Stereospecific Formation of E- and Z-Disubstituted Double Bonds by Dehydratase Domains from Modules 1 and 2 of the Fostriecin Polyketide Synthase

Article abstract of DOI:10.1021/jacs.7b08896

The dehydratase domain FosDH1 from module 1 of the fostriecin polyketide synthase (PKS) catalyzed the stereospecific interconversion of (3R)-3-hydroxybutyryl-FosACP1 (5) and (E)-2-butenoyl-FosACP1 (11), as established by a combination of direct LC-MS/MS and chiral GC-MS. FosDH1 did not act on either (3S)-3-hydroxybutyryl-FosACP1 (6) or (Z)-2-butenoyl-FosACP1 (12). FosKR2, the ketoreductase from module 2 of the fostriecin PKS that normally provides the natural substrate for FosDH2, was shown to catalyze the NADPH-dependent stereospecific reduction of 3-ketobutyryl-FosACP2 (23) to (3S)-3-hydroxybutyryl-FosACP2 (8). Consistent with this finding, FosDH2 catalyzed the interconversion of the corresponding triketide substrates (3R,4E)-3-hydroxy-4-hexenoyl-FosACP2 (18) and (2Z,4E)-2,4-hexadienoyl-FosACP2 (21). FosDH2 also catalyzed the stereospecific hydration of (Z)-2-butenoyl-FosACP2 (14) to (3S)-3-hydroxybutyryl-FosACP2 (8). Although incubation of FosDH2 with (3S)-3-hydroxybutyryl-FosACP2 (8) did not result in detectable accumulation of (Z)-2-butenoyl-FosACP2 (14), FosDH2 catalyzed the slow exchange of the 3-hydroxy group of 8 with [¹⁸O]-water. FosDH2 unexpectedly could also support the stereospecific interconversion of (3R)-3-hydroxybutyryl-FosACP2 (7) and (E)-2-butenoyl-FosACP2 (13).

Full text of DOI:10.1021/jacs.7b08896

Subscriber access provided by Oakland University Libraries
Article
Stereospecific Formation of E- and Z- Disubstituted
Double Bonds by Dehydratase Domains from Modules
1 and 2 of the Fostriecin Polyketide Synthase
Young-Ok You, Dhara D. Shah, and David E. Cane
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b08896 • Publication Date (Web): 13 Sep 2017
Downloaded from http://pubs.acs.org on September 17, 2017
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 free 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 accessible to all
readers and 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.
Journal of the American Chemical Society 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
Stereospecific Formation of E- and Z- Disubstituted
Double Bonds by Dehydratase Domains from Mod-
ules 1 and 2 of the Fostriecin Polyketide Synthase
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
,†
Dhara D. Shah,‡ Young-Ok You,‡ and David E. Cane*
Department of Chemistry, Brown University, Box H, Providence, Rhode Island 02912-9108, USA.
ABSTRACT: The dehydratase domain FosDH1 from module 1 of the fostriecin polyketide synthase (PKS) catalyzed
the stereospecific interconversion of (3R)-3-hydroxybutyryl-FosACP1 (5) and (E)-2-butenoyl-FosACP1 (11), as estab-
lished by a combination of direct LC-MS/MS and chiral GC-MS. FosDH1 did not act on either (3S)-3-hydroxybutyryl-
FosACP1 (6) or (Z)-2-butenoyl-FosACP1 (12). FosKR2, the ketoreductase from module 2 of the fostriecin PKS that nor-
mally provides the natural substrate for FosDH2, was shown to catalyze the NADPH-dependent stereospecific reduc-
tion of 3-ketobutyryl-FosACP2 (23) to (3S)-3-hydroxybutyryl-FosACP2 (8). Consistent with this finding, FosDH2 cata-
lyzed the interconversion of the corresponding triketide substrates (3R,4E)-3-hydroxy-4-hexenoyl-FosACP2 (18) and
(2Z,4E)-2,4-hexadienoyl-FosACP2 (21). FosDH2 also catalyzed the stereospecific hydration of (Z)-2-butenoyl-FosACP2
(14) to (3S)-3-hydroxybutyryl-FosACP2 (8). Although incubation of FosDH2 with (3S)-3-hydroxybutyryl-FosACP2 (8)
did not result in detectable accumulation of (Z)-2-butenoyl-FosACP2 (14), FosDH2 catalyzed the slow exchange of the
3-hydroxy group of 8 with [18O]-water. FosDH2 unexpectedly could also support the stereospecific interconversion of
(3R)-3-hydroxybutyryl-FosACP2 (7) and (E)-2-butenoyl-FosACP2 (13).
INTRODUCTION
bonds of 1 are predicted to be generated by the dehydra-
tase domains of fostriecin modules 1 and 6, FosDH1 and
FosDH6 respectively,4c while the ꢀ14,15 and ꢀ12,13 Z (cis)
Fostriecin (1), a polyketide phosphate monoester iso-
lated from Streptomyces pulveraceus,1 belongs to a class of
selective protein phosphatase inhibitors showing anti-
metastatic and antitumor activity that includes the close-
ly related natural products cytostatin (2) and phoslac-
tomycin B (3) (Figure 1).2 Fostriecin also exhibits antimy-
cotic activity.3 In the biosynthesis of fostriecin (1), the
parent nonaketide 4 is assembled by a polyketide syn-
thase (PKS) consisting of 8 modules, with a loading tri-
domain at the N-terminus of module 1 responsible for
priming by the acetyl starter unit, while a thioesterase
(TE) domain at the C-terminus of module 8 mediates
polyketide chain release and lactonization (Figure 2).4a A
series of tailoring reactions, including elimination of ma-
lonate to introduce the 2,3-double bond of the dihydro-
pyrone moiety, then converts 4 to the mature product
fostriecin (1).4
OH
H2O3PO
OH
OH
Figure 2. Biosynthesis of fostriecin (1) by a modular polyke-
tide synthase. PKS domains: ketosynthase (KS), acyl trans-
ferase (AT), acyl carrier protein (ACP), ketoreductase (KR),
dehydratase (DH), enoyl reductase (ER), thioesterase (TE).
O
O
Fostriecin (1)
H2O3PO
OH
H2O3PO
OH
O
O
double bonds are attributed to the action of the corre-
sponding FosDH2 and FosDH3 domains from fostriecin
modules 2 and 3.4a Although the majority of the thou-
sands of known polyketides harbor E double bonds,5
isomeric Z double bonds, while considerably less com-
mon, are nonetheless well represented, being found in
not only 1-3, but also in the microtubule stabilizer
O
O
OH
Phoslactomycin B (3)
H2N
Cytostatin (2)
Figure 1. Fostriecin (1) and related polyketides.
The acyclic chain of fostriecin harbors four disubsti-
tuted double bonds. The ꢀ16,17 and ꢀ6,7 E (trans) double
ACS Paragon Plus Environment

Journal of the American Chemical Society
epothilone,6 the microtubule polymerization inhibitor
Page 2 of 12
curacin,7 the anti-angiogenic agent borrelidin,8 and the
antitubercular rifamycins.9 Surprisingly, all reported
attempts to demonstrate in vitro formation of the rele-
vant Z-double bonds by PKS-derived DH domains have
been unsuccessful, resulting only in the generation of
products containing the isomeric E double bonds.8b,c,9d,
While PlmKR1, the ketoreductase from module 1 of the
phoslactomycin PKS, has been shown to generate a (3S)-
hydroxyacyl thioester that likely serves as the precursor
of the Z-3-cyclohexylpropenoate produced by phoslac-
tomycin module 1, direct biochemical evidence for the
function of the paired PlmDH1 domain is still lacking.10
We now report the elucidation of the biochemical func-
tion of both FosDH1 and FosDH2, thereby confirming
the predicted role of each dehydratase domain in the
stereospecific formation of their respective E- and Z-
disubstituted enoyl-ACP products.
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
RESULTS
Recombinant Fostriecin PKS Domains. FosDH1 and
FosDH2 as well as FosKR2 from module 2 of the fos-
triecin PKS were each expressed in Escherichia coli as N-
terminal His6-tagged proteins using codon-optimized
synthetic genes, based on well-precedented PKS domain
boundaries (Figures S1-S4). The purity and molecular
mass of each recombinant protein was verified by SDS-
PAGE and LC-ESI(+)-MS (Figures S6- S8). Although di-
rect expression in E. coli of synthetic genes for FosACP1
and FosACP2 gave predominantly insoluble protein in-
clusion bodies, the corresponding NusA-FosACP1 and
NusA-FosACP2 fusion proteins were obtained in soluble
form (Figures S5, S9, and S10).9d Cleavage of the N-
terminal NusA with HRV 3C protease gave apo-
FosACP1 and apo-FosACP2, which were each confirmed
to have the expected mass by LC-ESI(+)-MS (Figures S9
and S10).
