Characterization of FabG and FabI of the Streptomyces coelicolor dissociated fatty acid synthase

Article abstract of DOI:10.1002/cbic.201402670

Streptomyces coelicolor produces fatty acids for both primary metabolism and for biosynthesis of the secondary metabolite undecylprodiginine. The first and last reductive steps during the chain elongation cycle of fatty acid biosynthesis are catalyzed by FabG and FabI. The S. coelicolor genome sequence has one fabI gene (SCO1814) and three likely fabG genes (SCO1815, SCO1345, and SCO1846). We report the expression, purification, and characterization of the corresponding gene products. Kinetic analyses revealed that all three FabGs and FabI are capable of utilizing both straight and branched-chain β-ketoacyl-NAC and enoyl-NAC substrates, respectively. Furthermore, only SCO1345 differentiates between ACPs from both biosynthetic pathways. The data presented provide the first experimental evidence that SCO1815, SCO1346, and SCO1814 have the catalytic capability to process intermediates in both fatty acid and undecylprodiginine biosynthesis.

Full text of DOI:10.1002/cbic.201402670

DOI: 10.1002/cbic.201402670
Full Papers
Characterization of FabG and FabI of the Streptomyces
coelicolor Dissociated Fatty Acid Synthase
Renu Singh and Kevin A. Reynolds*[a]
Streptomyces coelicolor produces fatty acids for both primary
metabolism and for biosynthesis of the secondary metabolite
undecylprodiginine. The first and last reductive steps during
the chain elongation cycle of fatty acid biosynthesis are cata-
lyzed by FabG and FabI. The S. coelicolor genome sequence
has one fabI gene (SCO1814) and three likely fabG genes
(SCO1815, SCO1345, and SCO1846). We report the expression,
purification, and characterization of the corresponding gene
products. Kinetic analyses revealed that all three FabGs and
FabI are capable of utilizing both straight and branched-chain
b-ketoacyl-NAC and enoyl-NAC substrates, respectively. Fur-
thermore, only SCO1345 differentiates between ACPs from
both biosynthetic pathways. The data presented provide the
first experimental evidence that SCO1815, SCO1346, and
SCO1814 have the catalytic capability to process intermediates
in both fatty acid and undecylprodiginine biosynthesis.
Introduction
Streptomycetes, like many other bacteria, utilize a type II fatty
acid synthase to catalyze the formation of fatty acids. They
produce primarily branched-chain fatty acids with only a minor
proportion of straight-chain fatty acids.[1] The first step in
a type II fatty acid synthase (FAS) process is catalyzed by FabH
(b-ketoacyl synthase III), which catalyzes a decarboxylative con-
densation of an acyl-CoA primer with a malonyl-acyl carrier
protein (ACP; Scheme 1). For branched-chain fatty acids, the
primer is typically either isobutyryl-CoA or methylbutyryl-CoA.
For straight-chain fatty acids, acetyl-CoA and propionyl-CoA
serve as the most common primers. The resulting 3-ketoacyl-
ACP product is reduced by NADPH-dependent FabG to provide
3-hydroxyacyl-ACP, which is dehydrated by FabA to form
enoyl-ACP. The NADH-dependent FabI completes the cycle by
catalyzing a reduction to provide the corresponding acyl-ACP.
In subsequent rounds of elongation, the condensation step
with the acyl-ACP is catalyzed by FabF rather than FabH,
which only catalyzes the chain initiating step. The malonyl-ACP
utilized in each elongation step is generated from malonyl-
CoA by the action of a malonyl CoA:ACP transacylase (encoded
by fabD). In streptomycetes, fabH, fabC (encoding the ACP),
fabF, and fabD are clustered together as an operon.[2]
those made by the primary metabolic type II FAS (e.g., dapto-
mycin, frenolicin, R1128, tunicamycin, and undecylprodigi-
nine).[4–9] The biosynthesis of their alkane chain occurs by a
type II FAS pathway, which must differ from the primary meta-
bolic pathway in the type of acyl precursors and number of
cycles used. One such example is undecylprodiginine, a tripyr-
role red pigmented compound with a linear alkyl chain known
to exhibit a wide range of biological activities, such as antibac-
terial, immunosuppressive, antimalarial, and anticancer activi-
ties.[10,11]
Studies have revealed that the control of the two processes
providing fatty acids for primary and secondary metabolism
lies partially in the initial condensing enzymes.[1,12,13] For unde-
cylprodiginine biosynthesis in Streptomyces coelicolor, homo-
logues of the condensing enzymes (FabH and FabF) and the
ACP (FabC) are encoded by redP, redR, and redQ, respectively in
the red gene cluster. RedP is proposed to initiate biosynthesis
of undecylprodiginine’s alkane chain by condensing an acyl-
CoA with a malonyl-RedQ. The 3-keto group of the resulting
3-ketoacyl-RedQ is then reduced to provide butyryl-RedQ, pre-
sumably by the type II FAS enzymes FabG, FabA, and FabI.
RedR would then catalyze the subsequent condensation steps
with malonyl-RedQ, and type II FAS enzymes shared with pri-
mary metabolism would handle reduction of the 3-keto group
during each elongation cycle. Recently, we have demonstrated
that FabH and its homologue, RedP, which catalyze the initial
condensation in fatty acid and undecylprodiginine biosynthe-
sis, respectively, are highly selective for either straight- (RedP)
or branched-chain (FabH) acyl-CoA substrates.[14] Specifically,
RedP will only process acetyl-CoA, whereas FabH is more effi-
cient with branched acyl-CoA substrates. Additionally, both en-
zymes have been shown to possess differing ACP specificities:
RedP only processes RedQ, whereas FabH is more efficient
with FabC. This combination of acyl-CoA and ACP specificity
Streptomycetes also produce a vast array of biologically
active secondary metabolites widely used in human health,
such as for antibacterial, immunosuppressant, anticancer, and
antimalarial purposes,[3] and in a number of cases, these natu-
ral products use fatty acids as a building block. These fatty
acids often differ in chain length or extent of branching from
[a] R. Singh, K. A. Reynolds
Finance and Administration, Portland State University
P.O. Box 751, Portland, OR 97207 (USA)
E-mail: reynoldk@pdx.edu
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cbic.201402670.
