A KAS-III Heterodimer in Lipstatin Biosynthesis Nondecarboxylatively Condenses C8 and C14 Fatty Acyl-CoA Substrates by a Variable Mechanism during the Establishment of a C22 Aliphatic Skeleton

Article abstract of DOI:10.1021/jacs.8b12843

β-Ketoacyl-acyl carrier protein synthase-III (KAS-III) and its homologues are thiolase-fold proteins that typically behave as homodimers functioning in diverse thioester-based reactions for C-C, C-O, or C-N bond formation. Here, we report an exception obs

Full text of DOI:10.1021/jacs.8b12843

Subscriber access provided by WEBSTER UNIV
Article
A KAS-III Heterodimer in Lipstatin Biosynthesis Nondecarboxylatively
8
14
Condenses C and C Fatty Acyl-CoA Substrates by a Variable
22
Mechanism during the Establishment of a C Aliphatic Skeleton
Daozhong Zhang, Fang Zhang, and Wen Liu
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b12843 • Publication Date (Web): 14 Feb 2019
Downloaded from http://pubs.acs.org on February 14, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 12
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
A KAS-III Heterodimer in Lipstatin Biosynthesis Nondecarboxylatively
Condenses C8 and C14 Fatty Acyl-CoA Substrates by a Variable
Mechanism during the Establishment of a C22 Aliphatic Skeleton
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Daozhong Zhang1, Fang Zhang2 and Wen Liu1,3,*
1 State Key Laboratory of Bioorganic and Natural Products Chemistry, Center for Excellence in Molecular Synthesis,
Shanghai Institute of Organic Chemistry, University of Chinese Academy of Sciences, 345 Lingling Road, Shanghai
200032, China
2 Innovation Research Institute of Traditional Chinese Medicine, Shanghai University of Traditional Chinese Medicine,
1200 Cailun Road, Shanghai 201203, China.
3 Huzhou Center of Bio-Synthetic Innovation, 1366 Hongfeng Road, Huzhou 313000, China.
ABSTRACT: -Ketoacyl-acyl carrier protein (ACP) synthase-III (KAS-III) and its homologs are thiolase-fold proteins that
typically behave as homodimers functioning in diverse thioester-based reactions for C-C, C-O or C-N bond formation. Here,
we report an exception observed in the biosynthesis of lipstatin. During the establishment of the C22 aliphatic skeleton of this
-lactone lipase inhibitor, LstA and LstB, which both are KAS-III homologs but phylogenetically distinct from each other,
function together by forming an unusual heterodimer to catalyze a nondecarboxylating Claisen condensation of C8 and C14
fatty acyl-CoA substrates. The resulting C22 -alkyl -ketoacid, which is unstable and tends to be spontaneously
decarboxylated to a shunt C21 hydrocarbon product, is transformed by the stereoselective -ketoreductase LstD into a
relatively stable C22 -alkyl -hydroxyacid for further transformation. LstAB activity tolerates changes in the stereochemistry,
saturation degree and thioester form of both long-chain fatty acyl-CoA substrates. This flexibility, along with the
characterization of catalytic residues, benefits our investigations into the individual roles of the two KAS-III homologs in the
heterodimer-catalyzed reactions. The large subunit LstA contains a characteristic Cys-His-Asn triad, and likely reacts with C8
acyl-CoA to form an acyl-Cys enzyme intermediate. In contrast, the small subunit LstB lacks this triad but possesses a catalytic
Glu residue, which can act on the C8 acyl-Cys enzyme intermediate in a substrate-dependent manner, either as a base for C
deprotonation or as a nucleophile for a Michael-type addition-initiated cascade reaction, to produce an enolate anion for head-
to-head assembly with C14 acyl-CoA through an unidirectional nucleophilic substitution. Uncovering LstAB catalysis draws
attention to thiolase-fold proteins that are non-canonical in both active form and catalytic reaction/mechanism. LstAB
Homologs are widespread in bacteria and remain to be functionally assigned, generating great interest in their corresponding
products and associated biological functions.
(PKSs) (Figure S1A),4-6 which condense malonyl-CoA with
INTRODUCTION
variable acyl-CoA in a similar decarboxylative manner, and
the C-O or C-N bond-forming variants (e.g., ChlB6, CerJ,
PtmR, XclC and BomK), which essentially are
acyltransferases and utilize thioester-based acyl donors for
target molecule decoration (Figure S1B).7-11 Recently,
increasing evidence has indicated that KAS-III homologs
-Ketoacyl-acyl carrier protein (ACP) synthase-III (KAS-
III, e.g., FabH in Escherichia coli) catalyzes
a
decarboxylating Claisen condensation between acetyl-
coenzyme A (CoA) and malonyl-ACP, yielding acetoacetyl-
ACP (Figure 1A).1-3 In plants and bacteria, this reaction
initiates fatty acid biosynthesis and thus plays a role
fundamentally important in cellular metabolism. KAS-III
homologs (Figure S1), which share with the FabH archetype
a highly conserved thiolase-fold and a characteristic Cys-
His-Asn triad for catalysis, are widespread in plants, fungi
and bacteria and involved in the biosynthesis of many
bioactive secondary metabolites that hold promise for drug
development. These homologs include the C-C bond-
forming variants, such as Type III polyketide synthases
(e.g.,
OleA,
PpyS,
StlD
and
DarB)
catalyze
nondecarboxylating condensations (Figure S1C), which in
some cases can initiate a cyclization process in addition to
the coupling of two acyl thioester units.12-15 Despite
functioning in diverse reactions, these proteins generally
behave as homodimers during various catalytic processes.
However, the KAS-III homologs in the biosynthesis of
lipstatin represent an exception.
Lipstatin (Figure 1B), produced by Streptomyces
toxytricini, is a naturally occurring lipase inhibitor.16,17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 12
A
B
ketoreduction and subsequent -lactonization. The
O
1
2
3
4
5
6
7
8
biogenesis of lipstatin includes the genes coding for two
KAS-III homologs, i.e., LstA (374-aa) and LstB (288-aa),
which share 23% sequence identity with each other.22 Both
proteins are necessary for the biosynthesis of lipstatin,
because inactivating the gene lstA or lstB in S. toxytricini
exclusively abolished lipstatin production.22 Notably,
neither of the resulting mutant S. toxytricini strains
produces condensed C22 intermediate(s). Starting with
questioning whether LstA and LstB function together and
participate in the unusual condensation of two long-chain
(C8 and C14) fatty acyl-CoA substrates, we here focus on the
in vitro reconstitution of the activities of these two distinct
KAS-III homologs during the formation of the main carbon
skeleton of lipstatin.
ACP
ACP
ACP
FabH
CoAS
O-
S
S
S
CO2
O
O
O-
O
O
O
OH
O
O
OH
OH
OH
1 (C8)
2 (C14
O
C22
OH
9
)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
OHCHN
O
2
O
O
3
7
5
10
Lipstatin
(S) (S)
(S)
O
RESULTS
OHCHN
O
O
O
Orlistat
LstA and LstB form a heterodimer in solution. To
characterize the roles of LstA and LstB in the biosynthesis
of lipstatin, we overexpressed their encoding genes in E. coli.
