An evolutionary model encompassing substrate specificity and reactivity of type i polyketide synthase thioesterases

Article abstract of DOI:10.1002/cbic.201402475

Bacterial polyketides are a rich source of chemical diversity and pharmaceutical agents. Understanding the biochemical basis for their biosynthesis and the evolutionary driving force leading to this diversity is essential to take advantage of the enzymes as biocatalysts and to access new chemical diversity for drug discovery. Biochemical characterization of the thioesterase (TE) responsible for 6-deoxyerythronolide macrocyclization shows that a small, evolutionarily accessible change to the substrate can increase the chemical diversity of products, including macrodiolide formation. We propose an evolutionary model in which TEs are by nature non-selective for the type of chemistry they catalyze, producing a range of metabolites. As one metabolite becomes essential for improving fitness in a particular environment, the TE evolves to enrich for that corresponding reactivity. This hypothesis is supported by our phylogenetic analysis, showing convergent evolution of macrodiolide-forming TEs.

Full text of DOI:10.1002/cbic.201402475

CHEMBIOCHEM
COMMUNICATIONS
DOI: 10.1002/cbic.201402475
An Evolutionary Model Encompassing Substrate Specificity
and Reactivity of Type I Polyketide Synthase Thioesterases
Taylor P. A. Hari, Puneet Labana, Meaghan Boileau, and Christopher N. Boddy*[a]
Bacterial polyketides are a rich source of chemical diversity and
pharmaceutical agents. Understanding the biochemical basis
for their biosynthesis and the evolutionary driving force lead-
ing to this diversity is essential to take advantage of the en-
zymes as biocatalysts and to access new chemical diversity for
drug discovery. Biochemical characterization of the thioester-
ase (TE) responsible for 6-deoxyerythronolide macrocyclization
shows that a small, evolutionarily accessible change to the
substrate can increase the chemical diversity of products, in-
cluding macrodiolide formation. We propose an evolutionary
model in which TEs are by nature non-selective for the type of
chemistry they catalyze, producing a range of metabolites. As
one metabolite becomes essential for improving fitness in a
particular environment, the TE evolves to enrich for that corre-
sponding reactivity. This hypothesis is supported by our phylo-
genetic analysis, showing convergent evolution of macrodio-
lide-forming TEs.
Unique among these PKS domains is the thioesterase (TE).
Phylogenic analysis of TEs show that they do not cluster based
on substrate specificity or reactivity (Figure 1). Thus, predicting
and characterizing TE specificity and reactivity has relied on ex-
perimental characterization.[11–13] We characterized the in vitro
substrate specificity and reactivity of the TE from the erythro-
mycin biosynthetic pathway (DEBS TE), and herein we show
that it is both highly substrate-specific and capable of catalyz-
ing a high diversity of chemistries, including hydrolysis, esterifi-
cation, macrocyclization, and macrodiolide formation. We pro-
pose an evolutionary model using the screening hypothe-
sis[14,15] to account for these observations.
Type I PKS TEs play a key role in the biosynthesis of poly-
ketides, catalyzing release of the completed polyketide chain
from the PKS to which it is covalently linked.[16] This step is es-
sential for turnover of the biosynthesis machinery and provides
an opportunity to increase the chemical diversity and complex-
ity of the polyketide product. Type I PKS TEs can catalyze
hydrolysis of the acyl-PKS intermediate, generating free acids
as seen in ambruticin[17] and gephyronic acid[18] biosynthesis,
transesterification with an intermolecular alcohol to generate
esters as has been observed in vitro,[19] macrocyclization with
intramolecular alcohols to generate macrolactones as seen in
erythromycin[20] and epothilone[21,22] biosynthesis, or dimeriza-
tion with a second equivalent of acyl-PKS to generate macro-
diolides as is predicted in elaiophylin[23] and disorazol[24] biosyn-
thesis.
Modular polyketides, produced by bacteria, represent some of
the most chemically rich and biologically potent molecules
known. This diverse family of compounds includes clinical anti-
biotics, anticancer agents, and immunosuppressants. Although
highly diverse in structure and function, these compounds are
unified by their mechanism of biosynthesis.[1,2] All are pro-
duced from simple building blocks in an assembly line-like
fashion by multidomain proteins called type I polyketide syn-
thases (PKS). Understanding the detailed mechanism of their
biosynthesis is critical to the discovery of new polyketides and
to harnessing PKSs to produce new and designer polyketide
products.
