Structural Basis of a Broadly Selective Acyltransferase from the Polyketide Synthase of Splenocin

Article abstract of DOI:10.1002/anie.201802805

Polyketides are a large family of pharmaceutically important natural products, and the structural modification of their scaffolds is significant for drug development. Herein, we report high-resolution X-ray crystal structures of the broadly selective acyltransferase (AT) from the splenocin polyketide synthase (SpnD-AT) in the apo form and in complex with benzylmalonyl and pentynylmalonyl extender unit mimics. These structures revealed the molecular basis for the stereoselectivity and substrate specificity of SpnD-AT, and enabled the engineering of the industrially important Ery-AT6 to broaden its substrate scope to include three new types of extender units.

Full text of DOI:10.1002/anie.201802805

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Structural Basis of a Broadly Selective Acyltransferase from the
Polyketide Synthase of Splenocin
Authors: Yuan Li, Wan Zhang, Hui Zhang, Wenya Tian, Lian Wu,
Shuwen Wang, Mengmeng Zheng, Jinru Zhang, Chenghai
Sun, Zixin Deng, Yuhui Sun, Xudong Qu, and Jiahai Zhou
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201802805
Angew. Chem. 10.1002/ange.201802805
Link to VoR: http://dx.doi.org/10.1002/anie.201802805
http://dx.doi.org/10.1002/ange.201802805

10.1002/anie.201802805
Angewandte Chemie International Edition
COMMUNICATION
Structural Basis of a Broadly Selective Acyltransferase from the
Polyketide Synthase of Splenocin
Yuan Li+, Wan Zhang+, Hui Zhang, Wenya Tian, Lian Wu, Shuwen Wang, Mengmeng Zheng, Jinru
Zhang, Chenghai Sun, Zixin Deng, Yuhui Sun, Xudong Qu* and Jiahai Zhou*
Abstract: Polyketides are
a large family of pharmaceutically
important natural products, and the structural modification of their
scaffolds is significant for drug development. Herein, we report one
apo- and two complex structures (with a pentynylmalonyl and
benzylmalonyl, respectively) of the broadly selective acyltransferase
(AT) from the splenocin polyketide synthase. These structures
revealed the molecular basis of the stereo-selectivity and substrate
specificity of SpnD-AT and enabled engineer the industrially important
Ery-AT6 to broaden its substrate scope to include three new types of
extender units.
Polyketides are a large class of pharmaceutically important
natural products. Their natural forms or derivatives have been
widely utilized as medicines and agrochemicals.[1] To develop
these compounds into drugs, polyketides typically require
structural modification to improve their pharmacological
properties.[2] However, chemical efforts seeking to modify
polyketide carbon scaffolds have frequently been impeded by
their structural complexity and/or inert reactivity.[3] Biosynthetic
engineering, particularly alteration of the specificity of AT
enzymes (ATs), is now established as an effective alternative to
traditional chemical modification of polyketide skeletons.[4] Such
approaches use domain swapping or the direct engineering of
ATs.[4-5]
Classical ATs in an extension module selectively transfer
extender units (malonyl- or alkylmalonyl-CoA) onto polyketides.[4]
Currently there a few results have been reported to successfully
engineer canonical ATs to accept various extender units.[5] The
lack of such AT-substrate complex structures has made these
efforts challenging; previous attempts have relied on the
screening of large numbers of mutations and have resulted in
limited extension of substrate scope.[5] While it is known that
typical ATs only accept a very limited class of substrates (malonyl-
or methylmalonyl-CoA), we and others have recently identified
several new broadly selective AT domains from polyketide
Figure 1. (a) Sturcutures of splenocin and erythromycin whose AT domains
were studied in this work. The structural variety introduced by SpnD-AT and
Ery-AT6 are highlighted by red. (b) Extender unit mimics used in this study,
including the methylmalonyl-SNAC (1), propargylmalonyl-SNAC (2),
pentynylmalonyl-SNAC (3), heptenylmalonyl-SNAC (4) and benzylmalonyl-
SNAC (5).