ChemBioChem 0000, 00, 0 – 0
1
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Scheme 1. Interface between fatty acid and undecylprodiginine biosynthesis in S. coelicolor.
provides for a separation of the two processes and control of
the final product at the initiation step.
ploying the reverse reaction, that is, oxidation of 3-hydroxy-
acyl-ZhuG to 3-oxoacyl-ZhuG (ACP from R1128, an aromatic
polyketide). The activity of this enzyme had not been previous-
ly measured with the cognate ACP (FabC or RedQ), and the
specificity for the acyl group had not been determined. We
posited that 1) FabI (encoded by SCO1814) and the putative
FabG (encoded by SCO1815) would have enoyl-ACP reductase
and 3-ketoacyl-ACP reductase activity, respectively, and 2) that
neither would show selectivity for the ACP substrate (FabC or
RedQ) or the branching in the acyl chain. This activity would
permit them to process the acyl chains for both primary and
secondary metabolism and would contrast the initiation
enzyme selectivity. We also posited that SCO1815 would have
3-ketoacyl-ACP reductase activity, as it clusters with FabI similar
to other organisms such as Mycobacterium tuberculosis and
Mycobacterium smegmatis.[16] Accordingly, the catalytic activity
and substrate specificity of the FabG homologues (encoded by
SCO1345 and SCO1346) might be dramatically different if their
biological roles are unrelated to fatty acid biosynthesis.
Separate enzymes and dedicated ACPs thus appear to be
key for separating the primary and secondary fatty acid biosyn-
thetic processes. As only these genes are found in the secon-
dary metabolite gene cluster, the remaining enzymes to com-
plete each cycles are likely to be shared. To date, most work
on this metabolic crosstalk has focused on FabD (malonyl-
CoA:ACP transacylase), the most well-established and wide-
spread example of a crosstalk between fatty acid biosynthesis
(a primary metabolic process) and polyketide biosynthesis (a
secondary metabolic process).[15] The gene for fabD is clustered
with fabH, fabF, and fabC.[2] Studies have revealed that FabD
has no significant ACP selectivity and plays an important role
in both natural product and fatty acid biosynthetic processes.
The genes encoding 3-ketoacyl-ACP reductase (FabG), 3-hy-
droxyacyl-ACP dehydratase (FabA), and enoyl-ACP reductase
(FabI), putatively shared between primary and secondary fatty
acid synthesis, have not been characterized in Streptomyces
(Scheme 1). The corresponding genes are not present within
the streptomycetes’s FAS gene cluster. Analysis of the S. coeli-
color genome sequence revealed the presence of one fabI
gene (SCO1814, encoding an enoyl-ACP reductase), and three
potential fabG genes (SCO1815, SCO1345, and SCO1846, en-
coding b-ketoacyl-ACP reductase).[16] The putative FabI has not
been previously characterized, and of the three possible candi-
dates for fabG, only one (SCO1815, which lies immediately ad-
jacent to fabI) has been partially characterized.[17] In this study,
the structure and functional role of SCO1815 was determined.
The ketoreductase activity of SCO1815 was monitored by em-
Here we report the characterization of the three FabG and
FabI homologues. Kinetic studies demonstrated that S. coelicol-
or FabI and FabG, encoded by SCO1814 and SCO1815, process
both straight and branched acyl chains in their respective sub-
strates. In addition, both enzymes can utilize either FabC or
RedQ as the ACP components of the substrate. These findings
are consistent with predictions. Surprisingly, 3-ketoacyl-ACP ac-
tivity was also observed with the FabG homologues encoded
by SCO1345 and SCO1346. However, differences in cofactor
specificity, overall activity, and ACP selectivity were observed
over that for the SCO1815 FabG.
&
ChemBioChem 0000, 00, 0 – 0
www.chembiochem.org
2
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Results and Discussion
Table 1. Kinetic data of S. coelicolor SCO1815, SCO1345, and SCO1346,
and E. coli FabG
b-Ketoacyl-ACP reductase (FabG)
Substrate
k
cat/Km [mmÀ1 minÀ1
SCO1815 SCO1345 SCO1346 E. coli FabG
]
Analysis of the S. coelicolor genome sequence revealed the
presence of three likely fabG genes (SCO1815, SCO1345, and
SCO1346, encoding b-ketoacyl-ACP reductase). To probe the
physiological role of these three S. coelicolor fabG homologues,
the genes were expressed in Escherichia coli, and the corre-
sponding recombinant proteins were purified as a soluble N-
terminal His6-tag protein. S. coelicolor produces predominantly
branched-chain fatty acids, whereas E. coli produces only
straight chain fatty acids. In order to facilitate a comparison of
S. coelicolor ketoacyl-ACP reductase acyl specificity with that of
a straight-chain fatty acid producer, the E. coli fabG gene was
also expressed and purified. Analysis of affinity purified pro-
teins by SDS-PAGE showed a major band with an apparent
molecular weight of approximately 28 kDa consistent with the
expected molecular weight. On the basis of earlier studies
which demonstrated that FabH (b-ketoacyl-ACP synthase III)
from E. coli is specific for acetyl-CoA,[12] we hypothesized that
E. coli FabG would be specific for straight-chain substrates. De-
spite the important role of FabG in FAS, analysis of specificity
for straight- and branched-chain substrates has not been in-
vestigated. Thus, a range of straight- and branched-chain car-
boxylic acids were synthesized and activated as either a NAC
or ACP derivatives and were tested with the three S. coelicolor
FabG homologues and E. coli FabG.
Straight-chain
3-ketohexanoic-NAC
3-ketooctanoic-NAC
3-ketodecanoic-NAC
Branched-chain
5-methyl-3-ketohexanoic-
NAC
38Æ1.5 16Æ1.2 160Æ1.0
280Æ17.0
93Æ9.0 118Æ6.0 205Æ7.0 1912Æ133.0
75Æ3.0 46Æ1.0
11 Æ1.1 91Æ8.0
92Æ5.0 1375Æ32.0
15Æ3.0
36Æ4.0
6-methyl-3-ketoheptanoic- 92Æ5.0 123Æ6.0 292Æ5.0
703Æ14.0
NAC
7-methyl-3-ketooctanoic- 122Æ2.0 385Æ48.0 350Æ47.0 1293Æ20.0
NAC
acid by FabG and potentially reflect greater occupancy of the
appropriate acyl binding pocket in the overall catalytic process.