The best-studied type I PKS TE is the TE from the erythromy-
cin biosynthetic pathway, DEBS TE. In vitro biochemical charac-
terization of DEBS TE has shown that macrocyclization activity
is highly substrate-selective, with hydrolytic activity predomi-
nating in virtually all substrates investigated. For example,
a small change in oxidation state or stereochemistry of the
two substrates known to undergo macrocyclization completely
abolishes macrocycle formation.[11,25,26] The TE from the pikro-
mycin biosynthetic pathway(PIK TE) demonstrates similarly
high substrate specificity for macrocyclization, suggesting that
this is a general phenomenon of type I PKS TEs.[13,26]
Phylogenic analysis of type I PKS catalytic domains has
shown that substrate specificity or reactivity of the domains
often has a primary sequence determinant. For example, the
acyltransferase (AT) domains cluster into malonyl-CoA- and
methylmalonyl-CoA-specific clades.[3–6] b-ketosynthase (KS) do-
mains from trans AT pathways cluster based on the substitu-
tion at the a and b carbons of the upstream polyketide inter-
mediates.[7,8] Ketoreductase (KR) domains cluster into type A or
type B, based on the facial selectivity for delivery of the hy-
dride from NADPH.[9,10]
In our efforts to investigate the origin of substrate selectivity
for macrocyclization, we synthesized substrates that were bio-
synthetically related to a known DEBS TE macrocyclization-
competent substrate, 3 (Scheme 1). Compound 3 possessed
key design features we wished to retain in the panel of sub-
strates used for this study. These included the replacement of
the ethyl group from 1 with a benzyl group, as this was
known to be effectively processed by DEBS TE and provided
an easily detectable chromophore,[25,27] the amide linkage,
which facilitates synthesis, and the N-acetylcysteamine (NAC)
activation of the carboxylate, which has been extensively used
[a] T. P. A. Hari, P. Labana, M. Boileau, Prof. C. N. Boddy
Departments of Chemistry and Biology
Centre for Catalysis Research and Innovation, University of Ottawa
Ottawa, ON K1N 6N5 (Canada)
E-mail: cboddy@uottawa.ca
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cbic.201402475.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 2656 – 2661 2656

CHEMBIOCHEM
COMMUNICATIONS
www.chembiochem.org
Figure 1. Phylogenetic analysis of 138 TE domains shows that they do not cluster based on substrate specificity or function. Pathway type and origin appear
*
*
:
to be greater determinants for clustering of TEs. Sequences and phylogenetic methodology can be found in the Supporting Information. : macrolactone,
~
&
&
^
d-lactone, : pyrone, : macrodiolide, : macrolactam, : acid.
with DEBS TE[11,25,26,28–31] and mimics the cysteamine and C ter-
minus of b-alanine from the native substrate’s phosphopante-
theinyl group.[30,32] Compound 4 could be envisaged as arising
from a biosynthetic gain of dehydratase (DH) function, result-
ing in conversion of the b-hydroxy functionality of 3 into an
a,b-unsaturated carbonyl. The relationship between 3 and 4 is
similar to what is seen for 6-deoxyerythronolide B (1) and nar-
bonolide (2), which are generated by the related TEs DEBS TE
and PIK TE, respectively. The enantiomer, ent-4, was synthe-
sized as a negative control compound, as no d-configured al-
cohol has been shown to be macrocyclized in vitro or in vivo
by DEBS TE.[25,27,33–36] Substrate 5 was envisioned as being relat-
ed to 4 through modulation of the specificity of the acyltrans-
ferase domain, from malonyl-CoA-specific to methylmalonyl-
CoA-specific. This panel of substrates provided compounds
that could be expected to arise through the evolution of PKS
pathways, thus enabling us to provide some of the first in-
sights into the relationship between substrate selectivity and
PKS pathway evolution.