biosynthetic pathways.[6] The substrate scope of these enzymes,
and especially the AT domain (SpnD-AT) of the splenocin (Figure
1a) polyketide synthase (PKS), are highly variable and can
therefore accept a number of diverse types of extender units,
including very long aliphatic chains (up to C7-malonyl-CoA) and
even aromatic benzylmalonyl-CoAs.[6a] Structural characterization
of broadly selective ATs in complex with substrates or substrate
analogs will reveal key insights into the molecular basis of the
substrate flexibility of these key enzymes and will almost certainly
guide the engineering of other canonical AT domains. Herein, we
report the high-resolution X-ray crystal structures of SpnD-AT in
the apo-form and in complex with benzylmalonyl and
pentynylmalonyl extender unit mimics. Our structures enabled the
elucidation of both the stereo-selectivity and broad substrate
scope of SpnD-AT. Moreover, using the information from our
structural study, we were able to mutate key residues of the
canonical Ery-AT6 (Figure 1a) from the erythromycin PKS and
thereby change its substrate scope to accept diverse bulky
extender units.
A form of SpnD-AT comprised of its N-terminal ketosynthase
(KS)-AT linker (residues 1-80) and its post AT linker (396-402)
was overexpressed in Escherichia coli BL21 (DE3) cells. The
highly purified protein (43 KDa) was then crystallized in 1.5 M
lithium sulfate monohydrate and 0.1 M Bis-Tris propane (pH 7.0)
at 18 °C. Its apo-structure was solved by molecular replacement
with the atomic coordinates of Ery-AT3 (PDB ID: 2QO3)[7] as a
searching model, and was refined to 2.35 Å. Similar to other
canonical AT domains such as Ery-AT3, Zma-AT (PDB ID:
4QBU),[8] and Ery-AT5 (PDB ID: 2HG4),[9] the core structure of
SpnD-AT contains two subdomains, a small ferredoxin (residues
207-272) and a large / hydrolase fold (residues 81-206, 273-
395) (Figure 2a). To investigate the substrate binding mode of
[*] Y. Li, W. Zhang, W. Tian, S. Wang, M. Zheng, C. Sun, Prof. Z. Deng,
Prof. Y. Sun, Prof. X. Qu
Key Laboratory of Combinatorial Biosynthesis and Drug Discovery
(Wuhan University), Ministry of Education, Wuhan University School of
Pharmaceutical Sciences, 185 Donghu Road., Wuhan 430071, China.
E-mail: quxd@whu.edu.cn
L. Wu, J. Zhang, Prof. J. Zhou
State Key Laboratory of Bioorganic and Natural Products Chemistry,
Center for Excellence in Molecular Synthesis, Shanghai Institute of
Organic Chemistry, Chinese Academy of Sciences, 345 Lingling Road,
Shanghai 200032, China.
Shaanxi Key Laboratory of Natural Products & Chemical Biology,
College of Chemistry and Pharmacy, Northwest A&F University, 3
Taicheng Road, Yangling 712100, Shaanxi, China.
E-mail: jiahai@mail.sioc.ac.cn
[+] These authors contributed equally to this work
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/anie.201802805
Angewandte Chemie International Edition
COMMUNICATION
Figure 2. (a) The overall structure of SpnD-AT. The apo- structure of SpnD-AT is shown as a cartoon, and the secondary structures are numbered sequentially.
The KS-AT linker is colored in cyan, and the α-helix, β-sheet and loop of AT domain is in brightorange, yellow, green, respectively. The loop of 268AAAH271 is
colored in magenta, and the side chain of the catalytic residue Ser173 is shown as stick. (b) The superposed of active site in two complex and the detailed
interactions between SpnD-AT and the substrate. The extender unit formed a covalent link with the OH of the catalytic residues Ser173. The atoms of the
pentynylmalonyl and benzylmalonyl with corresponding active residues are colored in cyan and orange, respectively. (c) The active pocket of SpnD-AT which
is constituted by a number of hydrophobic residues, including A268, A269, A270, A203, I202, I199, M266 and F145. Hydrophobicity and hydrophilicity are
indicated by brown and blue respectively.