Initially, our numerous attempts failed to prepare either of
the KAS-III homologs in soluble form when expressing lstA
or lstB alone. Within the lst operon (Figure 1C), these two
genes are closely clustered for cotranscription. We thus
cloned and coexpressed them both in E. coli, with the
assumption that the resultant proteins can interact and
cause a solubilization effect. Soluble protein complexes
were observed when producing the two KAS-III homologs
simultaneously (Figure S2A). Affinity purification was
conducted on a Ni-NTA column by taking advantage of His-
tag, which was fused either at the N-terminus of LstA or at
the C-terminus of LstB. These complexes appeared to be
single products under nondenaturing conditions, whereas
under denaturing conditions, they decomposed exclusively
into two distinct subunits with a ratio of ~ 1:1 and the sizes
appropriate to N-terminally 6 x His-tagged LstA ( ~ 42.4
kDa) and untagged LstB ( ~ 29.3 kDa) or untagged LstA ( ~
40.2 kDa) and C-terminally 8 x His-tagged LstB ( ~ 30.6 kDa).
Further analysis using multi-angle light scattering (MALS)
indicated the heterodimerization state of these complexes
(Figure S2B), with a molecular weight (MW) of ~ 68.0 kDa.
Clearly, LstA and LstB noncovalently interact with each
other and form a heterodimer in solution.
C
A
B
C
D
F
lst:
E
Figure 1. KAS-III and its homologs in the biosynthesis of
lipstatin. (A) The canonical KAS-III FabH-catalyzed
decarboxylating Claisen condensation for the initiation of fatty
acid biosynthesis. (B) Retro-biosynthetic analysis of lipstatin
and its derivative orlistat. The C22 aliphatic skeleton is labeled
in bold and in color according to the C8 (red) and C14 (black)
fatty acid precursors. The same below. (C) Biosynthetic genes
of lipstatin, The genes in this study, i.e., lstA (blue) and lstB
(red) coding for two distinct KAS-III homologs and lstD
(yellow) for 3-HSD-eSDR-fold -ketoreductase, are colored.
The other genes include lstC, lstE and lstF that encode acyl-CoA
synthetase, NRPS and formyltransferase, respectively.
It can be hydrogenated to orlistat (tetrahydrolipstatin,
Figure 1B), a Food and Drug Administration-approved
antiobesity agent that now forms the largest proportion of
drugs consumed in obesity treatment.18 The supply of
orlistat depends primarily on the production of lipstatin by
S. toxytricini fermentation; however, the process through
which this precursor is biosynthesized remains poorly
understood. Lipstatin features a highly strained -lactone
ring central to two aliphatic side chains that vary in length.
In contrast to the short chain, which is saturated, the long
chain is unsaturated and further functionalized with an N-
formyl-leucine residue via an ester linkage. Isotope-labeling
experiments have previously established that the C22 -
lactone-containing olefinic skeleton of lipstatin arises from
an unusual head-to-head assembly of octanoic acid (1, C8)
and (3S,5Z,8Z)-3-hydroxytetradeca-5,8-dienoic acid (2,
C14).19,20 Both 1 and 2 can be derived from long-chain fatty
acids through incomplete -oxidation.21 During this process,
their CoA-based thioester forms (i.e., 1-CoA and 2-CoA) are
LstAB catalyzes nondecarboxylating condensations.
Using a complex composed of N-terminally 6 x His-tagged
LstA and untagged LstB, we examined the condensation
process in vitro to establish the main carbon skeleton of
lipstatin. This process has long been speculated to occur in
a decarboxylative manner (Figure S3).18 The acyl-CoA
carboxylase (ACCase) complex in S. toxytricini was
suggested to catalyze the -carboxylation of 1-CoA to
hexylmalonyl (3)-CoA (Figure 2A). This C9 product might
undergo -decarboxylation to form an enolate anion, which
facilitates the reaction with 2-CoA through nucleophilic
substitution
produced, allowing for the occurrence of
condensation to afford an -alkyl -ketoacid for -
a Claisen
ACS Paragon Plus Environment

Page 3 of 12
Journal of the American Chemical Society
A
O
preparation, and indeed, it led to an ~ 80% decrease in
lipstatin production and an ~4.5-fold increase in
1
2
3
4
5
6
7
8
SCoA
2
NADP+
COOH
accumulation.23 However, inconsistency was found when
isotope-labeled precursors were fed to S. toxytricini.
Compared with 2, 3 was incorporated into lipstatin at a
relatively low rate, arguing that it does not serve as the
direct condensing precursor.24 Following the -
3
-CoA
AntE
H+ + NADPH + CO2
CO2
O
O
SCoA
SCoA
5
NADP+
1
-CoA
-CoA
OH
O
+
+
SCoA
decarboxylation-associated
mechanism,
we
first
2
-CoA
synthesized 2-CoA and (2E)-octenyl (5)-CoA (Figures 2A
and S4). After in situ transforming the ,-unsaturated
precursor 5-CoA into the malonyl derivative3-CoA using
LstAB
LstAB
9
OH
O
O
OH
O
O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
AntE (Figure 2A),25
a highly flexible crotonyl-CoA
OH
OH
reductase/carboxylase (CCR)-like protein that catalyzes
reductive -carboxylation, we initiated reactions by adding
the LstAB complex into the mixture (Figure 2B).
8
4
CO2
CO2
OH
O
OH
O
With LstAB, 2-CoA and newly produced 3-CoA in the
reaction mixture were completely consumed over a 4-hr
incubation period at 30℃. No CoA-based products were
observed upon analysis by high-performance liquid
chromatography (HPLC). Instead, we observed two CoA-
free, hydrophobic small-molecule products, 6 and 7, which
differ from each other by 2 Da in MW. This reaction was
scaled up, leading to the accumulation of sufficient products
for structural elucidation by nuclear magnetic resonance
(NMR) and high resolution mass spectrometry (HR-MS).