Synthesis of 4 and ent-4 was accomplished in eight linear
steps (Scheme 2). An asymmetric Brown allylation[37,38] of phe-
nylacetaldehyde (6) set the configuration of the stereogenic
carbon, furnishing the homoallylic alcohols 7 and ent-7 in
moderate yields. These were silylated, and the allylic double
bonds were oxidized to yield aldehydes 8 and ent-8. Horner–
Wadsworth–Emmons coupling between 8 and phosphonate 9
was accomplished in moderate yield, furnishing the E-a,b-un-
saturated esters 10 and ent-10. Subsequent hydrolysis of the
ethyl ester gave the carboxylic acid, which was treated with
EDC and coupled with free amine 13 to yield the silylated
NAC-thioester substrates 11 and ent-11 in good yield. Depro-
tection in an acetonitrile solution of HF/pyridine[25] completed
the synthesis of enantio-enriched substrates 4 and ent-4
(Scheme 2 and Scheme S2 in the Supporting Information). Sub-
strate 5 was generated in racemic form through an analogous
route, starting with the racemic form of aldehyde
8
(Scheme S3). Wittig coupling with methyl 2-(triphenylphos-
phoranylidene)propionate, gave the methyl substituted a,b-un-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 2656 – 2661 2657

CHEMBIOCHEM
COMMUNICATIONS
www.chembiochem.org
Scheme 1. Substrate 3 is known to undergo macrocyclization with DEBS TE.
Substrates 4, ent-4, and racemic 5 were designed based on the relationship
between 6-deoxyerythronolide B (1) and narbonolide 2 to evaluate the
effect of evolutionarily accessible substrate modifications on the function of
the DEBS TE.
Scheme 2. Synthesis of l-enantio-enriched substrate 4: a) (À)-Ipc2BAllyl,
À1008C, 90% ee, 65%; b) TBSCl, imidazole; K3Fe(CN)6, K2OsO4; Pb(OAc)4,
70% over three steps; c) 9, KOH, 63%; d) NaOH; 13, EDC, DMAP, Et3N, 72%
over two steps; e) HF, pyridine, 64%; f) HCl/Dioxane, 100%. The d-enantio-
enriched substrate ent-4 was prepared in identical fashion by using
(+)-Ipc2BAllyl.
saturated ester, which was converted into the NAC-thioester
substrate following analogous chemistry to 4 and ent-4.
To evaluate the ability of DEBS TE to macrolactonize 4, ent-4,
and 5, purified recombinant DEBS TE[30] (15 mm) was treated
with each NAC-thioester substrate (2.5 mm) in 50 mm phos-
phate buffer (pH 7.4) with DMSO (10% v/v). After 18 h, the re-
actions were quenched and analyzed for product formation by
LC/MS. The results obtained for 4 and ent-4 were consistent
with the stereochemical hypothesis indicating that l-config-
Figure 2. LCMS analysis of the enzymatic reaction of 4 with DEBS TE. The UV (254 nm) trace between 8 and 26 min is shown. Major and minor products are
indicated on each trace with the observed m/z values.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 2656 – 2661 2658

CHEMBIOCHEM
COMMUNICATIONS
www.chembiochem.org
ured alcohols were necessary for macrocyclization (Figures 2
and S4). The l-configured alcohol 4 was clearly converted into
the 14-member ring macrolactone 16, whereas the d-config-
ured alcohol ent-4, did not appear to be converted. Although
trace 16 can be observed from prolonged incubation of ent-4
with DEBS TE (Figure S4), the level of 16 formation was consis-
tent with conversion of trace 4 found in ent-4 into macrocycle
(ent-4 ee 90%, see the Supporting Information). Incubation of
4 and ent-4 with DEBS TE D169A, which is folded but catalyti-
cally inactive,[30] showed no consumption of the substrates or
product formation (Figures S7–S12). Macrolactone 16 obtained
from the treatment of 4 with wild type DEBS TE was confirmed
