Biosynthesis of branched-chain fatty acid in bacilli: FabD (malonyl-CoA:ACP transacylase) is not essential for in vitro biosynthesis of branched-chain fatty acids.

Article abstract of DOI:10.1271/bbb.67.2106

It was found that the partially purified beta-ketoacyl-ACP synthase of Bacillus insolitus did not require the addition of FabD (malonyl-CoA:ACP transacylase, MAT) for the activity assay. This study therefore examined the necessity of FabD protein for in vitro branched-chain fatty acid (BCFA) biosynthesis by crude fatty acid synthetases (FAS) of Bacilli. To discover the involvement of FabD in the BCFA biosynthesis, the protein was removed from the crude FAS by immunoprecipitation. The His-tag fusion protein FabD of Bacillus subtilis was expressed in Escherichia coli and used for the preparation of antibody. The rabbit antibody raised against the expressed fusion protein specifically recognized the FabD in the crude FAS of B. subtilis. Evaluation of the efficacy of the immunoprecipitation showed that a trace of FabD protein was present in the antibody-treated crude FAS. However, this complete removal of FabD from the crude FAS did not abolish its BCFA biosynthesis, but only reduced the level to 50-60% of the control level for acyl-CoA primer and to 80% for alpha-keto-beta-methylvalerate primer. Furthermore, the FabD concentration did not necessarily correlate with the MAT specific activity in the enzyme fractions, suggesting the presence of another enzyme source of MAT activity. This study, therefore, suggests that FabD is not the sole enzyme source of MAT for in vitro BCFA biosynthesis, and implies the existence of a functional connection between fatty acid biosynthesis and another metabolic pathway.

Full text of DOI:10.1271/bbb.67.2106

This article was downloaded by: [Illinois State University Milner Library]
On: 09 December 2014, At: 13:04
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer
House, 37-41 Mortimer Street, London W1T 3JH, UK
Bioscience, Biotechnology, and Biochemistry
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/tbbb20
Biosynthesis of Branched-chain Fatty Acid in Bacilli:
FabD (malonyl-CoA:ACP transacylase) Is Not Essential
for In Vitro Biosynthesis of Branched-chain Fatty
Acids
Hirosuke OKUa, Naoya FUTAMORIa, Kenichi MASUDAa, Yumiko SHIMABUKUROa, Tomoyo
OMINEa & Hironori IWASAKIa
a Division of Molecular Biotechnology, Center of Molecular Bioscience, University of the
RyukyusNishihara, Okinawa 903-0213, Japan
Published online: 22 May 2014.
To cite this article: Hirosuke OKU, Naoya FUTAMORI, Kenichi MASUDA, Yumiko SHIMABUKURO, Tomoyo OMINE & Hironori
IWASAKI (2003) Biosynthesis of Branched-chain Fatty Acid in Bacilli: FabD (malonyl-CoA:ACP transacylase) Is Not Essential
for In Vitro Biosynthesis of Branched-chain Fatty Acids, Bioscience, Biotechnology, and Biochemistry, 67:10, 2106-2114,
DOI: 10.1271/bbb.67.2106
To link to this article: http://dx.doi.org/10.1271/bbb.67.2106
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of
the Content. Any opinions and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied
upon and should be independently verified with primary sources of information. Taylor and Francis shall
not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other
liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or
arising out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions

Biosci. Biotechnol. Biochem., 67 (10), 2106–2114, 2003
Biosynthesis of Branched-chain Fatty Acid in Bacilli
:
FabD (malonyl-CoA:ACP transacylase) Is Not Essential for In Vitro
Biosynthesis of Branched-chain Fatty Acids
Hirosuke OKU,† Naoya FUTAMORI, Kenichi MASUDA, Yumiko SHIMABUKURO
Tomoyo OMINE, and Hironori IWASAKI
,
Division of Molecular Biotechnology, Center of Molecular Bioscience, University of the Ryukyus,
Nishihara, Okinawa 903-0213, Japan
Received March 26, 2003; Accepted July 7, 2003
It was found that the partially puriˆed
b-ketoacyl-
by the E. coli system has been studied in detail, and
the biosynthesis pathway consists of a set of individ-
ual enzyme proteins (type II) that function in concert
to produce a variety of fatty acids.1,2) Although the
lower activity of BCFA biosynthesis in the cell-free
system of Bacilli has severely impeded the eŠorts to
establish a detailed synthetic mechanism by using the
individually isolated enzymes, the available evidence
strongly supports the view that BCFAs in bacteria are
ACP synthase of Bacillus insolitus did not require the
addition of FabD (malonyl-CoA:ACP transacylase,
MAT) for the activity assay. This study therefore exam-
ined the necessity of FabD protein for in vitro branched-
chain fatty acid (BCFA) biosynthesis by crude fatty acid
synthetases (FAS) of Bacilli. To discover the involve-
ment of FabD in the BCFA biosynthesis, the protein
was removed from the crude FAS by immunoprecipita-
tion. The His-tag fusion protein FabD of Bacillus subti-
lis was expressed in Escherichia coli and used for the
preparation of antibody. The rabbit antibody raised
against the expressed fusion protein speciˆcally recog-
nized the FabD in the crude FAS of B. subtilis. Evalua-
tion of the e‹cacy of the immunoprecipitation showed
that a trace of FabD protein was present in the an-
tibody-treated crude FAS. However, this complete
removal of FabD from the crude FAS did not abolish
its BCFA biosynthesis, but only reduced the level to
synthesized by a mechanism very similar to that of
3)
SCFA synthesis operating in E. coli
.
