Screening for enhanced triacetic acid lactone production by recombinant Escherichia coli expressing a designed triacetic acid lactone reporter

Article abstract of DOI:10.1021/ja402654z

Triacetic acid lactone (TAL) is a signature byproduct of polyketide synthases (PKSs) and a valuable synthetic precursor. We have developed an endogenous TAL reporter by engineering the Escherichia coli regulatory protein AraC to activate gene expression in response to TAL. The reporter enabled in vivo directed evolution of Gerbera hybrida 2-pyrone synthase activity in E. coli. Two rounds of mutagenesis and high-throughput screening yielded a variant conferring ～20-fold increased TAL production. The catalytic efficiency (k_cat/K_m) of the variant toward the substrate malonyl-CoA was improved 19-fold. This study broadens the utility of engineered AraC variants as customized molecular reporters. In addition, the TAL reporter can find applications in other basic PKS activity screens.

Full text of DOI:10.1021/ja402654z

Article
pubs.acs.org/JACS
Screening for Enhanced Triacetic Acid Lactone Production by
Recombinant Escherichia coli Expressing a Designed Triacetic Acid
Lactone Reporter
Shuang-Yan Tang,†,§ Shuai Qian,† Olubolaji Akinterinwa,†,∥ Christopher S. Frei,† Joseph A. Gredell,†
and Patrick C. Cirino*,†
†Department of Chemical Engineering, Pennsylvania State University, 226A Fenske Laboratory, University Park, Pennsylvania 16802,
United States
†Department of Chemical & Biomolecular Engineering, University of Houston, S222 Engineering Building 1, Houston, Texas
77204-4004, United States
S
* Supporting Information
ABSTRACT: Triacetic acid lactone (TAL) is a signature
byproduct of polyketide synthases (PKSs) and a valuable
synthetic precursor. We have developed an endogenous TAL
reporter by engineering the Escherichia coli regulatory protein
AraC to activate gene expression in response to TAL. The
reporter enabled in vivo directed evolution of Gerbera hybrida
2-pyrone synthase activity in E. coli. Two rounds of
mutagenesis and high-throughput screening yielded a variant
conferring ∼20-fold increased TAL production. The catalytic
efficiency (kcat/Km) of the variant toward the substrate
malonyl-CoA was improved 19-fold. This study broadens the
utility of engineered AraC variants as customized molecular reporters. In addition, the TAL reporter can find applications in
other basic PKS activity screens.
previously reported in controlled fermentations were 0.47 g/L
(3.7 mM) from E. coli expressing wild-type 2-PS and 1.8 g/L
(14.3 mM) from Saccharomyces cerevisiae expressing a variant of
6-MSAS.13
INTRODUCTION
■
4-Hydroxy-6-methyl-2-pyrone, also referred to as triacetic acid
lactone (TAL), is a natural compound of polyketide origin,1,2
commonly identified as a triketide derailment product during
polyketide biosynthesis (e.g., lovastatin3 and 6-methylsalicylic
acid4). Natural products bearing the 2-pyrone functionality
exhibit biological activities and are valuable pharmaceutical
precursors.5−7 TAL is also a precursor in the chemical synthesis
of phloroglucinol, used in the synthesis of the thermostable
energetic material 1,3,5-triamino-2,4,6-trinitrobenzene
(TATB), and resorcinol, used in resin and adhesive
formulations.8,9 Further catalytic upgrading of TAL has
demonstrated its potential as a biorenewable platform
chemical.10 In Gerbera hybrida, TAL is a precursor in the
biosynthesis of the pyrone glucosides gerberin and parasorbo-
side.11 The g2ps1-encoded type III PKS 2-pyrone synthase (2-
PS) from G. hybrida (having 73% identity to the chalcone
synthase (CHS) in this organism) uses acetyl-CoA as a starter
substrate and catalyzes two condensation reactions with
malonyl-CoA, resulting in TAL formation.11
Engineering of genetic components involved in the synthesis
of a target metabolite is often limited by the lack of sensitive
and rapid screening/selection methods for identifying desirable
candidates from large gene libraries.14 Several studies, including
those from our group, have sought to overcome this limitation
by engineering a regulatory protein with altered effector
specificity to enable direct coupling between recognition of
the metabolite of interest (effector) and changes in reporter
gene expression.15−18 Owing to the lack of a well-developed
high-throughput screen for TAL synthesis, the Escherichia coli
regulatory protein AraC was engineered to recognize TAL as an
effector. An endogenous TAL reporter system based on
expression of β-galactosidase (LacZ) from AraC’s cognate
promoter PBAD was subsequently used to screen for 2-PS
variants conferring elevated TAL production by E. coli.