FabG from E. coli showed an apparent efficiency several times
higher than that of the S. coelicolor FabGs. This difference in
catalytic activity could be innate or a result of assay conditions
with substrate mimics. Finally, the data demonstrate, for the
first time, that FabG from E. coli can utilize both straight- and
branched-chain b-ketoacyl substrates with comparable efficien-
cy. This observation is in contrast to the data with E. coli FabH,
which has been shown to have a strong preference for
straight-chain substrates such as acetyl and propionyl-CoA.[12]
Although E. coli does not generate branched-chain fatty acids,
production of small levels of them has been observed when
the natural FabH is replaced with FabH which can process
branched acyl-CoA starter units.[13,18] This production is pre-
sumably dependent upon the relaxed acyl group specificity of
E. coli FabG.
b-Ketoacyl-NAC specificity
The activity of FabGs with a series of straight- and branched-
chain b-ketoacyl-NAC substrates was determined by using a
NADPH-based spectrophotometric assay. The apparent Km
values for all NAC substrates were greater than 2 mm (the high
apparent Km values can be attributed to using NAC thioesters
in place of the natural ACP substrate), and limited substrate
solubility prevented determination of Vmax and kcat values for
these substrate mimics. Limited solubility also prevented
assays with substrate mimics with chain lengths greater than
C10. Therefore, the catalytic efficiency (kcat/Km) was determined
under substrate-limited conditions. Kinetic analysis revealed
that all three FabG homologues were capable of utilizing both
straight- and branched-chain substrates (Table 1). For the strep-
tomycetes FabG homologues, this result is consistent with a
FabG role in fatty acid and prodiginine biosynthesis, wherein it
processes branched-chain products of FabH for primary metab-
olism, as well as straight-chain products of RedP for undecyl-
prodiginine biosynthesis. Furthermore, all FabGs were also ob-
served to process b-ketoacyl-NAC substrates with a range of
alkyl chain lengths C6ÀC10, and catalytic efficiency generally in-
creased as substrate chain length increased. This general trend
was observed for all FabGs with branched-chain substrates. In
the case of the straight-chain substrates, the greatest activity
was observed with the C8 substrates. The higher catalytic effi-
ciency with longer chain substrates corresponds to the average
chain length of the substrates that are processed in a fatty
The cofactor preference of all of the enzymes was also deter-
mined by using both NADPH and NADH. SCO1815, SCO1345,
and E. coli FabG all demonstrated at least 10-fold higher cata-
lytic activity with NADPH than NADH, consistent with previous
studies of FabG proteins.[19–21] In contrast, SCO1346 was highly
specific for NADH and, in this case, the catalytic activity was at
least 100-fold higher than with NADPH (data not shown).
ACP specificity
As the acyl-NAC is a poor mimic of the natural substrate for
FabG, activity of the latter was also determined with acyl-ACPs
in an LC/MS assay. S. coelicolor SCO1815 was initially assayed
by employing the reverse reaction, that is, oxidation of 3-hy-
droxyacyl-ZhuG to 3-oxoacyl-ZhuG (ZhuG is the ACP used in
the biosynthesis of the polyketide R1128, a non-native path-
way which has been introduced into S. coelicolor). This ap-
proach was taken to avoid problems associated with 3-keto-
acyl-ACP preparation and stability.[17] Studies have shown that
the acyl group specificity of an enzyme can change when
using non-native ACPs.[22] Our approach was a) to generate the
3-ketoacyl-ACP enzymatically and then use it immediately to
assay the reaction in the forward direction, and b) to test the
S. coelicolor FabG homologues with the physiologically relevant
ChemBioChem 0000, 00, 0 – 0
www.chembiochem.org
3
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
acyl-ACPs (RedQ and FabC). The phosphopantetheine ejection
assay was carried out to monitor the reaction according to
a previously described method.[23] Although e a detailed kinetic
analysis was not possible because of limitations of acyl-ACP
substrate production, a broad assessment of acyl and ACP spe-
cificity was possible for the all the FabG proteins.
droxyacyl-FabC product (Figure 1D–F for SCO1815, SCO1345,
and SCO1346, respectively). The level of reduction with
SCO1345 could only be achieved under the same assay condi-
tions by using tenfold higher levels of the enzyme. This level
of difference in catalytic efficiency was not observed when the
acyl-NAC substrate mimics were used. Assays were also carried
out with branched-chain 3-ketoacyl-FabC. These were generat-
ed by using isobutyryl-CoA (in place of butyryl-CoA) and ma-
lonyl-FabC with FabH (Figure 1G–I). These assays revealed con-
version of the 3-ketoacyl-FabC with all three FabG homologues
(Figure 1J–L). The lower catalytic efficiency for SCO1345 was
observed with both the straight- and branched-chain 3-keto-
acyl-FabC substrates. The ability to process both straight and
branched acyl-ACP substrates and the cofactor specificity of
the FabG homologues are consistent with the observations on
3-ketoacyl- NAC substrate mimics.
Initially, FabH was incubated with butyryl-CoA and malonyl-
FabC to generate the straight-chain b-ketoacyl-FabC, and the
product was analyzed with LC/MS (Figure 1A). A major m/z of
952 was observed, corresponding to the major mass-charge
ratio for the condensed ACP bound product of FabH reaction
(Figure 1B). A MS2 of m/z 952 produced a major m/z signal of
373, consistent with the mass of a phosphopantetheine thio-
ester of the expected 3-keto condensed product of FabH (Fig-
ure 1C). After addition of NADPH (or NADH) and all three
FabGs to the b-ketoacyl-FabC product of FabH, the same anal-
ysis provided a major m/z of 375 (an increase in 2 atomic mass
units), which is consistent with the expected mass of the b-hy-
Additionally, activity of each FabG was assayed with the 3-
ketoacyl-ACP substrates by using both the FAS ACP (FabC) and
Figure 1. Coupled assay of FabH and FabG. Left: A) LC/MS extracted ion chromatogram of the FabH reaction with butyryl-CoA and malonyl-FabC. B) Total ion
chromatogram; m/z of 952 corresponds to the major mass-charge ratio observed for ACP-bound 3-keto condensed product of FabH reaction. C) MS2 spectra
of 952 (373 corresponds to a phosphopantetheine thioester of expected 3-keto condensed product of FabH), and of the hydroxy product with D) SCO1815,
E) SCO1345, and F) SCO1346. In comparison to SCO1815 and SCO1346, a tenfold higher concentration of SCO1345 was necessary to obtain the complete
conversion of 3-ketoacyl substrate to a 3-hydroxyacyl product. Right: G) LC/MS–MS extracted ion chromatogram of the FabH reaction with isobutyryl-CoA
and malonyl-FabC. H) Total ion chromatogram of the FabH reaction product. I) MS2 spectra of 952. Phosphopantetheine fragment of hydroxy product of
J) SCO1815, K) SCO1345, and L) SCO1346.
