Synthesis of non-natural flavanones and dihydrochalcones in metabolically engineered yeast

Article abstract of DOI:10.1016/j.molcatb.2010.05.017

Flavonoids are plant phenolic compounds that have many interesting medicinal properties. Therefore, there is interestin the synthesisof non-natural flavonoidsas they may possess neworenhanced biological activities. In this study, metabolically engineered Saccharomyces cerevisiae expressing 4-coumaroyl:CoA-ligase (4CL) and chalcone synthase (CHS)was explored as a platform for producingnon-natural flavanones and dihydrochalcones. By precursor addition of cinnamic acid analogues to the engineered yeast, numerous non-natural flavanones and dihydrochalcones were formed in vivo. Also, several CHS derailment products were formed. Of the isolated compounds, one flavanone and three derailment products were found to be novel compounds.

Full text of DOI:10.1016/j.molcatb.2010.05.017

Journal of Molecular Catalysis B: Enzymatic 66 (2010) 257–263
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Synthesis of non-natural flavanones and dihydrochalcones in
metabolically engineered yeast
Sean R. Werner, Hao Chen, Hanxiao Jiang, John A. Morgan∗
School of Chemical Engineering, Purdue University, West Lafayette, IN 47907, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Flavonoids are plant phenolic compounds that have many interesting medicinal properties. Therefore,
there is interest in the synthesis of non-natural flavonoids as they may possess new or enhanced biological
activities. In this study, metabolically engineered Saccharomyces cerevisiae expressing 4-coumaroyl:CoA-
ligase (4CL) and chalcone synthase (CHS) was explored as a platform for producing non-natural flavanones
and dihydrochalcones. By precursor addition of cinnamic acid analogues to the engineered yeast, numer-
ous non-natural flavanones and dihydrochalcones were formed in vivo. Also, several CHS derailment
products were formed. Of the isolated compounds, one flavanone and three derailment products were
found to be novel compounds.
Received 1 December 2009
Received in revised form 26 May 2010
Accepted 31 May 2010
Available online 8 June 2010
Keywords:
Phloretin
Polyketides
Galactose inducible promoter
Saccharomyces cerevisiae
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
incorporated in their structures) may also possess interest-
ing biological activity, and hence have potential pharmaceutical
Flavonoids are a large class of secondary metabolites derived
from the phenylpropanoid pathway. In plants, flavonoids have
several functions including act as pigments, as well as pro-
tection against UV radiation, microbial invasion, insects, and
animals [1–3]. It has been shown that many flavonoids bene-
fit human health, as they possess antioxidant, anti-inflammatory,
and anti-carcinogenic activity [4–9] In the first step of flavonoid
biosynthesis, phenylalanine and/or tyrosine are deaminated
by phenylalanine ammonia-lyase (PAL) to give cinnamic acid
and p-coumaric acid, respectively. Cinnamate-4-hydroxylase
(C4H) hydroxylates cinnamic acid at the para position, giving
p-coumaric acid. Phenylpropanoic acids are ligated with coen-
zyme A (CoA) by 4-coumaroyl:CoA-ligase (4CL), creating CoA
thioesters. Chalcone synthase (CHS) then catalyzes the conden-
sation of phenylpropanoid CoA thioesters with three acetate
units from malonyl-CoA to form chalcones. Chalcones can
either undergo ring closure, to give a racemic mixture of fla-
vanones, or a single enantiomer by chalcone isomerase (CHI).
Flavanones are the precursors to the further biosynthesis of other
flavonoids, such as flavones, flavonols, isoflavones, and antho-
cyanins [10–12].
uses. Therefore, there is an interest in synthesizing non-
natural flavonoids. Chemical-enzymatic synthesis of non-natural
flavonoids and polyketides have been pursued, but these meth-
ods require enzymatic or chemical synthesis of CoA thioesters
which are then used as substrates to the polyketide synthase
[13–15]. The enzymatic methods require the use of expensive
cofactors (e.g. ATP, CoA) limiting the ability to scale up the syn-
thesis. There has been considerable recent interest in the area of
metabolic engineering of microorganisms for the production of
flavonoids [12,16,17]. Flavonoids can be synthesized by expressing
enzymes involved in flavonoid biosynthesis in recombinant hosts
such as Saccharomyces cerevisiae and E. coli, where the cofactors
required by the enzymes are biosynthesized by the host. These
metabolically engineered microorganisms can be used to synthe-
size non-natural flavonoids by a precursor directed biosynthesis
approach [18]. First, a non-natural substrate is added to an organ-
ism expressing a biosynthetic pathway, and if the substrate and
the corresponding intermediates are accepted by the enzymes in
the pathway, this non-natural substrate is incorporated into the
final products. This method has shown promise for production of
non-natural flavonoids such as flavanones, flavonols, and stilbenes
[19,20].