MATERIALS AND METHODS
Few studies have focused on improving microbial TAL
production.12,13 Two different mutants of Penicillium patulum
6-methylsalycilic acid synthase (6-MSAS) exhibit TAL-synthe-
sizing activities, in contrast to their wild-type counterparts.13
TAL yields from different TAL-synthesizing enzymes were
compared by Zhao and co-workers. The highest TAL titers
■
Fluorescence-Based Positive and Negative Screening for
AraC Mutants. Construction and screening of the AraC mutant
Received: March 15, 2013
Published: June 20, 2013
© 2013 American Chemical Society
10099
dx.doi.org/10.1021/ja402654z | J. Am. Chem. Soc. 2013, 135, 10099−10103

Journal of the American Chemical Society
Article
library were carried out as previously described.18 Strain HF19
harboring PBAD-gfpuv reporter plasmid pPCC442 was transformed
with the araC gene library contained in plasmid pPCC423. In the first
round of screening, cells were grown in the absence of TAL and the
least fluorescent 1.5 × 108 cells were collected (representing 82% of all
cells sorted). This was done to eliminate AraC variants having high
levels of leaky expression. In the second round, the most fluorescent
1.5 × 106 cells were sorted from a total of 1.5 × 108 cells after growth
in the presence of 5 mM TAL. This dual screening procedure was
repeated two more times and then two additional times with the
concentration of TAL at 2.5 mM during positive screening, followed
by a final round of negative screening. Twenty clones were then
selected for rescreening. These were cultured separately and
characterized in green fluorescent protein (GFP) expression
fluorescence assays (described below). Three clones showing increased
fluorescence in the presence of 2.5 mM TAL were selected for
sequencing. The sequencing results revealed that all the clones
contained the same mutations.
2-PS Activity Assays. TAL production was quantified by
monitoring the corresponding increase in absorbance at 298 nm
when incubating purified 2-PS variants with acetyl-CoA and/or
malonyl-CoA using a molar extinction coefficient of 2540 M−1 cm−1 to
quantify TAL.13 The reaction mixture (200 μL) contained 100 mM
HEPES buffer (pH 6.0) and the indicated concentration of acetyl-CoA
and/or malonyl-CoA, and the reaction was initiated by the addition of
20 μL of purified enzyme. The increase in absorbance at 298 nm at 37
°C was measured using a NOVOstar microplate spectrophotometer.
No background absorbance increase in the absence of the enzyme was
detected, and the absorbance at 298 nm was verified to correspond to
the TAL product measured by HPLC. The protein concentration was
determined by the Quick Start Bradford Protein Assay (Bio-Rad
Laboratories). All assays were performed in triplicate, and standard
deviations are reported with the data. The mean values of the initial
TAL production rate divided by the enzyme concentration under
different malonyl-CoA concentrations are plotted in Figure S2
(Supporting Information). Kinetic parameters in Table 4 were
obtained by nonlinear regression of the Michaelis−Menten equation.