SpnD-AT, we used two SNAC substrate mimics for co-
crystallization experiments: one is pentynylmalonyl-SNAC (3),
and the other is benzylmalonyl-SNAC (5) (Figure 1b). The
groups of the extender units (Figure 2b). Since SNAC substrate
mimics in the AT cavity is identical to native CoA substrates[5c] and
have shown that the specificity of acylation of an extender unit
can be predictive of the substrate scope of a given AT, we used a
panel of malonyl-SNACs (Figure 1b), including propargylmalonyl-
SNAC (2, C3), pentynylmalonyl-SNAC (3, C5), heptenylmalonyl-
SNAC (4, C7), and benzylmalonyl-SNAC (5) to evaluate AT
complex structures
of
benzylmalonyl
SpnD-AT
and
pentynylmalonyl SpnD-AT were refined to 2.50 Å and 2.40 Å
resolution, respectively (Figure 2b). In both structures, the
substrate extender units were successfully tethered to S173 of
SpnD-AT via ester bonds, with clear electron densities (Figure S1). specificity based on monitoring free thiols released from SNAC
These structures are very similar to the apo-structure (with RMSD
of 0.225 for pentynylmalonyl SpnD-AT and 0.243 for
benzylmalonyl SpnD-AT), suggesting that substrate binding does
not cause any obvious conformational changes in the enzyme.
Both complex structures show that the extender units are
stabilized by (i) a bidentate salt bridge formed between the
guanidyl group of the R198 and its carboxylate group and by (ii)
two hydrogen bonds formed between the backbone nitrogens of
G92 and L174 and their respective carbonyl groups (Figure 2b).
These multiple stabilizing interactions tightly fasten the malonyl
plane and force the -substituted groups in an S-configuration
pointing towards the hinge region which was constituted by the
helix α9 and a loop with a 268AAAH271 motif (Figure 2a). Although
the S-preference of substrate by AT domain was biochemically
confirmed in experiments using the DEBS system,[10] the
possibility of R-form extender units in the AT cavity remain under
discussion.[5c, 11] With the help of our complex structures, it is now
clear that the rigidity of the malonyl plane makes the flexibility of
the substituted group much lower than previously assumed; we
anticipated that the occurrence of R-form substituents would
cause a significant repulsion with the loop of 268AAAH271, and
found this to be the case (data not shown).
The substituted groups of extender units in the complex
structures are surrounded by hydrophobic residues, including
A268, A269, A270, A203, I202, I199, M266, and F145 (Figure 2c).
I199, I202, and A203 form the bottom of the cavity, and together
function to confine the maximum acceptable length of the
substituted groups of substrates. Residues M266 and F145, as
well as the 268AAAH271 motif, encircle the substituted groups of the
extender units and form the side walls of the cavity. Interestingly,
F145 also forms a strong π-interaction with the phenyl and alkynyl
substrate mimics after reaction with the mBBr reagent (Figure 3).
Unlike the C5 and benzylmalonyl-SNAC, the terminal unsaturated
bonds of C3 and C7 are away from the phenyl plane of F145; thus,
the π-interactions between these substrates and F145 should be
negligible. As expected, activity assay results (Figure 3a)
revealed that the wild type SpnD-AT indeed prefers the C5 and
benzyl substituents, with about two-fold higher activity than for the
C3 and C7 substituents. Moreover, mutation of F145 into either
alanine or glutamine resulted in a significant reduction in the
acylation activity for the C5 and benzyl substrates (~60% and
80% reduction for either mutant) yet only a slightly (8% by F145A)
and medium reduction (33% by F145Q) reduction to the C7.