Since flavonoids have significant medicinal properties, non-
natural flavonoids (flavonoids with non-natural substituents
This study demonstrates the ability of metabolically engi-
neered S. cerevisiae, expressing 4CL and CHS from the flavonoid
biosynthesis pathway, to synthesize novel non-natural fla-
vanones from cinnamic acid analogues and dihydrochalcones from
phenylpropanoic acids by the precursor directed biosynthesis
approach.
∗
Corresponding author at: FRNY Hall of Chemical Engineering, West Lafayette,
IN 47907-2050, United States. Tel.: +1 765 494 4088; fax: +1 765 494 0805.
E-mail address: jamorgan@purdue.edu (J.A. Morgan).
1381-1177/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2010.05.017

258
S.R. Werner et al. / Journal of Molecular Catalysis B: Enzymatic 66 (2010) 257–263
2. Materials and methods
ysis. Product formation was determined by formation of new peaks
on the HPLC chromatogram.
2.1. Chemicals
2.5. Fermentation and product isolation
2-Aminocinnamic acid and 3-aminocinnamic acid was
purchased from City Chemical LLC (West Haven, CT). 2-
Chlorocinnamic acid was purchased from ACROS Organics
(Geel, Belgium). 3-Methoxycinnamic acid, 2-methylcinnamic acid,
2-methoxycinnamic acid, was purchased from Alfa Aesar (Ward
Hill, MA). 5-Bromo-2-methoxy cinnamic acid, 2-chloro-6-fluoro
cinnamic acid, 3-chlorocinnamic acid, trans-2,3-dimethoxy-
cinnamic acid, 2-ethoxycinnamic acid, 2-fluorocinnamic acid,
3-fluorocinnamic acid, hydrocinnamic acid (phenylpropanoic
acid), 2-hydroxycinnamic acid, 3-hydroxycinnamic acid, 3-
methylcinnamic acid, 4-methoxycinnamic acid, 2-nitrocinnamic
acid, 3-nitrocinnamic acid, phloretic acid (dihydro-p-coumaric
acid), urocanic acid (3-(4-Imidazolyl)acrylic acid), were of the
highest available purity and purchased from Sigma–Aldrich Co. (St.
Louis, MO). Phloretin was purchased from MP Biomedical (Solon,
OH).
A single colony of AH22 harboring pKS2␮Hyg-4CL-CHS from
a YPDH plate was inoculated in 50 mL of YPDH in a 250 mL flask
and grown for about 24 h at 30 ◦C with shaking at 300 rpm. Three
1-L flasks containing 300 mL of YPLH were inoculated with the
appropriate volume of cells from overnight culture to achieve an
OD600 = 0.05. After 5 h the cinnamic acid analogue was added to the
culture in powder form to achieve a calculated initial concentration
of 1 mM. Cultures were incubated at 30 ◦C with shaking at 300 rpm
for 3 days. After fermentation, the cells were removed by centrifu-
gation at 600 × g (2000 rpm using JA-14 rotor) for 10 min. Next,
the combined supernatant was extracted twice with equal volumes
of ethyl acetate. The ethyl acetate fractions were concentrated by
rotary evaporation and the isolated residue was resuspended in
methanol and stored at −20 ◦C.
2.6. Chromatography and product purification
2.2. Strain and plasmid
Flavonoids produced by AH22 expressing 4CL and CHS were
analyzed by high performance liquid chromatography (HPLC) using
an Agilent Zorbax SB-C18 column (4.6 mm × 75 mm) with the col-
umn temperature maintained at 30 ◦C. Solvent A was 1.5% (v/v)
acetic acid in water; solvent B was methanol. Solvent B was main-
tained at 10% for 4 min, then increased to 60% over 42 min, and
held for 4 min before increased to 70% in 2 min and maintained for
2 min. Solvent B was returned to 10% over 2 min. Phloretin produc-
tion was analyzed by HPLC using the following method. Solvent A
was 1.5% (v/v) acetic acid in water and solvent B was acetonitrile.