GFP Expression Fluorescence Assays. The fluorescence assays
were performed as described previously.17 Briefly, HF19 cells
harboring plasmids pPCC442 and either pFG1-AraC or pFG1-TAL
were grown overnight in lysogeny broth (LB) containing chlor-
amphenicol, apramycin, and 0.4 mM isopropyl β-D-1-thiogalactopyr-
RESULTS
■
Development of a TAL-Responsive AraC Variant. TAL
does not induce expression of green fluorescent protein gene
(gfpuv) from promoter PBAD in E. coli expressing wild-type
AraC (Figure 1). We previously described the use of multiple-
anoside (IPTG), then diluted to OD600 = 0.01 (Table 1) or OD600
=
0.2 (Figure 1) in the same medium containing the appropriate
concentration of inducer, and allowed to grow under inducing
conditions for 14 h. A total of 500 μL of the resulting culture was
centrifuged, and the cells were washed with phosphate-buffered saline
(PBS) (pH 7.4) and resuspended in 1000 μL of the same buffer. The
OD600 of the suspension was measured using a NOVOstar microplate
reader (BMG LABTECH), and the fluorescence emission was
measured using a SpectraMax Gemini EM microplate reader
(Molecular Devices Corp.) (400 nm excitation filter, 510 nm emission
filter). The fluorescence emission was normalized with respect to
OD600, and the background fluorescence due to buffer was subtracted
from all measurements. Reported data represent the average of three
independent data points, and the standard deviations are shown in
Figure 1. The coefficient of variation (CV) for all measurements was
less than 15% and less than 10% for most data.
β-Galactosidase Activity-Based Solid-Phase Screening of 2-
PS Random and Saturation Mutagenesis Libraries for
Increased TAL Production. Strain HF2218 was transformed with
plasmid pPCC704, which contains the 2-PS random or saturation
mutagenesis library. Cells were plated onto LB containing 400 μM
IPTG, 274 mM glycerol (as the carbon source for TAL production),
40 μg/mL X-Gal, and apramycin. The plates were incubated at 37 °C
for ∼20 h, and the colonies which were obviously darker blue (by the
eye) than those expressing wild-type 2-PS (for random mutagenesis
library screening) or the 2-PS E3 mutant (for saturation mutagenesis
library screening) were picked for quantifying TAL production in
liquid culture.
HPLC Quantification of TAL. A single colony of strain HF22
harboring wild-type or mutant plasmid pPCC704 (carrying wild-type
or mutant g2ps1 genes) was grown in 3 mL of LB containing
apramycin for 10 h. The cultures were then diluted to OD600 = 0.2 in
20 mL of LB containing 220 mM glycerol, 400 μM IPTG, and
apramycin. At different time points, 700 μL of culture was centrifuged
and the supernatant was passed through a 0.45 μm filter and acidified
with 1% acetic acid. A Shimadzu LC-20AD HPLC system with an
SPD-20A dual-wavelength UV−vis detector and a Phenomonex Luna
C18 column (25 cm × 4.6 mm, 5 μm) was used for TAL
quantification, following the protocol of Xie et al.13 The TAL elution
time was ∼15 min, and concentrations were determined from a TAL
standard curve. Reported data represent the average of three
independent data points, and standard deviations are shown in Figure
2. The compound eluted at 15 min was collected and verified as TAL
by comparison to an authentic standard using a Bruker MicroToF
mass spectrometer operated in the positive ESI mode. HPLC
chromatograms and mass spectra are available in the Supporting
Information.
Figure 1. AraC-TAL TAL dose response. GFP expression from strain
HF19 harboring PBAD-GFP reporter plasmid and expressing either
AraC-TAL (squares) or wild-type AraC (triangles) as a function of the
TAL concentration (added to the culture).
site saturation mutagenesis, a PBAD-gfpuv reporter plasmid, and
iterative fluorescence activated cell sorting (FACS) to isolate
AraC variants having altered effector specificity (e.g., specificity
switched from its natural inducer L-arabinose to D-arabinose or
mevalonate).17,18 In a similar fashion, an araC library in which
codons for five amino acid positions located in the effector
binding pocket (P8, T24, H80, Y82, and H93)18 were subjected
to simultaneous saturation mutagenesis was screened for AraC
variants responding to the presence of exogenous 5 or 2.5 mM
TAL. Variants showing high levels of constitutive (uninduced)
expression were eliminated by “negative” sorting, in which the
least fluorescent cells are collected in the absence of TAL. After
several rounds of screening, AraC variant “AraC-TAL” (P8V,
T24I, H80G, Y82L, H93R) was isolated. The exogenous TAL
dose response of GFPuv expression under control of AraC-
TAL at promoter PBAD is shown in Figure 1. TAL
concentrations exceeding ∼8 mM reduce growth and specific
GFP expression, making it difficult to quantify the response of
the AraC-TAL reporter system to a saturating exogenous TAL
dose. As shown in Table 1, phloroglucinol (chemically similar
10100
dx.doi.org/10.1021/ja402654z | J. Am. Chem. Soc. 2013, 135, 10099−10103

Journal of the American Chemical Society
Article
Table 1. Fluorescence of Strain HF19 Harboring PBAD-gfpuv Reporter Plasmid and Expressing AraC-TAL in the Presence of the
a
Indicated Effector Compound
effector
fluorescence
induction fold
effector
fluorescence
induction fold
none
1420 70
19660 680
2830 130
6 mM L-arabinose
60 mM mevalonate
1230 60
1830 50
0.9
1.3
6 mM TAL
13.9
2.0
6 mM phloroglucinol
a
The concentration indicated corresponds to the maximum fluorescence response.
