Structural and functional analysis of the loading acyltransferase from avermectin modular polyketide synthase

Article abstract of DOI:10.1021/cb500873k

The loading acyltransferase (AT) domains of modular polyketide synthases (PKSs) control the choice of starter units incorporated into polyketides and are therefore attractive targets for the engineering of modular PKSs. Here, we report the structural and biochemical characterizations of the loading AT from avermectin modular PKS, which accepts more than 40 carboxylic acids as alternative starter units for the biosynthesis of a series of congeners. This first structural analysis of loading ATs from modular PKSs revealed the molecular basis for the relaxed substrate specificity. Residues important for substrate binding and discrimination were predicted by modeling a substrate into the active site. A mutant with altered specificity toward a panel of synthetic substrate mimics was generated by site-directed mutagenesis of the active site residues. The hydrolysis of the N-acetylcysteamine thioesters of racemic 2-methylbutyric acid confirmed the stereospecificity of the avermectin loading AT for an S configuration at the C-2 position of the substrate. Together, these results set the stage for region-specific modification of polyketides through active site engineering of loading AT domains of modular PKSs.

Full text of DOI:10.1021/cb500873k

Articles
pubs.acs.org/acschemicalbiology
Structural and Functional Analysis of the Loading Acyltransferase
from Avermectin Modular Polyketide Synthase
Fen Wang,† Yanjie Wang,† Junjie Ji,† Zhan Zhou,† Jingkai Yu,† Hua Zhu,‡ Zhiguo Su,† Lixin Zhang,*,‡
and Jianting Zheng*,†
†National Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing
100190, P.R. China
‡CAS Key Laboratory of Pathogenic Microbiology and Immunology, Institute of Microbiology, Chinese Academy of Sciences, Beijing
100101, P.R. China
S
* Supporting Information
ABSTRACT: The loading acyltransferase (AT) domains of
modular polyketide synthases (PKSs) control the choice of starter
units incorporated into polyketides and are therefore attractive
targets for the engineering of modular PKSs. Here, we report the
structural and biochemical characterizations of the loading AT
from avermectin modular PKS, which accepts more than 40
carboxylic acids as alternative starter units for the biosynthesis of a
series of congeners. This first structural analysis of loading ATs
from modular PKSs revealed the molecular basis for the relaxed
substrate specificity. Residues important for substrate binding and
discrimination were predicted by modeling a substrate into the
active site. A mutant with altered specificity toward a panel of
synthetic substrate mimics was generated by site-directed mutagenesis of the active site residues. The hydrolysis of the N-
acetylcysteamine thioesters of racemic 2-methylbutyric acid confirmed the stereospecificity of the avermectin loading AT for an S
configuration at the C-2 position of the substrate. Together, these results set the stage for region-specific modification of
polyketides through active site engineering of loading AT domains of modular PKSs.
last extension module to control the release and cyclization of
the full-length polyketide intermediates.
Modular polyketide synthases (PKSs) are large multifunctional
enzyme complexes responsible for the assembly of structurally
diverse polyketide natural products, many of which are
endowed with a wide range of useful biological and
pharmacological properties including antibacterial, antitumor,
antifungal, antiparasitic, and immunosuppressant activities.1−3
Polyketide biosynthesis resembles fatty acid biosynthesis in that
an acyl starter unit is condensed with a series of substituted
malonyl extender units to form poly-β-keto intermediates.
Compared to fatty acid biosynthesis, modular PKSs utilize a
much wider variety of starter and extender units, leading to
enormous structural diversity of polyketides. A minimal
extension module has three domains: a ketosynthase (KS)
that catalyzes the decarboxylative condensation, an acyltransfer-
ase (AT) that recruits the extender unit, and an acyl carrier
protein (ACP) that shuttles the growing polyketide inter-
mediate between enzymatic domains. It may also contain a
ketoreductase (KR) to catalyze the reduction of the initially
formed β-ketoester to a β-hydroxyester, a dehydratase (DH) to
dehydrate the β-hydroxyester, and an enoylreductase (ER)
domain to reduce the double bond. An additional loading
module is located at the N-terminus of the first extension
module to prime the KS1 domain, whereas a dedicated
thioesterase (TE) domain is located at the C-terminus of the
The avermectin PKS from Streptomyces avermitilis is a
modular PKS that catalyzes the formation of avermectin
aglycons by the addition of seven acetate and five propionate
extender units to a starter unit.4 Under normal conditions in
vivo, the loading module of avermectin PKS recruits 2-
methylbutyryl-coenzyme A (CoA) or isobutyryl-CoA as the
starter unit to synthesize the avermectins of “a” series or “b”
series, respectively (Figure 1), whereas the substrate feeding of
a S. avermitilis mutant shows more than 40 alternative
carboxylic acids can be accepted as starter units to generate a
wide range of novel derivatives, including the commercially
important analogue, doramectin.5 Interestingly, this broad
specificity toward unnatural starter units could be transferred
to a heterologous modular PKS by the substitution of the
loading module. New erythromycin derivatives with branched
starter units are produced when the loading module of 6-
deoxyerythronolide B synthase (DEBS) is replaced by that of
the avermectin PKS.6 The loading module of avermectin PKS
contains an AT domain and an ACP domain and is thus also
Received: October 25, 2014
Accepted: January 12, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/cb500873k
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
The extension ATs that play a pivotal role in extender unit
selection during polyketide chain elongation have been
extensively investigated. The crystal structures of two extension
ATs from DEBS (EryAT3 and EryAT5, named by the
corresponding PKS and module of origin) have been solved
and facilitated the engineering of substrate specificity of
extension AT domains.12,13 Here, we report the 2.0 Å
resolution structure of the loading AT domain of avermectin
PKS, AveAT0, which represents a compelling target for the
study of substrate specificity because of its ability to accept
various unnatural starter units for the biosynthesis of a series of
avermectin derivatives. The structure helped reveal the
molecular basis for the broad substrate specificity. Residues
important for substrate binding and discrimination were
predicted by modeling a substrate into the active site and a
single mutation could significantly alter the specificity and
activity. The stereospecificity of AveAT0 was revealed by the
hydrolysis of racemic 2-methylbutyryl-SNAC (SNAC = N-
acetylcysteamine that mimics the end of the pantetheine moiety
of CoA) thioester. The structural and functional studies of
AveAT0 presented here highlight the possibility of the rational
engineering of loading ATs in modular PKSs.