These results confirm the significance of the π-interaction by
F145 in determining the substrate scope of SpnD-AT. Further,
F145A and F145Q also show a severe reduction (~70%) in the
acylation activity for the C3-substituted substrate, suggesting that
a large cavity does not promote this AT to take substrates with
small extender units efficiently, probably due to the loss of
hydrophobic stabilization from the binding cavity.
To gain further insight into the substrate specificity of ATs
more generally, we compared SpnD-AT with two other canonical
AT domains: Ery-AT3 (PDB ID: 2QO3)[7] and Ery-AT5 (PDB ID:
2HG4)[9] (Figure S3 and S4). Although SpnD-AT shares a low
degree of sequence identity with Ery-AT3 (31%) and Ery-AT5
(30%), their overall architectures are highly similar. By mapping
the substrate-surrounding residues onto the structures of SpnD-
AT, Ery-AT3, and Ery-AT5, we found that conserved residues for
substrate binding, including S173, R198, and H271, are all
clustered in the interior of the active site (Figure S5). Interestingly,
in Ery-AT3, we observed an unusual 56.3o rotation of the guanidyl
group of R208 (Figure S5); this can significantly change the
This article is protected by copyright. All rights reserved.

10.1002/anie.201802805
Angewandte Chemie International Edition
COMMUNICATION
bulky Y278 and S280. These two residues form direct repulsive
interaction with large substrates and hence restrict Ery-AT5 to
only accepting methylmalonyl-CoA as a substrate (Figure S4a
and S6). Significantly, these constraints are common among
canonical ATs; both Q150 and the bulky residues in the
278YASH281 (Y278 and S280) or 278HAFH281 (H278 and F280)
motifs are highly conserved in methylmalonyl-CoA specific or
malonyl-CoA specific ATs.[5d]
The constraint factors we have identified provide insight that
can inform the rational engineering of other canonical ATs to
expand their substrate scopes. Ery-AT6 has been widely studied
in the past; previous studies with site-directed mutagenesis forms
of the Ery-AT6 enzyme have shown that it is possible to alter the
ability of the enzyme to accommodate various C3 extender units
(e.g., propargylmalonyl-CoA),[5a-c] and demonstrated the use of
altered substrate scopes for the biosynthesis of macrolactones of
erythromycin.The structure of Ery-AT6 was modeled and
superimposed on the SpnD-AT complex structures (Figure S4b).
Like Ery-AT5, its residues for the stabilization of the malonyl plane
and for the pocket structure are identical (Figure S4b and S3); and
Q150, Y278, and S280 appear to form the major repulsive
interactions with bulky substrates (Figure S4b). Previously, both
of the Ery-AT6 and the two best mutant form Y278R, V276A-
Y278R have been evaluated with methylmalonyl (C1),
propargylmalonyl (C3) for their substrate scope.[5b] In order to test
their ability to take bulkier substrates, these enzymes were
prepared and assayed for acylation activity against the
aforementioned synthetic malonyl-SNACs (1-5). In vitro assays
(Figure 3b) revealed an identical result that i) wild-type Ery-AT6 is
able to effectively accept both of methylmalonyl and
propargylmalonyl extender units, actually showing a slight
preference for propargylmalonyl-SNAC (2); ii) Y278R mutant form
retains only weak activity to methylmalonyl-SNAC and other
substrates but reacts well with propargylmalonyl-SNAC; and iii)
Y278R-V276A mutant has weak but balanced activity (10%-20%
activity of Ery-AT6 on 1) to extender units. Mutation of Y278 into
arginine is able to alter the substrate binding conformation to
specifically take C3 extender units but this arginine residue
apparently prevents Ery-AT6 from accommodating larger
extender units. A previous study showed that the combination of
three mutations (Y278G, S280G, and V276A) is able to increase
the efficiency of Ery-AT6 to accommodate C3-extender units,[5c]
results which suggested that reducing the area occupied by 278
and 280 would be beneficial if the goal was to enable
accommodation of larger extender units. It was assumed that the
V276A mutation increased the flexibility of the YASH motif in a
way that facilitated accommodation of propargylmalonyl-CoA by
Ery-AT6.[5c] However, our findings in the present study indicate
that the V276A mutation may also cause a detrimental influence
on the AT activity of Ery-AT6-Y278R (by sacrificing the activity to
broaden its specificity). We therefore chose to keep the wild type
residue at position 276, and replaced two key Ery-AT6 residues
(Y278 and S280) with either the corresponding two residues from
SpnD-AT (both of which are alanines: “Y278A-S280A double
mutant”) or two more flexible glycine residues (“Y278G-S280G
double mutant”). Biochemical assays indicate that these
mutations dramatically broadened the substrate scope. Whereas
the Y278A-S280A double mutant showed a strong preference for
propargylmalonyl substrates, the Y278G-S2280G double mutant
Figure 3. Characterization of the AT specificity in vitro. Malonyl-
SNACs (1-5) were hydrolyzed by ATs; the catalytic efficiency was
measured based on the generation of SNAC-bimanes (Figure S2),
which were produced by the reaction of mBBr and released SNAC.