Table 2. Amino Acid Substitutions in Improved 2-PS Variants
variant
amino acid substitutions
variant
amino acid substitutions
E1
E3
E8
L261F, K58K (silent mutation)
L261F, G221G (silent mutation)
L261F, D248G
S1
S2
L202G, M259L, L261N
L202G, M259M, L261N
to TAL), L-arabinose (the native AraC inducer), and
mevalonate (which AraC was previously engineered to
recognize) are much poorer inducers of GFPuv expression,
indicating that the expression response of the AraC-TAL
reporter is relatively specific to TAL (presumably a result of
negative sorting in which endogenous metabolites can act as
decoy ligands). Complete dose responses for all ligands are
shown in Figure S1 (Supporting Information).
white. Thirteen of the darkest blue colonies (identified by the
eye) were selected (from a total of 4 × 104 colonies screened)
and rescreened by quantifying TAL production from liquid
cultures of these clones. Three clones, E1, E3, and E8,
produced ∼2-fold more TAL than wild-type 2-PS in liquid
culture (TAL produced by all 13 clones is provided in Table S3,
Supporting Information). Thus, the qualitative visual screen
rapidly identifies improved clones from a large library, as was
the case for our mevalonate reporter system.
Plasmids from these clones were purified, sequenced, and
confirmed to improve TAL production by HF22. All three
variants contained amino acid substitution L261F, with an
additional D248G substitution appearing in mutant E8 (Table
2). The time courses of TAL production and cell growth
(OD600) for clone E3 are presented in Figure 2 and Figure S3
(Supporting Information), respectively. At 12 h, E3 produced
2.3 mM TAL, compared to 0.8 mM by strain HF22 expressing
wild-type 2-PS.
Autodock19 was used to estimate binding energies and
identify potential AraC-TAL binding pocket interactions with
TAL, L-arabinose, and mevalonate (Supporting Information).
We were unable to rationalize inducer specificity from this
analysis (binding energies for the ligands varied by <0.7 kcal/
mol, while a typical error associated with such predictions is
2
kcal/mol20), though it should be noted that ligand binding does
not imply induced gene expression (a tight-binding ligand can
inhibit activation) and the determinants of AraC molecular
recognition resulting in activation remain unclear.21
The three-dimensional structure of G. hybrida 2-PS shows
L261 located in the enzyme’s polyketide initiation/elongation
cavity.22 To further improve TAL production in E. coli and
considering the results from the random mutagenesis study, the
2-PS Mutagenesis and Library Screening. Plasmid-
based expression of wild-type 2-PS from strong promoters in E.
coli resulted in production of ∼0.8 mM TAL in a shake-flask
culture with excess glycerol (data presented here) and ∼3.7
mM TAL in a glucose-limited fermentation (previously
reported13). Using AraC-TAL as an endogenous reporter of
TAL, we next sought 2-PS variants conferring improved TAL
production in E. coli. In our previous study involving an AraC-
based mevalonate sensor, a visual colony screen on agar plates
using strain HF22, which contains a chromosomal PBAD-lacZ
reporter, identified mutants that produce enhanced levels of
mevalonate in liquid culture.18 This screening approach reduces
unwanted reporter expression due to the presence of effector
(mevalonate or in this case TAL) that has been secreted by
other high-producing clones, e.g., if the clones were pooled and
screened directly from liquid culture via FACS.