The diŠerence between SCFA and BCFA biosyn-
thesis thus appeared to lie in the precursor speciˆcity
of the initiation step: the substrate speciˆcity of acyl-
CoA:ACP transacylase (step B in Fig. 1)4) or the ˆrst
step condensing enzyme of b-ketoacyl-ACP synthase
III (KAS III) FabH (steps C, D, and E in Fig. 1). It
has actually been shown by using a His-tag fusion
protein of FabH that the Bacillus FabH (bFabH) is a
determining factor in BCFA biosynthesis.5) There are
at least two FabH (KAS III) like protein and one
50–60
80% for
%
of the control level for acyl-CoA primer and to
-keto- -methylvalerate primer. Furthermore,
a
b
the FabD concentration did not necessarily correlate
with the MAT speciˆc activity in the enzyme fractions,
suggesting the presence of another enzyme source of
MAT activity. This study, therefore, suggests that FabD
is not the sole enzyme source of MAT for in vitro BCFA
biosynthesis, and implies the existence of a functional
connection between fatty acid biosynthesis and another
metabolic pathway.
FabB F (KAS I II) homologue encoded by the B.
W
W
subtilis genome.
The initiation of the condensing reaction (steps C,
D, and E in Fig. 1) usually requires the presence of
malonyl-CoA:ACP transacylase (MAT) FabD that
catalyzes the conversion of malonyl-CoA to malonyl
ACP (step A in Fig. 1). It was also shown that the B.
subtilis fabD gene complemented the MAT deˆciency
of an E. coli mutant.6) However, we found in the
course of enzyme puriˆcation work that partially
puriˆed native KAS from B. insolitus did not neces-
sarily require the addition of FabD for the activity
assay. Thus this study clariˆed the requirement of
FabD for BCFA biosynthesis in vitro and discussed
the necessity of FabD for the condensation reaction
Key words: branched-chain fatty acid; biosynthesis;
Bacillus; malonyl-CoA; ACP
Fatty acids are the most important component of
bacterial membrane lipids. The chemical structure of
fatty acids occurring in bacteria is classiˆed largely
into two types: straight-chain fatty acids (SCFAs)
and branched-chain fatty acids (BCFAs) of the iso-
and anteiso-series. The SCFA synthesis represented
in Bacilli
.
†
To whom correspondence should be addressed. TelWFax: 81-98-895-8972; E-mail: okuhiros comb.u-ryukyu.ac.jp
＋ ＠

Fatty Acid Biosynthesis in Bacilli
2107
transform a BL21(DE3) competent cell strain. The
plasmid harboring the fabD gene was isolated, and
sequenced to verify the absence of PCR artifacts.
Protein expression was induced according to the
manufacture's instruction manual. The His-tag fu-
sion protein was puriˆed by use of His Trap chelating
a‹nity column (Amersham Pharmacia Biotech).
Preparation of short-chain acyl-CoA. The bran-
ched short-chain fatty acids were converted into their
thioesters as described previously.8) Brie‰y, the short-
chain fatty acid (5
mmol) was reacted with carbox-
ydiimidazole (6 mol) in 0.2 ml of tetrahydrofuran to
m
give a rise to 1-acylimidazole. After 30 min of reac-
tion at room temperature, the solvent was evaporated
oŠ, and the reaction product was dissolved in 0.2 ml
tetrahydrofuran-H2O (2:1 by vol.). The solution was
Fig. 1. Schematic Representation of the Condensing Reaction in
the Fatty Acid Biosynthesis.
Both malonyl-CoA and acyl-CoA become available to the b-
allowed to react with 5 mmol of CoASH in 0.5 ml of
ketoacyl-ACP synthase (KAS) after their conversion into ACP
thioesters by steps A and B. FabD is thought to be responsible
for charging the ACP subunit with a malonyl group (step A).
The enzyme KAS catalyzes the condensation of acyl-ACP and
tetrahydrofuran-H2O (2:1 by vol.). The reaction was
continued for 4 hours at room temperature with a
nitrogen atmosphere. Tetrahydrofuran was evapo-
rated oŠ, and the residual aqueous solution was
acidiˆed to pH 3–4 by the addition of Dowex 50
(H＋ ). Dowex 50 was removed by ˆltration, and the
malonyl-ACP to produce b-ketoacyl-ACP. The ˆrst step is the
transacylation of the acyl-moiety from acyl-ACP to the active
site cysteine of KAS. The enzyme next reacts with malonyl-ACP
and releases CO2. The condensing enzyme assay in this study
thus measured the release of [14C]-CO2 from [14C]-malonyl-CoA
as described in the methods section.
ˆltrate was extracted with diethyl ether (0.5 ml
×
3
times) to remove unreacted fatty acid. The concen-
tration of acyl-CoA thioester was measured by
DTNB (5,5
described previously.9) Acyl-CoA thioester synthe-
sized by this method was usually over 90 pure on
the basis of free and esteriˆed SH assays.9)
?-dithiobis-2-nitrobenzoic acid) method as
Materials and Methods
z
Materials. [14C]-Malonyl-CoA and [14C]-isoleucine
were purchased from Muromachi Chemicals Co.
[14C]-isoleucine was converted into the corresponding
Bacterial strain and cell culture. The strains used
for the preparation of crude enzyme were Bacillus
subtilis (ATCC23857) and Bacillus insolitus (ATCC-
a
-ketoacid using -amino acid oxidase and catalase
L
(both from Sigma-Aldrich Japan.) as described
previously.7) Malonyl-CoA and acyl carrier protein
from E. coli were from Sigma-Aldrich Japan Co.
Other chemicals used were all the highest grade and
were obtained from the domestic suppliers.