Solvent A was then held at 90% for 4 min. Solvent A was decreased
from 90% to 40% over 42 min, and then held at 40% for 4 min. Sol-
vent A was decreased further to 30% over 2 min and held for 2 min
before returning back to 90%. The flow rate was 0.9 mL/min with an
injection volume of 20 ␮L. Preparative HPLC was performed using
an Agilent Zorbax SB-C18 column (9.4 mm × 100 mm) with the col-
umn temperature maintained at 30 ◦C. Solvent A was 1.5% (v/v)
acetic acid in water; solvent B was acetonitrile. The flow rate was
2.0 mL/min. Solvent A was held at 95% for 2 min and then decreased
from 95% to 35% over 23 min. Solvent A was then increased back to
95% over 2 min. Fractions were collected with a Foxy Jr. fraction
collector (Teledyne Isco, Lincoln, NE).
Saccharomyces cerevisiae strain AH22 (MATa leu2-3 leu2-112
his4-519 can1) (ATCC 38626) and S. cerevisiae strain, WAT11U,
a derivative of the W303-B strain (MAT a; ade2-1; his3-11, -15;
leu2-3, -112; ura3-1; canR; cyr+) were used as the host strains.
The plasmid pKS2␮Hyg containing 4CL and CHS open reading
frames (pKS2␮Hyg-4CL-CHS) with individual galactose inducible
promoters (GAL10) controlling expression was constructed as pre-
viously described [21,22]. The 4CL gene (GenBank accession no.
U18675) is from the plant Arabidopsis thaliana, and CHS (GenBank
accession no. AF315345) is from the plant Hypericum androsae-
mum.
2.3. E. coli and yeast manipulations
The plasmid pKS2␮Hyg-4CL-CHS was transformed into chem-
ically competent E. coli TOP10 F’ cells according to the
manufacturer’s protocol. Recombinant E. coli TOP10 F’ (Invitro-
gen) was grown at 37 ◦C in Luria Bertani broth supplemented with
100 ␮g/mL of ampicillin for maintenance of plasmid. The plasmid
pKS2␮Hyg-4CL-CHS was transformed into S. cerevisiae using the
lithium acetate/SS carrier DNA/PEG method [23]. Yeast transfor-
mants were selected on YPDH agar plates that were incubated at
30 ◦C for about 3 days. Yeast strains were grown at 30 ◦C in YPDH
and YPLH media. YPDH media contains 10 g/L yeast extract, 20 g/L
peptone, 20 g/L glucose, and 200 ␮g/mL Hygromycin (A.G. Scien-
tific, San Diego, CA). YPLH is similar to YPDH, but contains the same
quantity of galactose instead of glucose. Hygromycin was added
before inoculation of media.
2.7. LC/MS
Electrospray ionization analyses were carried out on a Waters
Micromass Q-TOF micro mass spectrometer system (Waters, Mil-
ford, MA). Samples were analyzed by HPLC on a C18 column
(2.1 mm × 150 mm, Waters) by the following method. A was 1.5%
(v/v) acetic acid in water; solvent B was acetonitrile. The flow rate
was 0.3 mL/min and the column temperature maintained at 30 ◦C.
After 20 ␮L of sample was injected, solvent B was 5% for 4 min, then
was increased to 50% over 56 min, and held for 5 min, before it was
returned to 5% in 5 min. UV absorbance at 290 nm was monitored
for detection of novel flavonoids. Negative ion mode was applied
with the mass spectrometer scan range from 50 to 600 m/z.
2.4. Substrate Screening
A single colony of AH22 or WAT11U harboring pKS2␮Hyg-4CL-
CHS from a YPDH plate was inoculated in 10 mL of YPDH and grown
for about 24 h at 30 ◦C with shaking at 300 rpm. Culture tubes con-
taining 10 mL of YPDH were inoculated to achieve an initial OD600
of 0.05. After approximately 20 h of growth, the cultures were cen-
trifuged at 1000 × g for 5 min. The cells were resuspended in 10 mL
of YPLH medium in order to induce gene expression. After 6 h
of induction, substrate dissolved in DMSO was added to the cul-
ture to give an initial concentration of 500–600 ␮M. Samples were
taken after about 20 h and were quenched with an equal volume
of methanol. The samples were then centrifuged at 18,000 × g for
1 min and the supernatant was stored at −20 ◦C prior to HPLC anal-
2.8. MS analysis
Isolated residues samples were dissolved in methanol in prepa-
ration for mass spectrometry analysis. Electrospray ionization MS
and Tandem MS (MS/MS) were carried out on a FinniganMAT LCQ
Classic (ThermoFinnigan Corp., San Jose, CA) mass spectrometer
system in either the negative or positive ion mode. Tandem MS

S.R. Werner et al. / Journal of Molecular Catalysis B: Enzymatic 66 (2010) 257–263
259
Table 1
results were obtained by selecting the ion of interest. The precursor
ion was then subjected to collision-induced dissociation, resulting
in the formation of product ions. The collision energy was set to
40% of the maximum available from the 5-V tickle voltage, with a
2-mass-unit isolation window.