Figure 1. Loading AT within avermectin PKS. The loading AT recruits
2-methylbutyryl or isobutyryl from coenzyme A and transfers it to
loading ACP to primer the biosynthesis of avermectins of “a” series or
“b” series, respectively.
called a loading didomain. The loading AT is the primary
determinant of starter unit specificity in polyketide chain
initiation, while the loading ACP accepts the starter unit
selected by the loading AT and transfers it to the KS of the first
extension module. The didomain architecture of loading
module is also observed in DEBS that catalyzes the formation
of 6-deoxyerythronolide B (6-dEB), the aglycon of eryth-
romycin. In the wild type strain, the loading AT of DEBS
recruits only propionate-CoA as starter unit, whereas in
recombinant cells or with the purified enzyme in vitro, several
other acyl-CoA thioesters are accepted as starter units to prime
the assembly of 6-dEB derivatives.7 The loading AT of
lipomycin PKS shows significant similarity to that of avermectin
PKS. As expected, in vitro characterization of LipPks1, a PKS
subunit containing both the loading didomain and the first
extension module, reveals that several acyl-CoA thioesters,
including isobutyryl-CoA, are accepted as the starter units.8
More than half of PKS loading modules contain an additional
N-terminal domain termed KSQ closely resembling a KS
domain except that the catalytic cysteine residue of the active
site is replaced by glutamine (Q).9 ATs of KSQ-type loading
modules are specific for CoA-esters of dicarboxylic acids rather
than monocarboxylic acids that are accepted by the ATs of
avermectin-type (AVE-type) loading modules. Decarboxylation
of dicarboxylic acid CoA-esters catalyzed by KSQs has been
invoked as the mechanism for chain initiation on these modular
PKSs. In contrast to the broad specificity of AVE-type loading
ATs, the KSQ-type loading ATs exhibit a strict substrate
specificity and select either malonyl-CoA or methylmalony-
CoA, which is decarboxylated in situ to provide acetyl or
propionate starter units for the polyketide initiation.10 The
strict substrate specificity of the KSQ-type loading ATs is
preserved when they are studied in vitro with purified enzymes.
Similar to ATs of extension modules, the ATs of the KSQ-type
loading modules contain a conserved arginine residue in their
active sites. The crystal structure of the Escherichia coli malonyl-
CoA:ACP acyltransferase (MCAT) crystallized with malonyl-
CoA reveals that this arginine residue lies in the active site and
interacts with the free carboxyl group of the substrate.11 The
oleandomycin PKS contains a KSQ in its loading module and
its loading AT can be replaced by an authentic extension AT
without interrupting the activity of the enzyme.10
RESULTS AND DISCUSSION
■
Expression of Discrete Loading AT. In the past decade,
the structural analysis has facilitated the dissection of an intact
PKS module into isolated domains that could be used for assays
of substrate specificity, catalytic activity, stereocontrol, and
interdomain interactions.14,15 The KS-to-AT linkers have been
shown to stabilize the extension ATs that are expressed as
isolated domains. The absence or incompleteness of KS-to-AT
linkers in AT constructs results in insoluble and/or inactive
proteins. The expression of discrete loading AT domains has
been a challenge. Thus far, kinetic assays are all carried out by
using multimodular enzymes or loading AT-ACP didomains
whose utility is limited due to the presence of the fused ACPs
and other catalytic domains.7,8,16 Here, we show that AveAT0
can be expressed as a standalone domain in E. coli BL21 (DE3),
providing the opportunity to investigate the detailed mecha-
nism of loading ATs. The 26 amino acids preceding the
AveAT0 were included in the construct since the 107 residues
N-terminal of EryAT0 were necessary for creating an active
loading AT, whereas the G354 that was 16 residues upstream of
the loading ACP was chosen as the C-terminal end by sequence
alignments with EryKS-AT3 and EryKS-AT5 didomains.7,12,13
The N-terminally histidine-tagged protein migrated at ∼35 kDa
on a size exclusion column as estimated by comparison to
molecular weight standards, consistent with the expected
monomer mass of 40 kDa (Supporting Information Figure 1).
Overall Structure. The AveAT0 purified to homogeneity
was entered into crystallization trials and diffraction-quality
crystals were obtained by sitting-drop method using sodium
formate as the precipitant. The protein crystallized in space
group P212121 with one monomer per asymmetric unit. The
crystal structure was solved by molecular replacement using the
structure of EryAT5 (PDB code 2HG4) as a search model and
refined to 2.0 Å resolution, with Rworking and Rfree values of 0.21
and 0.24, respectively (Table 1).13 The final refined model
contains 318 out of 354 residues of the monomer. No electron
density was observed for the 26 N-terminal residues, the last
two C-terminal residues, and an internal disordered region
(residue 319−326).
B
DOI: 10.1021/cb500873k
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
structurally homologous proteins using the Dali server revealed
similarity to the AT domains of PKSs (PDB code 4MZ0, 2.1 Å
Cα rmsd; PDB code 3TZY, 1.8 Å Cα rmsd; PDB Code 3SBM,
2.2 Å Cα rmsd; PDB code 4AMP, 2.5 Å Cα rmsd), the AT
domains of type I mammalian fatty acid synthases (FASs)
(PDB code 2VZ8, 1.9 Å Cα rmsd), and the discrete malonyl-
CoA:ACP transacylases of type II bacterial FASs (PDB code
2G2Z, 2.4 Å Cα rmsd; PDB code 1NM2, 2.6 Å Cα
rmsd).11,17−22
Table 1. Data Collection and Refinement Statistics
(Molecular Replacement)
AveAT0
Data Collection
space group
a, b, c (Å)
α, β, γ (deg)
resolution (Å)
Rmerge
P212121
47.21, 74.29, 82.42
90
50−2.0
0.073 (0.209)
14.8(8.1)
99.4 (99.9)
4.3 (4.9)
Active Site. The active site of AveAT0 is a predominantly
hydrophobic cavity located at the interface of the large α,β-
hydrolase subdomain and the small ferredoxin-like subdomain
(Figure 2A). The volume of the AveAT0 active site cavity (971
Å3 calculated by program CAST) is significantly smaller than
that of EryAT3 (1800 Å3) and EryAT5 (1588 Å3).12,13,23
Dynamics simulations of MonAT5 reveals that the ferredoxin-
like subdomain tends to move away from the α,β-hydrolase
subdomain to form a more open and exposed active site in the
absence of substrate.24 The active site cavity of AveAT0 is
formed by the residues from the β strands of the small
subdomain, the α helices of the large subdomain, and the loops
connecting these secondary structure elements (Figure 2B).
The catalytic S120 of the highly conserved Gly-X-Ser-X-Gly
motif is located at the N terminus of helix αE where its
nucleophilicity is enhanced by the α helix dipole effect, while
catalytic His227 that is proposed to act as a general base/acid
catalyst in the acyl-transfer reaction resides in the loop FI
connecting the two subdomains. The carbonyl groups of
Asn276 and Ser279 are at hydrogen bond distance from the δ-
nitrogen atom of His227 and help to position its side chain in
an optimal orientation where the ε-nitrogen atom establishes a
hydrogen bond with the side-chain hydroxyl group of the
catalytic Ser120 (Figure 3), thereby increasing its reactivity.