intermediate, releasing a linear ester. This type of transesterifi-
cation activity has been seen with glycerol and type I PKS
TEs,[19] as well as with the TE for the fungal pathway responsi-
ble for zearalenone biosynthesis (Zea TE).[44] In the case of the
Zea TE, a wide range of simple exogenous nucleophiles were
shown to be capable of accessing the acyl-enzyme intermedi-
ate. We thus hypothesized that if dimer formation was occur-
ring through exogenous nucleophile attack on the acyl-
enzyme intermediate by a second equivalent of monomer,
then other exogenous nucleophiles should also be able to
attack the acyl-enzyme intermediate. We therefore investigated
the ability of 4 to undergo DEBS TE-catalyzed transesterifica-
tion with exogenous alcohols. A panel of alcohols were incu-
1
by MS and H NMR analysis. Kinetic characterization of DEBS TE
activity with 4 and ent-4 displayed Michaelis–Menten kinetics
with kcat values of 0.228Æ0.015 minÀ1 and 0.114Æ0.008 minÀ1
,
Table 1. Relative velocities for DEBS TE-mediated crosscoupling with
exogenous O-nucleophiles.[a]
respectively, KM values of 2.8Æ0.4 mm and 1.3Æ0.2 mm, re-
spectively and kcat/KM values of 1.4Æ0.3mÀ1 sÀ1 and 1.5Æ
0.3mÀ1 sÀ1, respectively (Figures S13 and S14). These data are
consistent with the observed catalytic efficiencies of in vitro
characterized modular polyketide TEs.[11–13,25,26,28–31,39,40] For ex-
ample DEBS TE processed 3 with a specificity constant of 1.2Æ
0.5mÀ1 sÀ1, which is in excellent agreement with our data.
Analysis of the reaction products for 4 also demonstrated
that the C2- symmetrical, head-to-tail dimerized macrodiolide
product 19 ([M+H]+ 631 m/z) was also formed. In addition, the
glycerolysis adduct 14 ([M+H]+ 408 m/z), hydrolysis prod-
uct 15 ([M+H]+ 334 m/z), the linear dimer seco-acid 17
([M+H]+ 648 m/z), and linear dimer NAC-thioester 18 ([M+H]+
750 m/z) were also observed (Figure 2). The structures of the
Nucleophile Relative con-
version
Nucleophile Relative con-
version
Nucleophile
butan-1-ol
propan-1-ol
methanol
ethanol
1.00Æ0.08
0.85Æ0.16
0.38Æ0.03
0.36Æ0.04
0.34Æ0.02
pentan-3-ol
pentan-1-ol
hexan-1-ol
t-butanol
0.26Æ0.02
0.22Æ0.08
0.17Æ0.05
0.14Æ0.01
0.12Æ0.01
butan-1-ol
propan-1-ol
methanol
ethanol
cyclohexanol
propan-2-ol
cyclohexanol
[a] For 15 mm DEBS TE with 2.5 mm 4 in 50 mm phosphate buffer (pH 7.4)
at 238C, with DMSO (10%, v/v) and 100 mm nucleophile. Negative con-
trols lacking DEBS TE showed no reactivity.
1
dimeric compounds were consistent with H NMR and MS/MS
bated with 4, generating the corresponding ester adducts
(Table 1). The formation of the ester products was consistent
with our proposed mechanism for dimer formation, further
supporting this mechanism. Comparison of the relative conver-
sion to the transesterification product showed that primary al-
cohols were converted most effectively into esters as com-
pared to secondary and tertiary alcohols. As was seen with the
transesterification catalyzed by the fungal Zea TE, butan-1-ol
was the most competent alcohol substrate for DEBS TE.[44] This
similar substrate specificity in two diverse TEs suggests that
transesterification could be a broadly conserved activity for
PKS TEs; however, further examples are needed to test this hy-
pothesis.
data collected on these compounds (Supporting Information).
The formation of the macrodiolide and linear dimer structures
is unprecedented for the DEBS TE, though TEs from the bio-
synthesis of elaiophylin[23] and disorazol[24] are proposed to cat-
alyze this chemistry.