Screening of cinnamic acid analogues for metabolism by 4CL and CHS in recombinant
S. cerevisiae.
Substrate Name
New products
formed by
4CL-CHSa
# of new products
detected by HPLC.
Cinnamic acid analogues
2.9. 1H NMR analysis
2-Aminocinnamic acidb
3-Aminocinnamic acidb
5-Bromo-2-methoxy cinnamic acid
2-Chlorocinnamic acid b
3-Chlorocinnamic acid b
2-Chloro-6-fluoro cinnamic acid
2,3-Dimethoxycinnamic acid
2-Ethoxycinnamic acid
2-Fluorocinnamic acid b
3-Fluorocinnamic acid b
2-Hydroxycinnamic acidb
3-Hydroxycinnamic acid b
2-Methylcinnamic acid b
3-Methylcinnamic acid b
2-Methoxycinnamic acid
3-Methoxycinnamic acid
4-Methoxycinnamic acid
2-Nitrocinnamic acid
Yes
Yes
Yes
Yes
Yes
Yes
No
1
2
3
1
1
4
-
In preparation for proton NMR analysis, the isolated residues
were dried then resuspended in deuterated methanol. The 1H NMR
spectra were acquired on a Bruker DRX 500 MHz spectrometer.
Correlation spectroscopy (COSY) was used to find which protons
are spin–spin coupled to each other to determine molecular struc-
ture. Chemical shifts are referenced to the methanol solvent peak
at 3.3 ppm.
No
-
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
No
No
2
2
3
3
3
2
-
2
-
-
3. Results and discussion
3.1. Substrate screening for production of non-natural flavonones
In a previous study, Jiang et al. [22] constructed a metaboli-
cally engineered S. cerevisiae expressing PAL, 4CL and CHS from
the flavonoid biosynthetic pathway capable of producing the
flavanones naringenin and pinocembrin from the amino acids
phenylalanine and tyrosine. The dihydrochalcones, phloretin and
2ꢀ,4ꢀ,6ꢀ-trihydroxydihydrochalcone as well as CHS derailment prod-
ucts were also produced. Jiang et al. [21] also reported formation of
flavanones and dihydrochalcones by addition of cinnamic acid and
p-coumaric acid to S. cerevisiae coexpressing 4CL and CHS. There-
fore, we examined the ability of this yeast stain expressing 4CL
and CHS as a platform for production of non-natural flavonoids by
addition of cinnamic acid analogues to the yeast cultures.
As a screen for cinnamic acids that could be accepted through
the engineered pathway, 22 cinnamic acid analogues were indi-
vidually added to recombinant yeast cultures expressing 4CL and
CHS, and the products formed were compared against a control
where no substrate was added. Formation of new peaks on HPLC
chromatograms by monitoring 290 nm compared to the control
indicated the possible formation of non-natural flavonoids and
novel CHS derailment products. The majority of the substrates
tested showed formation of new products by HPLC (Table 1). This
could indicate 4CL and CHS have broad substrate specificity and
capable of metabolizing many non-natural cinnamic acids. To con-
firm the new products being formed in the substrate screening
assay were the corresponding non-natural flavanones or derail-
ment products from CHS, LC/MS analysis was performed on a select
number samples. For some of the screened analogues, only one or
two of the proposed products were detected (Table 2). Two types of
possible products, analogues of pinocembrin (1b–10b) and tetrake-
tide lactones (1c–10c), have the same molecular weight and could
not be differentiated by LC/MS. Due to lack of authentic standards, it
is not possible to definitely conclude which of these products have
been produced. Products from 10 of the 22 cinnamate analogues
added to the AH22 host coexpressing 4CL and CHS were analyzed
by LC/MS (Table 2).
3-Nitrocinnamic acid
Urocanic acid
Phenylpropanoic acids
Phenylpropanoic acid
Phloretic acid
Yes
Yes
2
1
Yes
Yes
1
2
a
As determined by new peaks on the HPLC chromatogram compared to control
by monitoring 290 nm.
b
Samples that were selected for LC/MS analysis to confirm formation of products
from CHS.
5a and 6a gave products with a reduced carbon-carbon double
bond, dihydrochalcone and the triketide derailment products. 7a
produced a tetraketide product but 8a produced only a triketide
lactone. 9a and 10a both produced tetraketide products. In gen-
eral, all substrates with R1 substitution regardless of size or polarity
produced tetraketide products except for 5a. Substrates with R2
substitutions tend to produce the reduced dihydrochalcone and
triketide lactone products.
3.2. Production of non-natural flavanones and polyketides
Precursor addition of cinnamic acid analogues to the S. cerevesiae
strain AH22 expressing 4CL and CHS was scaled up to 3× 300 mL
scale fermentations (900 mL combined) to isolate a significant
quantity of products for further characterization. The non-natural
flavanones and polyketides that were isolated in significant quan-
tity are shown in Fig. 1.
The product isolated from addition of 2-methylcinnamic acid
(1a, Fig. 1) to the engineered yeast occurred at a retention time
of 25.5 min and UV spectrum had ꢀmax = 289 nm. ESI-MS spectrum
gave a parent ion peak [M+H]+ at m/z 271 suggesting the conden-
sation of three malonyl-CoA by CHS. Analysis of 1H NMR and COSY
spectra revealed two aromatic protons of ring-A (ı 5.9), four cou-
pled aromatic protons of ring-B (ı 7.53 and 7.25), three methyl
protons (ı 2.38), and two ␣-carbonyl protons (ı 3.1 and 2.72)
coupled to one proton (ı 5.67), confirming ring closure and the
structure of 2ꢀ-methyl-5,7-dihydroxyflavanone (1b, Fig. 1). A quan-
tity of 3 mg of 1b was isolated from the fermentation. A chemical
synthesis of this flavanone has been reported in the literature, but
this is the first report using precursor directed biosynthesis [24,25].
The product isolated from addition of 2-aminocinnamic acid
(9a, Fig. 1) to the engineered yeast occurred at a retention time of
19.8 min and the UV spectrum had ꢀmax = 289 nm. ESI-MS spectrum
gave a parent ion peak [M–H]− at m/z 270 suggesting the conden-
sation of three malonyl-CoA by CHS. Tandem MS on the parent ion
The results showed that all ten of the tested compounds could
be metabolized by 4CL and CHS. Based on the structure of prod-
ucts from CHS and the molecular weight of metabolites detected
by negative ion mode LC/MS, the structures of major products
were proposed. The effects of functional group substitution on
the cinnamic acid aromatic ring gave interesting trends in prod-
uct distributions. Among the tested analogues, 1a produced only
tetraketide products but 2a produced only dihydrochalcone and
triketide lactone products. 3a produced only tetraketide products
but 4a produced both tetraketide products and triketide lactone.

260
S.R. Werner et al. / Journal of Molecular Catalysis B: Enzymatic 66 (2010) 257–263
Table 2
Products from incubation of cinnamic acid analogue substrates (1a–10a) with AH22 coexpressing 4CL and CHS. R1 and R2 are substituted functional group on the aromatic
ring of cinnamate; 1b–10b are pinocembrin analogues; 1c–10c are coumaroyltriacetic acid lactones (CTAL); 1d–10d are 2ꢀ,4ꢀ,6ꢀ-trihydroxydihydrochalcone products; 1e–10e
are triketidelactones.
.
Cinnamic acid analog substrates
Substitution
R1
Products
b or c
R2
d
e
1a
2a
3a
4a
5a
6a
7a
8a
9a
10a
Me
–
OH
–
F
–
Cl
–
NH2
–
–
Me
–
OH
–
F
–
Cl
–
Yes
No
Yes
Yes
No
No
Yes
No
No
Yes
No
Yes
Yes
Yes
No
No
No
Yes
No
Yes
No
Yes
Yes
Yes
No
Yes
No
Yes
Yes
NH2
No
gave a major fragment at m/z 252 corresponding to [M–H–H2O]−.
Analysis of 1H NMR and COSY spectra revealed two aromatic pro-
tons of ring-A (ı 5.9), four coupled aromatic protons of ring-B (ı
7.23, 7.1, 6.78 and 6.72), and two ␣-carbonyl protons (ı 3.3 and
2.75) coupled to one proton (ı 5.55), confirming ring closure and
the structure of 2ꢀ-amino-5,7-dihydroxyflavanone (9b, Fig. 1). A
quantity of 1.2 mg of 9b was isolated from the fermentation. This
flavanone has not been previously reported.
ring. In addition, positive ion mode ESI-MS spectrum gave a parent
ion peak [M+H]+ at m/z 274 consistent with the condensation of
three malonyl-CoA by CHS. Tandem MS on the parent ion in positive
ion mode gave a fragment at m/z 148 corresponding fragmentation
around the central ketone of a CHS derailment product. Based on
the above evidence, we hypothesize the residue is a derailment
product from CHS that could be identified as the structure 10f, the
dihydro form of 10c (Fig. 1). A quantity of <1 mg of product was iso-
lated from the fermentation. These derailment products have not
been reported in the literature.