The g2ps1 gene was first randomly mutated using error-
prone polymerase chain reaction (PCR), and strain HF22 was
transformed with the library. Colonies expressing AraC-TAL
and growing on LB agar supplemented with glycerol express
LacZ when TAL is produced, and the presence of 5-bromo-4-
chloro-3-indolyl β-D-galactopyranoside (X-GAL) results in blue
color formation. After ∼20 h colonies expressing wild-type 2-PS
were faintly blue, while the majority of library clones were
Figure 2. Time profiles of TAL concentration measured from cultures
(LB + 220 mM glycerol) of strain HF22 harboring pPCC704
expressing wild-type 2-PS or its indicated variant (E3, S1, and S2).
Final cell densities are in parentheses (OD600).
10101
dx.doi.org/10.1021/ja402654z | J. Am. Chem. Soc. 2013, 135, 10099−10103

Journal of the American Chemical Society
Article
g2ps1 gene was next subjected to simultaneous saturation
mutagenesis at three codons encoding amino acids in the 2-PS
initiation/elongation cavity (L202, M259, and L261) (refer to
Figure 6 in ref 22), corresponding to 8 × 103 2-PS variants.
AraC-TAL and the PBAD-lacZ reporter were again used to
screen this 2-PS library. Eight dark blue colonies were selected
(from a total of 4 × 104 colonies screened) and rescreened by
quantifying TAL production from liquid cultures of these
clones (Table S4, Supporting Information). Seven of the eight
clones produced more TAL than 2-PS variant E3. Plasmids
from the two top-producing clones, S1 and S2, were purified,
sequenced, and confirmed to improve TAL production by
HF22. The amino acid substitutions in variants S1 and S2 are
shown in Table 2. The time courses of TAL production and cell
growth (OD600) of clones S1 and S2 are presented in Figure 2
and Figure S3 (Supporting Information), respectively. At 12 h,
they produced significantly more TAL (14.4 1.5 and 16.4
0.5 mM) than wild-type 2-PS and variant E3. Improved TAL
production did not appreciably affect cell growth, and cultures
were near the saturating OD600 before TAL concentrations
reached potentially inhibiting levels (∼8 mM).
in exponential growth was reported to be 35 μM,29 and
previous measurements range from 4 to 90 μM30 (in contrast,
acetyl-CoA concentrations are 10-fold higher). Given the
similarity of these values to the malonyl-CoA Km of wild-type 2-
PS (75 μM), the reduced Km for the 2-PS variant S1 (26 μM)
may therefore contribute significantly to more efficient
utilization of this metabolite in E. coli (in addition to the
elevated kcat). L261N lies within the 2-PS initation/elongation
cavity, opposite the catalytic C169.22 This substitution may
enhance binding of malonyl-CoA and help to polarize malonyl-
CoA for decarboxylation (during elongation and/or initiation).
CONCLUSIONS
■
We have demonstrated the design and application of a novel
endogenous TAL reporter, which facilitated screening for
improved TAL production by engineered E. coli. Additional
improvements in TAL production are foreseeable through
further directed evolution of 2-PS or combinatorial host strain
engineering using the TAL reporter, rational strain engineering
(e.g., see ref 26), high-density fermentation, and optimized
medium formulation. Moreover, the TAL reporter could serve
as a useful tool for high-throughput identification of functional,
minimal PKSs. Further directed evolution of AraC-TAL toward
alternate 2-pyrone derivatives should also prove useful for
screening and engineering polyketide biosynthetic pathways
that incorporate alternate extender units.31 Finally, this study
further demonstrates the applicability of engineered AraC
variants as customized endogenous molecular reporters in
protein and metabolic engineering applications.
2-PS Variant Kinetic Parameters. His6-tagged wild-type
2-PS and variants E3 and S1 were purified for activity
measurements, which are summarized in Tables 3 and 4. The
Table 3. Initial Activity (mM TAL/min/mg of protein) of
Purified Wild-Type 2-PS and Its E3 and S1 Variants in the
Presence of Malonyl-CoA plus Acetyl-CoA
wild type
E3
S1
1 mM malonyl-CoA +
0.5 mM acetyl-CoA
0.046 0.004
0.16 0.02
0.23 0.01
ASSOCIATED CONTENT
* Supporting Information
■
S
Additional materials and methods, strain, plasmid, and primer
descriptions, dose response profiles of wild-type AraC and
AraC-TAL with phloroglucinol, L-arabinose, and mevalonate,
AraC-TAL ligand docking analyses, enzyme Michaelis−Menten
plots, HPLC traces, and TAL mass spectra. This material is
available free of charge via the Internet at http://pubs.acs.org.