The mechanism for related non-ribosomal peptide synthe-
tase TE-mediated dimer formation has been investigated in
vitro.[41–43] Enterobactin TE, which trimerizes a 2,3-dihydroxy-
benzoyl-l-serinyl monomer and macrocyclizes it to generate
a macrotriolide has been biochemically characterized. By using
top-down mass spectrometry, a mutant of this TE was shown
to accumulate monomeric and dimeric acyl-enzyme intermedi-
ates. It was postulated that the dimer was formed by the seryl
side chain alcohol of the monomer-TE attacking an activated
thioester of a second equivalent of monomer.[41] The formation
of 18 in the current study strongly suggests that a second
equivalent of monomer attacks the electrophilic monomeric TE
acyl-enzyme intermediate, generating the dimeric NAC-thioest-
er 18. This type of mechanism was also proposed for the for-
mation of the macrocyclic dimeric non-ribosomal peptide, gra-
micidin.[43] We thus propose that in our assay, formation of the
macrodiolide 19 and the dimeric free acid 17 occurs by load-
ing of 18 onto the TE to generate a dimeric acyl-enzyme inter-
mediate, which can undergo macrocyclization and hydrolysis,
respectively.
Incubation of DEBS TE with 5, the methyl-substituted deriva-
tive, led to formation of the hydrolysis product with no detect-
able macrocycle. Kinetic characterization of the enzymatic ac-
tivity fit the Michaelis–Menten model, providing a kcat value of
0.20Æ0.01 minÀ1 and a KM value of 0.75Æ0.13 mm (Fig-
ure S15), consistent with type I PKS TE-catalyzed chemis-
tries.[11–13,25,26,28–31,39,40] The addition of the a-methyl group sub-
stantially increases the allylic 1,3-strain[45] across the amide
bond, limiting access to the S-trans configuration required for
macrocyclization. The observed lack of macrocyclization with 5
thus suggests that DEBS TE was unable to overcome intrinsic
unfavorability of the reactive conformation of the substrate.
The screening hypothesis suggests that long-term evolution-
ary selection favors pathways that support chemical diversity
and can be readily “screened” in the face of new environ-
This mechanism requires an exogenous nucleophile (a
second equivalent of monomer) to attack the TE acyl-enzyme
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 2656 – 2661 2659

CHEMBIOCHEM
COMMUNICATIONS
www.chembiochem.org
ments.[14,15] In part, the diversity in a multistep pathway is de-
pendent on substrate tolerance of late-stage enzymes. This tol-
erance permits plasticity in the upstream components of the
pathway and fitness-based screening of the newly released
compounds. In addition, the off-loading enzymes should be
able to process diverse substrates and introduce substantial
new chemical diversity to amplify the new compounds accessi-
ble from upstream modifications.
Acknowledgements
This work was supported by the Natural Sciences and Engineer-
ing Research Council (NSERC) of Canada, the Canadian Founda-
tion for Innovation, the Ministry for Research and Innovation,
and the University of Ottawa.
Keywords: macrodiolide
·
macrolactone
·
screening
Our results, in light of the screening hypothesis, suggest a
model for the evolution of type I polyketide chemical diversity
in the context of TE-mediated off-loading. We propose that an-
cestral TEs were able to catalyze a variety of chemistries includ-
ing hydrolysis, esterification, amidation, macrocyclization, and
macrodiolide formation. This range of reactivity expands the
chemical diversity accessible from a PKS pathway. As one me-
tabolite, such as a macrodiolide, becomes essential for improv-
ing fitness in a particular environment, the TE evolves under
pressure to enrich for the corresponding reactivity. This is sup-
ported by the phylogenic analysis, which is consistent with
convergent evolution of macrodiolide-forming activity. We
hypothesize that the ability to access this diversity of reactivity
is retained as TE evolves to become selective for a particular
product. Based on this hypothesis, we predict that small
changes to the substrate can unlock access to this ancestral
activity. This has been observed in the in vitro characterization
of the DEBS TE domain, where subtle structural changes lead
to dramatic changes in reactivity, from hydrolysis, to macrocy-
cle formation, to macrodiolide formation.
hypothesis · substrate selectivity · thioesterase
[1] C. Khosla, D. Herschlag, D. E. Cane, C. T. Walsh, Biochemistry 2014, 53,
2875.
[2] T. A. M. Gulder, M. F. Freeman, J. Piel, Top. Curr. Chem. 2011; DOI:
10.1007/128_2010_113.
[3] H. Jenke-Kodama, A. Sandmann, R. Mꢁller, E. Dittmann, Mol. Biol. Evol.
2005, 22, 2027.
[4] C. P. Ridley, H. Y. Lee, C. Khosla, Proc. Natl. Acad. Sci. USA 2008, 105,
4595.