Each experiment was performed in triplicate. (a) Catalytic efficiency of
SpnD-AT and its related mutatants on the substrates 2-4. (b) Catalytic
efficiency of Ery-AT6 and its related mutatants on the substrates 1-5.
The reaction condition in control is same to AT reaction except for
omitting the enzyme.
orientation of the malonyl plane, and forces the substituent groups
to point toward a helix (D152-Y172). This conformation leaves
only a very limited space to accommodate the extender units.
Unlike Ery-AT3, the arginine residue in other solved extender AT
structures (PDB. 3IM8, 5DZ6, 3RGI, 3TZW, 4RR5, 4MZ0, 4QBU,
and 2HG4) are all identical to that of SpnD-AT, suggesting that
the binding mode of extender units is highly conserved in most
ATs.
The bottom of the Ery-AT5 binding pocket is formed by S204,
L207, and R208 (Figure S4a). The observed most influential
residues in SpnD-AT are A268 and A270 in the 268AAAH271 motif
and F145, where corresponding residues are changed as Y278,
S280 and Q150 in the Ery-AT5 (Figure S4a). Unlike the phenyl
group of F145, which is parallel to the substituent group of
extender units, the side chain of Q150 protrudes deeply into the
pocket. It pushes the substrate outward to force it closer to the
This article is protected by copyright. All rights reserved.

10.1002/anie.201802805
Angewandte Chemie International Edition
COMMUNICATION
form of Ery-AT6 can effectively accommodate C3, C5, C7, and
benzyl substrates, but not C1 substrate (Figure 3b). Introducing
additional V276A into the Y278G-S280G decrease the activity of
C3, C5, C7 and benzyl substrates (Figure 3b), suggesting Y278G-
S280G is better for engineering erythromycin structures.
Acknowledgements
We thank the staff of beamline BL19U1 of the National Center for
Protein Science Shanghai (China) for access and help with the X-
ray data collection. We also thank Profs. Yi Tang, Jianting Zheng
and Pinghua Liu for criticallly reading the manuscript. This work
was supported by the Strategic Priority Research Program (B) of
the CAS (XDB20000000 to J. Z.), NSFC grant (31570057,
31770063 to X. Q.), and the Science and Technology
Commission of Shanghai Municipality (15JC1400403 to J. Z.).