Consequently, the major product, 6, was characterized to be
(10S,12Z,15Z)-10-hydroxyhenicosa-12,15-dien-8-one
(Figure S12 and Table S3), a C21 hydrocarbon assumedly
arising from the decarboxylation of the resulting C22 -
branched -ketoacid 4, therefore validating the condensing
activity of LstAB. Distinct from 6, the minor C21 hydrocarbon,
7 (Figure S13 and Table S4), is 5,6-unsaturated, and could
be derived from the corresponding C22 product, 8, via
decarboxylation. The finding of 7 production, which is
unexpected (in both structure and producing mechanism,
see Discussion), indicates that the reductive -
carboxylation of 5-CoA to 3-CoA is unnecessary for
condensation with 2-CoA. Consistently, 7 was observed in
the reaction in which the CCR protein AntE was omitted,
confirming that LstAB catalyzes the direct condensation of
5-CoA and 2-CoA. Inspired by these findings, we next
simplified the above reaction mixture and assayed LstAB
activity using 1-CoA and 2-CoA (Figure 2B). Remarkably,
the condensation of these two precursors proceeded
effectively, resulting in 6 as the sole detectable product by
HPLC. Unambiguously, the LstAB complex constructs the
main carbon skeleton of lipstatin in a nondecarboxylative
manner, rather than in a decarboxylative manner as
previously proposed. With regard to the production of 6 in
the initial reaction mixture that contains AntE, we reasoned
that either the spontaneous -decarboxylation of 3-CoA or
the AntE-catalyzed side reaction for 5-CoA saturation can
produce the direct precursor 1-CoA in situ for the following
condensation with 2-CoA (Figure 2A).25
6
7
B
5
-CoA
2
-CoA
I
3
-CoA
II
AntE
7
III
IV
LstAB
6
LstAB + AntE
1
-CoA
V
LstAB
VI
13
(min)
10
11
12
16
18
20
Figure 2. In vitro assays of the nondecarboxylating
condensation activity of LstAB. (A) Transformations involving
AntE and LstAB activities, respectively, for the preparation of
-carboxylated 3-CoA and saturated 1-CoA from ,-
unsaturated 5-CoA and the individual reactions of these C8
acyl-CoA substrates with 2-CoA (C14) to produce C22 -alkyl -
ketoacids (4 and 8) and shunt C21 hydrocarbons (6 and 7). (B)
HPLC analysis of substrate consumption and C21 hydrocarbon
production during LstAB catalysis over a 4-hr incubation
period unless otherwise stated. For CoA-based products,  at
245 nm; and for CoA-free products,  at 210 nm. Using 5-CoA
and 2-CoA, reactions were conducted in the absence of enzyme
(I) and in the presence of AntE (II, 3 min), LstAB (III) and both
AntE and LstAB (IV), respectively; and using 1-CoA and 2-CoA,
reactions were performed in the absence (V) and presence (VI)
of LstAB, respectively.
to yield a C22 -branched -ketoacid (4, (5S,7Z,10Z)-2-
hexyl-5-hydroxy-3-oxohexadeca-7,10-dienoic acid, Figure
2A). Intriguingly, previous experimental evidence appeared
to be contradictory. Consistent with this hypothesis that the
-carboxylation of 1-CoA is necessary for condensation
with 2-CoA, decomposing the ACCase complex in vivo by
gene inactivation was supposed to negatively affect 3-CoA
Tracing the unstable C22 -alkyl -ketoacid. During the
condensation of 1-CoA and 2-CoA, we observed two
putative C22 -branched -ketoacids (e.g., 4-I and 4-II) with
the same MW in the HPLC-MS traces (Figure 3A). These
condensed products, which are unstable because of
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 12
A
Consistent with this conclusion, MS/MS analysis revealed
that 4-I and 4-II were identical in fragmentation, and
I
6
4-I 4-II
1
10 min
2
3
4
5
conducting this condensation reaction in D2O led to the
dominant production of the + 1 Da deuterium-labeled
derivatives of both C22 products and the shunt C21
hydrocarbon 6 (Figure S5).
II
30 min
60 min
III
6
7
8
9
LstD produces a relatively stable C22 -alkyl -
IV
V
hydroxyacid. To support the notion that
4 is an
120 min
intermediate in the biosynthesis of lipstatin, LstD, which
shares sequence homology with members of the 3-
hydroxysteroid dehydrogenase extended-short chain
dehydrogenase/reductase (3-HSD eSDR) superfamily,22
was added into the above LstAB-involving reaction mixture
for further transformation. This transformation led to the
production of the + 2 Da reduced product 9, with the
disappearance of both 4-I and 4-II (Figure 4A). Spectral
analyses revealed the identity of 9 to be (3S,5S,7Z,10Z)-2-
hexyl-3,5-dihydroxyhexadeca-7,10-dienoic acid, a C22 -
alkyl -hydroxyacid (Figure S14 and Table S5). Compared
with 4, 9 is relatively stable in solution because reducing -
keto to -hydroxyl lowers the acidity at C and the
decarboxylation tendency. These results confirmed the -
ketoreductase activity of LstD and supported the structural
assignment for 4. LstD could function stereoselectively and
prefer (2S)-4 as the substrate to drive the equilibrium with
(2R)-4 toward the production of 9 with an S-configuration
at C3 (Figure 4B). During the purification process, 9 was
found to be partially cyclized by lactonizing its carboxylate
group with the hydroxyl group at C5, yielding 10, a shunt 6-
membered -lactone product previously observed in
mutant S. toxytricini strains that lack the ability of N-formyl-
leucine incorporation (Figure S15).22 During the
biosynthesis of lipstatin, 9 may undergo a different
intracellular lactonization with the C3 hydroxyl group to
establish the aliphatic skeleton containing a more strained
-lactone, thereby leaving the C5 hydroxyl group for N-
formyl-leucine functionalization.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
240 min
8
10
12
14
16
18
20
(min)
B
route b
OH
O
O
SCoA
CoAS
2
1
-CoA
-CoA
route a
LstAB
route a
route b
OH
O
O
OH
O
O
OH
OH
4
4'
CO2
CO2
OH
O
OH
O
6
Figure 3. Characterization of the unstable C22 -alkyl -
ketoacid in LstAB catalysis. (A) A time-course analysis of the
LstAB-catalyzed condensation of 1-CoA and 2-CoA by HPLC-
MS. ESI m/z [M+H]+ for 4-I and 4-II, 367.0; and for 6, 323.0. (B)
Two possible routes for LstAB-catalyzed condensation to
achieve the C21 hydrocarbons 6 after decarboxylation. Either 1-
CoA (route a, right) or 2-CoA (route b, left) can be utilized as
the nucleophilic substituent.
Identification of catalytic residues. We aligned LstA
and LstB sequences with selected FabH proteins and KAS-
III homologs known to catalyze Claisen condensations
(Figure S6). LstA shows overall head-to-tail homology with
these proteins and contains a conserved triad composed of
Cys128, His271 and Asn305. In contrast, LstB has no
such a triad and appears to be a much shortened KAS-III
version that lacks the two particular sequences
corresponding to those for the formation of the 1-2-3,
3 and 8 structures of the archetypal FabH.26 We mutated
the triad residues to Ala, and each engineered LstA protein
was coproduced with LstB in E. coli (Figure S7). Intriguingly,
the coproduction of LstAN305A with LstB failed to form a
soluble heterodimer, suggesting that the residue Asn305
is evolutionarily alienated from a catalytic role and instead
plays a structural role in LstA or in its complex with LstB.
Both complexes LstAC128AB and LstAH271AB were soluble and
thus subjected to activity assays in the presence of 1-CoA
and 2-CoA (Figure 4C).
the low pKa value of the  hydrogen, reached maximum
yields after 30 min and then decreased gradually with
increase in the decarboxylated shunt C21 hydrocarbon (6).