The product isolated from addition of urocanic acid (11a, Fig. 1)
to the engineered yeast occurred at a retention time of 8.2 min and
UV spectrum had ꢀmax = 284 nm. ESI-MS spectrum gave a parent
ions peak [M–H]− at m/z 245 and [2M–H]− m/z 491 (dimer) sug-
gesting the condensation of three malonyl-CoA by CHS. Tandem MS
of the parent ion gave a major fragment at m/z 201 corresponding to
[M–H–CO2]−, indicating the presence of a pyrone ring. A quantity of
2.2 mg of product was isolated from the fermentation. Based on the
above evidence, we hypothesize the residue is a derailment prod-
uct from CHS that could be identified as the structure of 11c (Fig. 1).
This derailment product has not been reported in the literature.
When 3-aminocinnamic acid (10a, Fig. 1) was incubated with
the engineered yeast, two products were isolated by HPLC. The
first product occurred at a retention time of 14.5 min and UV spec-
trum had ꢀmax = 290 nm. ESI-MS spectrum gave a parent ion peak
[M–H]− at m/z 270 suggesting the condensation of three malonyl-
CoA by CHS. Tandem MS on the parent ion gave a major fragment
at m/z 226 corresponding to [M–H–CO2]−, suggesting the presence
of a pyrone ring. In addition, positive ion mode ESI-MS spectrum
gave a parent ion peak [M+H]+ at m/z 272 consistent with the con-
densation of three malonyl-CoA by CHS. Tandem MS on the parent
ion in positive ion mode gave the fragments at m/z 153 and m/z
146 corresponding to fragmentation around the central ketone of a
CHS derailment product. Based on the above evidence, we hypoth-
esize the residue is a derailment product from CHS that could be
identified as the structure of 10c (Fig. 1). A quantity of <1 mg of
the product was isolated from the fermentation. The second prod-
uct occurred at a retention time of 12.9 min and UV spectrum had
ꢀmax = 287 nm. ESI-MS spectrum gave a parent ion peak [M–H]−
at m/z 272 suggesting the condensation of three malonyl-CoA by
CHS. Tandem MS on the parent ion gave a fragment at m/z 228 cor-
responding to [M–H–CO2]−, suggesting the presence of a pyrone
3.3. Production of dihydrochalcones
Jiang et al. [22] hypothesized the dihydrochalcones formed
in addition to the flavanones naringenin and pinocembrin from
the metabolically engineered yeast is due to endogenous reduc-
tases capable of reducing the carbon–carbon double bonds of
cinnamic acid and p-coumaric acid. This suggested, in addition

S.R. Werner et al. / Journal of Molecular Catalysis B: Enzymatic 66 (2010) 257–263
261
Fig. 1. Biosynthesis of non-natural flavanones and CHS derailment products by precursor addition of cinnamic acid analogues to S. cerevisiae AH22 expressing 4CL and CHS.
to cinnamic acids being metabolized through 4CL and CHS, that
phenylpropanoic acids are used as precursors in the formation of
dihydrochalcones.
Precursor addition of phloretic acid (12a, Fig. 3) to S. cerevisiae
expressing 4CL and CHS resulted in the formation of phloretin (12d,
Fig. 3). The quantity of phloretin in the supernatant increased to
36 7 mg/L after 17 h. There was no further increase in phloretin
concentration after longer fermentation times (Fig. 2). When
phloretic acid (12a, Fig. 3) was incubated with the engineered yeast
on a larger scale, two products were isolated by HPLC. One product
occurred at the same retention time as phloretin (22.17 min) and
showed the same UV spectrum (ꢀmax = 286 nm). ESI-MS spectrum
gave a parent ion peak [M–H]− at m/z 273. Analysis of 1H NMR
and COSY spectra revealed two aromatic protons of ring-A (ı 5.80),
four coupled aromatic protons of ring-B (ı 7.03 and 6.68), and four
coupled alkyl protons (ı 3.25 and 2.84), confirming the structure of
phloretin (12d, Fig. 3). A quantity of 3.8 mg of phloretin was puri-
fied from the large scale fermentation. Phloretin was synthesized in
a similar manner using metabolically engineered E. coli [26]. Small
Fig. 2. Addition of the precursor phloretic acid to S. cerevisiae AH22 expressing 4CL
and CHS for the phloretin production. After the cells were grown in YPDH for 20 h
reaching an OD600 of 3.5, the cells were resuspended in YPLH and induced for 6 h
before addition of substrate. Time zero corresponds with addition of phloretic acid to
the induced culture. The data points and error bars correspond to the mean stdev
(n = 3).