Our structural analysis indicated that Q150 also constrains the
substrate scope of Ery-AT6. In the pocket, the side chain of Q150
points toward the middle, which pushes the substrate into the
vicinity of Y278 and S280 (Figure S4b). To verify the functional
role of Q150, two triple mutants were generated and assayed
(Q150A-Y278G-S280G and Q150A-Y278A-S280A). Compared
to their double mutant forms, both of these triple mutants had
significantly reduced abilities to accommodate extender unit
substrates, and these reductions were more pronounced for
larger substrates such as heptenylmaolonyl (4) and
benzylmalonyl SNACs (5). We suspect that mutation of Q150 into
alanine could have a powerful impact to increase the volume of
pocket; however this mutation might also sacrifice the
hydrophobic interaction to the substrate, which we earlier
confirmed to be very important for both the activity and substrate
specificity of SpnD-AT (see above). Thus, to further broaden the
substrate scope without decreasing the AT activity, an ideal
mutation at this position would cause reduced steric-hindrance
but would promote increased hydrophobic interaction; we are
currently exploring this area. Kinetic data revealed that Y278G-
S280G is indeed as efficient as SpnD-AT on the taking of
heptenylmaolonyl-SNAC (4) (kcat/Km are 0.47 and 0.50 × 10-3 ×
10-3 min-1μM-1 respectively) and benzylmalonyl-SNAC (5) (kcat/Km
are 1.48 and 1.42 × 10-3 min-1μM-1 respectively). More so, Y278A-
S278A shows a comparative activity on the propargylmalonyl-
SNAC (2) (kcat/Km = 0.54 × 10-3 min-1μM-1) (Figure S7). These
results show that the artificially engineered canonic ATs are highly
competent for accepting bulky substrates.
Conflict of interest
The authors declare no conflict of interest.
Keywords: Polyketide • acyltransferase • engineering •
biosynthesis • substrate scope
[1] D. J. Newman, G. M. Cragg, J. Nat. Prod. 2016, 79, 629-661.
[2] A. Bauer, M. Bronstrup, Nat. Prod. Rep. 2014, 31, 35-60.
[3] M.-C. Wu, B. Law, B. Wilkinson, J. Micklefield, Curr. Opin. Biotech. 2012, 23,
931-940.
[4] B. J. Dunn, C. Khosla, J. R. Soc. Interface 2013, 10, 20130297.
[5] a) U. Sundermann, K. Bravo-Rodriguez, S. Klopries, S. Kushnir, H. Gomez,
E. Sanchez-Garcia, F. Schulz, ACS Chem. Biol. 2013, 8, 443-450; b) I.
Koryakina, C. Kasey, J. B. McArthur, A. N. Lowell, J. A. Chemler, S. Li, D. A.
Hansen, D. H. Sherman, G. J. Williams, ACS Chem. Biol. 2017, 12, 114-123;
c) K. Bravo-Rodriguez, S. Klopries, Kyra R. M. Koopmans, U. Sundermann,
S. Yahiaoui, J. Arens, S. Kushnir, F. Schulz, E. Sanchez-Garcia, Chem. Biol.
2015, 22, 1425-1430; d) C. D. Reeves, S. Murli, G. W. Ashley, M. Piagentini,
C. R. Hutchinson, R. McDaniel, Biochemistry 2001, 40, 15464-15470E; e) M.
Musiol-Kroll, F. Zubeil, T. Schafhauser, T. Hartner, A. Kulik, J. McArthur, I.
Koryakina, W. Wohlleben, S. Grond, G. J. Williams, S. Y. Lee, T. Weber, ACS
Synth. Biol. 2017, 6, 421−427.
Understanding the physical interactions between AT enzymes
and their substrates is crucial for AT engineering, but this
structural knowledge has been previously restricted to a malonyl-
tethered transacylase (FabD) of E. coli type II fatty acid synthase
(FAS) family.[12] This structure not only has low identity with PKS
AT (e.g., 28% protein identity with SpnD-AT), but also lacks
information for substituted groups. Recently, an acyl-complex
structure of a transacylase (SAT) domain from a fungal iterative
PKS was also solved,[13] however this enzyme only select an acyl
substrate and shows a very different substrate binding mode from
those in classical PKS extension modules; hence the utility of
these complex structures for informing AT engineering has thus
been limited. Our apo-structure and two complex structures of the
highly promiscuous AT domain of SpnD-AT guided our
engineering of Ery-AT6, and our results suggest that future efforts
to alter AT activity and substrate scope should be focused on the
engineering of these three residues (Q150, Y278, S280 in
methylmalonyl-CoA specific AT or Q150, H278, F280 in malonyl-
CoA specific AT) to enable accommodation of various extender
unit substrates. To our knowledge, this is the first successful
example of the characterization of PKS extender AT-substrate
complex structure and demonstration of an engineered AT
domain to take substrates with highly bulky extender unit. These
new structures open the door for introducing diverse structural
modifications into many polyketide carbon scaffolds.