Chemoenzymatic synthesis of C4 and C6 acyl-
FosACP substrates. (3R)- and (3S)-3-Hydroxybutyryl-
FosACP1 (5 and 6) as well as (3R)- and (3S)-3-
hydroxybutyryl-FosACP2 (7 and 8) were chemo-
enzymatically prepared from (3R)- and (3S)-
hydroxybutyric acid (9 and 10), respectively, via the cor-
responding -SCoA esters, using apo-FosACP1 or apo-
FosACP2 and the surfactin phosphopantetheinyl trans-
ferase Sfp (Scheme 1). Similar procedures were also used
to prepare the individual unsaturated (E)- and (Z)-2-
butenoyl-FosACP1 (11 and 12) as well as (E)- and (Z)-2-
butenoyl-FosACP2 (13 and 14) from (E)- and (Z)-2-
butenoic acid (15 and 16). For the synthesis of ACP-
Scheme 1. Synthesis of C4 acyl-FosACP substrates.
bound
triketides,
synthetic
(3R,4E)-3-hydroxy-4-
hexenoic acid (17, 3R/3S 90:10) was chemo-enzymatically
converted via its -SCoA thioester to the corresponding
(3R,4E)-3-hydroxy-4-hexenoyl-FosACP2 (18, 80% d.e.)
(Scheme 2).11,12 In like manner, (2Z,4E)-2,4-hexadienoic
acid (19), prepared as previously described by ring-
closing metathesis–base-induced ring opening,13 and
(2E,4E)-2,4-hexadienoic acid (20) were each converted
via their -SCoA esters to the corresponding (2Z,4E)-2,4-
hexadienoyl-FosACP2 (21) and (2E,4E)-2,4-hexadienoyl-
FosACP2 (22).
Scheme 2. Synthesis of C6 acyl-FosACP2 derivatives.
ACS Paragon Plus Environment

Page 3 of 12
Journal of the American Chemical Society
Fostriecin Synthase Module 1. Substrate and Product
predicted to produce (2Z,4E)-2,4-hexadienoyl-FosACP2
(21), integrated into the parent module, as inferred from
the Z (cis) geometry of the derived ꢀ14,15 double bond of
fostriecin (1) (Figure 1). Within fostriecin synthase mod-
ule 2, FosKR2 is responsible for generating the bound 3-
hydroxyacyl-FosACP2 triketide that serves as the native
substrate of the paired FosDH2 domain in intact fos-
triecin module 2. The absence of the characteristic Leu-
Asp-Asp triad typical of 3R-hydroxy (D)-specific ketore-
ductases suggested that FosKR2 should generate the (L)-
hydroxy product typical of an A-type KR domain.14 To
establish the stereochemistry of this reduction, 3-
ketobutyryl-FosACP2 (23), chemoenzymatically pre-
pared by treatment of acetoacetyl-SCoA with Sfp and
apo-FosACP2, was reduced with FosKR2 in the presence
of NADPH (Scheme 4 and Figure S15). The exclusive
product was (3S)-3-hydroxybutyryl-FosACP2 (8), as es-
tablished by chiral GC-MS analysis of the corresponding
bis(TMS)-derivative.
1
2
3
4
5
6
7
8
Specificity of FosDH1. Module 1 of the fostriecin PKS is
predicted to produce (E)-2-butenoyl-FosACP1 (11), inte-
grated into the parent module, as inferred from the E
(trans) geometry of the derived ꢀ16,17 double bond of fos-
triecin (1) (Figure 1). The substrate for FosDH1 is ex-
pected to be (3R)-3-hydroxybutyryl-ACP1, as suggested
by the presence of the characteristic highly conserved
Leu-Asp-Asp motif in the paired ketoreductase domain
FosKR1 (Figure S1).14 Fully consistent with these predic-
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
tions,
incubation
of
FosDH1
with
(3R)-3-
hydroxybutyryl-FosACP1 (5) resulted in dehydration to
give exclusively (E)-2-butenoyl-FosACP1 (11) (Scheme
3). Thus treatment of the incubation mixture with PICS
TE, the thioesterase from the picromycin PKS, so as to
hydrolyze the ACP-bound diketide substrate and prod-
ucts,
followed
by
treatment
with
N,O-
bis(trimethylsilyl)trifluoroacetamide
(N,O-bis(TMS)-
TFA) gave the corresponding (E)-2-butenoyl-TMS de-
rivative whose geometry was confirmed by GC-MS
analysis and direct comparison with authentic standards
derived from both (E)-15 and (Z)-16. No dehydration
product from 5 was detected in the absence of added
FosDH1 (Figure S11).
Scheme 4. Stereochemistry of FosKR2-catalyzed reduc-
tion.
In a complementary set of incubations, FosDH1 cata-
lyzed the reverse hydration reaction by stereospecifically
converting (E)-2-butenoyl-FosACP1 (11) to (3R)-3-
hydroxybutyryl-FosACP1 (5) (Scheme 3). LC-ESI(+)-MS
analysis of the reaction mixture showed the expected
addition of water (M+H2O) for the parent ion while LC-
ESI(+)-MS/MS confirmed the predicted increase of
M+18 in the mass of the derived pantetheinyl ejection
fragments (329 to 347 Da) (Figures S12 and S13).15 Chiral
GC-MS analysis after hydrolysis of the incubation prod-
ucts with PICS TE and derivatization by (N,O-bis(TMS)-
TFA established the exclusive formation of bis(TMS)-
(3R)-3-hydroxybutyrate (bis(TMS)-(3R)-9) (rt 9.61 min,
identical with an authentic reference sample) (Figure
S14). Hydration of 11 was not detected in the absence of
FosDH2. Incubation of FosDH2 with the chemoen-
zymatically prepared form of the natural substrate ste-
reoisomer, (3R,4E)-3-hydroxy-4-hexenoyl-FosACP2 (18)
(~80% e.e.), resulted in stereospecific dehydration to
give (2Z,4E)-2,4-hexadienoyl-FosACP2 (21), as deter-
mined by GC-MS analysis of the derived methyl ester
(2Z,4E)-21-Me following PICS TE-catalyzed hydrolysis
and treatment with TMS-CHN2, (Scheme 5, Figures S16
and S17).16 Complementary incubation of (2Z,4E)-2,4-
hexadienoyl-FosACP2 (21) with FosDH2 resulted in the
reverse hydration reaction to give (3R,4E)-3-hydroxy-4-
hexenoyl-FosACP2 (18) as the exclusive product of DH-
catalyzed hydration (Scheme 5). This result was estab-
lished by chiral GC-MS analysis of the derived methyl
ester, including direct comparison with reference stand-
ards of synthetic methyl (3R,4E)-17-Me and the enanti-
omeric methyl (3S,4E)-3-hydroxy-4-hexenoate (Figures
S18 and S19). Hydration of 21 did not occur in the ab-
sence of added FosDH2, although (2Z,4E)-21 did under-
go 10-15% buffer-catalyzed isomerization to (2E,4E)-22
after 2 h incubation under the same conditions.
FosDH1.
Neither
the
stereoisomeric
(3S)-3-
hydroxybutyryl-FosACP1 (6) nor (Z)-2-butenoyl-
FosACP1 (12) underwent any detectable reaction in the
presence of FosDH1.
Scheme 3. Stereospecific dehydration and hydration by
FosDH1.
Scheme 5. FosDH2-catalyzed hydration and dehydra-
tion of FosACP2-bound triketide substrates.
We also carried out the analogous series of reactions
with FosDH2 and the corresponding FosACP2-bound
Fostriecin Synthase Module 2. Substrate and Product
Specificity of FosKR2. Module 2 of the fostriecin PKS is
ACS Paragon Plus Environment

Journal of the American Chemical Society
C4 substrate analogs. Thus (Z)-2-butenoyl-FosACP2 (14)
Page 4 of 12
1
underwent time-dependent, stereospecific hydration to
give (3S)-3-hydroxybutyryl-FosACP2 (8) when incubat-
ed with FosDH2 (Scheme 6, Figures 3 and S20-S22). On
the other hand, when FosDH2 was incubated with 8
(Figures S23 and S24) the expected dehydration product,
(Z)-14, could not be directly detected, presumably due to
the thermodynamically unfavorable Keq ~10-3 for dehy-
dration to the (Z)-enoyl-ACP product.17 Evidence for the
transient formation of 14 was obtained, however, by the
observation that incubation of 8 with FosDH2 in [18O]-
H2O for 90 min resulted in ~10% net exchange of the 3-
hydroxyl oxygen of 8, as revealed by the enzyme- and
time-dependent increase in the relative intensity of the
[M+2] peak of the derived pantetheinate ejection frag-
ment (349.23 Da) observed by LC-MS/MS analysis of
recovered 8 (Figure S26). Isotope exchange was not de-
tectable in the absence of FosDH2. FosDH2 was found to
be unexpectedly permissive in also being able to inter-
convert the unnatural pair of stereoisomeric C4 analogs,
(3R)-3-hydroxybutyryl-FosACP2 (7) and (E)-2-butenoyl-
FosACP2 (13) (Figures S22-1 and S27-S30). Consistent
with these observations, incubation of (3R)-7 with
FosDH2 in [18O]-H2O resulted in essentially complete
isotope exchange of the 3-hydroxyl oxygen of recovered
7 within 30 min (Figure S31).