Mechanistically, the condensation of 1-CoA with 2-CoA can
go through different nucleophilic substitutions to produce
two possible C22 -branched -ketoacids, i.e., 4 and its
isomer, 4’, both of which could undergo decarboxylation to
yield the same shunt C21 hydrocarbon 6 (Figure 3B). The
route (b) in which 2 is the precursor subjected to -
deprotonation for enolate anion formation is unlikely,
because the structure of the product 4’ contrasts the C22
carbon skeleton of lipstatin (Figure 1B).19,20 The two
observed C22 products can be related to only 4, which is
generated through the route (a) utilizing
1 as the
nucleophilic substituent (Figure 3B). Most likely, 4-I and 4-
II represent the two isomeric forms of 4, i.e., (2S)-4 and
(2R)-4, which can interconvert into each other rapidly in
solution due to the active  hydrogen that readily causes a
proton exchange-coupled configuration change at C2.
ACS Paragon Plus Environment

Page 5 of 12
Journal of the American Chemical Society
and LstABE60A for activity assays. Consequently, both
A
1
-CoA
1
2
3
4
LstAE103AB and LstAE154AB retained condensation activity, in
contrast to LstABE60A, which completely lost this ability and
cannot condense 2-CoA with either 1-CoA or 5-CoA to
produce 6 or 7, respectively (Figure 4C). Therefore, Glu60
is a residue necessary for LstAB activity, and the shortened
KAS-III homolog LstB is a catalytic unit in addition to
forming a complex with LstA.
2
-CoA
I
6
9
II
LstAB
5
III
6
LstAB & LstD
7
10
11
12
13
9
10
11
min
12 18
19
20
21
8
9
B
C
The flexibility of LstAB activity in substrate. By
examining the production of related shunt C21 hydrocarbon
products, we evaluated the substrate tolerance of the LstAB
heterodimer. In terms of the short C8 substrate, LstAB can
utilize ,-unsaturated 5-CoA in the reaction with 2-CoA to
produce 7 as mentioned above. Replacing 5-CoA with its
isomer 5’-CoA, which is ,-unsaturated, also led to the
production of 7 (Figure S4 and Figure S9). Focusing on the
long C14 substrate, we tested whether the commercially
available (3S,5Z,8E)-3-hydroxytetradeca-5,8-dienoic acid
(11) and 3-hydroxytetradecanoic acid (12) along with
tetradecanoic acid (13), which was synthesized in this
study, can be used as the precursors (Figure S4). These fatty
acids were chemically transformed into their CoA
derivatives, i.e., 11-CoA, 12-CoA and 13-CoA, which then
individually reacted with 1-CoA in the presence of LstAB
(Figure S9). Using 11-CoA led to the effective production of
(3S,12Z,15E)-10-hydroxyhenicosa-12,15-dien-8-one, 14,
which is a (15E)-stereoisomer of 6, indicating that LstAB
activity is not sensitive to the stereochemistry of double
bonds. Using 12-CoA resulted in 10-hydroxyhenicosan-8-
one, 15, a condensed C21 product that shares with orlistat a
similar saturated main carbon skeleton. Therefore,
saturation of the C14 aliphatic chain have little effect. C21
hydrocarbon failed to be produced when 13-CoA was used,
confirming that the -hydroxyl group of the C14 acyl-CoA
substrate is indispensable. In addition, both 2-CoA and 1-
CoA can be replaced by their mimics of S-(N-acetyl)
cysteamine (SNAC), i.e., 2-SNAC and 1-SNAC, and
thiophenol (Sph), i.e., 2-SPh and 1-SPh, albeit with a lower
yield in 6 production, indicating the changeability in the
thioester forms of the fatty acyl substrates (Figures S4 and
S9).
OH
O
O
I
6
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
LstAB
OH
4
4
(2S)- or (2R)-
II
LstAC128A
B
NADPH + H+
LstD
III
NADP+
LstAH271A
B
OH OH
O
IV
V
OH
LstAE103A
B
B
9
-Lactonization
LstAE154A
H2O
O
O
VI
LstABE60A
21
18
19
20
OH
min
10
Figure 4. In vitro assays of the -ketoreductase activity of LstD
and the condensation activity of LstAB variants. For CoA-based
products,  at 245 nm; and for CoA-free products,  at 210 nm.
(A) HPLC analysis of substrate consumption and (shunt)
product formation during LstD catalysis. Using 1-CoA and 2-
CoA, reactions were conducted in the absence of enzyme (I)
and in the presence of LstAB (II) and both LstAB and LstD (III),
respectively. (B) LstD-catalyzed -ketoreduction of 4 to the C22
-alkyl -hydroxyacid 9 and its cyclized derivative 10. (C)
HPLC analysis of C21 hydrocarbon production during the
catalysis of each LstAB variant. I, wild-type LstAB; II,
LstAC128AB; III, LstAH271AB; IV, LstAE103AB; V, LstAE154AB; and VI,
LstABE60A
.
Neither LstAC128AB nor LstAH271AB produced the shunt C21
hydrocarbon 6, validating the necessity of both Cys128
and His271 for LstAB catalysis. Most likely, Cys128 is the
residue responsible for covalently channeling an acyl group
in the initial step of the condensation reaction. To exclude
the possibility that LstB conducts this acyl carrier function,
Phylogenetic analysis. In contrast to LstA, which is
phylogenetically related to the KAS-III homologs OleA and
PpyS that catalyze nondecarboxylating condensations and
FabH proteins that function decarboxylatively for
acetoacetyl-ACP formation, LstB appears to be relatively
independently evolved and is distantly related to the clade
of the KAS-III homologs that catalyze acyl transfer reactions
for C-O or C-N bond formation (Figure 5A). Genome survey
using lstB sequence revealed that lstB counterparts are
widespread in various bacteria (Figure 5B), e.g., the species
of Streptomyces, Nocardia, Rhodococcus, Euyarchaeota and
Mycobacterium, leading to an interesting question
concerning their intrinsic biological functions. In general,
lstB counterparts are closely clustered with lstA
counterparts, and in many cases, they both cluster with lstD
counterparts.
we mutated the residues that potentially act as
a
nucleophile in this shortened KAS-III homolog, including
C33, C98, S21, S42, S88, T58 and T245. As
anticipated, all of the corresponding heterodimers, i.e.,
LstABC33A
,
LstABC98A
,
LstABS21A
,
LstABS42A
,
LstABS88A,
LstABT58A, and LstABT245A, were soluble and exhibited
condensation activity for 6 production (Figure S8).