The similarly arranged serine-histidine “‘catalytic dyad”’
indicates that AveAT0 utilizes a catalytic mechanism analogous
I/σI
completeness (%)
redundancy
Refinement
resolution (Å)
50−2.0
19000
No. reflections
R
work/Rfree
No. atoms
protein
water
B-factors
0.21/0.24
2392
58
protein
water
25
28
rmsd
bond lengths (Å)
bond angles (deg)
0.010
1.391
AveAT0 contains two subdomains: a large α,β-hydrolase
subdomain (residues 27−152 and 230−353) composed of a
mixed-stranded β-sheet surrounded by α-helices and a small
subdomain (residues 156−219) that adopts a typical
ferredoxin-like fold containing four-stranded antiparallel β-
sheets and two distal α-helices (Figure 2). AveAT0 most closely
resembles its counterparts in the extending modules of DEBS
(Protein Data Bank (PDB) codes 2HG4, 1.1 Å Cα rmsd; PDB
code 2QO3, 1.4 Å Cα rmsd).12,13 An automatic search for
Figure 2. Structure of a loading AT. (a) The loading AT AveAT0 contains two subdomains: a hydrolase subdomain and a ferredoxin subdomain.
The linkers between two subdomains are colored in gray. The substrate binding tunnel was shown as surface. Catalytic serine and histidine residues
are shown as stick while the N-terminal and C-terminal ends as spheres. (b) Structure based sequence alignment of AveAT0, EryAT3, and EryAT5.
The reported secondary structure is from AveAT0. The linkers connecting two subdomain are underlined in gray. Catalytic residues are labeled “*”.
Residues of active site cavity are labeled “+”.
C
DOI: 10.1021/cb500873k
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
to other PKS and FAS ATs. Transfer of the acyl group from
CoA to ACP occurs by a ping-pong bibi mechanism through an
acyl-AT intermediate that is attacked by the thiol residue of the
phosphopantetheine arm of ACP cosubstrates.25−27 In the
absence of holo-ACP, the acyl-enzyme intermediate will be
rapidly hydrolyzed due to the nucleophilic attack by the water
molecule.28
AveAT0 catalyzes the hydrolysis of a panel of synthetic
SNAC thioesters of monocarboxylic acids. The crystals could
be grown or soaked in crystallization buffer containing
millimolar concentration of these substrate mimics, but no
corresponding electron density was observed. To elucidate the
details of substrate binding, 2-methylbutyryl-SNAC was
modeled into the active site using the program Coot. The
substrate−enzyme interactions observed in the malonyl-FabD
intermediate structure complexed with CoA (PDB code 2Z2G)
were used to guide the modeling.11 The substrate was
positioned with the 2-methylbutyryl moiety inserted deep
into the cavity and the SNAC moiety toward the entrance. An
ordered water molecule forming hydrogen bonds to both the
backbone amides of the conserved Gln35 and the side chain
OH of the catalytic Ser120 was used to guide the placement of
the thioester carbonyl of the substrate while the SNAC portion
was orientated through a hydrogen bond to the side chain of
His119. The modeled results suggested that (2S)-methylbutyryl
was accepted by AveAT0 as the starter unit for biosynthesis of
the avermectin. The α-methyl in an R-orientation would
sterically clash with the catalytic histidine.
Figure 3. Active site of AveAT0. (a) The hydrogen bonds shown as
dash lines in yellow help the orientation of catalytic serine and
histidine. (b) Stereodiagram shows the 2Fo-Fc electron density map
(contoured at 1.0 δ) surrounding the residues in direct contact with
the acyl moiety of the substrate. The (2S)-methylbutyryl-SNAC
shown as yellow sticks was modeled into the cavity by program Coot.
(c) The surface representation of residues forming the cavity that
accommodates the acyl moiety of (2S)-methylbutyryl-SNAC.
The modeling of the substrate revealed 13 residues (Gln35,
Gln92, His119, Ser120, Leu121, Trp145, Gln149, Leu158,
Ile220, Val222, Val224, Ala226, and His227) within 6 Å of the
2-methylbutyryl moiety (Figure 3B, C). The hydrophobic
Figure 4. Comparisons of residues involved in enzyme−substrate recognition by sequence alignment. The alignment was numbered according to
AveAT0. The residues directly contact the acyl moiety of the substrate are labeled by “*”. The conserved HFAH or YASH motifs and arginine
residue were underlined.
D
DOI: 10.1021/cb500873k
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
interactions between these residues and the acyl moiety are the
major contributors to the enzyme−substrate recognition, in
agreement with the relaxed specificity of Ave-type loading ATs
since the hydrophobic interactions are less discriminatory
compared to the electrostatic and hydrogen bond interactions
that require specific arrangements of functional groups within a
substrate.29 A sequence alignment was performed by ClustalX
to compare the 13 residues of loading ATs (Figure 4). These
substrate-recognition residues of AVE-type loading ATs are less
conserved compared to the KSQ-type loading ATs, consistent
with the fact that monocarboxylic acids with diverse acyl
moieties are recruited. The majority of KSQ-type loading ATs
recruit malonyl or methylmalonyl as the starter unit. The active
site contains an arginine at the end of the substrate binding
pocket (position 145, AveAT0 numbering) to stabilize the
carboxylate group and a conserved “HAFH” or “YASH” motif
(positions 224−227) to mediate the binding of malonyl or
methylmalonyl, respectively, as observed in extension ATs.26
The ionic interaction of the arginine side chain with the
dicarboxylic acid thioester imposes a strong spatial constraint
on the binding of the substrate and accounts for the strict
specificity of the KSQ-type loading ATs. The BorAT0 (loading
AT of borrelidin PKS) contains a conserved arginine and
accepts trans1,2-cyclopentanedicarboxylic acid as the starter
unit that is extended without in situ decarboxylation.16 As a
result, the KSQ of borrelidin loading module has degenerated
due to the absence of selective pressure and performs no
catalytic function. Besides, the BorAT0 has a Thr224 and a
Gly226 to lower the sterical hindrance of the substrate binding
pocket. The Val222 has been proposed to constrain the size of
the malonate side chain and the V222A mutation of EryAT6
enables the incorporation of 2-propargylmalonyl, an extender
with a bulky side chain.30 Indeed, the loading AT of apoptolidin
PKS (ApoAT0) that recruits methoxymalonate has an alanine
residue at this position.31
Table 2. Kinetic Analysis of the Hydrolytic Reaction
Catalyzed by AveAT0 in the Absence of ACP Acceptor
kcat/Km
enzyme SNAC thioesters (mM−1 min−1
)
Kcat (min−1
)
Km (mM)
AveAT0 2-
methylbutyryl
29.6 3.8
14.8 0.7
0.5 0.06
a
isobutyryl
acetyl
10.4 4.5
6.0 3.2
6.0 3.0
4.1 1.1
2.2 0.8
1.3 0.4
1.1 0.6
0.5 0.4
0.1 0.01
1.7 0.3
4.8 0.8
2.3 1.1
0.2 0.2
2.6 0.3
1.5 0.7
0.1 0.01
0.03 0.01
1.8 0.2
0.8 0.5
5.2 0.9
1.2 0.2
0.6 0.07
3.3 0.4
2.6 0.4
1.0 0.1
5.1 1.9
1.0 0.4
0.5 0.2
0.2 0.1
0.1 0.05
0.8 0.2
1.2 0.4
0.8 0.2
4.6 2.3
2.0 1.4
N.D.