&
ChemBioChem 0000, 00, 0 – 0
www.chembiochem.org
4
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
the prodiginine ACP (RedQ; Figure 2). The expectation was
that both ACP substrates would be used if FabG is involved in
both processes. These assays were carried out in an analogous
fashion to that used to look at straight- and branched-chain
specificity. The phosphopantetheine fragment of the starting
substrate, malonyl-FabC, exhibited an expected mass of 347
(Figure 2A). The b-ketoacyl-chain substrates were generated
by using butyryl-CoA and malonyl-FabC with FabH and, after
addition of FabG, the expected reduced product was observed
(Figure 2C–F for SCO1815, SCO1345, and SCO1346 and E. coli
FabG, respectively). Similarly, the b-ketoacyl-RedQ substrate
was generated by the reaction of RedP with butyryl-CoA and
malonyl-RedQ and was analyzed by LC/MS. The expected
phosphopantetheine fragment of the starting substrate, ma-
lonyl-RedQ, was exhibited in the LC/MS assay (Figure 2G). Simi-
lar to the FabH product, a condensed product of RedP (b-keto-
acyl-RedQ) was observed after incubation with malonyl-RedQ
and butyryl-CoA (Figure 2H). Again, a reduced product was
obtained after addition of FabG (Figure 2I, K and L, SCO1815,
SCO1346 and E. coli FabG). In contrast, SCO1345 was not active
with the RedQ substrates (Figure 2J). These data demonstrate
that, of the three S. coelicolor FabGs enzymes, SCO1815 and
SCO1346 have capability to process the ACPs from both fatty
acid and undecylprodiginine biosynthesis. This is consistent
with one or both of these FabGs processing 3-ketoacyl-FabC
substrates for primary metabolism and 3-ketoacyl-RedQ sub-
strates for secondary metabolism. The ability of E. coli FabG to
process 3-ketoacyl-RedQ (Figure 2L) demonstrates that this
enzyme can tolerate changes in both the acyl and ACP compo-
nents of the 3-ketacyl-ACP substrate.
These analyses are consistent with the predictions for the
activity and likely role of the FabG encoded by SCO1815. Nota-
bly, it uses both straight- and branched-chain 3-keotacyl-ACP
substrates with either FabC or RedQ. These observations, in ad-
dition to the NADPH cofactor specificity and colocation of the
gene with SCO1814 (and, detailed below, the enoyl-ACP reduc-
tase activity required for fatty acid biosynthesis), all support
the hypothesis that this is the primary FabG used for both
fatty acid and undecylprodiginine biosynthesis. Colocation of
genes for FabG and FabI are also observed in other organisms,
Figure 2. Coupled assay of FabH and FabG with butyryl-CoA and malonyl-FabC (or malonyl-RedQ). Left: Phosphopantetheine fragment of A) starting substrate
malonyl-FabC, B) 3-keto product of FabH reaction with butyryl-CoA and malonyl-FabC, reduced product of C) SCO1815, D) SCO1345, and E) SCO1346, and
F) E. coli FabG. Right: Phosphopantetheine fragment of G) malonyl-RedQ, H) 3-keto product of FabH reaction with butyryl-CoA and malonyl-RedQ, reduced
product of I) SCO1815, K) SCO1346, and L) E. coli FabG; J) no product was observed with SCO1345.
ChemBioChem 0000, 00, 0 – 0
www.chembiochem.org
5
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
notably mycobacterium.[16] Given all of this and the literature,
which to date supports the role of a single FabG in a type II
dissociated fatty acid biosynthetic process,[16] we were sur-
prised to see 3-ketoacyl-NAC and 3-ketoacyl-ACP activity with
FabG encoded by SCO1345 and SCO1346. The markedly lower
activity of the FabG encoded by SCO1345 (compared to the
other two FabG homologues) when 3-ketoacyl-ACP substrates
were used indicates that these are not the physiological sub-
strates, and that this reductase is involved in another cellular
process. The same cannot be said for the FabG encoded by
SCO1346, which differs from the SCO1815 encoded FabG by
cofactor specificity. The physiological role of both of these
FabG homologues thus remains elusive.
value for all enoyl-NAC substrates was greater than 2 mm. Sub-
strate solubility prevented the use of a longer substrate (great-
er than C10) and a concentration well above the Km value.
Therefore, the apparent catalytic efficiency (kcat/Km) values were
obtained under substrate-limited conditions, as in the case of
FabG. The high Km value likely reflects the use of NAC thioes-
ters in place of ACPs. Nonetheless, these kinetic analyses dem-
onstrated that S. coelicolor FabI has the capability of utilizing
all of the straight- and branched-chain enoyl-NAC substrates
tested (C4ÀC10; Table 2). No clear pattern was observed be-
Table 2. Kinetic data of S. coelicolor and E. coli FabI.
Substrates
kcat/Km [mmÀ1 minÀ1
]
S. coelicolor FabI
E. coli FabI
Enoyl-ACP reductase (FabI)
Straight-chain
but-2-enoic-NAC
hex-2-enoic-NAC
oct-2-enoic-NAC
dec-2-enoic-NAC
Branched-chain
4-methylpent-2-enoic-NAC
5-methylhex-2-enoic-NAC
6-methylhept-2-enoic-NAC
7-methyloct-2-enoic-NAC
9-methyldec-2-enoic-NAC
Analysis of the S. coelicolor genome sequence has revealed the
presence of one putative fabI (SCO1814). In order to test the
hypothesis that SCO1814 encodes the enoyl-ACP reductase
used in fatty acid and undecylprodiginine biosynthesis, fabI
(SCO1814) was amplified from S. coelicolor genomic DNA, ex-
pressed in E. coli, and purified. The purified protein had the
expected molecular mass of approximately 30 kDa, as deter-
mined by SDS-PAGE. In addition to SCO1814, the previously
characterized E. coli fabI gene[24,25] was also expressed and puri-
fied (also with an expected molecular mass of approximately
30 kDa). As E. coli generates only straight-chain fatty acids, we
hypothesized that FabI from E. coli would have a strong prefer-
ence for straight-chain substrates (as in the case of E. coli
FabH). To date, no data has compared FabI activity with
straight- and branched-chain substrates. Therefore, it would be
useful to compare this FabI acyl specificity with that of the
S. coelicolor FabI. In this study, a series of straight- and
branched-chain substrates varying in chain length were syn-
thesized (Scheme 2) and used in the spectrophotometric and
LC/MS assay.
10Æ1.9
19Æ1.0
49Æ1.0
20Æ1.2
97Æ6.0
231Æ12.0
611Æ15.0
157Æ27.0
17Æ1.5
2Æ0.1
12Æ2.0
0.37Æ0.05
302Æ6.0
361Æ19.0
10Æ1.0
29Æ2.0
63Æ4.0
9.0Æ1.0
tween catalytic efficiencies for straight and branched enoyl-
NAC substrates of comparable chain length. These observa-
tions are consistent with the predicted activity of FabI and its
role in both fatty acid and prodiginine biosynthesis
(Scheme 1), wherein it processes both short- and long-chain
products of FabH (or RedP) or FabF (or RedR), respectively. The
catalytic efficiency of FabI with various straight-chain enoyl-
NAC (chain length of C4 to C10) generally increased as substrate
chain length increased, with a general trend of: oct-2-enoyl-
NAC>dec-2-enoic-NAC>hex-2-enoic-NAC>but-2-enoic-NAC.