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
Figure 3. Hydration of (2Z,4E)-3-hydroxy-4-hexenoyl-
FosACP2 (21) by FosDH2. Chiral GC-MS of methyl (3R,4E)-
3-hydroxy-4-hexenoate from incubation of 21 with FosDH2.
A-1: XIC m/z 71); A-2: Co-injection with added methyl
(3S,4E)-3-hydroxy-4-hexenoate, XIC m/z 71. See Figure S19
for comparisons of the TIC of reaction product, rt 36.89 min,
and authentic methyl (3R,4E)-3-hydroxy-4-hexenoate.
Scheme 6. FosDH2-catalyzed hydration and dehydra-
tion of FosACP2-bound diketide substrate analogs.
DISCUSSION
Dehydratases of both Type I and Type II fatty acid
synthases (FASs) catalyze the exclusive syn dehydration
of a (3R)-3-hydroxyacyl-ACP thioester to the corre-
sponding (E)-enoyl-ACP.18 The structure of the E. coli
dehydratase (FabA) displays a characteristic hotdog fold
that has been observed in all other DH structures from
both FAS and PKS systems.7a,19,20 The active site of each
DH harbors a universally conserved pair of His and Asp
residues. Schwab, in a critical review of research on
FabA and the closely related dehydratase-isomerase
FabZ, has discussed a one-base, two-step mechanistic
model in which the active site imidazole residue first
catalyzes the stereospecific removal of 2-Hsi of the 3-
hydroxyacyl-ACP substrate following which the transi-
ently-generated imidazolium species donates its proton
to the 3-hydroxyl group to promote C–O bond cleav-
age.18a,21,22 Although the distinct enoyl-CoA hydratase of
fatty acid oxidation differs significantly from DH en-
zymes in both protein fold and the presence of two es-
sential active site Glu residues in place of the His-Asp
dyad of DH domains, it catalyzes an analogous net syn
hydration of (E)-2-enoyl-CoA thioesters to yield the cor-
responding (3S)-3-hydroxyacyl-CoA products.17a Both
protons and the oxygen of the nucleophilic water are
incorporated into the product, consistent with a single-
base mechanism for enoyl-CoA hydratase in which one
Glu residue acts sequentially, first as base and then as
ACS Paragon Plus Environment

Page 5 of 12
Journal of the American Chemical Society
active site acid, while the second Glu side chain posi-
tions the active site water by an essential H-bond.23 In-
The finding that FosKR2 stereospecifically reduces 3-
ketobutyryl-FosACP2 (23) to -(3S)-3-hydroxybutyryl-
FosACP2 (8) establishes that FosDH2 never encounters
the unnatural epimer, -(3S,4E)-3-hydroxy-4-hexenoyl-
FosACP2, in its native modular context. Indeed, we have
now shown that FosDH2 stereospecifically interconverts
1
2
3
4
5
6
7
L
terestingly, while enoyl-CoA hydratase can also add
water to the isomeric (Z)-2-butenoyl-CoA to give the
(3R)-3-hydroxybutyryl-CoA product, the reverse dehy-
dration could not be observed, an observation that was
attributed to the calculated unfavorable Keq <10-3 for for-
mation of the (Z)-isomer.17a
D
L
-(3R,4E)-3-hydroxy-4-hexenoyl-FosACP2
(18)
and
(2Z,4E)-hexadienoyl-FosACP2 (21), the predicted prod-
uct of fostriecin synthase module 2. Interestingly, under
the incubation conditions tested, dehydration of the C4-
analog L-(3S)-8 to (Z)-14 could be detected only indirect-
ly by FosDH2-catalyzed isotopic exchange of the 3-
hydroxyl group with [18O]-water. This apparent discrep-
ancy between the results with C6 and the C4 substrates
is likely the consequence of the use of a 1.5:1 stoichio-
metric excess of FosDH2 to triketide alcohol 18, com-
pared to the 5-fold lower 0.3-0.5:1 ratio that was used for
the incubation of FosDH2 with the shorter chain diketide
analog (3S)-8. Thus although FosDH2 cannot alter the Keq
~10-3 for free substrates and products, enzymes can dif-
ferentially bind substrates and products so as to shift the
ratio of bound substrates species closer to 1:1.27 For the
intact fostriecin synthase module 2, the unfavorable
equilibrium of the FosDH2-catalyzed dehydration reac-
tion is overcome by metabolic coupling to the thermo-
dynamically favorable, metabolically irreversible chain
elongation reaction catalyzed by the KS domain of the
fostriecin synthase module 3.
8
9
EryDH4 (from module 4 of the erythromycin PKS) and
NanDH2 (from module 2 of the nanchangmycin PKS)
both catalyze the syn elimination of water from a
(2R,3R)-2-methyl-3-hydroxyacyl-ACP substrate to the
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
corresponding
(E)-2-enoyl-ACP
product,24
while
RifDH10 (from module 10 of the rifamycin PKS) catalyz-
es the syn dehydration of the diastereomeric (2S,3S)-2-
methyl-3-hydroxyacyl-RifACP10 substrate to the (E)-2-
enoyl-ACP product.9d The structures of both EryDH4
and RifDH10 display the characteristic DH double
hotdog fold as well as the conserved active His and Asp
residues.9d,20 While one or two of the four DH domains
from the curacin PKS are thought to be responsible for
formation of unsaturated intermediates possessing (Z)
double bonds, all four proteins exhibit the common
double hotdog fold and high levels of mutual structural
homology, while the actual enzymatic formation of (Z)-
enoyl-ACP products has not yet been reported.7 FosDH2
shows 71.0% mutual sequence identity (88.8% similarity)
over 276 aa to the closely related PlmDH1 from module
1 of the phoslactomycin PKS, which has also been impli-
cated in the formation of a cis double bond (Figure S32).
Interestingly, FosDH1 and FosDH2 themselves show a
more modest 40.8% mutual sequence identity (63.4%
similarity), while retaining each of the key conserved
amino acid motifs typified by the structurally character-
ized dehydratases EryDH4 and RifDH10.
Finally, the demonstration that FosDH2 catalyzes the
dehydration of L-(3R,4E)-18 to (2Z,4E)-21 does not vali-
date the common assumption that there is a requisite
correlation between an
L-hydroxy configuration of the
substrate and the cis geometry of the olefinic product of
DH-catalyzed dehydration.14a There are already suffi-
cient exceptions7b,8b,c,9d,23,25 to this superficially appealing
generalization to establish that it does not have reliable
predictive value.
In spite of the frequent occurrence of Z double bonds
in complex polyketides, the experimental demonstration
that FosDH2 can catalyze the interconversion of the
L
-3-
EXPERIMENTAL METHODS
hydroxyacyl-ACP and (Z)-2-enoylacyl-ACP thioesters is
the first documented in vitro confirmation of this nomi-
nally straightforward reaction directly catalyzed by a
PKS DH domain. Earlier unsuccessful attempts to ob-
serve DH-catalyzed formation of conjugated Z-enoyl
thioester double bonds have been summarized above.
One recent report has described the highly unusual de-
Materials. IPTG and kanamycin were purchased from
Thermo Fisher Scientific. (E)-2-Butenoyl-CoA ((E)-
Crotonyl-CoA), 3-ketobutyryl-CoA (acetoacetyl-CoA),
N,O-BSTFA, TCEP, (3R)- and (3S)-3-hydroxybutyric ac-
ids, 2-butenoic acid (crotonic acid), and (2E,4E)-2,4-
hexadienoic acid were purchased from Sigma-Aldrich
and utilized without further purification. (Z)-2-Butenoic
acid was prepared as previously described.17a,28 [18O]-
H2O was purchased from Cambridge Isotope Laborato-
ries. HRV 3C protease was obtained from Pierce. Pseu-
domonas fluorescens Amano lipase P was from Sigma-
Aldrich. Sfp and PICS TE were each expressed and puri-
fied as previously described.9d,24,29 DNA primers were
synthesized by Integrated DNA Technologies. Compe-
tent E. coli DH5α, DH10β and BL21(DE3) cloning and
expression strains were purchased from New England
Biolabs (NEB). Restriction enzymes, T4 DNA ligase and
Phusion High-Fidelity PCR Master Mix with HF Buffer
were purchased from NEB. Amicon Ultra Centrifugal
Filter Units (Amicon Ultra-15 and Amicon Ultra-0.5,
3000, 10000 and 30000 MWCO) were obtained from Mil-
hydration
of
(3S,4S)-3,4-dihydroxypentanoyl-N-
acetylcysteamine thioester to the (Z)-2-enoyl-SNAC cata-
lyzed by a modular PKS domain formally classified as a
TE on the basis of phylogenetic sequence comparisons.25
Curiously, the TE-catalyzed dehydration reaction could
not be detected with the corresponding ACP thioester.