3-methylbutyryl
pentanoyl
butyryl
propionyl
2-methylbutyryl
isobutyryl
acetyl
V222L
V222A
b
N.D.
3-methylbutyryl
pentanoyl
butyryl
1.0 0.09
6.3 0.4
11.2 3.2
2.1 0.8
6.7 0.4
2.6 0.6
N.D.
0.6 0.1
1.3 0.2
4.9 1.9
8.6 4.4
2.6 0.2
1.7 0.7
N.D.
propionyl
2-methylbutyryl
isobutyryl
acetyl
3-methylbutyryl
pentanoyl
butyryl
N.D.
N.D.
2.1 0.1
0.6 0.1
0.6 0.2
1.2 0.1
0.8 0.4
2.4 1.5
propionyl
0.3 0.2
a
b
Racemic mixture was used in assays. N.D., not determined.
10.4
4.5 mM−1 min−1), which correlated well with the
avermectin biosynthesis in S. avermitilis (Zhuo et al., 2014).4
The noncognate acyl-SNACs could be hydrolyzed by AveAT0
as expected, although the efficiency (kcat/Km) was 5−25 fold
lower. To get further insights into the specificity of AveAT0,
two residues (Val222 and Ala226) whose involvement in
substrate binding had been revealed by structural analysis were
examined by site-directed mutagenesis. The A226L mutant was
almost completely inactive (data not shown), probably due to
the bulky side chain of leucine hindering the binding of all
tested substrates, while the V222A mutation moderately
compromised the activity. Interestingly, the V222L mutation
that significantly decreased catalytic efficiency for the native
substrate slightly increased the catalytic efficiency for non-
native pentanoyl-SNAC and butyryl-SNAC, suggesting a single
mutation potentially shifted the specificity of AveAT0.
Hydrolytic Activity. The hydrolytic activity toward
incorrect acyl-CoA thioesters has been proposed to control
the specificity of an AT domain.28 However, kinetic assays of
EryAT3 toward methylmalonyl-CoA and malonyl-CoA reveal
the specificity (as reflected by kcat/Km) for the cognate substrate
is 200-fold higher in the hydrolysis and 30-fold higher in the
transacylation, suggesting that the hydrolytic rate actually
reflects the substrate specificity of an AT.25,27 The formation of
the acyl-enzyme intermediate instead of the selective hydrolytic
cleavage of incorrectly acylated protein is likely the most
important step in the substrate discrimination of ATs from
modular PKSs. The acyl-SNAC thioesters can be incorporated
into polyketides efficiently by modular PKSs.32 We therefore
sought to evaluate the ability of AveAT0 to hydrolyze a panel of
seven synthetic acyl-SNAC thioesters by monitoring the free
SH of the released SNAC with Ellman’s reagent, which has
been used for spectrophotometric measurement of the SH
groups in a protein.33 A series of control reactions were carried
out in parallel to ensure that the hydrolysis of acyl-SNAC was
due to enzymatic activity of AveAT0. The steady-state kinetic
parameter for each substrate is listed in Table 2. The catalytic
efficiency of the hydrolysis reaction of AveAT0 for 2-
Stereospecificity of AveAT0. During the biosynthesis of
avermectins, the 2-methylbutyryl-CoA starter unit produced by
the degradation of branched-chain amino acid isoleucine has an
S configuration at C-2 position.4 The active site architecture
also suggests the stereospecificity AveAT0 for the (2S)-
methylbutyryl thioester. Since the two isomers of 2-
methylbutyric acid can be readily resolved by capillary gas
chromatography (GC) and detected by mass spectrum (MS)
(Supporting Information Figure 2), we sought to determine the
in vitro stereospecificity of AveAT0 toward the racemic (2RS)-
methylbutyryl-SNAC thioesters.34 In a reaction catalyzed by
AveAT0, 7 times more (2S)-methylbutyric acid was observed
than the R isomer, which was obviously from the spontaneous
hydrolysis of the substrate by comparing to a control reaction
without enzyme (Figure 5). The stereospecificity of AveAT0
for (2S)-methylbutyryl-SNAC was further confirmed by its
incapability to hydrolyze the optical pure (2R)-methylbutyryl-
SNAC (data not shown).
methylbutyryl-SNAC (kcat/Km = 29.6
3.8 mM−1 min−1)
was about 30 times lower than that of EryAT3 for
methylmalonyl-CoA.25 AveAT0 showed the highest rate of
hydrolysis with 2-methylbutyryl-SNAC and isobutyryl-SNAC.
The specificity for 2-methylbutyryl-SNAC, as measured by kcat/
Km, was approximately 3-fold over isobutyryl-SNAC (kcat/Km =
E
DOI: 10.1021/cb500873k
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
hydrolytic activity of the mutants. The shift of the substrate
specificity of V222L mutant to pentanoyl was caused by the
drastic decrease in catalytic efficiency toward the natural
substrate rather than the enhanced catalytic efficiency for the
unnatural substrate, as observed in the biochemical assays of
EryAT3 mutants (Dunn et al., 2013).25 In addition to pinpoint
active residues that are involved in enzyme−substrate
recognition, the crystal structures of AT domains can also be
used in more global approaches such as molecular dynamics
simulations to help the identification of residues outside the
substrate binding pockets that may indirectly influence the
architecture of active sites and to predict the promiscuity of
wild-type enzymes toward unnatural substrates.24,30
In summary, understanding the molecular basis for the
substrate specificity of loading ATs is crucial to introduce
unnatural starter units into polyketide products by rational
engineering of modular PKSs. In this work, we focused on the
loading AT of avermectin modular PKS, which was shown to
accept a variety of starter units and provide the first structural
insights into the substrate specificity of loading ATs. We have
examined the substrate preference of isolated AveAT0 toward
natural and unnatural substrates, shifted its specificity by single
mutation of the active site residues, and confirmed its
stereospecificity for an S configuration at C-2 position of the
2-methylbutyryl substrate. The structural and functional studies
of loading ATs presented here combined with computational
modeling will help the engineering of avermectin modular PKS
for the generation of new derivatives with expected starter units
and to push the ratio of desired products.