A similar trend was also observed with the branched-chain
substrates. The preference for processing longer straight-chain
enoyl-substrates has previously been observed and is, again,
likely the result of greater occupancy in the enoyl-group bind-
ing pocket of the enzyme.[26,27] Similar observations of process-
ing both straight- and branched-chain enoyl-NAC substrates
were seen with the E. coli FabI (Table 2). Thus, in both organ-
isms that produce branched-chain fatty acids and those which
produce only straight-chain fatty acids, the reductive enzymes
FabI and FabG can process both straight- and branched-chain
substrates. The apparent catalytic efficiency of E. coli FabI was
several-fold higher than S. coelicolor FabI, an observation made
previously when E. coli FabI was compared with FabI from
other enzymes.[28] It has also been observed that enoyl-reduc-
tase (FabI) from other organisms prefers CoA over NAC thioest-
ers, even though CoA is not the physiological substrate, and
that NADH is the preferred cofactor.[28] Similar preferences
were observed with the S. coelicolor FabI in this study and, in
fact, no activity was observed with NADPH, even when
enzyme concentration was increased 20-fold.
Scheme 2. General scheme for synthesis of (E)-S-(2-acetamidoethyl) alk-2-
enethioate (enoyl-NAC). a) oxalyl chloride, DMSO, CH2Cl2, À788C, 2.5 h;
b) Methyl diethylphosphonoacetate, NaH, 1,2-dimethoxyethane, 08C to RT,
3 h; c) LiOH, THF/H2O (1:1), reflux, 3 h; d) N-acetyl cysteamine, EDCl, DMAP,
CH2Cl2, RT, 12 h.
Enoyl-NAC specificity
The activity of FabI with a series of straight- and branched-
chain enoyl-NAC substrates (C4ÀC10) was determined in an
NADH-dependent spectrophotometric assay. The apparent Km
&
ChemBioChem 0000, 00, 0 – 0
www.chembiochem.org
6
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
ACP specificity
These assays clearly demonstrate that SCO1814 (FabI) has
both enoyl-NAC and enoyl-ACP reductase activity. In both
cases, FabI utilizes various chain lengths of straight- and
branched-chain substrates. Furthermore, FabI does not discrim-
inate between ACPs from primary metabolism and secondary
metabolism. A gene encoding the enoyl-ACP reductase (FabI)
is not present in the red biosynthetic gene cluster, and thus, all
evidence support a role for SCO1814 in both fatty acid and
undecylprodiginine biosynthetic processes in S. coelicolor
(Scheme 1). These results provide an explanation for a previous
observation, wherein it was shown that FabI inhibitors de-
crease the biosynthetic yield of undecylprodiginine, containing
a fatty acid chain generated by primary metabolism. The bio-
synthetic yield was not decreased for secondary metabolites
not containing fatty acid-derived moieties.[29]
Similar to FabG, an LC/MS assay was carried out to determine
the FabI specificity with the ACPs from both metabolic pro-
cesses (Figure 3). In this study, an AcpP from E. coli fatty acid
synthase was used because of technical difficulties in generat-
ing a sufficient amount of apo-FabC. Crotonoyl-AcpP was gen-
erated from apo-AcpP (the E. coli fatty acid synthase ACP) and
crotonoyl-CoA by using Sfp (a phosphopantetheinyl transfer-
ase) and analyzed by LC/MS (Figure 3A). An m/z of 1115 was
observed, corresponding to the mass-charge ratio observed for
crotonoyl-AcpP (Figure 1B). An MS2 of the parent (m/z 1115)
provides a fragment of m/z 329, which is consistent with the
expected mass of a phosphopantetheine thioester of the cro-
tonoyl group (Figure 3C). The FabI catalyzed the reduction of
crotonoyl-AcpP by NADH to the saturated product, butyryl-
AcpP, with the predicted mass of 331 observed (Figure 3D).
Similarly, crotonoyl-RedQ was generated to evaluate the role of
FabI in secondary metabolism by the incubation of Sfp with
apo-RedQ and crotonoyl-CoA (Figure 3E–G). Consistent with
FabI’s role in secondary metabolism, it was able to process cro-
tonoyl-RedQ into the saturated product (Figure 3H).
Conclusions
In summary, undecylprodiginine biosynthesis involves numer-
ous intriguing enzymes and a fascinating interface with fatty
acid biosynthesis. We cloned all three 3-ketoacyl-ACP reductas-
es homologues (SCO1815, SCO1345, and SCO1346), along with
Figure 3. LC/MS analysis of FabI reaction with crotonoyl-AcpP (or crotonoyl-RedQ) and NADH. Left: A) Mass spectrum of the starting substrate, crotonoyl-
AcpP. B) Total ion chromatogram, m/z. C) MS2 spectra of 1115 (329 corresponds to a phosphopantetheine thioester of a crotonoyl group). D) Saturated prod-
uct of FabI with crotonoyl-AcpP and NADH. Right: E) Mass spectrum of crotonoyl-RedQ. F) m/z of crotonoyl-RedQ. G) MS2 spectra of 1016. H) Saturated prod-
uct of InhA.
ChemBioChem 0000, 00, 0 – 0
www.chembiochem.org
7
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
The resulting E. coli cell pellets were suspended in lysis buffer A
(50 mm sodium phosphate buffer pH 7.2, 300 mm NaCl, 5 mm 2-
mercaptoethanol, 10% glycerol, 0.05% (v/v) Tween-20) with 10 mm
imidazole and lysozyme (1 mgmLÀ1). The resulting cell suspension
was incubated on ice for 30 min, then the cell lysate was cleared
by centrifugation at 16000 g for 20 min. The crude cell extract was
loaded onto a Ni-NTA resin column. The N-terminal polyhistidine-
tagged protein was eluted by using buffer A with imidazole
(300 mm). Fractions containing pure protein were pooled, ex-
changed with 50 mm sodium phosphate buffer pH 7.2, and stored
in 20% glycerol at À808C.
the SCO1814 enoyl-ACP reductase homologue from S. coelicol-
or, and purified all proteins. We carried out the first kinetic
analyses of these streptomycetes enzymes and demonstrated
that all three FabGs homologues have 3-ketoacyl-ACP reduc-
tase activity, with the FabI homologue having enoyl-ACP re-
ductase activity. The activities of the SCO1815-encoded FabG
and the SCO1814 FabI provide compelling support for a role
of these enzymes in providing both fatty acids for primary
metabolism and dodecanoic acid for undecylprodiginine bio-
synthesis (Scheme 1). Thus, the data now provide experimental
evidence to support the role of FabG and FabI, in addition to
FabD, in both processes and suggests they might exert no
control over the products. In contrast, the condensing en-
zymes (FabH and RedP) are the key controlling factors.[14]
General procedures for synthesis of (E)-S-(2-acetamidoethyl) alk-
2-enethioate (enoyl-NAC thioester) substrates for FabI assays:
Straight- and branched-chain enoyl-NAC compounds, ranging from
C4 to C10 in alkyl chain length, were synthesized according to gen-
eral Scheme 2. Reactions were monitored by TLC, and all com-
pounds were characterized by 1H NMR spectroscopy.