The well-studied FabZ-catalyzed formation of the non-
conjugated
Z-3,4-decenoyl-ACP
from
(3R)-3-
hydroxydecanoyl-ACP involves the allylic isomerization
of the initially-formed (E)-2-decenoyl-ACP.18,21 Although
there is molecular genetic evidence that the characteristic
Z-double bond of epothilone is introduced by the DH
domain from the proximal downstream module acting
in trans, this transformation has not been verified at the
enzyme level.26
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 12
lipore. The 20 mL HisPrep FF 16/10 column, prepacked with 150 mL of wash buffer (50 mM sodium phosphate,
1
2
3
4
5
6
7
8
with precharged Ni Sepharose 6 Fast Flow, was pur-
chased from GE Healthcare.
500 mM NaCl, 10 mM imidazole, pH 7.5) to elute con-
taminating proteins. FosDH1 or FosDH2 were then elut-
ed with a linear gradient from 10 mM - 500 mM imidaz-
ole in the same buffer. The fractions containing FosDH1
or FosDH2 were pooled, buffer-exchanged, and concen-
trated to final buffer (50 mM sodium phosphate, 250 mM
NaCl, 10% glycerol, pH 7.5) using an Amicon Ultra-15
(30000 MWCO) centrifugal filter. The purity and MW of
FosDH1 and FosDH2 were analyzed by SDS-PAGE and
LC-ESI(+)-MS (Figures S6 and S7) Aliquots of the puri-
fied proteins were stored at -80 °C.
Methods. General methods were as previously de-
scribed.30 Growth media and conditions used for E. coli
strains and standard methods for handling E. coli in vivo
and in vitro were those described previously, unless oth-
erwise noted.31 All DNA manipulations were performed
following standard procedures.31 Plasmid DNA was pu-
rified using a Thermo Scientific GeneJET Plasmid mini-
prep kit. DNA sequencing was carried out by Genewiz,
South Plainfield, NJ. Synthetic genes, optimized for ex-
pression in E. coli, were prepared by DNA 2.0, Newark,
California. All proteins were handled at 4 °C unless oth-
erwise stated. Protein concentrations were determined
according to the method of Bradford, using a Tecan Infi-
nite M200 Microplate Reader with bovine serum albu-
min as standard. 32 Protein purity and size were estimat-
ed using SDS-PAGE, visualized using Coomassie Blue
stain, and analyzed with a Bio-Rad ChemiDoc MP Sys-
tem. Gas chromatography-mass spectrometry (GC−MS)
analyses were performed using an Agilent 5977A Series
GC/MSD instrument (70 eV, electron impact) with a 3
min solvent delay. Protein accurate molecular weight
was determined on an Agilent 6530 Accurate-Mass Q-
TOF LC-MS. A Thermo LXQ equipped with Surveyor
HPLC system and a Phenomenex Jupiter C4 column (150
mm × 2 mm, 5.0 ꢁm) was utilized for analysis of acyl-
ACP compounds. HPLC-ESI-MS/MS analysis was car-
ried out in positive ion mode for analysis of pan-
tetheinate ejection fragments, as previously described.15,33
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
Expression and Purification of His6-tag-NusA-
FosACP1 and His6-tag-NusA-FosACP2 and Preparation
of apo-FosACP1 and apo-FosACP2. Synthetic genes for
FosACP1 (V3130 to T3220 of Fos Module 1 (Figures S1
and S2) and FosACP2 (region from R1696 to T1820 of
Fos Module 2 (Figures S3 and S4) were each subcloned
into pET28a between the NdeI and XhoI restriction sites.
Since the resulting recombinant ACP proteins were pro-
duced predominantly as insoluble inclusion bodies
when expressed in E. coli BL21(DE3), the corresponding
NusA-FosACP1 and NusA-FosACP2 fusion proteins
were then generated. In brief, the DNA regions encoding
FosACP1 and FosACP2 were each amplified by PCR
from the above-described pET28a constructs using the
primers FosACP1/2-HRV-FP and pET-28a-RP which is
complementary to the region just 3′ of the native XhoI
site (Figure S5). The resultant amplified DNA harboring
FosACP1 or FosACP2 was digested with NheI and XhoI
before ligation into the corresponding sites of an
NheI/XhoI-digested vector, immediately downstream of
DNA encoding His6-NusA-HRV3C housed in a pET28
vector. The resultant expression plasmids encoded the
individual NusA-FosACP1 and NusA-FosACP2 fusion
proteins, each carrying an N-terminal His6-tag and an
HRV 3C protease site between the N-terminal NusA and
either FosACP1 or FosACP2.
Expression and Purification of FosDH1 and FosDH2.
A synthetic gene for FosDH1, optimized for expression
in E. coli and corresponding to the region from A1992 to
G2294 of Fos Module 1 (Figures S1 and S2), was sub-
cloned into pET28a between the NdeI and XhoI re-
striction sites. The FosDH2 expression plasmid corre-
sponding to the region from A947 to A1232 of Fos Mod-
ule 2 (Figure S2 and S3) was generated by subcloning the
synthetic gene optimized for expression in E. coli into
pET28a between the NdeI and XhoI restriction sites. Sin-
gle colonies of E. coli BL21(DE3) cells that had been
transformed with the individual FosDH1 or FosDH2
expression vectors were inoculated into 10 mL LB media
containing 50 mg/L kanamycin and incubated overnight
at 37 °C. This starter culture was then inoculated into 500
mL Super Broth (SB) containing 50 mg/L kanamycin.
The culture was then grown at 37 °C at 225 rpm until
OD600 = 0.5. At this point, the culture was cooled to 18 °C
and then induced with 0.2 mM IPTG. Cells were har-
vested after 20 h by centrifugation at 4000g for 40 min.
The purification was then carried out at 4 °C unless men-
tioned otherwise. Harvested cells were re-suspended in
50 mL start buffer (50 mM sodium phosphate, 500 mM
NaCl, pH 7.5). The cells were lysed by sonication and
cell debris was removed by centrifugation at 20000g for
50 min. The supernatant was loaded onto a HisPrep FF
16/10 (GE Healthcare Life Science) column pre-
equilibrated with start buffer. The column was washed
NusA-FosACP1 or NusA-FosACP2 were expressed in
E. coli BL21(DE3) as described above for FosDH1 and
FosDH2. Cells were harvested after 20 h by centrifuga-
tion at 4000g for 40 min. The recovered cells were re-
suspended in 50 mL start buffer and lysed by passage
three times through a French Press at 10000 psi, then
centrifuged at 20000g for 50 min. The pellet was discard-
ed and the supernatant was loaded on to a HisPrep FF
16/10 (GE Healthcare Life Science) column pre-
equilibrated with start buffer. The column was washed
with 150 mL of wash buffer to elute contaminating pro-
teins. The respective His6-NusA-FosACP1 and His6-
NusA-FosACP2 proteins were then eluted with a gradi-
ent from 10 mM - 500 mM imidazole in the same buffer.
The fractions containing His6-NusA-FosACP1or His6-
NusA-FosACP2 were pooled. Protein was concentrated
to 2-3 mL using an Amicon Ultra-15 (30000 MWCO) cen-
trifugal filter before further purification on a Hiload
16/600, Superdex 200 pg size exclusion column pre-
equilibrated with 50 mM sodium phosphate and 250
mM NaCl, pH 7.5. Fractions containing His6-NusA-
ACS Paragon Plus Environment

Page 7 of 12
Journal of the American Chemical Society
FosACP1 or His6-NusA-FosACP2 were pooled and
with 10% ethylacetate/90% hexane and then (3R,4E)-3-
acetoxy-4-hexenoate was eluted by 20%
1
treated with 1 ꢁL of HRV-3C protease (1U/ꢁL) per 200
ꢁg of purified protein. After overnight incubation at 4 °C
with continuous shaking, the mixture was loaded on to a
HisPrep FF 16/10 column pre-equilibrated with start
buffer. Both His6-NusA up to the HRV 3C protease
cleavage site, as well as the HRV-3C protease, which also
carries a His6-tag, along with any undigested His6-NusA-
FosACP were retained by the column while the cleaved
FosACP1 and FosACP2 eluted during the column wash
step. The fractions containing cleaved FosACP1 or
FosACP2 were pooled and buffer-exchanged with final
buffer (50 mM sodium phosphate, 250 mM NaCl, 10%
glycerol, pH 7.5) using an Amicon Ultra-15 (30000
MWCO) centrifugal filter. The purity and MW of
FosACP1 and FosACP2 were analyzed by SDS-PAGE
and LC-ESI(+)-MS (Figures S9 and S10). Aliquots of pu-
rified proteins were stored at – 80 °C.