Figure 5. Stereospecificity of AveAT0. (a) AveAT0 stereospeifically
catalyzes the hydrolysis of (2S)-methylbutyryl-SNAC. (b) Chiral GC-
MS of 2-methylbutyric acid produced by incubation of racemic
methylbutyryl-SNAC with AveAT0. The (2R)-methylbutyric acid was
produced by the spontaneous hydrolysis of the substrate by comparing
to the control reaction without enzyme. See also Supporting
Information Figure 2.
Engineering of Loading ATs. The intrinsic promiscuity of
loading ATs of modular PKSs has been widely utilized to
introduce noncognate starter units into the final polyketide
products.5,35 Although the relaxed substrate specificity provides
the potential to generate novel polyketide derivatives with
altered substituents by feeding unnatural carboxylic acids to the
producers, it sometimes causes serious problems for the
industrial production of polyketide compounds due to the
biosynthesis of mixtures of multiple chemical species. For
example, the commercially important avermectins are produced
as mixtures of “a” series that use the 2-methylbutyryl starter
unit and “b” series that use the isobutyryl starter unit (Figure
1).4 A better understanding of the specificity of loading ATs
toward different carboxylic acid starter units has the commercial
potential to increase the production ratios of the desired
polyketides and, consequently, to reduce the cost of the
purification process.
Compared to the swapping of entire catalytic domains, site
directed mutagenesis of the active site residues lining the
substrate binding pockets seems to be more reasonable for
engineering the specificity of ATs because the catalytic domains
remain in their native environments, thus minimizing the
deleterious perturbations to the protein−protein interactions
introduced by the manipulations.36 According to the recruited
starter units, the loading ATs are classified into two types. A
conserved arginine in the active site indicates a KSQ-type
loading AT that recruits a dicarboxylic acid starter unit while
the absence of arginine in the corresponding position suggests
an Ave-type loading AT that accepts a monocarboxylic acid
starter unit. The shift of specificity from malonyl to acetyl by an
arginine to alanine mutation in mammalian FAS inspires the
engineering of corresponding residue in loading ATs.37
However, neither the tryptophan to arginine mutation in an
Ave-type loading AT (EryAT0) nor the arginine to tryptophan
mutation in a KSQ-type AT (oleandomycin loading AT)
switches the specificity of the enzymes toward propionate and
malonate, suggesting the involvement of other active site
resides in the discrimination of starter units.7,10 By docking a
substrate mimic into the active site of AveAT0, the residues that
may directly contact the 2-methylbutyryl moiety were
predicted. The involvement in substrate binding of Val222
and Ala226 residues was confirmed by in vitro assays of the
METHODS
■
Protein Expression and Purification. The AveAT0 domain was
amplified from S. avermitilis genomic DNA using primers 5′-
ATCGTAATCCATATGCAGAGGATGGACGGCGGGGAAGAAC-
3′ and 5′-TGATTCGATGAATTCAACCATGCCCCCGTAG-
TTGGGCGAGAGAC-3′ (NdeI-EcoRI sites italicized, the stop codon
underlined) and cloned into pET28a. The resulting plasmid was
transformed into E. coli BL21(DE3). The transformed cells were
grown to an OD600 of 0.4 in LB medium at 37 °C and induced with
0.15 mM IPTG for 12 h at 16 °C. The cells were then harvested by
centrifugation at 4000 rpm for 20 min and resuspended with lysis
buffer containing 500 mM NaCl, 50 mM Tris (pH 7.5), and 10%
glycerol (v/v). The cell suspension was lysed by sonication on ice and
centrifuged at 15 000g for 45 min to remove cell debris. The cell-free
extract was loaded onto Nickel-NTA resin that had been washed with
lysis buffer. The column was washed with lysis buffer containing 10
mM imidazole (pH 7.5). His-tagged AveAT0 was then eluted with
lysis buffer containing 300 mM imidazole. The protein was further
polished using a Superdex 200 gel filtration column (GE Healthcare)
equilibrated with a buffer containing 150 mM NaCl, 10 mM Tris
(pH7.5), and 10% (v/v) glycerol. Using protein concentrators, the
resulting protein was moved into a buffer containing 25 mM NaCl, 10
mM Tris (pH 7.5), 10% (v/v) glycerol, and 1 mM DTT at a
concentration of 12 mg mL−1.
Crystallization and Structure Determination. The AveAT0
crystals were grown in sitting drops containing 2 μL protein solution
and 2 μL precipitant solution (pH6.75, 3.4 M sodium formate) at 20
°C. Crystals were soaked in crystallization solution containing 20% (v/
v) glycerol before being frozen in liquid nitrogen. Data were collected
at Shanghai Synchrotron Radiation Facility Beamline BL17U1 and
processed with Mosflm, Pointless, and Scala programs of the CCP4
suite.38 The structure of EryAT5 [PDB code 2HG4] was used as a
search model for molecular replacement in Phaser.38 The model was
built in Coot, refined through Refmac, and visualized by PyMol.38,39
Size Estimation. A 0.1 mL sample of AveAT0 was injected onto a
Superdex 200 gel-filtration column equilibrated with 150 mM NaCl,
F
DOI: 10.1021/cb500873k
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
10 mM Tris (pH 7.5), and 10% (v/v) glycerol. The molecular weight
of AveAT0 was estimated through a comparison to standards as
previously described.40
Aliquots (20 μL) were withdrawn at specific time points and then an
equal volume dimethyl sulfoxide (DMSO) was added to quench the
reaction. After the addition of 40 μL of 2 mM Ellman’s reagent, the
sample was immediately vortexed and incubated for 10 min. All
samples were analyzed spectrophotometrically at 412 nm. Kinetic
parameters were calculated from the average of at least three sets of
triplicates with the Michaelis−Menten equation using the curve fitting
software GraphPad Prism.
Chiral GC-MS. The above-described hydrolytic reactions of
racemic 2-methylbutyryl-SNAC were extracted with equal volume of
methyl tert-butyl ether. The organic extract was concentrated in vacuo
and the residue was dissolved in 100 μL of methanol. Chiral GC-MS
analyses were carried out isothermally at 100 °C (55 cm/s) on a GC-
MS system consisting of Thermo Trace 1300 gas chromatograph and
ISQ single-quadrupole mass spectrometer, using a 30 m × 0.25 mm
inner diameter, 0.25 μm film-thickness β-DEX 120 capillary column
(Supelco) in the split mode (20:1) with helium as carrier.