Synthesis of alkanal 1: Dimethyl sulfoxide (6.2 mL, 88 mmol) was
added dropwise to a stirred solution of oxalyl chloride (3.80 mL,
44 mmol) in anhydrous CH2Cl2 (100 mL) at À788C under argon.
After 30 min, a solution of the starting alcohol (3.0 g, 29 mmol)
was added dropwise, and the reaction mixture was stirred at
À788C for 1.5 h. Triethyl amine (20.45 mL, 147 mmol) was then
added dropwise, and the reaction mixture was allowed to warm to
room temperature. Saturated aqueous NH4Cl (30 mL) was added,
and the organic layer was separated. The resulting aqueous phase
was extracted with CH2Cl2 (3ꢁ40 mL), and the combined organic
layers were dried over anhydrous sodium sulfate (Na2SO4) and con-
centrated in vacuo to give aldehyde 1.
Experimental Section
Materials: b-NADH and b-NADPH were from Research Products In-
ternational Corp. All other chemicals, including unlabeled CoA de-
rivatives, were from Sigma. Cosmid 3F7 and 4A7, containing S. coe-
licolor genomic DNA, were kindly provided by the John Innes Insti-
tute. FabC, RedQ, and E. coli ACP were sourced as described previ-
ously.[1,30,31] Preparation of malonyl-ACPs and crotonoyl-ACPs was
carried out as described previously.[14]
S. coelicolor and E. coli fabG: The genes encoding SCO1815,
SCO1345, and SCO1346 (fabG) were amplified from the appropriate
S. coelicolor cosmids by PCR by using the primers 5’-CATATG
AGCCGC TCGGTT CTC-3’ and 5’-GGATCC TCAGTG ACCCAT TCCCAG
TCC-3’ (SCO1815); 5’-CATATG ACTGAA CTGCCC GAGCCC TCC-3’
and 5’-GGATCC TCAGGC CACGCC GGCGTT C-3’ (SCO1345); and 5’-
CATATG TCCACC ACTGAG CAGCG-3’ and 5’-GGATCC CTAGTC
GAGCGG TCCGCC-3’ (SCO1346). The E. coli fabG was amplified
from E. coli genomic DNA by PCR by using 5’-CATATG AATTTT
GAAGGA AAAATC GCACTG-3’ and 5’-GGATCC TCAGAC CATGTA
CATCCC GCCG-3’ (restriction sites underlined). The amplified inserts
were digested with NdeI and BamHI and cloned into the pET15b
vector to provide the expression plasmid for SCO1815 (pRS1),
SCO1345 (pRS2), and SCO1346 (pRS3), as well as E. coli fabG
(pRSE1).
Synthesis of (E)-methyl alk-2-enoate (2): Methyl diethylphospho-
noacetate (5.86 g, 27 mmol) was added dropwise to a stirred solu-
tion of the sodium hydride (0.66 g, 27 mmol) in 1.2-dimethoxy-
ethane (15 mL) at 08C under argon atmosphere and stirred for
30 min. After dropwise addition of the aldehyde (1; 2.0 g,
23 mmol), the reaction mixture was allowed to warm to room tem-
perature and stirred for 3 h. The reaction mixture was quenched
with cold water and extracted with EtOAc (3ꢁ15 mL). The com-
bined organic layers were washed with brine (20 mL), dried over
Na2SO4, and concentrated in vacuo. The crude product was purified
by column chromatography (EtOAc/hexane, 7:93) to yield the
desired ester 2.
Synthesis of (E)-alk-2-enoic acid (3): LiOH (1.0 g, 42.2 mmol) was
added to a stirred solution of ester 2 (2.0 g, 14 mmol) in tetrahy-
drofuran/H2O (1:1, 40 mL), and the mixture was heated at reflux.
After 2.5 h, the reaction mixture was cooled to 08C and acidified
with 1n HCl. The layers were separated, and the aqueous phase
was extracted with EtOAc (3ꢁ30 mL). The combined organic layers
were washed with brine, dried over Na2SO4, and concentrated in
vacuo to give the corresponding carboxylic acid (3).
S. coelicolor and E. coli fabI: The SCO1814 gene (encoding FabI)
was amplified from the appropriate S. coelicolor cosmid by using
the forward primer 5’-CATATG AGCGGA ATTCTC GAGGGC AAG-3’
(introducing a NdeI restriction site) and a reverse primer 5’-
GGATCC TCAGGC GCCGAT GGCGTG C-3’ (introducing a BamHI re-
striction site). The E. coli fabI was amplified from E. coli genomic
DNA by using the forward primer 5’-CATATG GGTTTT CTTTCC
GGTAAG-3’ (introducing a NdeI restriction site) and reverse primer
5’-GGATCC TTATTT CAGTTC GAGTTC G-3’ (introducing a BamHI re-
striction site). The resulting PCR products were digested with NdeI
and BamHI and ligated into the corresponding sites in pET15b to
provide the S. coelicolor fabI expression plasmid, pRS4, and the
E. coli fabI expression plasmid, pRSE2.
Synthesis of (E)-S-(2-acetamidoethyl) alk-2-enethioate (4): N-
Acetyl cysteamine (0.94 mL, 8.98 mmol) was added to a stirred so-
lution of carboxylic acid 3 (1.0 g, 7.81 mmol) in anhydrous CH2Cl2
(20 mL) at 08C under nitrogen atmosphere, followed by 4-(N,N-di-
methylamino)pyridine (DMAP; 0.24 mg, 1.95 mmol), and N-(3-dime-
thylaminopropyl)-N’-ethylcarbodiimide hydrochloride (EDCl; 1.5 g,
7.81 mmol). The reaction mixture was allowed to warm to room
temperature and stirred overnight. The reaction was quenched
with saturated aqueous NH4Cl (15 mL) and extracted with CH2Cl2
(3ꢁ20 mL). The combined organic layers were dried over Na2SO4,
concentrated in vacuo, and the residue was purified by column
chromatography (hexane/EtOAc, 40:60) to yield desired product 4.
Protein expression and purification: All plasmids were transform
into E. coli BL21(DE3) cells. The resulting transformants were grown
at 378C in LB medium containing ampicillin (100 mgmLÀ1) until
A600 =0.6 was reached, induced with isopropyl-b-d-thiogalactopyra-
noside (0.1 mm), and incubated for 3 h at 378C. Cells were harvest-
ed by centrifugation for 10 min at 12000g and 48C and stored at
À808C.