Table 4. Kinetic Parameters of Wild-Type 2-PS and Its E3
and S1 Variants Using the Substrate Malonyl-CoA
enzyme
Km (mM)
kcat (min−1
)
kcat/Km (mM−1 min−1
)
wild type
0.075 0.012
0.040 0.006
0.026 0.001
0.32 0.02
0.28 0.03
2.18 0.07
4.4 0.5
7.1 0.4
84.1 6.4
E3
S1
AUTHOR INFORMATION
Corresponding Author
pccirino@uh.edu
Present Addresses
■
acetyl-CoA starter molecule for TAL synthesis can be supplied
by decarboxylation of malonyl-CoA by 2-PS, although in vitro
activity using malonyl-CoA alone is 2−3-fold lower than in the
presence of acetyl-CoA and malonyl-CoA.11 In the presence of
1 mM malonyl-CoA plus 0.5 mM acetyl-CoA, variants E3 and
S1 showed 3.5- and 5-fold higher initial activity than wild-type
2-PS, respectively. The kinetic parameters of wild-type 2-PS
and variants E3 and S1 using malonyl-CoA as the substrate are
reported in Table 4 (Michaelis−Menten curves are plotted in
Figure S2, Supporting Information). Variant E3 has a nearly 2-
fold lower Km compared to wild-type 2-PS, while kcat is not
significantly different. Variant S1 has a further reduced Km
relative to the wild type and E3 and also a kcat that is 6.7-fold
higher than that of wild-type 2-PS, resulting in 19-fold
increased catalytic efficiency (kcat/Km). Improved activities of
the variants on malonyl-CoA therefore reflect the resulting
improvements in TAL production.
§S.-Y.T.: CAS Key Laboratory of Microbial Physiological and
Metabolic Engineering, Institute of Microbiology, Chinese
Academy of Sciences, No. 1 West Beichen Rd., Chaoyang
District, Beijing 100101, China.
∥O.A.: The Procter and Gamble Co., Cincinnati, OH 45241,
United States.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the Penn State Center for Quantitative Cell Analysis
for cell sorting and the Rice University Bioscience Research
Collaborative for mass spectrometry determination. Financial
support was provided by the National Science Foundation
(CAREER Award No. 0644678 and CBET Award No.
1159398).
Malonyl-CoA is a major building block for natural products
such as fatty acids, polyketides, and flavonoids, and increasing
intracellular malonyl-CoA levels using metabolic engineering
strategies have improved malonyl-CoA-dependent production
of a variety of these compounds in E. coli.23−28 The intracellular
malonyl-CoA concentration in an E. coli K-12 “wild-type” strain
REFERENCES
■
(1) Bentley, R.; Zwitkowits, P. M. J. Am. Chem. Soc. 1967, 89, 681.
10102
dx.doi.org/10.1021/ja402654z | J. Am. Chem. Soc. 2013, 135, 10099−10103

Journal of the American Chemical Society
Article
(2) Bentley, R.; Zwitkowits, P. M. J. Am. Chem. Soc. 1967, 89, 676.
(3) Ma, S. M.; Li, J. W.; Choi, J. W.; Zhou, H.; Lee, K. K.; Moorthie,
V. A.; Xie, X.; Kealey, J. T.; Da Silva, N. A.; Vederas, J. C.; Tang, Y.
Science 2009, 326, 589.
(4) Maia, T.; Richardson, N. L. P. Metab. Eng. 1999, 1, 180.
(5) Groutas WC, S. M.; Brubaker, M. J.; Huang, T. L.; Moi, M. K.;
Carroll, R. T. J. Med. Chem. 1985, 28, 1106.