2
3
4
5
6
7
8
9
ethylacetate/80% hexane. If the resolution was incom-
plete and the acetate ester contained more than 10%
(3S,4E) isomer, a second round of Amano lipase P reac-
tion was performed. Hydrolysis of the recovered ethyl
(3S,4E)-3-hydroxy-4-hexenoate with LiOH (1.0 eq) at 0
°C for 1-2 h generated (3S,4E)-3-hydroxy-4-hexenoic ac-
id. Methanolysis of (3R,4E)-3-acetoxy-4-hexenoate (700
mg, 3.5 mmol) was effected by treatment with K2CO3
(966 mg, 2.0 eq)in 12 mL MeOH with vigorous stirring at
room temperature for 30 min. The solution was diluted
with EtOAc (240 mL) and washed with 0.1 M aq. NaOH
(240 mL). The aqueous layer was back-extracted with
EtOAc (3 x 240 mL). The combined organic fractions
were dried over MgSO4 and concentrated to give methyl
(3R,4E)-3-hydroxy-4-hexenoate (22-Me) which was puri-
fied by flash column chromatography on silica gel (20%
ethylacetate/80% hexane). Hydrolysis with LiOH (1.0
eq) at 0 °C for 1-2 h generated (3R,4E)-3-hydroxy-4-
hexenoic acid (22). By chiral GC-MS analysis (Method B)
of the methyl esters, the major component (~90%) of the
(3R,4E)-22-Me eluted ca. 0.9 min earlier than the (3S,4E)-
methyl ester (~10%), corresponding to ~80% ee (Figure
S18). Chiral GC-MS analysis also established that the
preparation of the enantiomeric methyl (3S,4E)-3-
hydroxy-4-hexenoate was obtained in ~90% ee.
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
Expression and Purification of FosKR2. A synthetic
gene for FosKR2, optimized for expression in E. coli and
corresponding to the region from F1239 to L1683 of Fos
Module 2 (Figures S3 and S4) was subcloned into
pET28a between the NdeI and XhoI restriction sites.
FosKR2 was expressed and purified by the same proce-
dures described for FosDH1 and FosDH2 (Figure S8).
(2Z,4E)-2,4-Hexadienoic Acid (18). (2Z,4E)-2,4-
Hexadienoic acid (18) was prepared as previously de-
scribed.13 The cis-olefin geometry was verified by 1H
NMR and chiral GC-MS by which the (2Z,4E)-2,4-
hexadienoic acid (18) was readily separated from com-
mercially available (2E,4E)-17 either as the carboxylic
acids or derived methyl esters 17-Me and 18-Me. The
carboxylic acid forms were analyzed with an Agilent
ChiraSil-Dex capillary GC column, (0.32 mm ID x 25 m
length x 0.25 ꢁm film) using a temperature program (GC
Method A) with a 1 min hold at 50 °C, followed by a 7.5
°C/min increment to 200 °C. Separation of the methyl
esters 17-Me and 18-Me forms used a temperature pro-
gram (GC Method B) with a 1 min hold at 50 °C,
followed by a 1.00 °C/min increase to 90 °C, a 2 min
hold at 90 °C, and then a further increment of 20.00
(3R)- and (3S)-Hydroxybutyryl-CoA. Using the previ-
ously described method,24a 25 mg (0.24 mmol) of (3R)- or
(3S)-3-hydroxybutyric acid in 1 mL THF were treated by
dropwise addition of 60 mg (0.36 mmol, 1.5 eq) of 1,1′-
carbonyldiimidazole (CDI) in 0.5 mL of THF. After reac-
tion at 0 °C for 60-90 min, a solution of CoASH (20 mg in
1.5-2.0 mL H2O, 0.024 mmol, 0.1 eq) was added drop-
wise. The reaction was continued at room temperature
and under nitrogen for 4 h. Organic solvent was re-
moved by rotary evaporation. The aqueous phase, after
extraction with ethyl acetate to remove byproducts, was
purified through by HPLC (Agilent) using a Phenom-
enex Gemini semi-preparative C18 column (150 x 10
mm, 10 ꢁm) equilibrated with 2% CH3CN/H2O. Elution
was carried out with a linear gradient from 2% to 90%
CH3CN/H2O. Collected peaks were checked for purity
by LC-ESI(+)-MS using an Agilent Zorbax Extend C18
column (100 x 2.1 mm, 3.5 ꢁm) with a linear gradient
from 2% to 70% CH3CN/H2O. Peaks containing product
of the desired mass, LC-ESI(+)-MS [M+H]+ 854, were
collected, lyophilized and stored at -80 °C.
1
°C/min to 200 °C (Figure S16). The H NMR data for 18
matched the literature values:13 1H NMR (400 MHz,
CDCl3): δ 7.35 (ddd, 1 H), 6.63 (dd, 1 H), 6.04 (dq, 1 H),
5.57 (d, 1 H), 1.90 (d, 3 H).
(3R,4E)-3-Hydroxy-4-hexenoic Acid (22) and (3S,4E)-
3-Hydroxy-4-hexenoic Acid. ( )-Ethyl (3RS,4E)-3-
Hydroxy-4-hexenoic acid was synthesized as previously
described.34 The enantiomers were kinetically resolved
using Amano lipase P (Scheme S1).12b,c In brief, ( )-ethyl
(4E,3RS)-3-hydroxy-4-hexenoate (1.58 g, 10 mmol) was
dissolved in 20 mL of hexane. Amano lipase P (1 g) was
added with vinyl acetate (20 mmol) and the reaction
mixture was stirred at room temperature for 24 h with
monitoring by GC-MS. The lipase was removed by filtra-
tion and the solvent was evaporated. Ethyl (3S,4E)-3-
hydroxy-4-hexenoate and ethyl (3R,4E)-3-acetoxy-4-
hexenoate were separated by SiO2 flash column chroma-
tography. The ethyl-3-hydroxy-4-hexenoate was eluted
(E)- and (Z)-2-Butenoyl-CoA ((E)- and (Z)-Crotonyl-
CoA. Using the previously described method,24a 20 mg
(0.23 mmol) of (E)- or (Z)-2-butenoyl-CoA was dissolved
in 2 mL anhydrous CH2Cl2 under nitrogen, then treated
with 65 ꢁL of triethylamine (47 mg, 0.47 mmol, 2 eq) fol-
lowed after 10 min by 44 ꢁL of ethylchloroformate (50
mg, 0.46 mmol, 2 eq) . The reaction mixture was stirred
at 0 °C for 2 h. After removal of the organic solvent by
rotary evaporation, the residue was dissolved in 2 mL
THF. Insoluble salts were removed by centrifugation
and the mixed anhydride was added to a round bottom
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 12
Zorbax C18 column (2.1 × 50 mm, 3.5 ꢁm) and a linear
flask containing 20 mg of CoASH dissolved in 2 mL of
1
2
3
4
5
6
7
8
50 mM aq. NaHCO3 (pH 8.0) and the resultant mixture
was stirred for 1-3 h at room temperature under N2. Af-
ter removal of excess starting materials by extraction
with ethyl acetate, the aqueous phase was purified by
HPLC (Agilent) using the Phenomenex Gemini semi-
preparative C18 column (150 x 10 mm, 10 µm) equili-
brated with 2% CH3CN/H2O. Elution was carried out
with a linear gradient from 2% to 10% CH3CN/H2O.
Collected peaks were checked for purity by LC-ESI(+)-
MS as described above for 3-hydroxybutyryl-CoA. Peaks
exhibiting the desired mass, LC-ESI(+)-MS [M+H]+ 836
were collected, lyophilized and stored at -80 °C.
gradient from 10% to 100% of CH3CN/H2O.
Monitoring of FosDH1 and FosDH2 Activity Using
Acyl-CoA Substrates. The activity of various protein
preparations of FosDH1 and FosDH2 was conveniently
checked using the surrogate –SCoA substrates (E)-2-
butenoyl-CoA and 3-hydroxybutyryl-CoA. Assay mix-
tures contained 1 mM 3-hydroxybutyryl-CoA or (E)-2-
butenoylCoA in 50 mM sodium phosphate, pH 7.5 Buff-
er in a total volume of 100 ꢁL. The assay mixture was
divided into 50-ꢁL portions. To one, 100 ꢁM FosDH1 or
FosDH2 was added while the blank used an equivalent
volume of 50 mM sodium phosphate, pH 7.5 buffer. The
assay mixtures were incubated at room temperature for
2 h, then diluted with 200 ꢁL of H2O and passed through
a Millipore 30 kDa MWCO 500 ꢁL filter and centrifuged
at 14000g to remove FosDH1 or FosDH2 by buffer ex-
change. The assay mixtures were analyzed by LC-ESI(+)-
MS using an Agilent Zorbax Extend C18 column (100 x
2.1 mm, 3.5 ꢁm) with a linear gradient from 2% to 70%
CH3CN/H2O to monitor for the expected increase or
decrease of 18 amu in the mass of the acyl-CoA compo-
nents as a result of hydration or dehydration.
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
L
-(3R,4E)-3-Hydroxy-4-hexenoyl-CoA.