Mutation of AveAT0. Generation of the mutants was accom-
plished using the GeneTailor Site-Directed Mutagenesis System
(Invitrogen) following the manufacturer’s instructions. The oligonu-
cleotides used for mutagenesis were as follows: V222L:5′-GCATG-
ATCCCGCTGGACGTTCCCGCCCACTCCCCCCTGAT-
GTACGCCA-3′and 5′-CGGGAACGTCCAGCGGGATCATG-
CGCGTGCGCACCTGCGCGGCGGTGA-3′; V222A: 5′-
GCATGATCCCGGCCGACGTTCCCGCCCACTCCCCC-
CTGATGTACGCCA-3′ and 5′-CGGGAACGTCGGCCGGGA-
TCATGCGCGTGCGCACCTGCGCGGCGGTGA-3′; and
A226L:5′-TGTGGACGTTCCCCTGCACTCCCCCCTGATGTAC-
GCCATCGAGGAACGGGT-3′ and 5′-TCAGGGGGGAGTGCAG-
GGGAACGTCCACCGGGATCATGCGCGTGCGCACCT-3′. All
mutants were verified by DNA sequencing.
Syntheses of acyl-SNACs. The synthesis of SNAC has been
described previously.41 All acyl-SNAC thioesters used for functional
assays were synthesized by similar protocols. Briefly, to the stirred
solution of a monocarboxylic acid (0.84 mmol, 1.0 equiv) in 10 mL
THF at room temperature (RT) was added DCC (174 mg, 0.84
mmol, 1.0 equiv), HOBt (114 mg, 0.84 mmol, 1.0 equiv), and SNAC
(0.1 g, 0.84 mmol, 1.1eq ). After 45 min, K2CO3 (58 mg, 0.42 mmol,
0.5 equiv) was added. The reaction was filtered and concentrated after
another 3 h. The residue was redissolved in ethyl acetate and washed
with sat. aqueous NaHCO3. The organic layer was dried over sodium
sulfate and concentrated. The crude material was chromatographed on
silica gel eluting with 1:5 petroleum ether/ethyl acetate to yield the
desired compound as a clear solid.
ASSOCIATED CONTENT
■
S
* Supporting Information
This material is available free of charge via the Internet at
http://pubs.acs.org.
Accession Codes
The final structure is deposited in the PDB with accession code
4RL1.
AUTHOR INFORMATION
■
Corresponding Authors
*E-mail: jtzheng@ipe.ac.cn.
*E-mail: lzhang03@gmail.com.
(2RS)-Methylbutyryl-SNAC. 1H NMR (CDCl3, 600 MHz): δ (ppm)
6.51 (s, 1H), 3.42 (q, J = 6.3 Hz, 2H), 2.99 (t, J = 6.5 Hz, 2H), 2.59 (q,
J = 6.9 Hz, 1H), 1.98 (s, 3H), 1.75−1.70 (m, 1H), 1.51−1.46 (m, 1H),
1.17 (d, J = 7.0 Hz, 3H), 0.92 (t, J = 7.4 Hz, 3H). ESI-MS [M + Na]+,
226.1.
Notes
The authors declare no competing financial interest.
1
2-Methylpropionyl-SNAC. H NMR (CDCl3, 600 MHz): δ (ppm)
6.09 (s, 1H), 3.43 (q, J = 6.2 Hz, 2H), 3.01 (t, J = 6.5 Hz, 2H), 2.79−
2.74 (m, 1H), 1.97(s, 3H), 1.19 (d, J = 6.9 Hz, 6H). ESI-MS [M +
Na]+, 212.1.
ACKNOWLEDGMENTS
■
The authors acknowledge the Analytical Instrumentation
Center of Institute of Process Engineering, Chinese Academy
of Sciences. The diffraction data were collected at the Shanghai
Synchrotron Radiation Facility (SSRF, China) Beamline
BL17U1. This work was supported in part by the National
Program on Key Basic Research Project (973 program,
2013CB734000), National Natural Science Foundation of
China (31370101, 31400051), Beijing Natural Science
Foundation (5144031), the Recruitment Program of Global
Experts of China, and the starting fund of the Chinese Academy
of Sciences. L.Z. is an Awardee for National Distinguished
Young Scholar Program in China.
Acetyl-SNAC. 1H NMR (CDCl3, 600 MHz): δ (ppm) 6.29 (s, 1H),
3.43 (t, J = 6.1 Hz, 2H), 3.04 (t, J = 6.1 Hz, 2H), 2.36 (d, J = 5.0 Hz,
3H), 1.99 (s, 3H). ESI-MS [M + Na]+, 184.0.
1
Propionyl-SNAC. H NMR (CDCl3, 600 MHz): δ (ppm) 6.14 (s,
1H), 3.39 (m, 2H), 2.99 (t, J = 6.6 Hz, 2H), 2.56 (m, 2H), 1.94 (s,
3H), 1.15 (t, J = 7.5 Hz, 3H). ESI-MS [M + Na]+, 198.1.
3-Methylbutyryl-SNAC. 1H NMR (CDCl3, 600 MHz): δ (ppm)
6.41 (s, 1H), 3.42 (q, J = 6.2 Hz, 2H), 3.03 (t, J = 6.6 Hz, 2H), 2.45 (d,
J = 7.1 Hz, 2H), 1.97 (s, 3H), 2.18−2.13 (m, 1H), 0.96 (t, J = 6.7 Hz,
6H). ESI-MS [M + Na]+, 226.1.
Butyryl-SNAC. 1H NMR (CDCl3, 600 MHz): δ (ppm) 6.40 (s, 1H),
3.42 (q, J = 6.3 Hz, 2H), 3.03 (t, J = 6.6 Hz, 2H), 2.56 (t, J = 7.4 Hz,
2H), 1.97 (s, 3H), 1.69 (q, J = 7.4 Hz, 2H), 0.96 (t, J = 7.4 Hz, 3H).
ESI-MS [M + Na]+, 212.1.
REFERENCES
1
■
Pentanoyl-SNAC. H NMR (CDCl3, 600 MHz): δ (ppm) 6.75 (s,
(1) Khosla, C., Herschlag, D., Cane, D. E., and Walsh, C. T. (2014)
Assembly line polyketide synthases: Mechanistic insights and unsolved
problems. Biochemistry 53, 2875−2883.
(2) Xu, W., Qiao, K., and Tang, Y. (2013) Structural analysis of
protein−protein interactions in type I polyketide synthases. Crit. Rev.