The KAS-III homologs catalyzing nondecarboxylating
condensations usually utilize a conserved Glu residue for -
deprotonation/C
corresponding Glu residues in the LstAB complex (i.e.,
Glu103 and Glu154 in LstA and Glu60 in LstB) were
identified and subjected to Ala-scanning, leading to the
activation.14,27,28
Accordingly,
preparation of the soluble complexes LstAE103AB, LstAE154A
B
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 12
FabH Proteins
A
B
1
2
Plant
Type-III PKSs
Streptomyces
3
4
5
6
7
I
IV
Nocardia
III
Fungal
Type-III PKSs
Rodococcus
8
9
V
Mycobacterium
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
II
Bacterial
Type-III PKSs
Euryarchaeota
Others
Figure 5. Phylogenetic analysis. (A) LstA and LstB (shown in red) with selected FabH proteins and KAS-III homologs (referred to
Figure S1) in the phylogenetic tree. The evolutionary distances were computed using the Poisson correction method. (B) Sequence
similarity networks of LstAB homologs (287) in bacteria. The chosen sequences share ˃25% identity with LstB. Members of Groups
I, II and V (partial) are the LstAB homologs that are associated with LstD homologs. Bacterial species are indicated in color.
These findings indicate that in related bacteria, LstAB
homologs might behave as heterodimers for long aliphatic
chain assembly in a similar nondecarboxylative manner and
particularly be involved in the formation of various -alkyl
-hydroxylacids with the association of LstD homologs
(Figure S16A). Notably, mycolic acids, which are major lipid
components of the cell envelope that are necessaryfor the
survival of the genus Mycobacterium (including the
causative agents of both tuberculosis and leprosy), are such
long-chain -alkyl -hydroxylacids.29-30 Whether the
reactions associated with LstAB and LstD activities
complement the pathway known for mycolic acid
biosynthesis in related Mycobacterium species as an
alternative for -alkyl -hydroxylacid supply remains to be
determined. Some species share an entire lst-like cluster
(lstABCDEF), implying that lipstatin-like -lactones play an
uncharacterized common biochemical role in addition to
being recognized as a lipase inhibitor for human being
(Figure S16B).
size, i.e., ketosynthase (KS) and chain length factor (CLF),
which form a heterodimer capable of iteratively catalyzing
decarboxylating Claisen condensation reactions.31 While
the KS component is responsible for polyketide chain
assembly, the CLF component, which regulates polyketide
chain length through an interaction with KS, does not have
this synthetic ability. The other thiolase-fold proteins
typically function as homodimers. Until this study, the only
exception was recently observed in the biosynthesis of 2-
heptyl-4(1H)-quinolone.32 During the production of this
Pseudomonas aeruginosa signal molecule, the KAS-III
homologs PqsC (348-aa) and PqsB (283-aa) mimic the FabH
homodimer by forming a heterodimer that possesses a
pseudo-2-fold symmetry and catalyzes the decarboxylating
condensation of octanoyl-CoA and 2-amino-benzoylacetate
(2-ABA, a CoA-free -keto acid).33 Although LstA and LstB
share moderate sequence homology to PqsC (19% identity)
and PqsB (18% identity), respectively (Figure 5A), the two
KAS-III homologs form a heterodimer differing from the
PqsCB complex in the mechanism of carbon anion
formation and the roles of subunits during catalytic process
(Figure 6).
DISCUSSION
Given that no CoA-based products were observed, the
large subunit LstA () was proposed to employ the triad
residue Cys128 as the nucleophile and react with the C8
acyl-CoA substrate, forming a C8 acyl-Cys128 LstAB
intermediate (although other mechanisms, e.g., the
formation of a C14 acyl-Cys128 LstAB intermediate, cannot
be excluded at this stage), and hydrolysis can follow the
condensation of this enzyme intermediate with C14 acyl-CoA
substrate to produce a C22 -alkyl -ketoacid
KAS-III and its homologs belong to the thiolase
superfamily, along with the condensing domains or
components of KAS-I and KAS-II systems for aliphatic chain
elongation in fatty acid biosynthesis and Type I and Type II
PKS systems for polyketide biosynthesis, 3-hydroxy-3-
methylglutaryl-CoA synthases involved in mevalonate-
dependent isoprenoid biosynthesis and various thiolases
catalyzing C-C bond formation or cleavage in both
biosynthetic and degradative pathways.1-2 Members of this
superfamily share a common 3-dimensional fold but do not
have to have recognizable sequence similarity. In Type II
PKS systems, condensing activity comes from a complex
composed of two different thiolase-fold proteins similar in
ACS Paragon Plus Environment

Page 7 of 12
Journal of the American Chemical Society
unsaturated and nonenolizable. The occurrence of this
Cys128
1
2
3
4
5
6
7
8
reaction had initially allowed us to consider C14 acyl
substrate being subjected to C deprotonation. The fact that
the nucleophilic substitution in LstAB catalysis is
unidirectional excludes this possibility, otherwise the
condensation would produce a distinct C22 -alkyl -
ketoacid 8’, which then undergoes decarboxylation to result
in 7’, a C21 hydrocarbon different from the truly-observed
shunt product 7 in double-bond position (Figure S10). We
proposed that during the condensation of 2-CoA and 5-CoA,
the catalytic residue Glu60 in LstB plays a distinct role by
acting as a nucleophile. The addition of this nucleophile
onto the (2E)-octenyl-Cys128 LstAB intermediate might
lead to enolate anion formation through Michael-type
addition (Figure 6B), which then can initiate an unusual
cascade reaction, i.e., nucleophilc substitution and
subsequent elimination, to produce the C22 acyl-Cys128
LstAB intermediate.
OH
O
S
SCoA
O
HOOC
Glu60
C
-Deprotonation

Cys128
Cys128
Nucleophilic substitution
S
O
S
OH
O
H
O
OOC
Glu60
9
HOOC
Glu60
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
LstA (
)

HOOC
SH
LstB ( )

Cys128
Cys128
S
OH
O
S
O
H
O
B
OOC
Michael-type addition
The functional characterization of LstAB as a unique KAS-
III heterodimer, along with LstD as a 3-HSD eSDR-fold -
ketoreductase, provides great insights into the biosynthesis
of the -lactone natural product lipstatin. In S. toxytricini,
after hydrolysis and release from the LstAB heterodimeric
complex, the highly labile C22 -alkyl -ketoacid
intermediate 4, which tends to be spontaneously
decarboxylated to the C21 hydrocarbon 6, can be
immediately reduced by LstD in a stereoselective manner,
yielding the relative stable C22 -alkyl -hydroxyacid
intermediate 9. This intermediate then undergoes -
lactonization, which could be catalyzed by the acyl-CoA
synthetase-like protein LstC, and N-formyl-leucine
substitution, which likely involves the both activities of the
formyltransferase LstF and the nonribosomal peptide
synthetase LstE, to achieve the structure underlying lipase
inhibitor activity.
OOC
Glu60
Glu60
Cys128
OH
O
Elimination
S
SCoA
O
OOC
Nucleophilic substitution
Glu60
Figure 6. Proposed mechanisms of LstAB catalysis.