23299). Cells were suspended in 15
z
glycerol and
－
stored at 70
9C. The culture medium for B. subtilis
contained: glucose, 10 g; yeast extract, 1 g; Bacto
peptone, 1 g; K2HPO4, 1.5 g; NH4H2PO4, 0.5 g;
MgSO4
･
7H2O, 0.1 g; ZnSO4
･
7H2O, 10 mg; CuSO4
･
･
･
5H2O, 0.5 mg; MnCl2 4H2O, 4.5 mg; and FeSO4
Construction of expression vectors and puriˆca-
tion of His-tag protein. The malonyl-CoA:ACP
transacylase gene (fabD) of B. subtilis was ampliˆed
with a genomic DNA as PCR template. The forward
and reverse primers designed based on the FabD gene
sequence in the DNA database ``SubtiList'' respec-
tively created a restriction site for NdeI and BamHI
7H2O, 10 mg, in 1 liter of distilled water.10) The
culture medium for B. insolitus was Soybean-Casein-
Digest broth supplied by Nissui Pharmaceutical Co.
(Tokyo, Japan). The indicated concentration of the
medium was peptone 17 g, soybean-peptone 3 g,
NaCl 5 g, glucose 2.5 g, and K2HPO4 2.5 g per litter.
The seed culture in 20 ml was inoculated with the
at the 5
? terminus, and consisted of 5?-CAT-
GCATATGATTGCATTTTTATTCCCGG and 5
?-
stock cells, and was incubated 7 hours at 30
B. subtilis, and 16 hours at 20 C for B. insolitus. One
9C for
TTTAGGATCCAGCATTATCATTCTCCTCCT.
The PCR products were ˆrst ligated into pCRII
(Invitrogen) by use of a TOPO-TA cloning kit
(Invitrogen), and used to transform TOPO10-compe-
tent E. coli cells. A plasmid was isolated and digested
with NdeI and BamHI. The released insert DNA
fragment were separated by agarose gel electrophore-
sis, and ligated into pET-15b digested previously with
NdeI and BamHI. The ligation mixture was used to
9
litter of medium was inoculated with 10 to 20 ml of
seed culture, and cells were harvested at the late
exponential growth phase by centrifugation. The
yield of cells was usually in the range of 5–6 g wet
weight l culture medium.
W

2108
H. OKU et al.
Preparation of crude enzyme. A typical procedure
for the preparation of crude enzyme was as follows.
Pelleted cells were suspended in 3 vol. of 10 m
phosphate buŠer (pH 7.0) containing 2 m 2-mer-
captoethanol and 10 glycerol, and were sonicated
ture contained 50 m
1 m DTT, 2.5 ACP, 40
mol) in 0.5 ml. The reaction was in-
M
triethanolamine (pH 7.5),
M
m
M
m
M
[1,3-14C]-malonyl-
M
CoA (37 kBq
cubated at 30
m
W
M
9
C for 30 min, and stopped by addition
z
of 0.2 ml 5
z
HClO4. The [14C]-CO2 produced was
with an ultrasonic generator (Nihon Seiki Seisakusho
collected into 0.2 ml of 2 N NaOH for 60 min in a
closed reaction tube, and was counted with 8 ml of
scintillation ‰uid (ACS II).
×
Co., Tokyo) for 3 3 min with 1-min cooling inter-
vals. The irradiated cell suspension was centrifuged
×
for 28,000 G for 30 min, and the supernatant was
again centrifuged at 156,000
×
G for 2 hours. To the
Fatty acid synthetase (FAS) activity. Fatty acid
supernatant was added streptomycin sulfate to make
its concentration equivalent to protein in the extract.
The precipitate formed was then removed by cen-
synthetase (FAS) activity was measured by incorpo-
ration of [U-14C]-
a-keto-b-methylvalerate (KMV)
(740 kBq
m
mol) or [1,3-14C]-malonyl-CoA (37 kBq
W W
×
trifugation at 28,000 G for 20 min. The super-
mmol) into fatty acid. The radiolabeled fatty acid was
natant obtained after streptomycin treatment was
isolated from the reaction mixture by a solvent ex-
traction procedure. The standard reaction mixture
mixed with 1 10 vol. of 1
M
phosphate buŠer
saturation. The
precipitate formed was collected by centrifugation,
W
(pH 7.0) and with (NH4)2SO4 to 60
z
contained, in m : triethanolamine, 50 (pH 7.5);
M
DTT, 2; thiaminepyrophosphate, 0.5; NADPH, 0.5;
and dissolved in a minimal volume of 10 m
M
phos-
malonyl CoA, 0.1; and [U-14C]-KMV 0.01, in a total
phate buŠer (pH 7.0). The ˆnal protein concentra-
volume of 500
was the precursor, [U-14C]-KMV was replaced with
cold acyl-CoA (0.1 m ), and cold malonyl-CoA was
replaced with [1,3-14C]-malonyl CoA (0.04 m
). The
reaction mixture was incubated at 30 C for 30 min.
ml. In the case where cold acyl-CoA
tion was usually from 20 to 60 mg ml.
W
M
Puriˆcation of the condensing enzyme. Puriˆca-
tion of the condensing enzyme was done by a use of
fast protein liquid chromatography (FPLC) system
(Amersham-Pharmacia Biotech) or perfusion chro-
matography system (Perceptive Biosystems). The
M
9
The enzyme reaction was stopped by the addition of
0.025 ml of 6 N HCl, and the acidiˆed reaction mix-
×
g for 5 min. The super-
ture was centrifuged at 600
crude enzyme (0–60
diluted 5 to 9 times with the starting buŠer containing
20 m histidine-HCl (pH 5.5), 2 m dithiothreitol
(DTT) and 10
z
(NH4)2SO4 precipitate was
natant was mixed with 0.35 ml of H2O, 0.5 ml of
methanol and 3 ml of CCl4 in a glass-stoppered test
tube. The lower layer (CCl4) was washed twice with
M
M
z
glycerol, and was put on a Mono Q
z
2.5 ml of 50 methanol containing 0.05 N HCl. The
HR 10 10 column (Amersham-Pharmacia Biotech)
washed CCl4 layer was evaporated to dryness by
‰ushing with N2, and its radioactivity was counted
with a toluene scintillator.