[5] G. Yadav, R. S. Gokhale, D. Mohanty, J. Mol. Biol. 2003, 328, 335.
ˇ ˚
[6] P. Hill, J. Piel, S. Aris-Brosou, V. Kristufek, C. N. Boddy, L. Dijkhuizen, J.
Ind. Microbiol. Biotechnol. 2014, 41, 75.
[7] T. Nguyen, K. Ishida, H. Jenke-Kodama, E. Dittmann, C. Gurgui, T. Hoch-
muth, S. Taudien, M. Platzer, C. Hertweck, J. Piel, Nat. Biotechnol. 2008,
26, 225.
[8] R. Teta, M. Gurgui, E. J. N. Helfrich, S. Kꢁnne, A. Schneider, G. Van Echt-
en-Deckert, A. Mangoni, J. Piel, ChemBioChem 2010, 11, 2506.
[9] P. Caffrey, ChemBioChem 2003, 4, 654.
[10] R. Reid, M. Piagentini, E. Rodriguez, G. Ashley, N. Viswanathan, J. Carney,
D. V. Santi, C. R. Hutchinson, R. McDaniel, Biochemistry 2003, 42, 72.
[11] R. S. Gokhale, D. Hunziker, D. E. Cane, C. Khosla, Chem. Biol. 1999, 6, 117.
[12] C. N. Boddy, T. L. Schneider, K. Hotta, C. T. Walsh, C. Khosla, J. Am. Chem.
Soc. 2003, 125, 3428.
Our fitness-based model is sufficient to account for the ob-
served experimental TE activity results in the literature. It can
also be used predictively. For example, polyketide natural
products have been isolated with functional groups derived
from TE-mediated hydrolysis, macrolactonization, macrolactam-
ization, and dimerization. However, no transesterification-de-
rived polyketide products have been isolated. Having observed
a butan-1-ol kinetic preference in two diverse TEs, we predict
that a butan-1-ol ester would likely be natively produced by
a yet unidentified PKS pathway. Furthermore, as the presence
of butan-1-ol is required for this activity, we predict that it will
be observed in a native producer of butanol, such as the
genus Clostridium, when at least three of the forty sequenced
genomes (Clostridium kluyveri DSM 555, Clostridium botulinum
A2 BoNT, and C. botulinum H04402 065) encode a TE[46] and
possess a butanol dehydrogenase A (bdhA) orthologue.[47]
In summary, we have used phylogenic analysis and in vitro
biochemical characterization to develop and support an evolu-
tionary model in which the plasticity of TEs increases long-
term fitness through the retained ability to generate and
access chemical diversity. We biochemically showed with re-
combinant DEBS TE that small, evolutionarily accessible
changes to the substrate can lead to multiple different chemis-
tries, including hydrolysis, macrocyclization, dimerization, mac-
rodiolide formation, and transesterification.
[13] C. C. Aldrich, L. Venkatraman, D. H. Sherman, R. A. Fecik, J. Am. Chem.
Soc. 2005, 127, 8910.
[14] R. D. Firn, C. G. Jones, Nat. Prod. Rep. 2003, 20, 382.
[15] R. D. Firn, C. G. Jones, J. Exp. Bot. 2009, 60, 719.
[16] R. M. Kohli, C. T. Walsh, Chem. Commun. 2003, 297.
[17] B. Julien, Z.-Q. Tian, R. Reid, C. D. Reeves, Chem. Biol. 2006, 13, 1277.
[18] J. Young, D. C. Stevens, R. Carmichael, J. Tan, S. Rachid, C. N. Boddy, R.
Mꢁller, R. E. Taylor, J. Nat. Prod. 2013, 76, 2269.
[19] A. J. Hughes, M. R. Tibby, D. T. Wagner, J. N. Brantley, A. T. Keatinge-Clay,
Chem. Commun. 2014, 50, 5276–5278.
[20] S. Donadio, M. J. Staver, J. B. McAlpine, S. J. Swanson, L. Katz, Science
1991, 252, 675.
[21] L. Tang, S. Shah, L. Chung, J. Carney, L. Katz, C. Khosla, B. Julien, Science
2000, 287, 640.