Nondecarboxylating condensation can proceed in a substrate-
dependent manner, either by -deprotonation (top) or by a
Michael-type addition-initiated cascade reaction (bottom, in
the case that ,-unsaturated substrate is used) to produce an
enolate anion for an unidirectional nucleophilic substitution
during the assembly of two long-chain fatty acyl-CoA
substrates.
(Figure 6). While Asn305 likely evolves to play a structural
role, the triad residue His271 is catalytically necessary.
This residue, similar to its counterparts functioning in KAS-
III homodimer-catalyzed non-decarboxylating reactions,33
might be involved in acyl-enzyme formation or in an
acylation-dependent increase in affinity for the second acyl-
CoA substrate. Unlike PqsB, which essentially is a structural
subunit in its complex with PqsC,33 the small subunit LstB
() is a catalytic component in addition to forming a
heterodimer with LstA. Sequence alignment revealed that
Glu60 in LstB corresponds to the catalytic Glu residues of
OleA and PpyS, the KAS-III homodimers catalyzing non-
decarboxylating condensations (Figure S1), in line with the
necessity of this residue for LstAB activity. Similarly, Glu60
may act as a base at the active site to deprotonate LstA-
bound C8 acyl at C and form an enolate anion, thereby
facilitating head-to-head assembly with C14 acyl-CoA
substrate (Figure 6). The LstAB-catalyzed condensation of
2-CoA with 1-CoA or ,-unsaturated 5’-CoA may share this
mechanism, yielding the C22 -alkyl -ketoacid 4 or 8 and
the corresponding shunt C21 hydrocarbon 6 or 7.
CONCLUSION
In this study, we characterized the initial key steps for
establishing the C22 aliphatic skeleton of lipstatin in S.
toxytricini. The two distinct KAS-III homologs LstA and LstB
function together and form an unusual heterodimer to
catalyze a nondecarboxylating Claisen condensation of two
long-chain fatty acyl-CoA substrates, yielding a highly labile
C22 -alkyl -ketoacid intermediate, which tends to be
spontaneously decoarboxylated to a C21 hydrocarbon. This
intermediate is reduced immediately by the 3-HSD eSDR-
fold protein LstD in a stereoselective manner, yielding a
relatively
stable
C22
-branched
-hydroxyacid
intermediate for -lactone formation and N-formyl-leucine
substitution. LstAB activity is flexible and tolerates changes
in stereochemistry, saturation degree and thioester form of
both fatty acyl-CoA substrates. Uncovering LstAB catalysis
adds attention to KAS-III homologs in the thiolase-fold
superfamily, where to our knowledge, these proteins form
the most diverse clade in both active form and catalytic
reaction/mechanism. LstAB-like heterodimers are
widespread in bacteria and remain to be functionally
Notably, the LstAB-catalyzed reaction of 2-CoA with 5-
CoA cannot proceed by the above mechanism for enolate
anion formation, because the C8 acyl substrate is ,-
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 12
assigned. Characterizing their products and associated Product examination. For CoA-based products, related
1
2
3
4
5
6
7
8
biological functions would be extremely interesting.
reactions were quenched by addition of 1% TCA (v/v). After
centrifugation at 12,000 rpm for 5 min, the supernatant was
subjected to HPLC or HPLC-ESI-MS analysis on an Agilent
Zorbax column (SB-C18, 5 µm, 4.6 x 250 mm, Agilent
Technologies Inc., USA) by gradient elution of solvent A (5
mM NH4OAc) and solvent B (CH3CN) at a flow rate of 1 mL
min-1 over a 30-min period as follows: t = 0-5 min, 5% B; t =
5-25 min, 5%-100% B; and t = 25-30 min, 100% B ( at 245
nm).
MATARIALS AND METHODS
Sequence analysis. Target proteins were compared with
other known proteins in the databases using available
BLAST methods (http://blast.ncbi.nlm.nih.gov/Blast.cgi).
Sequence alignments were performed using Vector NT1
and ESPript 3.0 (http://espript.ibcp.fr/ESPript/ESPript/).
Sequence similarity networks were generated using the
EFI-EST webtool (http://efi.igb.illinois.edu/efi-est/).
9
For CoA-free products, each reaction (100 µL) was
quenched using an equal volume of CH3CN. After freeze-
drying, residues were dissolved in 60 µL of CH3CN and then
subjected to HPLC or HPLC-ESI-MS analysis on the same
column by gradient elution of solvent A (water containing
0.1% formic acid) and solvent B (CH3CN containing 0.1%
formic acid) at a flow rate of 1 mL min-1 over a 30-min
period as follows: t = 0-10 min, 80% B; t = 11-20 min, 90%
B; and t = 21-30 min, 100% B (( at 210 nm).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Site-specific
mutagenesis.
Rolling-cycle
PCR
amplification of (using the primers listed in Table S2)
followed by subsequent DpnI digestion was conducted to
produce the genes coding for LstA or LstB variants,
according to the standard procedure of the QuikChange
Site-Directed Mutagenesis Kit purchased from Stratagene
(USA) or Mut Express II (Vazyme Biotech Co. Ltd, China).
Each mutation was confirmed by sequencing.
Scale-up for product preparation. The C21 hydrocarbon
6 was produced in a 10 mL reaction mixture containing 100
mM Tris-HCl (pH 7.5), 5 mM NADPH, 2 mM MgCl2, 1 mM
TCEP, 30 mM NaHCO3, and 1 mM 5-CoA, 1 mM 2-CoA, 5 µM
AntE and 10 µM LstAB. The C21 hydrocarbon 7 was prepared
in an 80 mL reaction mixture containing 100 mM Tris-HCl
(pH 7.5), 2 mM MgCl2, 1 mM TCEP, 10% DMSO, 1 mM 2-CoA,
2 mM 5-CoA and 100 µM LstAB. The C22 -alkyl -
hydroxyacid 9 was produced in a 20 mL reaction mixture
containing 100 mM Tris-HCl (pH 7.5), 2 mM MgCl2, 1 mM
TCEP, 10% DMSO, 1 mM 1-SNAC, 2 mM 2-SNAC, 5 mM
NADPH, 100 µM LstAB and 10 µM LstD. Reaction mixtures
were incubated at 30℃ overnight and then quenched by
adding an equal volume of CH3CN. After centrifugation at
12,000 for 10 min, supernatants were lyophilized overnight.