Biochem. Mol. Biol. 48, 98−122.
(3) Dutta, S., Whicher, J. R., Hansen, D. A., Hale, W. A., Chemler, J.
A., Congdon, G. R., Narayan, A. R., Hakansson, K., Sherman, D. H.,
Smith, J. L., and Skiniotis, G. (2014) Structure of a modular polyketide
synthase. Nature 510, 512−517.
(4) Zhuo, Y., Zhang, T., Wang, Q., Cruz-Morales, P., Zhang, B., Liu,
M., Barona-Gomez, F., and Zhang, L. (2014) Synthetic biology of
avermectin for production improvement and structure diversification.
Biotechnol. J. 9, 316−325.
(5) Dutton, C. J., Gibson, S. P., Goudie, A. C., Holdom, K. S., Pacey,
M. S., Ruddock, J. C., Bu’Lock, J. D., and Richards, M. K. (1991)
1H), 3.42 (q, J = 6.3 Hz, 2H), 3.02 (t, J = 6.7 Hz, 2H), 2.57 (t, J = 7.6
Hz, 2H), 1.98 (s, 3H), 1.64 (t, J = 7.6 Hz, 2H), 1.35 (t, J = 7.5 Hz,
2H), 0.91 (t, J = 7.4 Hz, 3H). ESI-MS [M + Na]+, 226.1.
1
(2R)-Methylbutyryl-SNAC. H NMR (CDCl3, 600 MHz): δ (ppm)
6.29 (s, 1H), 3.43 (q, J = 6.2 Hz, 2H), 3.02 (t, J = 6.5 Hz, 2H), 2.59 (q,
J = 6.9 Hz, 1H), 1.97 (s, 3H), 1.75−1.70 (m, 1H), 1.50−1.45 (m, 1H),
1.17 (d, J = 6.9 Hz, 3H), 0.91 (t, J = 7.5 Hz, 3H); ESI-MS [M + Na]+,
226.1.
Kinetic Analysis of Hydolytic Activity. The Ellman’s reagent
DTNB was used to monitor the free SH of the released SNAC (Riener
et al., 2002).33 A series of control reactions were conducted in parallel
to eliminate the absorbance caused by two cysteine residues in the
enzyme, the spontaneous hydrolysis of the substrates, and the
reagents. All reactions were carried out at 25 °C in the presence of
25 mM NaCl, 10 mM Tris (pH 7.5), 10% (v/v) glycerol, varying
concentration (0−4 mM) of acyl-SNAC and 5−10 μM of enzymes.
G
DOI: 10.1021/cb500873k
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
Novel avermectins produced by mutational biosynthesis. J. Antibiot.
(Tokyo) 44, 357−365.
(6) Marsden, A. F. A., Wilkinson, B., Cortes, J., Dunster, N. J.,
Staunton, J., and Leadlay, P. F. (1998) Engineering broader specificity
into an antibiotic-producing polyketide synthase. Science 279, 199−
202.
(7) Lau, J., Cane, D. E., and Khosla, C. (2000) Substrate specificity of
the loading didomain of the erythromycin polyketide synthase.
Biochemistry 39, 10514−10520.
(8) Yuzawa, S., Eng, C. H., Katz, L., and Keasling, J. D. (2013) Broad
substrate specificity of the loading didomain of the lipomycin
polyketide synthase. Biochemistry 52, 3791−3793.
(9) Bisang, C., Long, P. F., Cortes, J., Westcott, J., Crosby, J.,
Matharu, A. L., Cox, R. J., Simpson, T. J., Staunton, J., and Leadlay, P.
F. (1999) A chain initiation factor common to both modular and
aromatic polyketide synthases. Nature 401, 502−505.
(10) Long, P. F., Wilkinson, C. J., Bisang, C. P., Cortes, J., Dunster,
N., Oliynyk, M., McCormick, E., McArthur, H., Mendez, C., Salas, J.
A., Staunton, J., and Leadlay, P. F. (2002) Engineering specificity of
starter unit selection by the erythromycin-producing polyketide
synthase. Mol. Microbiol. 43, 1215−1225.
(11) Oefner, C., Schulz, H., D’Arcy, A., and Dale, G. E. (2006)
Mapping the active site of Escherichia coli malonyl-CoA-acyl carrier
protein transacylase (FabD) by protein crystallography. Acta
Crystallogr., Sect. D: Biol. Crystallogr. 62, 613−618.
(12) Tang, Y., Chen, A. Y., Kim, C. Y., Cane, D. E., and Khosla, C.
(2007) Structural and mechanistic analysis of protein interactions in
module 3 of the 6-deoxyerythronolide B synthase. Chem. Biol. 14,
931−943.
(13) Tang, Y., Kim, C. Y., Mathews, II, Cane, D. E., and Khosla, C.
(2006) The 2.7-Å crystal structure of a 194-kDa homodimeric
fragment of the 6-deoxyerythronolide B synthase. Proc. Natl. Acad. Sci.
U.S.A. 103, 11124−11129.
(14) Chen, A. Y., Cane, D. E., and Khosla, C. (2007) Structure-based
dissociation of a type I polyketide synthase module. Chem. Biol. 14,
784−792.
(15) Castonguay, R., He, W., Chen, A. Y., Khosla, C., and Cane, D. E.
(2007) Stereospecificity of ketoreductase domains of the 6-
deoxyerythronolide B synthase. J. Am. Chem. Soc. 129, 13758−13769.
(16) Hagen, A., Poust, S., de Rond, T., Yuzawa, S., Katz, L., Adams, P.
D., Petzold, C. J., and Keasling, J. D. (2014) In vitro analysis of
carboxyacyl substrate tolerance in the loading and first extension
modules of Borrelidin PKS. Biochemistry 53, 5975−5977.
(17) Bergeret, F., Gavalda, S., Chalut, C., Malaga, W., Quemard, A.,
Pedelacq, J. D., Daffe, M., Guilhot, C., Mourey, L., and Bon, C. (2012)
Biochemical and structural study of the atypical acyltransferase domain
from the mycobacterial polyketide synthase Pks13. J. Biol. Chem. 287,
33675−33690.
(18) Keatinge-Clay, A. T., Shelat, A. A., Savage, D. F., Tsai, S. C.,
Miercke, L. J., O’Connell, J. D., Khosla, C., and Stroud, R. M. (2003)
Catalysis, specificity, and ACP docking site of Streptomyces coelicolor
malonyl-CoA:ACP transacylase. Structure 11, 147−154.