&
ChemBioChem 0000, 00, 0 – 0
www.chembiochem.org
8
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
General procedures for the synthesis of S-(2-acetamidoethyl) 3-
oxoalkanethioate (3-ketoacyl-NAC thioester) substrates for FabG
assays: Straight and branched 3-keto acyl-NAC compounds, with
alkyl chain lengths ranging from C4 to C10, were synthesized ac-
cording to general Scheme 3. Reactions were monitored by TLC,
was added to precipitate the chromium salts. After filtration
through a silica gel pad, the solvent was evaporated to yield the
crude keto compound, which was purified by column chromatog-
raphy (hexane/EtOAc 50:50) to afford desired compound 9.
FabG and FabI assays: Spectrophotometric assays were used to
assess the substrate specificity of FabG and FabI. FabG activity was
determined by monitoring the conversion of the 3-ketoacyl-NAC
substrate to 3-hydroxyacyl-NAC. The conversion of the enoyl-NAC
substrate to acyl-NAC was monitored to determine the activity of
FabI. Assays were performed in sodium phosphate buffer (50 mm,
pH 7.2), NADH (or NADPH 0.4 mm),substrate (1.0 mm, suspended in
DMSO to give a final concentration of 1%), and protein (0.05 mg) in
a final volume of 100 mL. The reaction mixture was incubated at
room temperature for 10 min, and enzymatic activity was mea-
sured by monitoring the decrease in absorbance at 340 nm due to
the consumption of NADH (or NADPH). Steady-state kinetic meas-
urements were obtained by determining FabG or FabI activity at
various concentrations (0.125–16.0 mm) of either 3-ketoacyl-NAC or
enoyl-NAC and a constant concentration (400 mm) of either NADPH
or NADH, respectively. All reactions were performed at least in trip-
licate, and nonlinear regression analysis with Grafit version 4.012
(Erithacus Software, Horley, United Kingdom) was used to deter-
mine kcat and Km values.
1
and all compounds were characterized by H NMR spectroscopy.
FabG and FabI activities were also determined with 3-ketoacyl-ACP
and enoyl-ACP, respectively, in an LC-MS-based assay under similar
reaction conditions as described above. Briefly, FabI was incubated
with NADH and either crotonoyl-AcpP (the E. coli fatty acid syn-
thase ACP) or crotonoyl-RedQ in a final volume of 20 mL for
10 min. Coupled assays were carried out to obtain the FabG activi-
ty. First FabH (or RedP) was incubated with butyryl-CoA (or iso-
butyryl-CoA) and with malonyl-FabC (or malonyl-RedQ) in a similar
way as described before[14] (50 mm sodium phosphate buffer
(pH 7.2); total volume of 20 mL). The reaction mixture was incubat-
ed at 258C for 10 min (leading to generation of the 3-ketoacyl
ACP) and then treated with FabG and either NADPH or NADH for
10 min. All reactions were quenched with 10% formic acid (20 mL),
and products were analyzed by LC-ESI-MS, performed with a high-
resolution mass spectrometer externally mass-calibrated prior to
analysis to obtain a mass accuracy within Æ5 ppm. Chromatogra-
phy was performed with an Accela HPLC system (Thermo Fisher
Scientific) by using a Discovery (3 mm, 15 cmꢁ2.1 mm, Supelco) re-
versed-phase column at a flow rate of 200 mLminÀ1. Solvent A was
water with 0.05% formic, and solvent B was acetonitrile with 0.5%
formic acid. The solvent gradient initiated with solvent A (99%) for
5 min and then went to 99% of solvent B over 25 min. A final
2 min at 99% Solvent B was used before re-equilibration with sol-
vent A. The HPLC eluent was directed to a ThermoElectron LTQ-Or-
bitrap XL Discovery instrument (San Jose, CA, USA), equipped with
an ESI ion max source. The ionization interface was operated in the
positive mode with the following settings: source voltage: 4 kV;
sheath and aux gas flow rates: 50 and 10 units respectively; tube
lens voltage: 100 V; capillary voltage: 14 V; and capillary tempera-
ture: 3008C. LTQ with collision-induced dissociation (CID) values of
30 was used for LC-ESI-MS/MS experiments.
Scheme 3. General scheme for synthesis of S-(2-acetamidoethyl) 3-oxoalka-
nethioate (3-ketoacyl-NAC). a) Lithium diisopropylamide, THF, À788C to RT,
3 h; b) NaBH4, CH3OH, 08C, 15 min; c) LiOH, THF/H2O (1:1), reflux, 3 h; d) N-
acetyl cysteamine, EDCl, DMAP, CH2Cl2, RT, 12 h; e) pyridinium chlorochro-
mate, CH2Cl2, RT, 1 h.
Synthesis of methyl 3-oxoalkanoate (5): Methyl acetate (7.50 mL,
93.8 mmol) was added to a stirred solution of lithium diisopropyla-
mide (LDA; 39.0 mL of 1.8m solution in THF, 70.3 mmol) in THF
(100 mL) at À788C. After stirring for 15 min, a solution of the start-
ing acyl chloride in THF was added dropwise, and the reaction mix-
ture was stirred at 788C for 1 h before being warmed to room tem-
perature and stirred for 3 h. The reaction mixture was quenched
with 1 n HCl and extracted with EtOAc (3ꢁ100 mL). The combined
organic extracts were washed with brine, dried over Na2SO4, and
concentrated in vacuo. The crude product was purified by column
chromatography (hexane/EtOAc, 95:5) to afford the desired ester
(5).
Synthesis of methyl 3-hydroxyalkanoate (6): NaBH4 (1.90 g,
50 mmol) was added slowly to a stirred solution of 5 (3.0 g,
50 mmol) in anhydrous methanol (20 mL) at 08C, and the reaction
mixture was stirred at 08C for 20 min. The resulting solution was
quenched with 1 n HCl and extracted with EtOAc (3ꢁ20 mL). The
combined organic layers were dried over Na2SO4, concentrated in
vacuo, and purified with column chromatography to give the cor-
responding methyl 3-hydroxyalkanoate (6).
Synthesis of S-(2-acetamidoethyl) 3-hydroxyalkanoate (8): Hy-
drolysis of 6 to provide 7, and subsequent conversion to 8, was
analogous to the steps for formation of formation of enoyl-NAC
(described above).
