, E., Tumlinson, J.H., 2003.
Quantitative relationships between induced jasmonic acid levels and
volatile emission in Zea mays during Spodoptera exigua herbivory.
Planta 216, 665–673.
Underwood, B.A., Tieman, D.M., Shibuya, K., Dexter, R.J., Loucas,
H.M., Simkin, A.J., Sims, C.A., Schmelz, E.A., Klee, H.J., Clark,
D.G., 2005. Ethylene-regulated floral volatile synthesis in petunia
corollas. Plant Physiol. 138, 255–266.
Vuralhan, Z., Morais, M.A., Tai, S.-L., Piper, M.D.W., Pronk, J.T., 2003.
Identification and characterization of phenylpyruvate decarboxylase
genes in Saccharomyces cerevisiae. Appl. Environ. Microbiol. 69,
4534–4541.
Watanabe, S., Hayashi, K., Yagi, K., Asai, T., MacTavish, H., Picone, J.,
Turnbull, C., Watanabe, N., 2002. Biogenesis of 2-phenylethanol in
rose flowers: incorporation of [2H8]L-phenylalanine into 2-phenyleth-
anol and its b-D-glucopyranoside during the flower opening of Rosa
‘Hoh-Jun’ and Rosa damascena Mill. Biosci. Biotechnol. Biochem. 66,
943–947.
Zhu, J., Obrycki, J., Ochieng, S., Baker, T., Pickett, J., Smiley, D., 2005.
Attraction of two lacewing species to volatiles produced by host plants
and aphid prey. Naturwissenschaften 92, 277–281.
Schmelz, E.A., Alborn, H.T., Tumlinson, J.H., 2001. The influence of
intact-plant and excised-leaf bioassay designs on volicitin- and
jasmonic acid-induced sesquiterpene volatile release in Zea mays.
Planta 214, 171–179.