W
which had previously been equilibrated with the same
buŠer. The column was washed with 3 ml of starting
buŠer and was eluted with a linear gradient of 0–
0.65
M
NaCl in 27 ml of the buŠer with the ‰ow rate
Malonyl-CoA:ACP transacylase (MAT) assay.
The MAT activity assay was done essentially as de-
scribed elsewhere.12) Brie‰y, the reaction mixture
of 1.5 ml min.
W
The fractions containing the condensing enzyme
were mixed with 3
M
(NH4)2SO4 to make 1.8
M
, and
containing 7.5
mol), 10 g ACP (23.2
imidazole-HCl (pH 7.0), and up to 1
ˆnal volume of 50 l was incubated at 30
min. The reaction was stopped with 0.4 ml of 5
z
m
M
[1,3-14C]-malonyl-CoA (2035 kBq
put on a reversed-phase column PE M (4.6 mm ID
×
m
W
m
m
M
), 1 m
M
DTT, 100 m
g of FabD in a
C for 5
M
W
100 mm, Perceptive Biosystem) equilibrated with
10 m phosphate buŠer (pH 7.0) containing 1.8
(NH4)2SO4, 2 m DTT and 10 glycerol. The
sample was eluted with a negative linear gradient of
1.8–0 (NH4)2SO4 in a 15 ml of eluent. The active
m
M
M
m
9
M
z
perchloric acid, followed by the addition of 370 mg
M
of carrier crude ACP prepared from a B. subtilis
enzyme fractions were combined and concentrated
by ultraˆltration (Centricon 30 or 50, Amicon Inc.)
homogenate. The precipitate formed was collected by
ˆltration with a 0.2 mm pore size membrane ˆlter,
to 0.2 ml, and put onto Superose 12 HR 10 30
and was further washed twice with 0.5 ml of 5
z
W
column (Amersham-Pharmacia Biotech) that had
been previously equilibrated with the elution buŠer
perchloric acid. The malonyl-ACP present in the
precipitate was re-dissolved in 0.5 ml 0.1 N NaOH,
and counted for its radioactivity.
containing 50 m
M
potassium phosphate (pH 6.0),
0.15 NaCl, 2 m
M
M
DTT, and 10z glycerol. The
sample was eluted with a ‰ow rate of 0.2 ml min.
Preparation of antibody. Antibody against the
puriˆed His-tag fusion protein FabD was raised in a
male Japanese white rabbit with a mixture of equal
volume of Freud's complete adjuvant and 203 mg an-
tigen FabD protein. The animal received the injection
twice a week for one month, and was bled after the
W
Condensing enzyme assay. [14C]-CO2 liberated by
the action of the condensing enzyme from [1,3-14C]-
malonyl-CoA was collected in NaOH, and used for
measurement of radioactivity.11) The reaction mix-

Fatty Acid Biosynthesis in Bacilli
2109
enough antibody production was conˆrmed by
double immunodiŠusion analysis.13)
ECL Western blotting detection reagent (Amersham-
Pharmacia Biotech).
Enrichment of IgG was made by (NH4)2SO4 frac-
tionation as follows: the rabbit serum was mixed with
Results
an equal volume of 50
z
(NH4)2SO4 in 50 m
M
phos-
phate buŠer (pH 7.0). After 20 min, the precipitate
Figure 2 shows the SDS-PAGE of the His-tag
fusion protein FabD. The expressed His-tag protein
×
formed was collected by centrifugation (12000 G,
10 min), and washed with 25 (NH4)2SO4 in 50 m
phosphate buŠer (pH 7.0). The washed precipitate
was dissolved in 20 m phosphate buŠer, (pH 7.0)
＋
z
M
was puriˆed to homogeneity by Ni2 chelate a‹nity
chromatography, and thereafter desalted with a
PD-10 column (Amersham-Pharmacia Biotech). The
estimated molecular mass on SDS-PAGE was 41
kDa, which was slightly higher than the calculated
molecular size of 37 kDa. This preparation was used
for the condensing enzyme assay and for the prepara-
tion of antibody.
M
and dialyzed overnight against the same buŠer. The
dialyzed solution was centrifuged, and the super-
－
9
natant was stored at 70 C until needed.
Immunoprecipitation. The crude FAS was mixed
with an equal amount of IgG to give the ˆnal enzyme
In order to choose the bacterial strain for the
puriˆcation of the condensing enzyme, we ˆrst stud-
ied in vitro biosynthesis of fatty acids by the cell-free
homogenate of B. subtilis and B. insolitus (Fig. 3).
concentration of 19.6 mg ml. Since the crude FAS
W
contained 1.04
mg FabD mg total protein, the ˆnal
W
concentration of FabD in the mixture was 20
m
g ml.
W
The mixture containing 3 mg of enzyme protein was
used with or without centrifugation for the BCFA
biosynthesis or MAT assay. In the case of control,
IgG was replaced by bovine serum albumin to adjust
the total protein concentration.
Western blotting. Samples separated on SDS-
PAGE were transblotted to nitrocellulose mem-
branes as described elsewhere. The membrane was
washed with phosphate-buŠered saline containing
Tween 20 (PBS-T): 80 m
M
Na2HPO4, 20 m
Tween 20, and sub-
skim milk blocking
M
NaH2PO4, 100 m
M
NaCl, 0.1
z
sequently incubated in a 10
z
solution for 30 min with occasional shaking. The
membrane was washed with PBS-T and incubated
with primary antibody (anti FabD) for 60 min at
room temperature. The un-bound antibody was
washed out with several changes of PBS-T. The anti-
FabD IgG bound to the sample was made visible by
Fig. 2. SDS-PAGE of Expressed His-tag Fusion Protein FabD.
The expressed protein was puriˆed to homogeneity and sepa-
rated on a polyacrylamide gel.