L-(3R,4E)-3-
Hydroxy-4-hexenoic acid (17, 26 mg, 0.2 mmol) in 1 mL
anhydrous CH2Cl2 was treated with 80 ꢁL of 2M oxalyl
chloride solution (0.16 mmol, 0.8 eq) and 4 drops of an-
hydrous DMF in a flask fitted with a glass funnel filled
with Drierite. (Note: It is important not to use an excess
of oxalyl chloride in order to avoid unwanted reaction of
the allylic alcohol group.) After vigorous stirring at room
temperature for 2-3 h, the solvent was evaporated after
dilution with ethyl acetate, resulting in co-evaporation of
any unreacted oxalyl chloride. The residue was dis-
solved in 2 mL THF and the solution was added to 20
mg of CoASH dissolved in 1 mL of 0.4 M NaHCO3, pH
8.0. After stirring for 1-2 h at room temperature, excess
starting materials were removed by extraction with ethyl
acetate. The resulted aqueous crude acyl-CoA mixture
was purified by HPLC using a Phenomenex Gemini
semi-preparative C18 column, 150 × 10 mm, equilibrated
with 10% CH3CN/H2O. The sample was eluted with a
linear gradient from 10% to 100% of CH3CN/H2O. HPLC
peaks were collected and lyophilized. Each fraction was
analyzed by HPLC-ESI(+)-MS using an Agilent Zorbax
C18 column (2.1 × 50 mm, 3.5 ꢁm) and a linear gradient
from 10% to 100% of CH3CN/H2O.
C4 Acyl-FosACP Substrates, 5-8 and 11-14. Small
scale acylation reactions for LC-MS analysis. For prep-
aration of acyl-FosACP substrates, each reaction con-
tained 100-150 ꢁM apo-FosACP1 or apo-FosACP2, and
250-300 ꢁM (3R)- or (3S)-3-hydroxybutyryl-CoA or (E)-
or (Z)-2-butenoyl-CoA, plus 2 ꢁM Sfp, 10 mM MgCl2,
and 1 mM TCEP in 50 mM sodium phosphate, pH 7.5, in
a total volume of 100 ꢁL. The reactions were incubated at
room temperature for 10 min to form the corresponding
acyl-FosACP1 (5-8) or acyl-FosACP2 (11-14). The reac-
tion mixture was then concentrated using a Amicon Ul-
tra-0.5 (3000 MWCO) centrifugal filter and centrifuged at
14000g to remove unreacted acyl-CoA by buffer ex-
change, with recovery of the acyl-FosACP retentate
which was diluted to a total volume of 200-300 ꢁL.
(2Z,4E)- and (2E,4E)-2,4-Hexadienoyl-CoA. (2Z,4E)-
2,4-Hexadienoic acid (19) or (2E,4E)-2,4-hexadienoic acid
(20) (0.2 mmol) was dissolved in 1 mL of anhydrous
CH2Cl2 under N2. Triethylamine (70 ꢁL, 47 mg, 4.0 eq)
was added, followed after 10 min at 0 °C by 50 ꢁL of
ethylchloroformate (45.8 mg, 3.0 eq). The reaction mix-
ture was stirred for 2 h at 0 °C. After evaporation of the
solvent, the residue was dissolved in 2 mL THF and the
insoluble salts were removed by centrifugation. The
mixed anhydride was then slowly added to a separate
round bottom flask containing 20 mg of CoASH in 1 mL
of NaHCO3 buffer (pH 8.0). The reaction mixture was
stirred 1-3 h at room temperature with monitoring by
LC-MS. After removal of excess starting material by ex-
traction with EtOAc, the aqueous crude acyl-CoA mix-
ture was purified by HPLC using the Phenomenex Gem-
ini semi-preparative C18 column, 150 × 10 mm, equili-
brated with 10% CH3CN/H2O. The sample was eluted
with a linear gradient from 10% to 100% of CH3CN/H2O.
HPLC peaks were collected and lyophilized. Each frac-
tion was analyzed by HPLC-ESI(+)-MS using an Agilent
FosDH1 and FosDH2-Catalyzed Dehydration and
Hydration of C4 Acyl-FosACP Substrates. LC-ESI(+)-
MS and LC-ESI(+)-MS/MS Analysis. The above-
described samples of acyl-FosACP were divided into 2
equal 100-150 ꢁL portions. To one, FosDH1 or FosDH2
was added to a final concentration of 50 ꢁM while the
blank was supplemented with an equivalent volume of
50 mM sodium phosphate, pH 7.5. After parallel 60-min
incubations at room temperature, each reaction mixture
was diluted with formic acid/H2O and centrifuged for 5
min at 14000g. These samples were then analyzed by LC-
ESI(+)-MS and LC-ESI(+)-MS/MS using an analytical
Aeris widepore-C4 column (3.6 ꢁm, 2.1 x 150 mm) from
Phenomenex using a linear gradient from 30% to 70%
CH3CN/H2O on a Thermo-LXQ mass spectrometer. For
LC-ESI(+)-MS/MS analysis15 the M11+ ion was selected
for MS/MSs such that, both the hydrated and dehydrat-
ed pPant ejection fragments could be observed together
(Figures S12, S13, S23, S24, S20, S21, and S27-S30).
ACS Paragon Plus Environment

Page 9 of 12
Journal of the American Chemical Society
C4 Acyl-FosACP Substrates, 5-8 and 11-14. Larger
hydrolysis, each sample was cooled on ice and then
acidified to pH 3.0-3.5 by addition of 54 ꢁL of 1 M HCl.
Each sample was centrifuged at 14000g for 5 min to
remove precipitated protein. The supernatant was then
extracted with 4 x 800 ꢁL of CH2Cl2. The solvent was
removed by rotary evaporation. The residue dissolved in
100 ꢁL of CH2Cl2 was derivatized by treatment with 10
ꢁL BSTFA. The samples were directly analyzed by chiral
GC-MS on a ChiraSil_Dex capillary GC column using a
temperature program (GC Method D) with a 1 min hold
at an initial temperature of 50 °C followed by an incre-
ment of 8 °C/min to 150 °C, a 1 min hold at 150 °C, an
increase of 20 °C/min to 210 °C, and a final 3 min hold at
210 °C (Figure S15).
1
2
3
4
5
6
7
8
scale acylation reaction for GC-MS analysis of FosDH-
catalyzed dehydration or dehydration: Each reaction
consisted of 100 ꢁM apo-FosACP1 or apo-FosACP2 and
250-300 ꢁM (3R)- or (3S)-3-hydroxybutyryl-CoA E)- or
(Z)-2-butenoyl-CoA, plus 2 ꢁM Sfp, 10 mM MgCl2, and 1
mM TCEP in 50 mM sodium phosphate, pH 7.5, in a
total volume of 500 ꢁL. (TCEP was omitted from the re-
action with (Z)-2-butenoyl-CoA.) Each reaction was in-
cubated at room temperature for 10 min to form the cor-
responding acyl-FosACP product. The reaction mixture
was then concentrated using an Amicon Ultra-0.5 (3000
MWCO) centrifugal filter with centrifugation at 14000g
to remove unreacted acyl-CoA by buffer exchange, with
recovery of the acyl-FosACP retentate, which was dilut-
ed to a total volume of 500 ꢁL.
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
FosDH2-Catalyzed Dehydration and Hydration of
C6 Acyl-FosACP2 Substrates, (3R,4E)-18 and (2Z,4E)-21.
(2Z,4E)-2,4-Hexadienoyl-CoA or (3R,4E)-3-hydroxy-4-
hexenoyl-CoA (200-500 ꢁM) was incubated with 50 ꢁM
apo-FosACP2 and 40 ꢁM Sfp, 2 mM DTT, and 15 mM
MgCl2 in reaction buffer (350 mM NaCl 50 mM
phosphate pH 7.5, total vol 2.5 mL) for 30 min at 30 °C.
The residual CoA substrate, MgCl2, and DTT were
removed by passage of 2.5 mL of the incubation mixture
through a PD-10 column that had been equilibrated
with 25 mL of the reaction buffer. FosACP2-bound
product, 18 or 21, was eluted by 3.5 mL of the reaction
buffer. The PD-10 eluate was concentrated using an
Amicon centrifuge unit 10,000 MWCO (30 min at 1,600g
at 4 °C reduced the volume to 200-300 ꢁL. The acylated
FosACP2 was analyzed by LC-MS The low
concentration of FosACP2 was used to avoid
precipitation during the procedure. FosDH2 was added
to a final concentraion of 300 ꢁM along with 2 mM DTT
and the reaction buffer was added to adjust the total
volume to 500 ꢁL. The reaction mixture was incubated at
room temperature for 1-2 h. At the end of the reaction,
the product was released from the FosACP2 by PICS TE
(20 ꢁL per reaction, 20 min at room temperature). The
reaction mixture was acidified to pH <3 by addtion of 1
M HCl. The C6 acids were extracted with 2 x 800 ꢁL
ethyl acetate. The solvent was removed by rotary
evaporation. The product was taken up in 80 ꢁL of
MeOH, 20 ꢁL of TMS-CHN2 was added, and the solution
incubated at room temperature for 5 min. The resulting
methyl esters were analyzed by chiral GC-MS on a
ChiraSil-Dex capillary GC column using GC Method B
(Figures S16-S19).