(19) Liew, C. W., Nilsson, M., Chen, M. W., Sun, H., Cornvik, T.,
Liang, Z. X., and Lescar, J. (2012) Crystal structure of the
acyltransferase domain of the iterative polyketide synthase in enediyne
biosynthesis. J. Biol. Chem. 287, 23203−23215.
(20) Maier, T., Leibundgut, M., and Ban, N. (2008) The crystal
structure of a mammalian fatty acid synthase. Science 321, 1315−1322.
(21) Wong, F. T., Jin, X., Mathews, II, Cane, D. E., and Khosla, C.
(2011) Structure and mechanism of the trans-acting acyltransferase
from the disorazole synthase. Biochemistry 50, 6539−6548.
(22) Whicher, J. R., Smaga, S. S., Hansen, D. A., Brown, W. C.,
Gerwick, W. H., Sherman, D. H., and Smith, J. L. (2013)
Cyanobacterial polyketide synthase docking domains: A tool for
engineering natural product biosynthesis. Chem. Biol. 20, 1340−1351.
(23) Liang, J., Edelsbrunner, H., and Woodward, C. (1998) Anatomy
of protein pockets and cavities: measurement of binding site geometry
and implications for ligand design. Protein Sci. 7, 1884−1897.
(24) Bravo-Rodriguez, K., Ismail-Ali, A. F., Klopries, S., Kushnir, S.,
Ismail, S., Fansa, E. K., Wittinghofer, A., Schulz, F., and Sanchez-
Garcia, E. (2014) Predicted incorporation of non-native substrates by
a polyketide synthase yields bioactive natural product derivatives.
ChemBioChem 15, 1991−1997.
(25) Dunn, B. J., Cane, D. E., and Khosla, C. (2013) Mechanism and
specificity of an acyltransferase domain from a modular polyketide
synthase. Biochemistry 52, 1839−1841.
(26) Dunn, B. J., and Khosla, C. (2013) Engineering the
acyltransferase substrate specificity of assembly line polyketide
synthases. J. R. Soc. Interface 10, 20130297.
(27) Dunn, B. J., Watts, K. R., Robbins, T., Cane, D. E., and Khosla,
C. (2014) Comparative analysis of the substrate specificity of trans-
versus cis-acyltransferases of assembly line polyketide synthases.
Biochemistry 53, 3796−3806.
(28) Bonnett, S. A., Rath, C. M., Shareef, A. R., Joels, J. R., Chemler,
J. A., Hakansson, K., Reynolds, K., and Sherman, D. H. (2011) Acyl-
CoA subunit selectivity in the pikromycin polyketide synthase PikAIV:
Steady-state kinetics and active-site occupancy analysis by FTICR-MS.
Chem. Biol. 18, 1075−1081.
(29) Babtie, A., Tokuriki, N., and Hollfelder, F. (2010) What makes
an enzyme promiscuous? Curr. Opin. Chem. Biol. 14, 200−207.
(30) Sundermann, U., Bravo-Rodriguez, K., Klopries, S., Kushnir, S.,
Gomez, H., Sanchez-Garcia, E., and Schulz, F. (2012) Enzyme-
directed mutasynthesis: A combined experimental and theoretical
approach to substrate recognition of a polyketide synthase. ACS Chem.
Biol. 8, 443−450.
(31) Du, Y., Derewacz, D. K., Deguire, S. M., Teske, J., Ravel, J.,
Sulikowski, G. A., and Bachmann, B. O. (2011) Biosynthesis of the
apoptolidins in Nocardiopsis sp. FU 40. Tetrahedron 67, 6568−6575.
(32) Klopries, S., Sundermann, U., and Schulz, F. (2013)
Quantification of N-acetylcysteamine activated methylmalonate
incorporation into polyketide biosynthesis. Beilstein J. Org. Chem. 9,
664−674.
(33) Riener, C. K., Kada, G., and Gruber, H. J. (2002) Quick
measurement of protein sulfhydryls with Ellman’s reagent and with
4,4′-dithiodipyridine. Anal. Bioanal. Chem. 373, 266−276.
(34) Zhang, A., Amalin, D., Shirali, S., Serrano, M. S., Franqui, R. A.,
Oliver, J. E., Klun, J. A., Aldrich, J. R., Meyerdirk, D. E., and Lapointe,
S. L. (2004) Sex pheromone of the pink hibiscus mealybug,
Maconellicoccus hirsutus, contains an unusual cyclobutanoid mono-
terpene. Proc. Natl. Acad. Sci. U.S.A. 101, 9601−9606.
(35) Moss, S. J., Carletti, I., Olano, C., Sheridan, R. M., Ward, M.,
Math, V., Nur, E. A. M., Brana, A. F., Zhang, M. Q., Leadlay, P. F.,
Mendez, C., Salas, J. A., and Wilkinson, B. (2006) Biosynthesis of the
angiogenesis inhibitor borrelidin: Directed biosynthesis of novel
analogues. Chem. Commun. (Camb), 2341−2343.
(36) Kushnir, S., Sundermann, U., Yahiaoui, S., Brockmeyer, A.,
Janning, P., and Schulz, F. (2012) Minimally invasive mutagenesis
gives rise to a biosynthetic polyketide library. Angew. Chem., Int. Ed.
Engl. 51, 10664−10669.
(37) Rangan, V. S., and Smith, S. (1997) Alteration of the substrate
specificity of the malonyl-CoA/acetyl-CoA:acyl carrier protein S-
acyltransferase domain of the multifunctional fatty acid synthase by
mutation of a single arginine residue. J. Biol. Chem. 272, 11975−11978.
(38) Potterton, E., Briggs, P., Turkenburg, M., and Dodson, E.
(2003) A graphical user interface to the CCP4 program suite. Acta
Crystallogr., Sect. D: Biol. Crystallogr. 59, 1131−1137.
(39) Emsley, P., and Cowtan, K. (2004) Coot: Model-building tools
for molecular graphics. Acta Crystallogr., Sect. D: Biol. Crystallogr. 60,
2126−2132.
(40) Zheng, J., Taylor, C. A., Piasecki, S. K., and Keatinge-Clay, A. T.
(2010) Structural and functional analysis of A-type ketoreductases
from the amphotericin modular polyketide synthase. Structure 18,
913−922.
(41) Piasecki, S. K., Taylor, C. A., Detelich, J. F., Liu, J., Zheng, J.,
Komsoukaniants, A., Siegel, D. R., and Keatinge-Clay, A. T. (2011)
Employing modular polyketide synthase ketoreductases as biocatalysts
H
DOI: 10.1021/cb500873k
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
in the preparative chemoenzymatic syntheses of diketide chiral
building blocks. Chem. Biol. 18, 1331−1340.
I
DOI: 10.1021/cb500873k
ACS Chem. Biol. XXXX, XXX, XXX−XXX