Fig. 3. Fatty Acid Biosynthesis from Short-chain Acyl-CoAs in B. insolitus (A) and B. subtilis
The assay mixture contained 50 m triethanolamine (pH 7.5), 2 m thiaminepyrophosphate, 0.5 m
malonyl-CoA, 0.1 m acyl-CoA, and 0.5–2 mg of crude FAS in 500 l. The reaction was done at 30
product of radiolabeled fatty acid was extracted into CCl4 for measurement of its radioactivity.
.
M
M
M
NADPH, 0.04 m
M
[1,3-14C]-
M
m
9
C for 30 min, and the reaction

2110
H. OKU et al.
Table 1. Component Requirements of Condensing Reactions
Component omitted
Activity
nmol
W
mg
z
±
±
±
±
±
None (complete)
FabD
ACP
Hexanoyl-CoA
DTT
67.3
76.8
0.0
±
3.73
16.7
0.0
100
114
0.0
30.6
47.6
5.5
24.8
0.0
18.4
18.4
±
±
±
±
20.6
32.1
12.4
12.4
Data are mean
SE (n 3).
± ＝
In contrast to the case of E. coli, the cell-free
homogenate of B. subtilis and B. insolitus accepted a
variety of acyl-CoA derivatives as a precursor for fat-
ty acid biosynthesis. Acyl-CoAs of chain-lengths C4
to C12 were good precursors for fatty acid biosynthe-
sis by the crude fatty acid synthetase (FAS) of B. sub-
tilis and B. insolitus. On the whole, the crude FAS
of B. insolitus showed a rather broad spectrum of
precursor speciˆcity, accepting both straight-chain
and branched-chain acyl-CoA (BCAC) to an almost
comparable extent. The relatively higher biosynthetic
activity of B. insolitus was also notable (see the
diŠerence in the scale of X-axis). The crude FAS of
B. subtilis surprisingly manifested a preference for
straight-chain acyl-CoA, except for acetyl-CoA,
rather than toward the branched-chain counterpart.
Combination of an ion-exchange column (Mono
Fig. 4. Western Blotting Proˆle of Crude FAS from B. subtilis
and B. insolitus
The sample size was; 0.005
FabD; 5 g crude FAS from B. subtilis; and 10
FAS from B. insolitus. Densitometric analysis found that the
concentration of FabD was 1.04 mg for B. subtilis crude
FAS, and 0.37 mg for B. insolitus enzyme.
.
m
g for His-tag fusion protein
g for crude
m
m
mgW
mgW
Q), a hydrophobic column (PE M), and a gel-ˆltra-
ˆle of the crude FAS from B. subtilis and B. insoli-
tus. The antibody raised against the His-tag fusion
protein FabD recognized essentially a single protein
in both crude FASs. The densitometric analysis
found that crude FAS from B. subtilis and B. insoli-
W
tion column (Superose 12), puriˆed the condensing
enzyme 5-fold from the cell-free homogenate of B.
insolitus. On the course of this puriˆcation, the assay
component requirement of the partially puriˆed con-
densing enzyme was studied (Table 1). Based on the
substrate speciˆcity of the crude FAS, hexanoyl-CoA
was the acyl-CoA substrate for the condensing
reaction. The reaction unexpectedly did not require
the FabD, while was completely dependent on the
presence of ACP. Removal of hexanoyl-CoA and
DTT respectively lowered the activity to 30.6 and
tus respectively contained 1.04
m
g and 0.37
m
g of
FabD mg total protein. The crude FAS from B.
W
insolitus contained the FabD in concentrations from
one sixth to one third of B. subtilis, and contrasted
with its higher BCFA biosynthetic activity compared
with that of B. subtilis. As was expected, the native
FabD in the crude synthetase moved slightly faster
than the His-tag fusion protein FabD. It is thus now
possible to remove only the FabD protein from the
crude FAS by immunoprecipitation.
47.6
z
of the control value. The incomplete block of
the condensation reaction by withdrawal of the sub-
strate hexanoyl-CoA may show that the decarboxyla-
tion reaction at the malonyl-ACP binding site
partially proceeded even in the absence of acyl-CoAs,
but not to the completion of the reaction.
A preliminary experiment on double immunodiŠu-
sion found that the equivalence was reached with
0.25 mg anti-FabD IgG against 1 mg of crude FAS.
To ensure the complete removal of FabD by the im-
munoprecipitation method, the concentration course
of antibody was studied against the His-tag fusion
protein FabD (Fig. 5A). In this case, the ˆnal concen-
tration of His-tag FabD in the reaction mixture was
comparable to the following immunoprecipitation
experiment. As shown in Fig. 5A, addition of
antibody itself had little eŠect on FabD activity.
However, removal of the FabD-antibody complex by
the centrifugation completely abolished the activity
in the supernatant. Based on this experiment, an
The above observations prompted us to suspect the
necessity of the FabD for the condensation reaction
and also for the BCFA biosynthesis. There was
actually a 5-fold increase in the speciˆc activity of the
condensing enzyme with the progress of the puriˆca-
tion, even when the assay was done without FabD.
Puriˆcation of KAS from B. subtilis encountered
di‹culties due to its low activity and instability, as
we mentioned in the foregoing section. We therefore
prepared the antibodies against FabD, and studied its
eŠect on in vitro BCFA biosynthesis by crude FAS of
B. subtilis. Figure 4 shows the Western blotting pro-
excess of antibody (1 mg against 1 mg FabD) was used

Fatty Acid Biosynthesis in Bacilli
2111
Fig. 5. EŠects of Rabbit Anti-FabD on MAT Activity of His-tag Fusion Protein FabD (A) and Crude FAS from B. subtilis (B).