262
S.R. Werner et al. / Journal of Molecular Catalysis B: Enzymatic 66 (2010) 257–263
Fig. 3. Biosynthesis of dihydrochalcones by precursor addition of phenylpropanoic acids to S. cerevisiae AH22 expressing 4CL and CHS.
amounts of phloretin were detected after 24 h of cultivation, but an
absolute quantity was not report.
and isolation steps. To improve the yield, further metabolic engi-
neering of the yeast strain could be used to increase malonyl-CoA
available for CHS as used for a polyketide synthase for production
of a fungal polyketide [28].
Besides phloretin, a second product eluted at a retention time of
17.6 min and UV spectrum had ꢀmax = 282 nm. ESI-MS gave a parent
ion peak [M–H]− at m/z 231 and [2M–H]− at m/z 463 (dimer), sug-
gesting the condensation two malonyl-CoA by CHS. Tandem MS of
the parent ion gave the major fragment at m/z 187 corresponding
to [M–H–CO2]−, indicating the presence of a pyrone ring. Analy-
sis of 1H NMR and COSY spectra revealed two protons consistent
with a pyrone ring (ı 5.80 and 4.80), four coupled aromatic pro-
tons of ring-B (ı 6.99 and 6.67), and four coupled alkyl protons (ı
2.85 and 2.70), confirming the structure of dihydrobisnoryangonin
(12c, Fig. 3). A quantity of 2.6 mg of product 12c was isolated from
the fermentation. Product 12c is a derailment product from CHS
and has been previously synthesized enzymatically with type III
polyketide synthase from Wachendorfia thyrsiflor [27].
The product isolated from addition of phenylpropanoic acid
(13a, Fig. 3) to the engineered yeast occurred at a retention time
of 22.6 min and UV spectrum had ꢀmax = 281 nm. ESI-MS gave a
parent ion peak [M–H]− at m/z 257, and tandem MS of the par-
ent ion resulted in no fragment. Analysis of 1H NMR and COSY
spectra revealed two aromatic protons of ring-A (ı 5.79), five cou-
pled aromatic protons of ring-B (ı 7.22 and 7.15), and four coupled
alkyl protons (ı 3.31 and 2.94), confirming the structure of 2ꢀ,4ꢀ,6ꢀ-
Trihydroxydihydrochalcone (13d, Fig. 3). A quantity of 2.1 mg of
13d was isolated from the fermentation. These studies confirm
phenylpropanoic acid can be metabolized by the phenylpropanoid
pathway for the production of dihydrochalcones (Fig. 3). This is the
first time trihydroxydihydrochalcone has been synthesized using
precursor directed biosynthesis starting from phenylpropanoic
acid.
4. Conclusion
In this study, a metabolically engineered yeast expressing 4CL
and CHS was explored as a platform for producing non-natural
flavanones and dihydrochalcones. By precursor addition of cin-
namic acid analogues to the engineered yeast, several non-natural
flavanones and dihydrochalcones were formed. Also, several prod-
ucts from CHS derailment were formed. Of the eight isolated
compounds, one flavanone and three derailment products are
novel compounds. It is conceivable that introducing non-natural
substituents into the flavonoid structure may result in novel com-
pounds with potentially valuable medicinal properties compared.
This study indicates metabolically engineered organisms could be
used to manufacture such non-natural compounds.
Acknowledgments
The authors thank Kristina Perlas for help with fermentations
and product isolation, Dr. John S. Harwood (Purdue Interdepart-
mental NMR Facility) for help with NMR, and Dr. Karl V. Wood and
Deb Miles (Campus-wide Mass Spectrometry Center at Purdue) for
help with MS analysis. This work and authors were supported by
the School of Chemical Engineering, Purdue University.
Appendix A. Supplementary data
There are several explanations for the low yields of prod-
uct achieved from this system and there are possibilities for
improvement. First, the pathway most likely does not accept the
non-natural substrates as well as the natural substrates. Use of
enzymes with broader substrate specificity could be used or engi-
neered. Second, the non-natural substrates could be inhibiting
growth of the yeast cells, impacting product yield. Third, there
could be loss of product due to the low efficiency of the extraction
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.molcatb.2010.05.017.
References
[1] J.B. Harborne, C.A. Williams, Phytochemistry 55 (2000) 481–504.
[2] B.W. Shirley, Trends Plant Sci. 1 (1996) 377–382.
[3] O. Benavente-Garcia, J. Castillo, F.R. Marin, A. Ortuno, J.A. Del Rio, J. Agr. Food
Chem. 45 (1997) 4505–4515.

S.R. Werner et al. / Journal of Molecular Catalysis B: Enzymatic 66 (2010) 257–263
263
[4] C.A. Rice-Evans, N.J. Miller, Biochem. Soc. Trans. 24 (1996) 790–795.
[5] G. Di Carlo, N. Mascolo, A.A. Izzo, F. Capasso, Life Sci. 65 (1999) 337–353.
[6] J.A. Manthey, N. Guthrie, K. Grohmann, Curr. Med. Chem. 8 (2001) 135–153.
[7] W.Y. Ren, Z.H. Qiao, H.W. Wang, L. Zhu, L. Zhang, Med. Res. Rev. 23 (2003)
519–534.
[17] F. Ververidis, E. Trantas, C. Douglas, G. Vollmer, G. Kretzschmar, N. Panopoulos,
Biotechnol. J. 2 (2007) 1235–1249.
[18] J. Kennedy, Nat. Prod. Rep. 25 (2008) 25–34.
[19] J.A. Chemler, Y.J. Yan, E. Leonard, M.A.G. Koffas, Org. Lett.
1855–1858.
9 (2007)
[8] L.G. Korkina, Cell. Mol. Biol. 53 (2007) 15–25.
[9] F. Ververidis, E. Trantas, C. Douglas, G. Vollmer, G. Kretzschmar, N. Panopoulos,
Biotechnol. J. 2 (2007) 1214–1234.
[10] B. Buchanan, W.G. Russell, L. Jones, B. Buchanan, W. Gruissem, R.L. Jones, Bio-
chemistry and Molecular Biology of Plants, John Wiley and Sons, Somerset, NJ,
2000.
[20] Y. Katsuyama, N. Funa, I. Miyahisa, S. Horinouchi, Chem. Biol. 14 (2007)
613–621.
[21] H. Jiang, Metabolic engineering of the phenylpropanoid pathway in Saccha-
romyces cerevisiae, PhD Thesis in School of Chemical Engineering, Purdue
University, 2005.
[22] H. Jiang, K.V. Wood, J.A. Morgan, Appl. Environ. Microb. 71 (2005) 2962–2969.
[23] R.D. Gietz, R.A. Woods, Transformation of yeast by lithiumacetate/single
stranded carrier DNA/polyethylene glycol method, in: C. Guthrie, G. Fink (Eds.),
Methods in Enzymology, vol. 350, Academic Press, San Diego, 2002, pp. 87–96.
[24] X. He, F. Yang, X. Lei, J. Chen, Y. Min, Yiyao Gongye 19 (1988) 447–451.
[25] Y. Yao, F. Yang, F. Gao, Zhongguo Yiyao Gongye Zazhi 23 (1992) 211–310.
[26] K.T. Watts, P.C. Lee, C. Schmidt-Dannert, Chembiochemistry 5 (2004) 500–507.
[27] S. Brand, D. Hoelscher, A. Schierhorn, A. Svatos, J. Schroeder, B. Schneider, Planta
224 (2006) 413–428.
[11] O. Yu, J.M. Jez, Plant J. 54 (2008) 750–762.
[12] Z.L. Fowler, M.A.G. Koffas, Appl. Microbiol. Biotechnol. 83 (2009) 799–808.
[13] H. Morita, Y. Takahashi, H. Noguchi, I. Abe, Biochem. Biophys. Res. Commun.
279 (2000) 190–195.
[14] H. Morita, H. Noguchi, Y. Takahashi, H. Noguchi, I. Abe, Eur. J. Biochem. 268
(2001) 3759–3766.
[15] I. Abe, H. Morita, A. Nomura, H. Noguchi, J. Am. Chem. Soc. 122 (2000)
11242–11243.
[16] I. Limem, E. Guedon, A. Hehn, F. Bourgaud, L.C. Ghedira, J. Engasser, M. Ghoul,
Process. Biochem. 43 (2008) 463–479.
[28] S. Wattanachaisaereekul, A.E. Lantz, M.L. Nielsen, J. Nielsen, Metab. Eng. 10
(2008) 246–254.