[6] a) C. Chang, R. Huang, Y. Yan, H. Ma, Z. Dai, B. Zhang, Z. Deng, W. Liu, X.
Qu, J. Am. Chem. Soc. 2015, 137, 4183-4190; b) Y. Yan, L. Zhang, T. Ito, X.
Qu, Y. Asakawa, T. Awakawa, I. Abe, W. Liu, Org. Lett. 2012, 14, 4142-4145;
c) J. Liu, X. Zhu, R. F. Seipke, W. Zhang, ACS Synth. Biol. 2015, 4, 559-565;
d) S. Takahashi, A. Toyoda, Y. Sekiyama, H. Takagi, T. Nogawa, M. Uramoto,
R. Suzuki, H. Koshino, T. Kumano, S. Panthee, T. Dairi, J. Ishikawa, H. Ikeda,
Y. Sakaki, H. Osada, Nat. Chem. Biol. 2011, 7, 461-468; e) L. Laureti, L. Song,
S. Huang, C. Corre, P. Leblond, G. L. Challis, B. Aigle, Proc. Natl. Acad. Sci.
U.S.A. 2011, 108, 6258-6263.
[7] Y. Tang, A. Y. Chen, C.-Y. Kim, D. E. Cane, C. Khosla, Chem. Biol. 2007,
14, 931-943.
[8] H. Park, B. M. Kevany, D. H. Dyer, M. G. Thomas, K. T. Forest, PLOS ONE
2014, 9, e110965.
[9] Y. Tang, C.-Y. Kim, I. I. Mathews, D. E. Cane, C. Khosla, Proc. Natl. Acad.
Sci. U.S.A. 2006, 103, 11124-11129.
[10] A. Marsden, P. Caffrey, J. Aparicio, M. Loughran, J. Staunton, P. Leadlay,
Science 1994, 263, 378-380.
[11] F. Bergeret, S. Gavalda, C. Chalut, W. Malaga, A. Quémard, J.-D. Pedelacq,
M. Daffé, C. Guilhot, L. Mourey, C. Bon, J. Biol. Chem. 2012, 287, 33675-
33690.
[12] C. Oefner, H. Schulz, A. D'Arcy, G. E. Dale, Acta Crystallogr. D 2006, 62,
613-618.
[13] J. M. Winter, D. Cascio, D. Dietrich, M. Sato, K. Watanabe, M. R. Sawaya,
J. C. Vederas, Y. Tang, J. Am. Chem. Soc. 2015, 137, 9885−9893.
This article is protected by copyright. All rights reserved.

10.1002/anie.201802805
Angewandte Chemie International Edition
COMMUNICATION
Entry for the Table of Contents (Please choose one layout)
Layout 1:
COMMUNICATION
An apo- and two complex structures
of a broadly selective AT domain from
the splenocin polyketide synthase
have been solved. These structures
enable to identify three residues
critical for building the promiscuous
substrate specificity. Engineering the
corresponding sites in the canonical
Ery-AT6 allowed to significantly
Yuan Li+, Wan Zhang+, Hui Zhang,
Wenya Tian, Lian Wu, Shuwen Wang,
Mengmeng Zheng, Jinru Zhang,
Chenghai Sun, Zixin Deng, Yuhui Sun,
Xudong Qu* and Jiahai Zhou*
Page No. – Page No.
Structural Basis of a Broadly
Selective Acyltransferase from the
Polyketide Synthase of Splenocin
expand or alter the substrate scope.
This article is protected by copyright. All rights reserved.