The MAT assay contained 100 m
M
imidazole-HCl (pH 7.0), 1 m
M
DTT, 10
m
g ACP, 7.5
C for 30 min. The resultant mixtures with (closed cir-
m [
M
14C]-malonyl-CoA, His-tag fusion pro-
tein FabD (A), or crude FAS (B) both non-treated or treated with antibody at 30
9
cle) or without centrifugation (open circle) were the enzyme source. The ˆnal concentration of FabD in the reaction mixture was adjust-
ed to approximate the following FAS assay experiment shown in Table 2. Data are expressed as the percentages of the control activity
(non-treaed enzyme). Control activity was 645 nmolWmg for His-tag fusion protein FabD (A) and was 1.89 nmolWmg for crude FAS (B).
for the later immunoprecipitation experiment to
ensure the complete removal of FabD in the crude
FAS.
In analogy to the constructed immunoprecipitation
experiment shown in Fig. 5A, the eŠect of antibody
on MAT activity of crude FAS was studied (Fig. 5B).
The X-axis shows the amount of antibody added
against 1 mg of crude FAS containing 1.04 mg FabD.
Thus, the ˆnal concentration of FabD in the mixture
was also similar to the following immunoprecipita-
tion experiment. In contrast to the case of Fig. 5A,
removal of FabD from the crude specimen did not
signiˆcantly reduce its MAT activity.
Fig. 6. Western Blotting Proˆle of Crude FAS Treated with
Anti-FabD IgG.
Crude FAS treated with pre-immune IgG or anti-FabD IgG
were analyzed by Western blotting. The sample size for crude
FAS and IgG was 10
Lane 1, control crude FAS (non-treated); lane 2, crude FAS
pre-immune IgG; lane 3, crude FAS anti-FabD IgG; lane 4,
mg, and was 0.03 mg for His-tag FabD.
The removal of FabD by immunoprecipitation was
further assessed by Western blotting. Based on
the results of concentration course experiments
(Fig. 5A), an excess of antibody (1 mg) was used
＋
＋
pre-immune IgG; lane 5, anti-FabD IgG; lane 6, His-tag fusion
protein FabD.
against 1 mg of FabD in the crude enzyme prepara-
tion. Figure 6 shows the Western blot proˆle of crude
enzyme treated with the anti-FabD. Consistent with
the foregoing observation, only a single band was
present in the crude B. subtilis enzyme (lane 1).
Treatment of the crude enzyme with non-immunized
control IgG had essentially no eŠect on the intensity
of the FabD band (lane 2). However, treatment with
anti-FabD IgG signiˆcantly decreased the band inten-
sity almost to the blank level (compare lanes 3, and
5). As shown in Fig. 6, the background level of the
faint band was also detected in the antibody itself at
the molecular size of FabD (lanes 4 and 5). Densito-
metric analysis of the band intensity and subtraction
of the background level revealed that the crude
enzyme contained a trace of FabD protein. The im-
munoreactive bands of 44 k and 68 k both originated
in the IgG (lanes 4 and 5).
by centrifugation, suggesting the presence of another
enzyme source for FabD (Fig. 5B). We therefore at-
tempted to identify another MAT source in the crude
FAS of B. subtilis, and examined the concentration
of FabD and MAT activity in the enzyme fractions
separated on Mono Q column. The highest concen-
tration of FabD was noted around fraction 25, while
the highest MAT speciˆc activity was seen with the
fractions of 20 or 21 (Fig. 7). The MAT activity thus
did not necessarily correlate with the concentration
of FabD in the enzyme fractions, supporting the view
that there was an another enzyme source of MAT
activity in the cell homogenate of B. subtilis. This is
particularly true for the fraction 19 which lacked
FabD and showed relatively higher speciˆc activity.
The gel-ˆltration chromatography of the combined
specimens of fraction 19 and 20 found that the
molecular mass for the alternate MAT activity was
58 kDa.
The crude FAS retained more than 80
z of its
original activity even after removal of FabD protein

2112
H. OKU et al.
EŠects of immunoprecipitation on the in vitro
BCFA biosynthesis were studied with several precur-
sors. BCFAs are synthesized from branched-chain
keto acids (BCKAs) of valine, leucine, and isoleucine
as the primer source: -ketoisovalerate (KIV),
-keto- -methylvalerate
a precursor, the treatment lowered the activity only
to 80 . Thus the removal of FabD from the crude
synthetase system did not necessarily eliminate the
BCFA biosynthesis from both BCKA and BCAC
primers.
Addition of the pre-immune serum IgG slightly
elevated the level of fatty acid biosynthesis from
BCACs, but not from KMV.
z
a
-
a
a-
ketoisocaproate (KIC), and
a
b
(KMV).14) These BCKAs are successively decarbox-
ylated either by BCKAs dehydrogenase to give a rise
to branched-chain acyl-CoAs (BCACs): isobutyryl-
CoA from KIV, isovaleryl-CoA from KIC, and
We observed an anti-FabD reactive band in the
crude FAS of B. insolitus as seen in Fig. 4. However,
no precipitation line between anti-FabD and the
crude FAS of B. insolitus was formed on double
immunodiŠusion analysis. For this reason, no
immunoprecipitation experiment was done with the
a
(2)-methylbutyryl-CoA from KMV, or possibly by
BCKAs decarboxylase to give rise to the corre-
sponding aldehyde derivatives.3,15) Although the
crude FAS accepts both BCKAs and BCACs for in
vitro BCFA biosynthesis, the latter seemed to be a
better substrate for the current purpose because these
precursors directly react with malonyl-ACP, the
reaction product of FabD, to produce fatty acids. As
KMV among the BCKAs was the best precursor for
crude FAS,15) this precursor was also included as the
primer in this experiment.
crude FAS of B. insolitus
.
Discussion
The ˆnding that FabD was not essential for the
assay of the partially puriˆed condensing enzyme
from B. insolitus prompted us to study the necessity
of FabD in the BCFA biosynthesis. This observation
may simply be interpreted as that the partially puri-
ˆed KAS contained contaminant FabD. However,
the increase in the speciˆc activity of KAS with the
progress of puriˆcation may argue against this expla-
nation. The puriˆcation procedure usually decreases
the contaminant concentration in the specimens.