Residues were dissolved in 5 mL of CH3CN, followed by
semi-preparative HPLC on an Agilent Zorbax column (SB-
C18, 5 µm, 9.4 x 250 mm, Agilent Technologies Inc., USA)
under the conditions described above. Purified product
were dissolved in the CDCl3 for NMR analysis (in CDCl3,
Assays of LstAB activity. Initial assays that follow the
decarboxylating mechanism for enolate anion formation
were carried out over a 4-hr incubation period at 30℃ each
in a 100 µL reaction mixture containing 100 mM Tris-HCl
(pH 7.5), 5 mM reduced form of nicotinamice-adenine
dinucleotide phosphate (NADPH), 2 mM MgCl2, 1 mM
trichloroethyl phosphate (TCEP), 30 mM NaHCO3, and 1 mM
5-CoA and 1 mM 2-CoA. 5 µM AntE was added into the
reaction mixture, and the in situ transformation of 5-CoA to
-carboxylated 3-CoA was nearly completed after 3 min at
30℃ (for details in determining the conditions of AntE-
catalyzed reductive -carboxylation, see SI Materials and
Methods). 10 µM LstAB was then incorporated to initiate the
condensation process. The reactions in which AntE, LstAB
or both were omitted were used as the controls.
The assays that follow the nondecarboxylating
mechanism for enolate anion formation were carried out at
30℃ each in a 100 µL reaction mixture containing 100 mM
Tris-HCl (pH 7.5), 2 mM MgCl2, 1 mM TCEP, 1 mM 1-CoA and
1 mM 2-CoA. 10 µM LstAB was added into the reaction
mixture to initiate the condensation process. The reaction
conducted in the absence of LstAB was used as the control.
For a time-course analysis of (shunt) product formation, the
reactions were performed at 30℃ for 10 min, 30 min, 60
min, 120 min and 240 min, respectively.
+
Tables S3-5). Compound 6, HR-ESI-MS m/z for C21H39O2 :
calcd. [M+H]+ = 323.2945, found 323.2946; compound 7,
+
HR-ESI-MS m/z for C21H37O2 : calcd. [M+H]+ = 321.2788,
found 321.2795; and compound 9, HR-ESI-MS m/z for
C21H38O2Na+: calcd. [M+Na]+ = 391.2819, found 391.2824.
ASSOCIATED CONTENT
For examining the proton exchange of the C22 -alkyl -
ketoacid 4, assays were carried out at 30℃ for 30 min in a
D2O-complemented reaction mixture containing 100 mM
Tris-HCl (pH 7.5), 2 mM MgCl2, 1 mM TCEP, 1 mM 1-CoA, 1
mM 2-CoA and 10 µM LstAB.
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: xxxxx.
Supplementary materials and methods (i.e.. general materials
and methods, protein expression and purification, chemical
synthesis and compound characterization), results, Figures S1-
S14, and Tables S1-S6.
Assays of LstD activity. To prepare the C22 -alkyl -
ketoacid substrate in situ, assays were carried out at 30℃
each in a 100 µL reaction mixture containing 100 mM Tris-
HCl (pH 7.5), 2 mM MgCl2, 1 mM TCEP, 1 mM 1-CoA, 1 mM
2-CoA and 10 µM LstAB. After incubation for 30 min, 5 mM
NADPH and 5 µM LstD were added into the reaction mixture
before further incubation at 30℃ for 2 hr. The reactions in
which LstD or both LstD and LstAB were omitted were used
as the controls.
AUTHOR INFORMATION
Corresponding Author
*wliu@sioc.ac.cn
ACS Paragon Plus Environment

Page 9 of 12
1
Journal of the American Chemical Society
12.
Frias, J. A.; Richman, J. E.; Erickson, J. S.; Wackett, L. P.,
Purification and characterization of OleA from Xanthomonas
campestris and demonstration of a non-decarboxylative Claisen
condensation reaction. J. Biol. Chem. 2011, 286, 10930-8.
Funding Sources
2
3
4
5
6
7
This work was supported in part by grants from NSFC
(21750004, 31430005, 21520102004, 21472231 and
21621002), CAS (QYZDJ-SSW-SLH037 and XDB20020200),
STCSM (17JC1405100), MST (2018ZX091711001-006-010)
and K. C. Wang Education Foundation.
13.
Brachmann, A. O.; Brameyer, S.; Kresovic, D.; Hitkova, I.;
Kopp, Y.; Manske, C.; Schubert, K.; Bode, H. B.; Heermann, R.,
Pyrones as bacterial signaling molecules. Nat. Chem. Biol. 2013, 9,
573-578.
14.
Mori, T.; Awakawa, T.; Shimomura, K.; Saito, Y.; Yang, D.;
Morita, H.; Abe, I., Structural insight into the enzymatic formation
of bacterial stilbene. Cell Chem. Biol. 2016, 23, 1468-1479.
8
9
Notes
The authors declare no competing financial interest.
15.
Fuchs, S. W.; Bozhuyuk, K. A.; Kresovic, D.; Grundmann,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
F.; Dill, V.; Brachmann, A. O.; Waterfield, N. R.; Bode, H. B.,
Formation of 1,3-cyclohexanediones and resorcinols catalyzed by
a widely occurring ketosynthase. Angew. Chem. Int. Ed. Engl. 2013,
52, 4108-4112.
ACKNOWLEDGMENT
The authors thank Yanping Qiu and Heng Guo for collecting
NMR and HR-MS data. The authors show our gratitude to Zhi
Lin and Zhijun Tang for reviewing manuscript and other people
in our lab for their practical support.
16.
Weibel EK, H. P., Hochuli E, Kupfer E, Lengsfeld H,
Lipstatin, an inhibitor of pancreatic lipase, produced by
Streptomyces toxytricini. I. Producing organism, fermentation,
isolation and biological activity. J. Antibiot. (Tokyo) 1987, 40, 1081-
1085.
17.
Hochuli, E.; Kupfer, E.; Maurer, R.; Meister, W.; Mercadal,
ABBREVIATIONS
CCR2, CC chemokine receptor 2; CCL2, CC chemokine ligand 2;
CCR5, CC chemokine receptor 5; TLC, thin layer
chromatography.
Y.; Schmidt, K., Lipstatin, an inhibitor of pancreatic lipase,
produced by Streptomyces toxytricini. II. Chemistry and structure
elucidation. J Antibiot (Tokyo) 1987, 40, 1086-1091.
18.
Kumar, P.; Dubey, K. K., Current trends and future
prospects of lipstatin: a lipase inhibitor and pro-drug for obesity.
RSC Advances 2015, 5, 86954-86966.
REFERENCES
19.
Eisenreich, W.; Kupfer, E.; Weber, W.; Bacher, A., Tracer
studies with crude U-13C-lipid mixtures: biosynthesis of the lipase
inhibitor lipstatin. J. Biol. Chem. 1997, 272, 867-874.
1.
biology. Nat. Prod. Rep. 2002, 19, 581-596.
2. Haapalainen, A. M.; Merilainen, G.; Wierenga, R. K., The
Heath, R. J.; Rock, C. O., The Claisen condensation in
20.
Goese, M.; Eisenreich, W.; Kupfer, E.; Stohler, P.; Weber,
W.; Leuenberger, H. G.; Bacher, A., Biosynthesis of Lipstatin.
Incorporation of multiply deuterium-labeled (5Z,8Z)-tetradeca-
5,8-dienoic acid and octanoic acid. J. Org. Chem. 2001, 66, 4673-
4678.
thiolase superfamily: condensing enzymes with diverse reaction
specificities. Trends Biochem. Sci 2006, 31, 64-71.
3.
Zhang, H.; Machutta, C. A.; Tonge, P. J., Fatty Acid
21.
Goese, M.; Eisenreich, W.; Kupfer, E.; Weber, W.; Bacher,
Biosynthesis and Oxidation. In Comprehensive Natural Products II,
Liu, H.-W.; Mander, L., Eds. Elsevier: Oxford, 2010; Vol. 8, pp 231-
275.
A., Biosynthetic origin of hydrogen atoms in the lipase inhibitor
lipstatin. J. Biol. Chem. 2000, 275, 21192-21196.
22.
Bai, T.; Zhang, D.; Lin, S.; Long, Q.; Wang, Y.; Ou, H.; Kang,
4.
Sukovich, D. J.; Seffernick, J. L.; Richman, J. E.; Hunt, K. A.;
Q.; Deng, Z.; Liu, W.; Tao, M., Operon for biosynthesis of lipstatin,
the beta-lactone inhibitor of human pancreatic lipase. Appl.
Environ. Microbiol. 2014, 80, 7473-7483.
Gralnick, J. A.; Wackett, L. P., Structure, function, and insights into
the biosynthesis of a head-to-head hydrocarbon in Shewanella
oneidensis strain MR-1. Appl. Environ. Microbiol. 2010, 76, 3842-9.
23.
Demirev, A. V.; Khanal, A.; Sedai, B. R.; Lim, S. K.; Na, M.
5.
polyketide synthases. Nat. Prod. Rep. 2014, 31, 1306-17.
6. Shimizu, Y.; Ogata, H.; Goto, S., TypeꢀIII polyketide
Hashimoto, M.; Nonaka, T.; Fujii, I., Fungal type III
K.; Nam, D. H., The role of acyl-coenzyme A carboxylase complex in
lipstatin biosynthesis of Streptomyces toxytricini. Appl. Microbiol.
Biotechnol. 2010, 87, 1129-1139.
synthases: functional classification and phylogenomics.
ChemBioChem 2017, 18, 50-65.
24.
Schuhr, C. A.; Eisenreich, W.; Goese, M.; Stohler, P.;
Weber, W.; Kupfer, E.; Bacher, A., Biosynthetic precursors of the
lipase inhibitor lipstatin. J. Org. Chem. 2002, 67, 2257-2262.
7.
He, Q.-L.; Jia, X.-Y.; Tang, M.-C.; Tian, Z.-H.; Tang, G.-L.; Liu,
W., Dissection of two acyl-transfer reactions centered on acyl-s-
carrier protein intermediates for incorporating 5-chloro-6-methyl-
o-methylsalicyclic acid into chlorothricin. ChemBioChem 2009, 10,
813-819.
25.
Yan, Y.; Chen, J.; Zhang, L.; Zheng, Q.; Han, Y.; Zhang, H.;
Zhang, D.; Awakawa, T.; Abe, I.; Liu, W., Multiplexing of
combinatorial chemistry in antimycin biosynthesis: expansion of
molecular diversity and utility. Angew. Chem. Int. Ed. 2013, 52,
12308-12312.
8.
Bretschneider, T.; Zocher, G.; Unger, M.; Scherlach, K.;
Stehle, T.; Hertweck, C., A ketosynthase homolog uses malonyl
units to form esters in cervimycin biosynthesis. Nat. Chem. Biol.
2012, 8, 154-161.
26.
White, S. W.; Zheng, J.; Zhang, Y.-M.; Rock, C. O., the
structural biology of type II fatty acid biosynthesis. Annu. Rev.
Biochem 2005, 74, 791-831.
9.
Lv, M.; Zhao, J.; Deng, Z.; Yu, Y., Characterization of the
27.
Goblirsch, B. R.; Jensen, M. R.; Mohamed, F. A.; Wackett, L.
biosynthetic gene cluster for benzoxazole antibiotics A33853
reveals unusual assembly logic. Chem. Biol. 2015, 22, 1313-1324.
P.; Wilmot, C. M., Substrate trapping in crystals of the thiolase OleA
identifies three channels that enable long chain olefin biosynthesis.
J. Biol. Chem. 2016, 291, 26698-26706.
10.
Abugrain, M. E.; Brumsted, C. J.; Osborn, A. R.; Philmus, B.;
Mahmud, T., A Highly Promiscuous ß-Ketoacyl-ACP Synthase (KAS)
III-like protein is involved in pactamycin biosynthesis. ACS Chem.
Biol. 2017, 12, 362-366.
28.
Kresovic, D.; Schempp, F.; Cheikh-Ali, Z.; Bode, H. B., A
novel and widespread class of ketosynthase is responsible for the
head-to-head condensation of two acyl moieties in bacterial
pyrone biosynthesis. Beilstein J Org. Chem. 2015, 11, 1412-1417.
11.
Proschak, A.; Zhou, Q.; Schöner, T.; Thanwisai, A.;
Kresovic, D.; Dowling, A.; ffrench-Constant, R.; Proschak, E.; Bode,
H. B., Biosynthesis of the Insecticidal Xenocyloins in Xenorhabdus
bovienii. ChemBioChem 2014, 15, 369-372.
29.
Marrakchi, H.; Lanéelle, M.-A.; Daffé, M., Mycolic acids:
structures, biosynthesis, and beyond. Chem. Biol. 2014, 21, 67-85.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 12
30.
Quémard, A., New insights into the mycolate-containing
molecules 4-hydroxy-2-alkylquinolines derive from fatty acids, not
3-ketofatty acids. Chem. Biol. 2013, 20, 1481-1491.
33. Drees, S. L.; Li, C.; Prasetya, F.; Saleem, M.; Dreveny, I.;
compound biosynthesis and transport in mycobacteria. Trends
Microbiol. 2016, 24, 725-738.
1
2
3
4
5
6
7
31.
Keatinge-Clay, A. T.; Maltby, D. A.; Medzihradszky, K. F.;
Williams, P.; Hennecke, U.; Emsley, J.; Fetzner, S., PqsBC, a
condensing enzyme in the biosynthesis of the pseudomonas
aeruginosa quinolone signal: crystal structure, inhibition, and
reaction mechanism. J. Biol. Chem. 2016, 291, 6610-6624.
Khosla, C.; Stroud, R. M., An antibiotic factory caught in action. Nat.
Struct. Mol. Biol. 2004, 11, 888-93.
32.
Groleau, M.-C.; Doucet, N.; Rahme, Laurence G.; Lépine, F.; Déziel,
E., The End of an Old Hypothesis: The Pseudomonas signaling
Dulcey, Carlos E.; Dekimpe, V.; Fauvelle, D.-A.; Milot, S.;
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 11 of 12
Journal of the American Chemical Society
1
2
3
Insert Table of Contents artwork here
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 12
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
85x40mm (300 x 300 DPI)
ACS Paragon Plus Environment