FosDH1-Catalyzed Dehydration and Hydration of
C4 Acyl-FosACP Substrates. Chiral GC-MS Analysis.
The above-described samples of acyl-FosACP1 (5, 6, 11,
and 12 were divided into 2 equal 250 ꢁL portions. To
one, FosDH1 was added to a final concentration of 40
ꢁM while the blank was supplemented with an equiva-
lent volume of 50 mM sodium phosphate, pH 7.5. Both
mixtures, with and without FosDH1, were incubated at
room temperature for 30 min before addition to each of
200 ꢁM of PICS TE. After hydrolysis by PICS TE for 15
min, the reaction was quenched and the pH was adjust-
ed to 3.0-3.5 by addition of ~9 ꢁL of 1 M HCl. After cen-
trifugation at 13000g for 5 min to remove precipitated
protein, the supernatant was extracted with 4 x 800 ꢁL of
CH2Cl2. After removal of solvent by rotary evaporation,
the residue was dissolved in 130 ꢁL CH2Cl2 and then
treated with 20 ꢁL BSTFA for derivatization of the or-
ganic acids (Total 150 µl). The derivatized samples were
directly analyzed by chiral GC-MS (HP GCD system) on
a Varian CP ChiraSil_DEX column (25 m length, 0.25 ꢁm
diameter) using a temperature program (GC Method C)
with a 1 min hold at 50 °C, followed by an increment of 8
°C/min to 150 °C, a 1 min hold, and then an increase of
20 °C/min to 210 °C, and a final hold at this temperature
for 3 min (Figures S11 and S14).
FosKR2-Catalyzed Reduction of 3-Ketobutyryl-
FosACP2 (23). 3-Ketobutyryl-CoA (400 ꢁM) was reacted
for 10 min at room temperature with 150 ꢁM apo-
FosACP2 in the presence of 2 ꢁM Sfp in 50 mM sodium
phosphate, pH 7.2 containing 10 mM MgCl2 and 1 mM
TCEP (total vol 400 ꢁL) to form 3-ketobutyryl-FosACP2
(23). The reaction mixture was passed through a
Millipore 3 KDa MWCO 500 ꢁL filter by centrifugation
at 14000g to remove unreacted acyl-CoA by buffer
exchange. The retentate containing 3-ketobutyryl-
FosACP2 (23) was divided into two separate 200 ꢁL
aliquots. To each of these, 1 mM NADPH and 40 µM
FosKR2 was added. One reaction was quenched
immediately by treatment with 150 mM NaOH (18 ꢁL of
2 M NaOH) at 65 °C for 20 min. The other portion was
incubated for 30 min before being quenched in the same
manner with aq.with 150 mM of NaOH. After
FosDH2-Catalyzed Dehydration and Hydration of
C4 Acyl-FosACP Substrates. Chiral GC-MS Analysis.
The above-described samples (1 mL) of acyl-FosACP2 (7,
8, 13, and 14 were mixed with FosDH2 (final concentra-
tion 50 ꢁM). Aliquots of 250 ꢁL were collected at 0, 15, 30
and 45 min incubation time and added to Eppendorf
tubes containing 200 ꢁM of PICS TE and incubated for
15 min at room temperature. After hydrolysis by PICS
TE, the reaction was quenched and the pH was adjusted
to 3.0-3.5 by addition of ~9 ꢁL of 1 M HCl. Samples were
centrifuged at 13000g for 5 min to remove precipitated
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 12
†Department of Chemistry and Biochemistry, Institute of
Advanced Biomedical Research, George Mason University,
Manassas, Virginia 20110, United States
protein. The supernatant was extracted with 4 x 800 ꢁL
of CH2Cl2. The solvent was then removed by rotary
evaporation and the residue was dissolved in 100 ꢁL
CH2Cl2 to which 10 ꢁL BSTFA was added for derivatiza-
tion of acids (Total 110 µl). These samples were directly
analyzed by chiral GC-MS (HP GCD system) with Vari-
an CP ChiraSil_DEX column (25 m length, 0.25 ꢁm di-
ameter) using a temperature program (GC Method D)
with a 1 min hold at 55 °C, followed by an increment of
0.5 °C/min to 65 °C, a 1 min hold, and then an increase
of 15 °C/min to 90 °C, a 1-min hold at 90 °C, an increase
of 20 °C/min to 200 °C, and a final 1 min hold at this
temperature (Figures S22, and S25).
1
2
3
4
5
6
7
8
Author Contributions
‡These authors contributed equally. All authors have given
approval to the final version of the manuscript.
Notes
The authors declare no competing financial interest.
9
ACKNOWLEDGMENTS
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
This work was supported by a grant from the U. S. National
Institutes of Health, GM022172, to D.E.C. We thank Will
Furuyama and Colin Gould for valuable experimental assis-
tance.
REFERENCES
(1) Lewy, D. S.; Gauss, C. M.; Soenen, D. R.; Boger, D. L.
Curr. Med. Chem. 2002, 9, 2005-2032.
(2) a) Amemiya, M.; Someno, T.; Sawa, R.; Naganawa, H.;
Ishizuka, M.; Takeuchi, T. J. Antibiot (Tokyo) 1994, 47, 541-544.
b) Fushimi, S.; Furihata, K.; Seto, H. J. Antibiot. 1989, 42, 1026-
1036. c) Palaniappan, N.; Kim, B. S.; Sekiyama, Y.; Osada, H.;
Reynolds, K. A. J. Biol. Chem. 2003, 278, 35552-35557.
(3) Mamber, S. W.; Okasinski, W. G.; Pinter, C. D.; Tunac, J.
B. J. Antibiot (Tokyo) 1986, 39, 1467-1472.
(4) a) Kong, R.; Liu, X.; Su, C.; Ma, C.; Qiu, R.; Tang, L. Chem.
Biol. 2013, 20, 45-54. b) Palaniappan, N.; Alhamadsheh, M. M.;
Reynolds, K. A. J. Am. Chem. Soc. 2008, 130, 12236-12237. c)
FosER1, the enoyl reductase domain of module 1, is apparently
catalytically silent.
(5) Shiomi, K.; Omura, S. In Macrolide Antibiotics. Chemistry,
Biology, and Practice; Second ed.; Omura, S., Ed.; Academic
Press: San Diego, CA, 2002, p 1-56.
(6) Tang, L.; Shah, S.; Chung, L.; Carney, J.; Katz, L.; Khosla,
C.; Julien, B. Science 2000, 287, 640-642.
(7) a) Akey, D. L.; Razelun, J. R.; Tehranisa, J.; Sherman, D.
H.; Gerwick, W. H.; Smith, J. L. Structure 2010, 18, 94-105. b)
Fiers, W. D.; Dodge, G. J.; Sherman, D. H.; Smith, J. L.; Aldrich,
C. C. J. Am. Chem. Soc. 2016, 138, 16024-16036.
(8) a) Olano, C.; Wilkinson, B.; Sanchez, C.; Moss, S. J.; Sheri-
dan, R.; Math, V.; Weston, A. J.; Brana, A. F.; Martin, C. J.; Oli-
ynyk, M.; Mendez, C.; Leadlay, P. F.; Salas, J. A. Chem. Biol.
2004, 11, 87-97. b) Vergnolle, O.; Hahn, F.; Baerga-Ortiz, A.;
Leadlay, P. F.; Andexer, J. N. Chembiochem 2011, 12, 1011-1014.
c) Hahn, F.; Kandziora, N.; Friedrich, S.; Leadlay, P. F. Beilstein
J. Org. Chem. 2014, 10, 634-640.
FosDH2-Catalyzed
Hydroxybutyryl-FosACP2.
Isotope
Exchange
and
of
(3S)-3-
3-
(3R)-
Hydroxybutyryl-FosACP2 (7 and 8) were generated by
incubation of 800-1000 ꢁM (3R)- or (3S)-3-
hydroxybutyryl-CoA and 200 ꢁM apo-FosACP2 with 5
ꢁM Sfp in the presence of 10 mM MgCl2 and 1 mM TCEP
in 50 mM sodium phosphate, pH 7.5 in a total volume of
100 ꢁL. After 15 min incubation at room temperature to
form (3R)- or (3S)-3-hydroxybutyryl-FosACP2 (7 or 8),
the reaction mixture was passed through a Millipore 3
KDa MWCO 500 ꢁL filter with centrifugation at 14000g
to remove unreacted acyl-CoA by buffer exchange. The
retentate (100 ꢁL) containing 7 or 8 was mixed with 300
ꢁL of [18O]-H2O-based buffer (final 18O enrichment 75
atom%.) To this solution, 50 ꢁM (final concentration)
FosDH2 was added. Samples of 95 ꢁL were withdrawn
after 0, 30 and 90 min for (3S)-3-hydroxybutyryl-
FosACP2 (8) and 0, 30, 90 and 300 min for (3R)-3-
hydroxybutyryl-FosACP2 (7). Each sample was added to
an Eppendorf tube containing 3.5% formic acid (final
concentration) and then directly analyzed by LC-ESI(+)-
MS/MS with monitoring of the pPant ejection fragment
from only the 3-hydroxybutyryl-FosACP and measure-
ment of the relative abundance of the 347 Da (16O) and
349 Da (18O) species (Figure S26 and S31.
ASSOCIATED CONTENT
The Supporting Information is available free of charge on
the ACS Publications website at DOI:10.1021/jacs.*******
Sequence alignments and protein structure comparisons,
PKS domain boundaries, SDS-PAGE and LC-MS analysis of
recombinant proteins, LC-MS/MS and chiral GC-MS analy-
sis.