Thus, if FabD is essential for the condensing reac-
tion, one can expect that the speciˆc activity of KAS
assayed without addition of FabD should be lowered
by the puriˆcation. However, the puriˆcation of
KAS increased the speciˆc activity to 5-fold, which
may suggest its independence of this enzyme on
FabD reaction. This may in turn suggest that KAS of
Bacilli itself has the intrinsic MAT activity.
Removal of FabD by immunoprecipitation lo-
wered the incorporation of [14C]-malonyl-CoA into
fatty acid 50 to 60
z of control (none), but did not
completely diminish the BCFA biosynthesis from
BCAC primers (Table 2). In the case of [14C]-KMV as
It has been proposed that the B. subtilis KAS III
homolog fabHA (yjaX) or fabHB (yhfB) is a decid-
ing factor for the preference of bulky BCACs as
primers of BCFA biosynthesis. His-tag fusion pro-
tein of these gene products preferentially accepted
iso- and anteiso-BCAC primers as substrates rather
than acetyl-CoA, which is the starter unit for the
SCFA biosynthesis.5) The nature of the condensing
enzyme from Bacilli seemed to diŠer from that for
SCFA biosynthesis. This explanation may further be
Fig. 7. MAT Speciˆc Activity (open circle) and FabD Concentra-
tion (closed circle) in the Mono Q Column Fractions.
The concentrations for FabD were estimated by the densito-
metric analysis of the band intensity on Western blotting.
Table 2. EŠects of Anti-FabD on BCFA Biosynthesis by the Crude Fatty Acid Synthetase of B. subtilis
Precursorsa
Antibody added
[
14C]-KMV
Isovaleryl-CoA
pmol mg
2-Methylbutyryl-CoA
pmol mg
Isobutyryl-CoA
pmol mg
z
pmol
W
mg
z
W
±
z
W
z
W
None
Preimune IgG
Anti-FabD IgG
4.9
4.6
4.0
±
0.3
0.6
0.2
100.0
94.7
82.6
±
±
±
6.3
162.7 13.4 100.0
13.2 205.6
3.6 101.9
±
±
±
8.21
1.44
2.27
216.8
±
3.5
100.0
N.D.b
49.5
±
1.6
186.4
±
6.6
100.0
N.D.b
56.2 3.5
±
3.6
±
±
±
±
2.3
3.7
126.3
62.6
N.D.b
N.D.b
107.3
±
2.2
±
1
104.8
±
6.5
±
±
＝
Data are mean SE (n 3).
For precusors of CoA derivatives, the radiolabeled tracer was 14C-malonyl-CoA.
Not determined.
a
b

Fatty Acid Biosynthesis in Bacilli
2113
supported by their resistance to the antibiotic thiolac-
tomycin, which selectively inhibits dissociated Type
II, but not Type I, FAS.16) Thus the studies on KAS
entities in Bacilli may unveil the novel and unique
features of this enzyme, and may address the ques-
tions raised by this study.
protein may alternatively agree in part with the previ-
ous observation that the reconstituted polyketide
synthase (PKS) from Streptomyces coelicolor pro-
duced polyketides without MAT when holo ACP is
present in excess. The polyketides are synthesized by
a similar mechanism that resembles fatty acid biosyn-
thesis. In that case, the PKS-ACP catalyses sel-
fmalonylation, and provides malonyl-ACP for the
successive condensing reaction.18–20) As is the case of
PKS, it also may be possible that the ACP of B. sub-
tilis undergoes selfmalonylation, and may serve as a
chain extension unit of fatty acid biosynthesis. To
address this question, we studied the selfmalonyla-
tion of B. subtilis ACP under the same conditions as
the MAT assay, but in the absence of a MAT enzyme
source. The puriˆed B. subtilis ACP alone, however,
showed no selfmalonylation in this study (data not
shown), and this may negate the possibility above
that the ACP selfmalonylation was responsible for
the FabD independent BCFA biosynthesis.
We herein ˆrst demonstrated that the FabD pro-
tein is not essential for in vitro BCFA biosynthesis by
the crude FASs of B. subtilis. We applied the im-
munoprecipitation method to remove FabD protein
from the crude BCFA synthetase. The antibody pre-
pared in this study speciˆcally recognized the FabD
protein in the crude FAS (Figs. 4 and 6), and was
found to be eŠective to remove FabD protein from
the crude system. By use of a similar method, our
previous study has demonstrated the requirement of
BCKA decarboxylase for the BCFA biosynthesis,
and conˆrmed the usefulness of the immunoprecipi-
tation method.15) The supernatant after im-
munoprecipitation in this study thus speciˆcally
lacked FabD protein retaining with activities of all
other members of FASs. However, the removal of
FabD did not necessarily eliminate the in vitro BCFA
biosynthetic activity from the crude FAS, but only
Although this study ˆrst demonstrated that BCFA
biosynthesis by crude FAS did not necessarily require
FabD protein, the current in vitro results should be
reasonably related to in vivo experiments. One
experimental approach for this purpose is to study
BCFA biosynthesis in an FabD-deˆcient mutant. The
FabD-deˆcient mutant of E. coli was an auxotroph
requiring both saturated and unsaturated fatty acid.1)
The authors hope that this paper triggers some
studies to clarify the metabolic link between fatty
acid biosynthesis and other pathways including
reduced the level to 50–60
z
for the case of BCACs,
and to 80 for KMV primer. This ˆnding clearly
z
indicates that the FabD in the crude system is not
essential for the BCFA biosynthesis.
The point to mention is that there may be an
alternate enzyme source that can catalyze the MAT
reaction. We attempted to identify another enzyme
source in the crude FAS, and detected alternate MAT
activity apart from FabD protein (Fig. 7). Thus, the
observation that the removal of FabD from the crude
FAS did not eliminate the MAT activity in the crude
enzyme may be explained by the presence of another
enzyme source for FabD. A search of the B. subtilis
genome database found three MAT homologues de-
scribed as involved in polyketide synthesis: PksC,
PksE, and PksD with the similarity of 56, 54, and
polyketide synthesis in B. subtilis
.