[22] I. Molnꢂr, T. Schupp, M. Ono, R. Zirkle, M. Milnamow, B. Nowak-Thomp-
son, N. Engel, C. Toupet, A. Stratmann, D. D. Cyr, J. Gorlach, J. M. Mayo,
A. Hu, S. Goff, J. Schmid, J. M. Ligon, Chem. Biol. 2000, 7, 97.
[23] M. Kopp, H. Irschik, S. Pradella, R. Mꢁller, ChemBioChem 2005, 6, 1277.
[24] S. F. Haydock, T. Mironenko, H. I. Ghoorahoo, P. F. Leadlay, J. Biotechnol.
2004, 113, 55.
[25] A. Pinto, M. Wang, M. Horsman, C. N. Boddy, Org. Lett. 2012, 14, 2278.
[26] W. He, J. Wu, C. Khosla, D. E. Cane, Bioorg. Med. Chem. Lett. 2006, 16,
391.
[27] J. R. Jacobsen, C. R. Hutchinson, D. E. Cane, C. Khosla, Science 1997, 277,
367.
[28] H. Lu, S.-C. Tsai, C. Khosla, D. E. Cane, Biochemistry 2002, 41, 12590.
[29] K. K. Sharma, C. N. Boddy, Bioorg. Med. Chem. Lett. 2007, 17, 3034.
[30] M. Wang, C. N. Boddy, Biochemistry 2008, 47, 11793.
[31] M. Wang, P. Opare, C. N. Boddy, Bioorg. Med. Chem. Lett. 2009, 19, 1413.
[32] Y. Liu, T. Zheng, S. D. Bruner, Chem. Biol. 2011, 18, 1482.
[33] C. M. Kao, G. Luo, L. Katz, D. E. Cane, C. Khosla, J. Am. Chem. Soc. 1995,
117, 9105.
[34] C. M. Kao, M. McPherson, R. N. McDaniel, H. Fu, D. E. Cane, C. Khosla, J.
Am. Chem. Soc. 1997, 119, 11339.
[35] Q. Xue, G. Ashley, C. R. Hutchinson, D. V. Santi, Proc. Natl. Acad. Sci. USA
1999, 96, 11740.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 2656 – 2661 2660

CHEMBIOCHEM
COMMUNICATIONS
www.chembiochem.org
[36] R. McDaniel, A. Thamchaipenet, C. Gustafsson, H. Fu, M. Betlach, G.
Ashley, Proc. Natl. Acad. Sci. USA 1999, 96, 1846.
[42] J. W. Trauger, R. M. Kohli, H. D. Mootz, M. A. Marahiel, C. T. Walsh, Nature
2000, 407, 215.
[43] K. M. Hoyer, C. Mahlert, M. A. Marahiel, Chem. Biol. 2007, 14, 13.
[44] M. Wang, H. Zhou, M. Wirz, Y. Tang, C. N. Boddy, Biochemistry 2009, 48,
6288.
[37] H. C. Brown, P. K. Jadhav, J. Am. Chem. Soc. 1983, 105, 2092.
[38] P. K. Jadhav, K. S. Bhat, P. T. Perumal, H. C. Brown, J. Org. Chem. 1986, 51,
432.
[45] R. W. Hoffmann, Chem. Rev. 1989, 89, 1841.
[46] A.-C. Letzel, S. J. Pidot, C. Hertweck, Nat. Prod. Rep. 2013, 30, 392.
[47] D. J. Petersen, R. W. Welch, F. B. Rudolph, G. N. Bennett, J. Bacteriol.
1991, 173, 1831.
[39] J. B. Scaglione, D. L. Akey, R. Sullivan, J. D. Kittendorf, C. M. Rath, E.-S.
Kim, J. L. Smith, D. H. Sherman, Angew. Chem. Int. Ed. 2010, 49, 5726;
Angew. Chem. 2010, 122, 5862.
[40] Z. Q. Beck, C. C. Aldrich, N. A. Magarvey, G. I. Georg, D. H. Sherman, Bio-
chemistry 2005, 44, 13457.
[41] C. A. Shaw-Reid, N. L. Kelleher, H. C. Losey, A. M. Gehring, C. Berg, C. T.
Walsh, Chem. Biol. 1999, 6, 385.
Received: August 22, 2014
Published online on October 29, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 2656 – 2661 2661