(9) a) Schupp, T.; Toupet, C.; Engel, N.; Goff, S. FEMS Micro-
biol. Lett. 1998, 159, 201-207. b) August, P. R.; Tang, L.; Yoon, Y.
J.; Ning, S.; Muller, R.; Yu, T. W.; Taylor, M.; Hoffmann, D.;
Kim, C. G.; Zhang, X.; Hutchinson, C. R.; Floss, H. G. Chem.
Biol. 1998, 5, 69-79. c) Tang, L.; Yoon, Y. J.; Choi, C. Y.;
Hutchinson, C. R. Gene 1998, 216, 255-265. d) Gay, D.; You, Y.
O.; Keatinge-Clay, A.; Cane, D. E. Biochemistry 2013, 52, 8916-
8928.
AUTHOR INFORMATION
(10) a) Alhamadsheh, M. M.; Palaniappan, N.; Daschouduri,
S.; Reynolds, K. A. J. Am. Chem. Soc. 2007, 129, 1910-1911. b)
Bonnett, S. A.; Whicher, J. R.; Papireddy, K.; Florova, G.; Smith,
J. L.; Reynolds, K. A. Chem. Biol. 2013, 20, 772-783.
(11) (3R,4E)-3-Hydroxy-4-hexenoic acid (17) has the same
configuration (L) at C-3 as the corresponding (3S)-3-
hydroxybutyric acid (10), reflecting the difference in priority of
the respective propenyl and methyl substituents in the Cahn-
Ingold-Prelog nomenclature.
Corresponding Author
*David_Cane@brown.edu
ORCID
David E. Cane: 0000-0002-1492-9529
Dhara D. Shah: 0000-0002-7533-7586
Young-Ok You: 0000-0002-8036-1345
(12) To prepare 17 and it (3S)-enantiomer, synthetic ( )-ethyl-
(3RS,4E)-3-hydroxy-4-hexenoate was kinetically resolved by
incubation with Amano lipase P and vinyl acetate to give a
Present Address
ACS Paragon Plus Environment

Page 11 of 12
Journal of the American Chemical Society
mixture consisting predominantly of ethyl (3R,4E)-3-acetoxy-4-
(19) Kimber, M. S.; Martin, F.; Lu, Y.; Houston, S.; Vedadi,
1
2
3
4
5
6
7
8
hexenoate and ethyl (3S,4E)-3-hydroxy-4-hexenoate, which
were readily separated by SiO2 chromatography (Scheme S1).
Chiral GC-MS analysis of the derived methyl 3-hydroxy-4-
hexenoate esters established that (3R,4E) alcohol had ~80% e.e.,
while the (3S,4E) enantiomer had ~90% e.e. The absolute con-
figuration of 17 was assigned based on the well-established
preference of Amano lipase for (S)/[L] alcohols. Cf. a) Brem, J.;
Paizs, C.; To•a, M. I.; Vass, E.; Irimie, F. D. Tetrahedron: Asym-
metry 2009, 20, 489-496. b) Ghanem, A.; Aboul-Enein, H. Y. Chi-
rality 2005, 17, 1-15.
M.; Dharamsi, A.; Fiebig, K. M.; Schmid, M.; Rock, C. O. J. Biol.
Chem. 2004, 279, 52593-52602.
(20) Keatinge-Clay, A. J. Mol. Biol. 2008, 384, 941-953.
(21) Schwab, J. M.; Henderson, B. S. Chem. Rev. 1990, 90,
1203-1245.
(22) An alternative two-base mechanism, also discussed by
Schwab, and that has been widely invoked, posits that the Asp
residue serve as the active site acid for the cleavage of the C–O
bond. This mechanism, however, would require that the pKa of
the active site Asp anomalously exceed that of the catalytic His.
(23) Bahnson, B. J.; Anderson, V. E.; Petsko, G. A. Biochemis-
try 2002, 41, 2621-2629.
9
(13) Schmidt, B.; Kunz, O. Eur. J. Org. Chem. 2012, 2012, 1008-
1018.
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
(14) a) Reid, R.; Piagentini, M.; Rodriguez, E.; Ashley, G.;
Viswanathan, N.; Carney, J.; Santi, D. V.; Hutchinson, C. R.;
McDaniel, R. Biochemistry 2003, 42, 72-79. b) Caffrey, P. Chem-
BioChem 2003, 4, 654-657. c) Keatinge-Clay, A. T. Chem. Biol.
2007, 14, 898-908.
(24) a) Valenzano, C. R.; You, Y. O.; Garg, A.; Keatinge-Clay,
A.; Khosla, C.; Cane, D. E. J. Am. Chem. Soc. 2010, 132, 14697-
14699. b) Guo, X.; Liu, T.; Valenzano, C. R.; Deng, Z.; Cane, D.
E. J. Am. Chem. Soc. 2010, 132, 14694-14696.
(25) He, H. Y.; Tang, M. C.; Zhang, F.; Tang, G. L. J. Am.
Chem. Soc. 2014, 136, 4488-4491.
(26) Tang, L.; Ward, S.; Chung, L.; Carney, J. R.; Li, Y.; Reid,
R.; Katz, L. J. Am. Chem. Soc. 2004, 126, 46-47.
(27) a) Albery, W. J.; Knowles, J. R. Biochemistry 1976, 15,
5627-5631. b) Albery, W. J.; Knowles, J. R. Biochemistry 1976, 15,
5631-5640.
(28) Rappe, C. Org. Syn. 1973, 53, 123-127.
(29) Lu, H.; Tsai, S. C.; Khosla, C.; Cane, D. E. Biochemistry
2002, 41, 12590-12597.
(30) a) Valenzano, C. R.; Lawson, R. J.; Chen, A. Y.; Khosla,
C.; Cane, D. E. J. Am. Chem. Soc. 2009, 131, 18501-18511. b) Xie,
X.; Garg, A.; Khosla, C.; Cane, D. E. J. Am. Chem. Soc. 2017, 139,
3283-3292.
(31) Sambrook, J.; Russell, D. W. Molecular Cloning: A Labora-
tory Manual, Third Edition; Cold Spring Harbor Laboratory
Press: Cold Spring Harbor, New York, 2001.
(15) a) Dorrestein, P. C.; Bumpus, S. B.; Calderone, C. T.;
Garneau-Tsodikova, S.; Aron, Z. D.; Straight, P. D.; Kolter, R.;
Walsh, C. T.; Kelleher, N. L. Biochemistry 2006, 45, 12756-12766.
b) Meluzzi, D.; Zheng, W. H.; Hensler, M.; Nizet, V.;
Dorrestein, P. C. Bioorg. Med. Chem. Lett. 2008, 18, 3107-3111.
(16) Analysis of the stereochemistry of the FosDH2-catalyzed
dehydration of (3R,4E)-3-hydroxy-4-hexenoyl-FosACP2 (18)
was complicated by competing buffer-catalyzed dehydration of
(18) to the isomeric (2E,4E)-22, as well buffer-catalyzed isomer-
ization of initially formed (2Z,4E)-21 to 22, as established by
control incubations carried out in the absence of added FosDH2
(cf Figure S17C). Thus the formation of (2Z,4E)-21 was accom-
panied by a time-dependent increase in the proportion of the
isomeric dienoic ester (2E,4E)-22 .
(17) a) Wu, W.-J.; Feng, Y.; He, X.; Hofstein, H. A.; Raleigh,
D. P.; Tonge, P. J. J. Am. Chem. Soc. 2000, 122, 3987-3994. b)
Fausto, R.; Tonge, P. J.; Carey, P. R. J. Chem. Soc. Faraday Trans.
1994, 90, 3491-3503.
(18) a) Schwab, J. M.; Habib, A.; Klassen, J. B. J. Am. Chem.
Soc. 1986, 108, 5304-5308. b) Sedgwick, B.; Morris, C.; French, S.
J. J. C. S. Chem. Commun. 1978, 193-194.
(32) Bradford, M. Anal. Biochem. 1976, 72, 248-254.
(33) Garg, A.; Khosla, C.; Cane, D. E. J. Am. Chem. Soc. 2013,
135, 16324-16327.
(34) Collins, I.; Nadin, A.; Holmes, A. B.; Long, M. E.; Man, J.;
Baker, R. J. Chem. Soc. Perkin Trans. 1 1994, 2205-2215.
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
TOC Graphic
12
ACS Paragon Plus Environment