Acknowledgements
Synthesis of S-(2-acetamidoethyl) 3-oxoalkanethioate (9): A sus-
pension of pyridinium chlorochromate (1.44 g, 6.72 mmol) in anhy-
drous CH2Cl2 (50 mL) was treated with 8 (1.0 g, 4.48 mmol). The
mixture was stirred at room temperature for 2 h. Et2O (100 mL)
We gratefully acknowledge Chris Hazzard and Dr. Galina Florova
for their invaluable help. This work was supported by a grant
from the National Institutes of Health (GM 077147).
ChemBioChem 0000, 00, 0 – 0
www.chembiochem.org
9
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
[17] Y. Tang, H. Y. Lee, Y. Tang, C.-Y. Kim, I. Mathews, C. Khosla, Biochemistry
2006, 45, 14085–14093.
Keywords: beta-ketoacyl-acyl carrier protein reductase · enoyl-
ACP reductase FabG · FabI · InhA · Streptomyces coelicolor
[18] N. Smirnona, K. A. Reynolds, J. Bacteriol. 2001, 183, 2335–2342.
[19] K. Bloch, Adv. Enzymol. Relat. Areas Mol. Biol. 1977, 45, 1–84.
[20] H. Marrakchi, S. Ducasse, G. Labesse, H. Montrozier, E. Margeat, L. Emor-
ine, X. Charpentier, M. Daffe, A. Quelmard, Microbiology 2002, 148, 951–
960.
[21] M. P. Patel, W. S. Liu, J. West, D. Tew, T. D. Meek, S. H. Thrall, Biochemistry
2005, 44, 16753–16765.
[22] A. K. Brown, S. Sridharan, L. Kremer, S. Lindenberg, L. G. Dover, J. C. Sac-
chetini, G. S. Besra, J. Biol. Chem. 2005, 280, 32539–32547.
[23] D. Meluzzi, W. H. Zheng, M. Hensler, V. Nizet, P. C. Dorrestein, Bioorg.
Med. Chem. Lett. 2008, 18, 3107–3111.
[24] H. Berglert, P. Wallnert, A. Ebelingt, B. Leitingert, S. Fuchsbichlers, H.
Aschaued, G. Kollenz, G. Hogenauert, F. Turnowskyt, J. Biol. Chem. 1994,
269, 5493–5496.
[25] R. J. Heath, C. O. Rock, J. Biol. Chem. 1995, 270, 26538–26542.
[26] J. Schiebel, A. Chang, H. Lu, M. V. Baxter, P. J. Tonge, C. Kisker, Structure
2012, 20, 802–813.
[27] A. Chang, J. Schiebel, W. Yu, G. R. Bommineni, P. Pan, M. V. Baxter, A.
Khanna, C. A. Sotriffer, C. Kisker, P. J. Tonge, Biochemistry 2013, 52,
4217–4228.
[28] R. J. Heath, N. Su, C. K. Murphy, C. O. Rock, J. Biol. Chem. 2000, 275,
40128–40133.
[1] L. Han, S. Lobo, K. A. Reynolds, J. Bacteriol. 1998, 180, 4481–4486.
[2] W. P. Revill, M. J. Bibb, D. A. Hopwood, J. Bacteriol. 1996, 178, 5660–
5667.
[3] D. A. Hopwood, Philos. Trans. R. Soc. London Ser. B 1989, 324, 549–562.
[4] M. A. Mchenney, T. J. Hosted, B. S. Dehoff, P. R. Rosteck, R. H. Baltz, J.
Biol. Chem. 1998, 180, 143–151.
[5] M. J. Bibb, D. H. Sherman, S. Omura, D. A. Hopwood, Gene 1994, 142,
31–39.
[6] T. Marti, Z. Hu, N. L. Pohl, A. N. Shah, C. Khosla, J. Biol. Chem. 2000, 275,
33443–33448.
[7] F. J. Wyszynski, A. R. Hesketh, M. J. Bibb, B. G. Davis, Chem. Sci. 2010, 1,
581–589.
[8] F. J. Wyszynski, S. S. Lee, T. Yabe, H. Wang, J. P. Gomez-Escribann, M. J.
Bibb, S. J. Lee, G. J. Davies, B. G. Davis, Nat. Chem. 2012, 4, 539–546.
[9] A. M. CerdeÇo, M. J. Bibb, G. L. Challis, Chem. Biol. 2001, 8, 817–829.
[10] N. R. Williamson, P. C. Fineran, T. Gristwood, S. R. Chawrai, F. J. Leeper,
G. P. Salmond, Future Microbiol. 2007, 2, 605–618.
[11] K. Papireddy, M. Smilkstein, J. X. Kelly, S. M. Salem, M. Alhamadsheh,
S. W. Haynes, G. L. Challis, K. A. Reynolds, J. Med. Chem. 2011, 54, 5296–
5306.
[29] A. Craney, C. Ozimok, S. M. Pimentel-Elardo, A. Capretta, J. R. Nodwell,
Chem. Biol. 2012, 19, 1020–1027.
[12] J. T. Tsay, W. Oh, T. J. Larson, S. Jackowski, C. O. Rock, J. Biol. Chem. 1992,
267, 6807–6814.
[30] S. Lobo, G. Florova, K. A. Reynolds, Biochemistry 2001, 40, 11955–11964.
[31] J. R. Whicher, G. Florova, P. K. Sydor, R. Singh, M. M. Alhamadsheh, G.
Challis, K. A. Reynolds, J. L. Smith, J. Biol. Chem. 2011, 286, 22558–
22569.
[13] K. H. Choi, R. J. Heath, C. O. Rock, J. Bacteriol. 2000, 182, 365–370.
[14] R. Singh, S. Mo, G. Florova, K. A. Reynolds, FEMS Microbiol. Lett. 2012,
328, 32–38.
[15] G. Florova, G. Kazanina, K. A. Reynolds, Biochemistry 2002, 41, 10461–
10471.
[16] G. Gago, L. Diacovich, A. Arabolaza, S. Tsai, H. Gramajo, FEMS Microbiol.
Rev. 2011, 35, 475–497.
Received: November 21, 2014
Published online on && &&, 0000
&
ChemBioChem 0000, 00, 0 – 0
www.chembiochem.org
10
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

FULL PAPERS
Functional crosstalk: b-Ketoacyl-acyl
carrier protein (ACP; FabG) and enoyl-
ACP reductase (FabI) from Streptomyces
coelicolor were identified and character-
ized. Kinetic analysis demonstrated that
these enzymes process straight and
branched-chain substrates along with
the ACPs from both fatty acid and un-
decylprodiginine biosynthetic pathways.
This relaxed substrate specificity allow
these enzymes to be involved in both
processes.
R. Singh, K. A. Reynolds*
&& – &&
Characterization of FabG and FabI of
the Streptomyces coelicolor
Dissociated Fatty Acid Synthase
ChemBioChem 0000, 00, 0 – 0
www.chembiochem.org
11
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