(6) Ridley, C. P.; Khosla, C. ACS Chem. Biol. 2007, 2, 104.
(7) Fairlamb, I. J.; Marrison, L. R.; Dickinson, J. M.; Lu, F. J.;
Schmidt, J. P. Bioorg. Med. Chem. 2004, 12, 4285.
(8) Hansen, C. A.; Frost, J. W. J. Am. Chem. Soc. 2002, 124, 5926.
(9) Agrawal, J. P. Prog. Energy Combust. Sci. 1998, 24, 1.
(10) Chia, M.; Schwartz, T. J.; Shanks, B. H.; Dumesic, J. A. Green
Chem. 2012, 14, 1850.
(11) Eckermann, S.; Schroder, G.; Schmidt, J.; Strack, D.; Edrada, R.
A.; Helariutta, Y.; Elomaa, P.; Kotilainen, M.; Kilpelainen, I.; Proksch,
P.; Teeri, T. H.; Schroder, J. Nature 1998, 396, 387.
(12) Zha, W. J.; Shao, Z. Y.; Frost, J. W.; Zhao, H. M. J. Am. Chem.
Soc. 2004, 126, 4534.
(13) Xie, D.; Shao, Z.; Achkar, J.; Zha, W.; Frost, J. W.; Zhao, H.
Biotechnol. Bioeng. 2006, 93, 727.
(14) Gredell, J. A.; Frei, C. S.; Cirino, P. C. Biotechnol. J. 2012, 7, 477.
(15) van der Meer, J. R.; Belkin, S. Nat. Rev. Microbiol. 2010, 8, 511.
(16) Galvao, T. C.; de Lorenzo, V. Curr. Opin. Biotechnol. 2006, 17,
34.
(17) Tang, S. Y.; Fazelinia, H.; Cirino, P. C. J. Am. Chem. Soc. 2008,
130, 5267.
(18) Tang, S. Y.; Cirino, P. C. Angew. Chem., Int. Ed. 2011, 50, 1084.
(19) Morris, G. M.; Huey, R.; Lindstrom, W.; Sanner, M. F.; Belew,
R. K.; Goodsell, D. S.; Olson, A. J. J. Comput. Chem. 2009, 30, 2785.
(20) Cosconati, S.; Forli, S.; Perryman, A. L.; Harris, R.; Goodsell, D.
S.; Olson, A. J. Expert Opin. Drug Discovery 2010, 5, 597.
(21) Schleif, R. FEMS Microbiol. Rev. 2010, 34, 779.
(22) Jez, J. M.; Austin, M. B.; Ferrer, J.; Bowman, M. E.; Schroder, J.;
Noel, J. P. Chem. Biol. 2000, 7, 919.
(23) Lu, X.; Vora, H.; Khosla, C. Metab. Eng. 2008, 10, 333.
(24) Davis, M. S.; Solbiati, J.; Cronan, J. E., Jr. J. Biol. Chem. 2000,
275, 28593.
(25) Zha, W. J.; Rubin-Pitel, S. B.; Shao, Z. Y.; Zhao, H. M. Metab.
Eng. 2009, 11, 192.
(26) Cao, Y.; Jiang, X.; Zhang, R.; Xian, M. Appl. Microbiol.
Biotechnol. 2011, 91, 1545.
(27) Leonard, E.; Lim, K. H.; Saw, P. N.; Koffas, M. A. Appl. Environ.
Microbiol. 2007, 73, 3877.
(28) Xu, P, R. S.; Fowler, Z. L.; Maranas, C. D.; Koffas, M. A. G.
Metab. Eng. 2011, 13, 578.
(29) Bennett, B. D.; Kimball, E. H.; Gao, M.; Osterhout, R.; Van
Dien, S. J.; Rabinowitz, J. D. Nat. Chem. Biol. 2009, 5, 593.
(30) Takamura, Y.; Nomura, G. J. Gen. Microbiol. 1988, 134, 2249.
(31) Koryakina, I.; McArthur, J. B.; Draelos, M. M.; Williams, G. J.
Org. Biomol. Chem. 2013, 11, 4449.
10103
dx.doi.org/10.1021/ja402654z | J. Am. Chem. Soc. 2013, 135, 10099−10103