References
1) Magnuson, K., Jackowski, S., Rock, C. O., and
Cronan, J. E., Regulation of fatty acid biosynthesis
in Escherichia coli. Microbiol. Rev., 57, 522–542
(1993).
2) Rock, C. O., and Cronan, J. E., Escherichia coli as a
model for the regulation of dissociable (type II) fatty
acid biosynthesis. Biochim. Biopys. Acta, 1302, 1–16
(1996).
3) Kaneda, T., Iso- and anteiso-fatty scids in bacteria:
biosynthesis, function and taxonomic signiˆcance.
Microbiol. Rev., 55, 288–302 (1991).
40
z
to FabD, respectively. The polyketide synthase
(PKS) and FAS are related not only in their mechan-
ism of carbon chain extension but also in their prima-
ry sequences. It thus appeared hold true that these
two systems evolved from a common origin after ear-
ly gene duplication. As the many biochemical steps
of FAS and PKS are similar, it seems quite reasona-
ble that these two system share common enzymatic
activities up to the point at which the pathways
diverge. In this context, it is noteworthy that FabD
has been implicated as functioning to charge the
ACP subunit of FAS and PKS prior to the condensa-
tion reaction.17) Thus the gene product of MAT
homologues in the crude extract may be a potential
alternate enzyme source, and may compensate the
deˆciency of the FabD activity for BCFA biosynthe-
sis.
4) Butterworth, P. H. W., and Bloch, K., Comperative
aspects of fatty acid synthesis in Bacillus subtilis and
Escherichia coli
.
Eur. J. Biochem., 12, 496–501
(1970).
5) Choi, K.-H., Heath, R. J., and Rock, C. O.,
b
-Keto-
acyl carrier protein synthase III (FabH) is a determin-
ing factor in branched-chain fatty acid biosynthesis.
J. Bacteriol., 182, 365–370 (2000).
6) Morbidoni, H. R., Mendoza, D., and Cronan, J. E.,
Bacillus subtilis acyl carrier protein is encoded in a
cluster of lipid biosynthesis genes. J. Bacteriol., 178,
4794–4800 (1996).
7) Kaneda, T., Biosynthesis of branched long-chain
fatty acids from the related short-chain
a-keto acid
The independence of BCFA biosynthesis on FabD

2114
H. OKU et al.
substrates by a cell-free system of Bacillus subtilis
Can. J. Microbiol., 19, 87–96 (1973).
8) Kawaguchi, A., Yoshimura, T., and Okuda, S., A
new method for the preparation of acyl-CoA
thioesters. J. Biochem., 89, 337–339 (1981).
9) Ellman, G. L., Tissue sulfhydryl groups. Arch.
Biochem. Biophys., 82, 70–77 (1950).
10) Kaneda, T., Biosynthesis of branched chain fatty
acids I. Isolation and identiˆcation of fatty acids
from Bacillus subtilis (ATCC 7059). J. Biol. Chem.
238, 1222–1228 (1963).
11) Heath, R. J., and Rock, C. O., Regulation of
malonyl-CoA metabolism by acyl-acyl carrier protein
.
factors aŠecting chain length. Can. J. Microbiol., 20,
1701–1708 (1974).
15) Oku, H., and Kaneda, T., Biosynthesis of branched-
chain fatty acids in Bacillus subtilis. A decarboxylase
is essential for branched-chain fatty acid synthetase.
J. Biol. Chem., 263, 18386–18396 (1988).
16) Arimura, N., and Kaneda, T., Type selective inhibi-
tion of microbial fatty acid synthases by thiolactomy-
cin. Arch. Microbiol., 160, 158–161 (1993).
17) Summers, R. G., Ali, A., Shen, B., Wessel, W. A.,
and Hutchinson, C. R., Malonyl-coenzyme A:acyl
carrier protein acyltransferase of Streptomyces glau-
cescens: a possible link between fatty acid and poly-
,
and
b
-ketoacyl-acyl carrier protein synthases in
ketide biosynthesis. Biochemistry, 34, 9389–9402
(1995).
Escherichia coli
.
J. Biol. Chem., 270, 15531–15538
(1995).
18) Hitchman, T. S., Crosby, J., Byrom, K. J., Cox, R.
J., and Simpson, T. J., Catalytic self-acylation of
type II polyketide synthase acyl carrier proteins.
Chem. Biol., 5, 35–47 (1998).
19) Matharu, A.-L., Cox, R. J., Crosby, J., Byrom, K.
J., and Simpson, T. J., MCAT is not required for in
vitro polyketide synthesis in a minimal actinorhodin
12) Han, L., Lobo, S., and Reynolds, K. A., Characteri-
zation of b-ketoacyl-acyl carrier protein synthase III
from Streptomyces glaucescens and its role in initia-
tion of fatty acid biosynthesis. J. Bacteriol., 180,
4481–4486 (1998).
13) Johnstone, A., and Thorpe, R., Precipitation tech-
niques in agar and agarose. In ``Immunochemistry in
Practice'', Blackwell Scientiˆc Publications, Oxford,
pp. 120–140 (1982).
polyketide synthase from Streptomyces coelicolor
Chem. Biol., 5, 699–711 (1998).
.
20) Zhou, P., Florova, G., and Reynolds, K. A., Polyke-
tide synthase acyl carrier protein (ACP) as a substrate
and a catalyst for malonyl ACP biosynthesis. Chem.
Biol., 6, 577–584 (1999).
14) Naik, D. N., and Kaneda, T., Biosynthesis of bran-
ched long-chain fatty acids by species of Bacillus
relative activity of three a-keto acid substrates and
: