Isolation, structural elucidation, and biosynthesis of 15-norlankamycin derivatives produced by a type-II thioesterase disruptant of Streptomyces rochei

Article abstract of DOI:10.1016/j.tet.2011.05.047

Lankamycin, a 14-membered macrolide antibiotic, contains a 3-hydroxy-2-butyl side chain at C-13. To analyze the function of lkmE, which encodes type-II thioesterase in the lankamycin cluster, we carried out a gene disruption experiment. Disruption of lkmE

Full text of DOI:10.1016/j.tet.2011.05.047

Tetrahedron 67 (2011) 5199e5205
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Isolation, structural elucidation, and biosynthesis of 15-norlankamycin derivatives
produced by a type-II thioesterase disruptant of Streptomyces rochei
*
Kenji Arakawa , Zhisheng Cao, Natsumi Suzuki, Haruyasu Kinashi
Department of Molecular Biotechnology, Graduate School of Advanced Sciences of Matter, Hiroshima University, 1-3-1 Kagamiyama, Higashi-Hiroshima, Hiroshima 739-8530, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Lankamycin, a 14-membered macrolide antibiotic, contains a 3-hydroxy-2-butyl side chain at C-13. To
analyze the function of lkmE, which encodes type-II thioesterase in the lankamycin cluster, we carried
out a gene disruption experiment. Disruption of lkmE resulted in a 70% decrease of lankamycin pro-
duction concomitant with an accumulation of novel lankamycin derivatives (LM-NS01A and LM-NS01B),
in which the C-13 side chain is replaced by a 1-carboxyethyl group. The biosynthetic origin of
1-carboxyethyl group was confirmed by incorporation of deuterium in [3-2H]3-methyl-2-oxobutyrate
into the C-14 position. These results indicate that the biosynthesis of LM-NS01A and LM-NS01B starts
from isobutyryl CoA in place of (S)-2-methylbutyryl CoA and LkmE removes the aberrantly loaded starter
unit and restores lankamycin production.
Received 29 March 2011
Received in revised form 10 May 2011
Accepted 10 May 2011
Available online 17 May 2011
Keywords:
Antibiotics
Biosynthesis
Thioesterase-II
Starter unit
Streptomyces
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
biosynthesis.10,11 Thus, the function of TE-II is different from species
to species and should be studied further.
Complex polyketides including macrolides are an important
family of antibiotics that possess a wide range of biological activi-
ties. Macrolides are assembled by type-I polyketide synthases
(PKSs) that contain sets of catalytic domains in giant polypeptides.1
Each module is responsible for one round of condensation and
reduction reactions in polyketide biosynthesis. During polyketide
assembly, acyl building substrates are tethered to acyl carrier pro-
teins (ACPs) as thioesters. After the completion of polyketide
extention, the polyketide chain bound to PKS is released and
cyclized by a thioesterase domain (TE-I) that is usually integrated
into the final PKS module.
Streptomyces rochei strain 7434AN4 carries three linear plasmid
pSLA2-L, -M, and -S and produces two structurally unrelated pol-
yketide antibiotics, lankamycin (1; Fig. 1) and lankacidin.12 Lanka-
mycin is a 14-membered macrolide antibiotic containing two
deoxysugars,
D L
-chalcose and 400-O-acetyl- -arcanose.13 The com-
plete nucleotide sequencing of the largest plasmid pSLA2-L
(210,614 bp) has revealed that the lankamycin biosynthetic gene
(lkm) cluster (orf24-orf53) is located on this plasmid.14 We pre-
viously determined the origin of a starter unit and the order of the
post-PKS modification steps including two hydroxylation and two
glycosylation steps in lankamycin biosynthesis.15,16 It was sug-
Additional thioesterases called type-II thioesterases (TE-IIs) are
coded as discrete proteins in many polyketide biosynthetic clusters.
They are thought to have an editing function to remove the aber-
rantly loaded acyl groups from the PKS and NRPS system.2e4 Dis-
ruption of TE-II genes resulted in a decrease of the overall titer of
tylosin (tylO),5 rifamycin (rifR),6 erythromycin (ery-ORF5),7 and
FR008/candicidin (fscTE).8 While, disruption of pikAV showed no
affect on picromycin production in Streptomyces venezuelae.9 On the
other hand, TE-IIs function for a release of the polyketide chain
from the PKS module in several polyether antibiotic
gested that L-isoleucine is oxidized to (S)-3-methyl-2-oxopentanoic
acid, which in turn is decarboxylated to (S)-2-methylbutyrate. Its
CoA ester, (S)-2-methylbutyryl CoA, is loaded on the loading ACP
domain (ACPL) in LkmAI, and then condensed with six extender
units of methylmalonyl CoA by three PKSs encoded by lkmAI-
lkmAIII (orf35-orf33) to synthesize 8,15-dideoxylankanolide. This
aglycon then receives the following modifications to afford 1; C-15
hydroxylation by LkmK, attachment of L-arcanose at C-3 hydroxyl
by LkmL, C-8 hydroxylation by LkmF, and attachment of
at C-5 hydroxyl by LkmI.
D-chalcose
The lkm cluster contains a TE-II gene, lkmE (orf25), which might
be involved in lankamycin biosynthesis. To reveal the function of
lkmE, we disrupted the lkmE gene, analyzed novel metabolites
produced by a disruptant, and carried out a feeding experiment, the
results of which are described in this paper.
* Corresponding author. Tel./fax: þ81 82 424 7767; e-mail address: karakawa@
hiroshima-u.ac.jp (K. Arakawa).
0040-4020/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2011.05.047

5200
K. Arakawa et al. / Tetrahedron 67 (2011) 5199e5205
Fig. 1. Chemical structures of lankamycin (1), LM-NS01A (2a), LM-NS01B (2b), LM-NS01A-Me (3a), and LM-NS01B-Me (3b). Me, methyl; Ac, acetyl.
2. Results and discussion
involved in catalysis, is the histidine residue indicated in a bold
letter in Fig. 2A.
2.1. Disruption of a type-II thioesterase gene lkmE
To reveal the function of LkmE in lankamycin biosynthesis, its
gene disruptant was constructed from a strain KA07, a lankamycin
high producer prepared by the disruption of the transcriptional
repressor gene srrB.17 The 424-bp ApaI fragment containing the
conserved motif mentioned above was deleted from lkmE (Fig. 2B).
As shown in Fig. 2C, the 4.6-kb and 1.0-kb BsiWI fragments in the
parent KA07 were replaced by a 5.2-kb BsiWI fragment in the
mutant NS01, which confirmed the gene disruption.
The lkm cluster comprises 30 open reading frames (ORFs) (orf24-
orf53), in which the type-II thioesterase (TE-II) gene lkmE (orf25) is
located on the left end. LkmE shows considerable similarities to TE-
II, such as SACE0729 (Gene Bank accession number, CAM00070;
57% identity and 66% similarity) for erythromycin biosynthesis and
TylO (AAA21345; 51% and 59%) for tylosin biosynthesis. Partial
amino acid sequences of LkmE and other bacterial TE-IIs are aligned
and compared in Fig. 2A. All of them contain the conserved GxSxG
motif (x¼any amino acid) characteristic of acyltransferases and
thioesterases, in which the serine residue functions as an active-site
center. A second conserved amino acid residue, which might be
2.2. Metabolites from the lkmE disruptant NS01
Metabolites of the lkmE disruptant NS01 were compared with
those of the parent KA07 by thin-layer chromatography (TLC)
Fig. 2. Function of type-II thioesterase (TE-II) LkmE. (A) Partial amino acid alignment of TE-IIs among bacterial PKSs. Conserved motif was boxed. Bold letters indicate catalytic
serine and histidine residues. LkmE (Accession number BAC76483, S. rochei 7434AN4); SACE0729 (Ery-orf5) (CAM00070, Saccharopolyspora erythraea NRRL2338); ScoT (CAC37888,
Streptomyces coelicolor A3(2)); AveG (BAC68663, Streptomyces avermitilis MA-4680); MonAX (AAO65810, Streptomyces cinnamonensis); PicAV (AAC69333, S. venezuelae); TylO
(AAA21345, Streptomyces fradiae); ChmI (AAS79448, Streptomyces bikiniensis). Conserved amino acid residues were marked as asterisk (identical) and colon (well conserved). (B)
Construction of the lkmE disruptant NS01 by homologous recombination. Ap, ApaI; Bs, BsiWI; Bg, BglII; Kp, KpnI. (C) Southern blot analysis of total DNA. Lane 1,
strain KA07 (parent)/BsiWI; lane 3, strain NS01 ( lkmE)/BsiWI. (D) TLC analysis of metabolites partially purified by Sephadex LH-20 chromatography. Metabolites were passed
through Sephadex LH-20 chromatography (1.5ꢀ40 cm) with methanol. Fractions were collected as 60 drops. TLC plates were developed with CHCl3/MeOH¼20:1. LC, lankacidin C.
l/HindIII; lane 2,
D

K. Arakawa et al. / Tetrahedron 67 (2011) 5199e5205
5201
(Fig. 2D). Strain NS01 produced 30% lankamycin (1) (4.5 mg/l)
compared with the parent strain KA07 (15.5 mg/l). In addition,
strain NS01 accumulated two novel lankamycin derivatives LM-
NS01A (2a) and LM-NS01B (2b) (1.2 mg/l), which were not pro-
duced by the parent strain KA07. Compounds 2a and 2b were
obtained as inseparable mixture in the ratio of 3:2 (Rf¼0.25 in
CHCl3/MeOH¼20:1). The molecular ion peaks of 2a and 2b
([MþNa]þ) were observed at 839 and 867, respectively.
The major ester LM-NS01A-Me (3a) was subjected to NMR and
ESI-MS analyses. The 1H and 13C NMR spectral data were summa-
rized in Table 1. The molecular formula of 3a was determined by
high resolution ESI-MS to be C42H70O16. In its 13C NMR, 42 carbons
were classified into 15 methyl, 3 methylene, 18 methine, and
6 quaternary carbons. The quaternary carbon corresponding to the
C-15 carboxyl group (dC¼173.7) showed long range HMBC corre-
lations with C-13 methine, C-14 methyl, and C-15 methoxy protons,
therefore, the C-13 side chain in 3a was determined to be
(1-methoxycarbonyl)ethyl. In addition, the C-8 quaternary carbon
in 1 was changed to a methine carbon (dC¼44.3) in 3a. This methine
proton signal (dH¼2.84) showed HMBC correlations with C-9 car-
bonyl and C-8 methyl carbons, indicating the absence of a hydroxyl
group at C-8 in 3a. Other signals of 3a were almost identical to
To separate 2a and 2b and elucidate their structures, the mix-
tures were methylated with diazomethane. Methyl derivatives 3a
and 3b were separable on TLC (Rf¼0.3 for 3a and 0.4 for 3b in
hexane/EtOAc¼1:1). The molecular ion peaks of 3a and 3b
([MþNa]þ) were observed at 853 and 881, respectively, which
suggested the presence of one carboxyl group in 2a and 2b.
Table 1
1H and 13C NMR data of lankamycin and its derivatives
No.
Lankamycin (1)
LM-NS01A-Me (3a)
LM-NS01B-Me (3b)
a
b
a
b
a
b
dC
dH
dC
dH
dC
dH
1
2
3
4
5
6
7
8
9
10
11
12
13
14
176.7 (s)
44.8 (d)
77.8 (d)
44.1 (d)
84.5 (d)
33.9 (d)
39.4 (t)
80.2 (s)
214.4 (s)
38.2 (d)
71.0 (d)
39.4 (d)
73.0 (d)
42.7 (d)
69.1 (d)
102.5 (d)
75.3 (d)
80.2 (d)
37.1 (t)
67.3 (d)
21.0 (q)c
96.6 (d)
30.6 (t)
72.6 (s)
73.9 (d)
62.6 (d)
16.8 (q)
21.0 (q)c
56.9 (q)
49.3 (q)
d
d
174.9 (s)
44.7 (d)
79.3 (d)
43.2 (d)
84.4 (d)
36.1 (d)
33.4 (t)
44.3 (d)
214.8 (s)
41.4 (d)
72.2 (d)
38.8 (d)
73.5 (d)
42.5 (d)
173.7 (s)
102.9 (d)
75.5 (d)
80.1 (d)
37.1 (t)
67.4 (d)
20.8 (q)c
97.9 (d)
31.1 (t)
72.4 (s)
73.7 (d)
62.6 (d)
16.9 (q)
20.7 (q)c
56.9 (q)
49.3 (q)
52.0 (q)
15.3 (q)
9.9 (q)c
18.5 (q)
18.9 (q)
9.2 (q)
d
174.8 (s)
44.7 (d)
79.6 (d)
43.0 (d)
84.8 (d)
36.2 (d)
33.5 (t)
43.0 (d)
214.7 (s)
41.5 (d)
72.2 (d)
38.8 (d)
73.5 (d)
42.5 (d)
173.7 (s)
102.1 (d)
72.4 (d)
80.6 (s)
41.3 (t)
66.9 (d)
20.9 (q)c
97.9 (d)
31.6 (t)
72.4 (s)
73.5 (d)
62.8 (d)
16.8 (q)
20.7 (q)c
d
d
2.80 (br)
3.95 (d, 4.9)e
1.83 (m)
3.51 (m)
2.21 (br)
1.93d
2.86 (m)
3.78 (d, 8.8)
1.78 (m)
3.49 (m)
2.00 (br)
1.34,1.80 (m)
2.84 (m)
d
2.86 (m)
3.76 (m)
1.80 (m)
3.53 (d, 8.5)f
2.00 (br)
1.33,1.82 (m)
2.85 (m)
d
d
d
3.15 (q, 6.7)g
4.87 (d, 8.9)g
1.93d
3.00 (m)
4.86 (dd, 9.5, 1.6)
1.84 (m)
5.15 (d, 10.1)h
2.78 (m)
d
2.98 (m)
4.88 (d, 9.8, 2.0)
1.88 (m)
5.15 (d, 10.4)h
2.80 (m)
d
4.83 (d, 6.4)h
1.82 (m)
3.71 (m)
4.33 (d, 7.7)
3.34 (dd, 7.7,8.9)
3.25 (m)
1.25,2.07 (m)
3.51 (m)
1.22 (d, 6.1)
5.05 (d, 4.6)
1.67,2.09 (m)
d
15
10
4.24 (d, 7.6)
3.31 (m)
3.23 (m)
1.26,2.03 (m)
3.47 (m)
1.21 (d, 6.1)
4.97 (d, 4.6)
1.70,2.07 (m)
d
4.69 (s)i
4.47 (q, 6.7)i
1.10d
1.06 (s)
3.44 (s)
4.53 (d, 7.7)
3.74 (d, 7.6)
d
20
30
40
1.55,1.73 (m)
3.94 (m)
1.18 (d, 7.0)
4.96 (d, 4.6)
1.69,2.07 (m)
d
4.72 (s)i
4.49 (q, 6.8)i
1.12 (d, 6.8)
1.06 (s)
50
60
100
200
300
400
4.67 (s)i
4.47 (q, 6.6)i
1.08d
1.05 (s)
3.44 (s)
3.30 (s)
d
500
600
300-CH3
30-OCH3
300-OCH3
15-OCH3
2-CH3
d
3.28 (s)
3.71 (s)
49.3 (q)
52.0 (q)
15.2 (q)
9.8 (q)
18.3 (q)
18.8 (q)
9.2 (q)
9.7 (q)
14.6 (q)
d
20.7 (q)c
169.9 (s)
20.9 (q)c
170.5 (s)
23.9 (q)
209.5 (s)
3.27 (s)
3.71 (s)
20.7 (q)
20.9 (q)c
14.4 (q)
27.0 (q)
10.1 (q)
9.8 (q)
11.2 (q)
19.7 (q)
20.9 (q)c
170.1 (s)
21.0 (q)c
170.7 (s)
d
1.17 (d, 7.4)
1.06d
1.20 (d, 7.0)
1.19 (d, 7.3)
1.10 (d, 7.1)
1.16 (d, 7.0)
1.13d
0.97 (d, 7.1)
0.99 (d, 7.9)
1.13d
4-CH3
6-CH3
8-CH3
1.12d
1.16 (d, 6.8)
1.34 (s)
1.10d
1.15 (d, 6.7)
1.13d
0.97 (d, 7.1)
1.00 (d, 7.0)
1.13d
10-CH3
12-CH3
14-CH3
15-CH3
11-OC(]O)CH3
11-OC(]O)CH3
400-OC(]O)CH3
400-OC(¼O)CH3
30-C(]O)CH3
30-C(]O)CH3
1.02d
9.8 (q)c
14.6 (q)
d
0.83 (d, 7.0)
1.15 (d, 6.4)
2.08 (s)
d
d
d
20.9 (q)c
169.9 (s)
21.0 (q)c
d
2.06 (s)
d
2.06 (s)
d
2.13 (s)
d
2.15 (s)
d
2.14 (s)
d
d
d
d
2.31 (s)
d
d
d
d
d
a
Multiplicity is shown in parenthesis.
Multiplicity and J value in Hz are shown in parenthesis.
Assignments are exchangeable.
b
c
d
e
f
Obscured by overlapping.
Vicinal proton coupling constant J3,4 is w0 Hz.
Vicinal proton coupling constant J5,6 is w0 Hz.
Vicinal proton coupling constant J10,11 is w0 Hz.
Vicinal proton coupling constant J12,13 is w0 Hz.
g
h
i
00,500
Vicinal proton coupling constant J4
is w0 Hz.

5202
K. Arakawa et al. / Tetrahedron 67 (2011) 5199e5205
those of 1. From these data, compound 3a was determined to be 8-
deoxylankamycin analog, which contained a (1-methoxycarbonyl)
ethyl side chain at C-13 in place of a 3-hydroxy-2-butyl side chain
(Fig. 1).
8,15-dideoxylankamycin compared with 1),15 these results show
the importance of hydroxyl groups for antimicrobial activity of
lankamycin. The MIC values of 3a and 3b indicated that
D
-chalcose
is a better component than 40,60-dideoxy-30-C-acetyl-
D-ribo-hex-
The molecular formula of LM-NS01B-Me (3b) was established to
be C43H70O17. The 1H and 13C assignments of 3b (Table 1) were al-
most identical to those of 3a, except for the signals corresponding
apyranose for antimicrobial activity.
2.3. Biosynthesis of LM-NS01A and LM-NS01B
to D-chalcose attached to the C-5 hydroxyl. The methoxy group at
the C-30 in 3a disappeared, while one acetyl moiety (dC¼209.5 for
a carbonyl and dC¼23.9 for a methyl group) appeared in 3b. The C-30
methine carbon in 3a was changed to a quaternary carbon
The biosynthetic origin of a starter unit for 2a and 2b may be
-valine. -Valine is oxidized to 3-methyl-2-oxobutyric acid, and
L
L
then decarboxylated to isobutyrate. To analyze the origin of the
starter unit for 2a and 2b, we carried out a feeding experiment of
deuterium labeled 3-methyl-2-oxobutyric acid. Deuterium labeled
compound, [3-2H]3-methyl-2-oxobutyric acid (5), was synthesized
as shown in Fig. 3A. Hydrogen atom at C-3 of 3-methyl-2-
oxobutyric acid (4) was replaced with deuterium atom in the
presence of deuterium oxide and pyridine. The 2H NMR of 2a and
2b obtained by feeding of 5 showed a distinct signal at 2.8 ppm,
which corresponds to the C-14 proton (Fig. 3B). This result indicates
that valine, a precursor of 3-methyl-2-oxobutyric acid, is a bio-
synthetic origin of the C-13 side chain of 2a and 2b.
(
dC¼80.6) in 3b. This quaternary carbon showed HMBC correlations
with methylene protons at C-40 (dH¼1.55 and 1.73) and a methyl
proton of the newly appeared acetyl group (dH¼2.31). These spec-
tral data allowed us to assign the structure of 3b, which contained
a novel deoxysugar, 40,60-dideoxy-30-C-acetyl-
D-ribo-hexopyranose,
instead of -chalcose in 3a (Fig. 1). The configuration of the new
D
asymmetric center at C-30 of 3b was established to be R by differ-
ential NOE experiments. The axialeaxial orientations of H-10
(
dH¼4.53) and H-20 (dH¼3.74) were confirmed by their coupling
constant of 7.7 Hz. This axial proton H-20
(dH¼3.74) showed
a moderate NOE (5%) on the methyl proton of the acetyl group
(
dH¼2.31), suggesting the equatorial orientation of the acetyl
branched-chain at C-30. This assignment was further strengthened
by the following biosynthetic consideration; 40,60-dideoxy-30-C-
acetyl-
D-ribo-hexopyranose is a possible biosynthetic intermediate
of D-
D
-aldgarose (3,10-O-carbo-4,6-dideoxy-3-C-[10-hydroxyethyl]-
ribo-hexopyranose), the configuration of whose branched-chain is
also R, in the biosynthesis of a neutral macrolide antibiotic aldga-
mycin E.18e20
The spectroscopic analysis revealed that the lkmE disruptant
accumulated novel 15-norlankamycin derivatives 2a and 2b, both
of which harbored a 1-carboxyethyl group at C-13 instead of
a 3-hydroxy-2-butyl group in 1. Compounds 2a and 2b lack
a hydroxyl group at C-8, suggesting that the 15-nor derivatives are
poor substrates of the P450 hydroxylase LkmF. To our surprise,
compound 2b contains
dideoxy-30-C-acetyl-
-ribo-hexapyranose, at the C-5 hydroxyl
instead of -chalcose in 2a. It is noteworthy that Streptomyces sp.
KMA001 produced two 16-membered macrolides, chalcomycin and
aldgamycin I, which contain -chalcose and decarboxylated
-aldgarose, respectively.21 Isolation of the lankamycin derivatives
a
novel branched-chain sugar, 40,60-
D
D
D
D
with altered sugar components suggests that glycosyltransferase
LkmI has a broad substrate specificity.
Deletion of the TE-II genes generally leads to a drastic reduction
of polyketide production. Gene inactivation of tylO in Streptomyces
fradiae resulted in a 85% loss of tylosin production, which was re-
stored by introduction of the intact tylO into the mutant.5 Removal
of rifR from Amycolatopsis mediterranei resulted in a 60% decrease in
rifamycin yield.6 Disruption of ery-ORF5 gene (SACE0729) in S.
erythraea caused a severe loss of erythromycin production and an
accumulation of 15-norerythromycin, which is synthesized from an
aberrant acetate starter unit instead of propionate.7 Similarly, in-
activation of the lkmE gene resulted in a 70% loss of lankamycin
production and an accumulation of 15-norlankamycin derivatives
2a and 2b. This result suggests that the LkmE protein controls
specific incorporation of a starter unit.
Fig. 3. Analysis of starter unit biosynthesis of 2a and 2b. (A) Synthetic scheme of
[3-2H]3-methyl-2-oxobutyrate (5). Reagents: a, D2O, pyridine. Me, methyl. (B) 2H NMR
spectrum of labeled 2a and 2b obtained by feeding of compound 5 (I), and 1H NMR
spectrum of unlabeled 2a and 2b (II).
During the biosynthesis of 2a and 2b, either of the two methyl
groups in valine is oxidized to carboxylate. To assign the C-14
configuration in 2a and 2b, we carried out feeding experiments of
stereospecifically labeled isobutyrate. As shown in Fig. 4A, the
synthesis of (2R)-[3-2H]isobutyrate 9 was started from the com-
mercially available methyl (S)-(þ)-3-hydroxyisobutyrate 6. The
hydroxyl of 6 was protected with a tetrahydropyranyl (THP) group,
and then reduction of the methyl ester with lithium aluminum
hydride gave alcohol 7 in 99% yield. Tosylate of 7, obtained by
treatment with p-toluenesulfonyl chloride, was treated with
LiAl2H4 to afford the deuterium labeled compound 8 in 68% yield
(two steps). Finally, compound 8 was converted to (2R)-[3-2H]iso-
butyrate 9 by acidic hydrolysis, Jones’ oxidation, and neutralization
with NaOH. Compound (2S)-[3-2H]isovalerate 13 was obtained
from methyl (R)-(ꢁ)-3-hydroxyisobutyrate 10 in the same manner
for 9.
The antimicrobial activities of 15-norlankamycin derivatives
were determined by the agar dilution methods against Micrococcus
luteus. The mixture of 2a and 2b exhibited moderate antimicrobial
activity with a MIC value of 4.0
that of lankamycin 1 (MIC value of 1.0
ester derivatives 3a and 3b showed antimicrobial activities with the
MIC values of 4.0 g/ml and 16 g/ml, respectively. Together with
mg/ml, which was four-fold less than
mg/ml). In addition, methyl
m
m
previous results of deoxylankamycins (50% relative activity for
8-deoxylankamycin, 18% for 15-deoxylankamycin, 2.5% for

K. Arakawa et al. / Tetrahedron 67 (2011) 5199e5205
5203
(1) in 30% yield compared with the parent, and in addition accu-
mulated the 15-norlankamycin derivatives 2a and 2b, both of
which harbored a 1-carboxyethyl group at C-13 instead of a
3-hydroxy-2-butyl group in 1. Based on these results, we propose
the early biosynthetic pathway of lankamycin including the func-
tion of LkmE in Fig. 5. During lankamycin biosynthesis, LkmE hy-
drolyzes the aberrantly loaded isobutyryl CoA starter unit from ACP
domain of the loading module in LkmAI. Then the vacant ACP do-
main accepts the specific starter unit, (S)-2-methylbutyryl CoA, to
afford 1. Disruption of lkmE caused an accumulation of 2a and 2b in
some extent by utilization of an isobutyryl CoA instead of (S)-2-
methylbutyryl CoA starter unit, and resulted in a severe decrease
of lankamycin production compared with the parent strain. Thus,
the main function of LkmE was shown to remove an aberrantly
loaded starter unit(s) and restore the enzymatic activity of lanka-
mycin PKSs that prefer to use (S)-2-methylbutyryl CoA.
4. Experimental
4.1. Strains and culture conditions
S. rochei wild type strain 7434AN4 and strain 51252 that carries
only pSLA2-L were described previously.12 Strain KA07, a disruptant
of the transcriptional repressor gene srrB, was used as a parent
strain.17 Cosmid B10 (nt 3341-48,756 of pSLA2-L), which carries the
lkmE gene, was constructed previously.14 YM medium (0.4% yeast
extract, 1.0% malt extract, and 0.4% D-glucose, pH 7.3) was used for
antibiotic production. YEME liquid medium22 was used for prepa-
ration of Streptomyces protoplasts. Protoplasts were regenerated on
R1M solid medium.23 Escherichia coli XL1-Blue was used for rou-
tine cloning and constructing of targeting plasmids. E. coli strains
were grown in Luria Bertani (LB) medium supplemented with
ampicillin (100
mg/ml). Lithium aluminum deuteride (98 atom%
enriched) was purchased from Cambridge Isotope Laboratory Inc.,
USA. CDCl3 (99.8 atom% enriched) and D2O (99.8 atom% enriched)
were purchased from Acros, USA.
4.2. Spectroscopic instruments
NMR spectra were recorded on a JEOL LA-500 spectrometer
equipped with a field gradient accessory. For 1H and 13C NMR,
CDCl3 and D2O were used as solvents. Chemical shifts were recor-
ded as a
d
value based on resident solvent signals (dC¼77.0 in CDCl3
and dH¼4.65 in D2O), or internal standard signals of tetrame-
thylsilane (dH¼0) or dioxane (dC¼66.5). For 2H NMR, CHCl3 and H2O
were used as solvents and internal references. Electrospray
ionization-mass (ESI-MS) spectra were measured by an LTQ Orbi-
trap XL mass spectrometer (Thermo Fisher Scientific, USA). Optical
rotations were measured with a JASCO DIP-370 polarimeter. IR
spectra were recorded on a JASCO FT/IR-8000 spectrometer.
Fig. 4. Determination of stereochemistry at C-14 position in 2a and 2b. (A) Synthetic
scheme of (2R)-[3-2H]isobutyrate (9) and (2S)-[3-2H]isobutyrate (13). Reagents: a,
3,4-dihydro-2H-pyran, p-toluenesulfonic acid, CH2Cl2; b, lithium aluminum hydride, THF;
c, p-toluenesulfonyl chloride, Et3N, 4-dimethylaminopyridine, CH2Cl2; d, lithium alumi-
num deuteride, THF; e, Jones’ reagent, acetone, and then NaOH. THP, tetrahydropyranyl.
(B) 2H NMR spectrum of labeled 2a and 2b obtained by feeding of compounds 9 (I),13 (II),
and 1H NMR spectrum of unlabeled 2a and 2b (III).
4.3. Construction of the lkmE disruptant NS01
The synthesized (2R)-[3-2H]isobutyrate 9 and (2S)-[3-2H]iso-
butyrate 13 was fed to the cell culture of strain NS01. Analysis of the
mixture 2a and 2b by 2H NMR showed that the deuterium atom of
9 was specifically incorporated into the C-14 methyl of 2a and 2b.
On the other hand, deuterium atom of 13 was not incorporated at
all. These results demonstrated that a pro-S methyl group of iso-
butyrate moiety in 8,15-dideoxy-15-norlankanolide (Fig. 5) was
stereospecifically oxidized to a carboxylate group during the bio-
synthesis of 2a and 2b.
The 5.0-kb KpnI-BglII fragment from cosmid B10 was cloned into
pUC19 predigested with KpnI and BamHI to give pSI001. The 424-bp
ApaI fragment in the middle of the lkmE gene was eliminated from
pSI001 to afford pSI002. The vector part of pSI002 was replaced
with an E. coli-Streptomyces shuttle vector pRES1824 to give a tar-
geting plasmid pSI003.
This plasmid was transformed into protoplasts of S. rochei strain
KA07. The transformant was grown in YEME liquid medium sup-
plemented with thiostrepton to give a plasmid-integrated (single
crossover) strain. The thiostrepton-resistant colonies were sub-
jected to a sequential cultivation in YEME liquid medium in the
absence of thiostrepton to give the double crossover strain. Gene
disruption was confirmed by Southern hybridization using a DIG
3. Conclusion
We have analyzed the function of the TE-II enzyme LkmE in
lankamycin biosynthesis. The lkcE disruptant produced lankamycin

5204
K. Arakawa et al. / Tetrahedron 67 (2011) 5199e5205
Fig. 5. Proposed biosynthetic pathway of 1, 2a, and 2b. Upper panel shows a loading of (S)-2-methylbutyryl CoA, a metabolite of L-isoleucine, for the synthesis of 1. Lower panel
shows a loading of non-specific isobutyryl CoA, a metabolite of L-valine, for the synthesis of 2a and 2b. The isobutyryl CoA starter unit is hydrolyzed by LkmE to reload the specific
(S)-2-methylbutyryl CoA.
DNA Labeling and Detection Kit (Roche Diagnostics, Germany)
4.7. Synthesis of the labeled compounds
according to the manufacture’s protocol.
4.7.1. [3-2H]3-methyl-2-oxobutyric acid, calcium salt (5). A solution
of calcium 3-methyl-2-oxobutyrate (4) (1.0 g) in pyridine (20 ml)
and deuterium oxide (10 ml) was stirred at room temperature for
1 day. The mixture was concentrated in vacuo, and this cycle was
repeated further three times. More than 90% deuterium atom was
incorporated at the C-3 position in 5.
4.4. Isolation of metabolites
The strains were incubated in YM liquid medium at 28 ꢂC for
2 days. The culture broth was extracted with equal volume of ethyl
acetate twice. The combined organic phase was dried (Na2SO4),
filtered, and concentrated to dryness. The resulting residue was
purified by Sephadex LH-20 (GE Healthcare) with methanol. The
fractions containing 2a and 2b were combined, and further purified
by silica gel chromatography with CHCl3/methanol¼50:1e10:1.
Yields in strain KA07 (per liter): 1, 15.5 mg; lankacidin C, 25.2 mg.
Yields in strain NS01 (per liter): 1, 4.5 mg; 2a and 2b, 1.2 mg; lan-
kacidin C, 22.7 mg.
1H NMR (D2O)
NMR (D2O)
d C
1.00 (6H, s), 2.91 (<0.1H, m, trace of CHMe2); 13
d
17.0, 37.7 (t, J¼18.5 Hz), 172.8, 212.1; 2H NMR (H2O)
d
2.85; High resolution ESI-MS: observed m/z 116.0463 [MꢁH]ꢁ
(calcd for C5H62HO3, 116.0463).
4.7.2. (2R)-2-Methyl-3-(tetrahydro-2H-pyran-2-yloxy)propan-1-ol
(7). A solution of methyl (S)-(þ)-3-hydroxyisobutyrate 6 (980 mg,
8.30 mmol), 3,4-dihydro-2H-pyran (2.00 ml, 21.9 mmol), and
p-toluenesulfonic acid (18 mg) in CH2Cl2 (25 ml) was stirred at
room temperature for 2 h. Saturated aqueous NaHCO3 (15 ml) was
added, and the mixture was extracted with EtOAc twice. The
combined organic phase was washed with brine, dried (Na2SO4),
filtered, and concentrated to dryness. The residue was purified over
silica gel with hexane/EtOAc (10:1, v/v) to give tetrahydropyranyl
ether (1.68 g, quant.) as an oil. This compound (1.68 g, 8.30 mmol)
in THF (25 ml) was treated with lithium aluminum hydride
(730 mg, 19.2 mmol) at 0 ꢂC for 12 h. The mixture was quenched by
4.5. Esterification of LM-NS01
To a solution of 2a and 2b (12 mg) in ether (8 ml) was added an
ethereal solution of diazomethane at 0 ꢂC until the mixture turns
yellow. The mixture was stirred at room temperature for 1 h, and
concentrated in vacuo. The residue was purified by silica gel chro-
matography with hexane/EtOAc (2:1e1:1, v/v) to give LM-NS01A-
Me (3a) (4.5 mg; Rf¼0.3 in hexane/EtOAc¼1:1) and LM-NS01B-Me
(3b) (3.3 mg; Rf¼0.4 in hexane/EtOAc¼1:1). The 1H and 13C NMR
spectral data of 1, 3a, and 3b are summarized in Table 1.
successive addition of water (800 ml), 5 N NaOH (800 ml), and water
(2.4 ml). The resulting mixture was filtered through a pad of Celite
with ether, and the filtrate and washings were concentrated in
vacuo. The residue was purified over silica gel with hexane/EtOAc
(2:1, v/v) to give alcohol 7 (1.45 g, quant.) as a colorless oil.
Compound 3a: high resolution ESI-MS; observed m/z 853.4537
[MþNa]þ (calcd for C42H70O16Na, 853.4556). [
a
]
ꢁ74.6 (c 0.932,
D25
CHCl3).
Diastereomer mixture: 1H NMR (CDCl3)
d 0.90 (1.5H, d,
Compound 3b: high resolution ESI-MS; observed m/z 881.4482
J¼6.8 Hz), 0.91 (1.5H, d, J¼6.8 Hz), 1.52e1.60 (4H, m), 1.70e1.82 (2H,
m), 2.01e2.08 (1H, m), 2.62 (0.5H, dd, J¼5.2 and 6.1 Hz), 2.67 (0.5H,
dd, J¼5.2 and 6.4 Hz), 3.35 (0.5H, dd, J¼7.3 and 9.5 Hz), 3.48e3.70
[MþNa]þ (calcd for C43H70O17Na, 881.4505). [
a
]
D25
ꢁ80.5 (c 0.950,
CHCl3).
(4H, m), 3.84e3.90 (1.5H, m), 4.58 (1H, br); 13C NMR (CDCl3)
d 13.4,
13.5, 19.6, 25.3, 25.3, 30.5, 30.6, 35.4, 35.6, 62.5, 62.6, 67.4, 67.5, 72.1,
72.2, 99.1, 99.4; High resolution ESI-MS: observed m/z 197.1146
[MþNa]þ (calcd for C9H18O3Na, 197.1148).
4.6. Antimicrobial activity
The antimicrobial activities of compounds 2a, 2b, 3a, and 3b
were measured by the agar dilution methods. The growth of test
microorganism Micrococcus luteus was judged after 24 h incubation
at 28 ꢂC.
4.7.3. 2-[(2R)-[3-2H]Isobutoxy]tetrahydro-2H-pyran (8). p-Toluene-
sulfonyl chloride (2.68 g, 14.1 mmol) was added to a solution
of
7 (1.38 g, 7.92 mmol), Et3N (5.0 ml, 36.1 mmol), and

K. Arakawa et al. / Tetrahedron 67 (2011) 5199e5205
5205
4-dimethylaminopyridine (30 mg) in CH2Cl2 (30 ml) at 0 ꢂC, and the
mixture was stirred at room temperature overnight. Water (20 ml)
was added, and the mixture was extracted with EtOAc twice. The
combined organic phase was washed with brine, dried (Na2SO4),
filtered, and concentrated to dryness. The residue was purified over
silica gel with hexane/EtOAc (4:1, v/v) to give tosylate (2.59 g,
quant.) as an oil. A mixture of tosylate (2.60 g, 7.92 mmol) and
lithium aluminum deuteride (500 mg, 11.9 mmol) in THF (35 ml)
was stirred at 0 ꢂC for 12 h. The mixture was quenched by suc-
4.7.7. Sodium (2S)-[3-2H]isobutyrate (13). Compound 12 (590 mg,
3.71 mmol) was treated in the same manner as described for the
preparation of 9 to give sodium (2S)-[3-2H]isobutyrate 13 (410 mg,
quant.) as a white powder.
1H NMR (D2O)
m); 13C NMR (D2O)
d
0.90 (2H, m), 0.92 (3H, d, J¼7.0 Hz), 2.26 (1H,
d
20.0 (t, J¼18.5 Hz), 20.3, 37.7, 184.1; 2H NMR
(H2O)
d 0.90; High resolution ESI-MS: observed m/z 88.0516
[MꢁH]ꢁ (calcd for C4H62HO2, 88.0514).
cessive addition of water (800 ml), 5 N NaOH (800 ml), and water
4.8. Feeding of labeled compounds
(2.4 ml). The resulting suspension was filtered through a pad of
Celite with ether, and the filtrate and washings were concentrated
in vacuo. The residue was chromatographed over silica gel with
hexane/EtOAc (9:1, v/v) to give the labeled compound 8 (862 mg,
68%) as a colorless oil.
One hundred milligram of the labeled compound (5, 9, or 13)
was added at 8 h period to 1 L culture of strain NS01. The culture
was incubated at 28 ꢂC for additional 2 days and analyzed.
Diastereomer mixture: 1H NMR (CDCl3)
d 0.87e0.91 (2H, m),
0.91 (1.5H, d, J¼6.8 Hz), 0.93 (1.5H, d, J¼6.8 Hz), 1.51e1.62 (4H, m),
1.68e1.74 (1H, m), 1.81e1.90 (2H, m), 3.15 (1H, dd, J¼6.4 and
9.5 Hz), 3.50 (1H, dd, J¼7.1 and 9.2 Hz), 3.51 (1H, m), 3.85 (0.5H, dd,
J¼3.7 and 8.3 Hz), 3.88 (0.5H, dd, J¼3.1 and 8.0 Hz), 4.58 (1H, t,
Acknowledgements
We thank to Mr. D. Kajiya and Ms. T. Amimoto (N-BARD, Hir-
oshima University) for the measurement of high resolution mass
spectra. This work was supported by a Grant-in-Aid for Scientific
Research from the Ministry of Education, Culture, Sports, Science
and Technology of Japan.
J¼4.3 Hz); 13C NMR (CDCl3)
d
19.1 (dt, J¼7.3 and 19.6 Hz), 19.4 (d,
J¼7.3 Hz), 19.6, 25.5, 28.4, 30.7, 62.1, 74.3, 98.9; 2H NMR (CHCl3)
d
0.93; High resolution ESI-MS: observed m/z 182.1260 [MþNa]þ
(calcd for C9H172HO2Na, 182.1262).
4.7.4. Sodium (2R)-[3-2H]isobutyrate (9). Jones’ reagent (3 ml) was
dropwisely added to a solution of 8 (576 mg, 3.62 mmol) in ac-
etone (30 ml) at 0 ꢂC, and the mixture was stirred at room tem-
perature for 2 h. Excess oxidant was destroyed by addition of
2-propanol at 0 ꢂC, and the resulting greenish solution was
passed through a pad of Celite while washing with ether. The
filtrate and washings were concentrated in vacuo. The residue
was dissolved in water, and titrated by 1 N NaOH. The solution
was washed with ether, and the aqueous phase was lyophilized to
give sodium (2R)-[3-2H]isobutyrate 9 (400 mg, quant.) as a white
powder.
Supplementary data
Supplementary data associated with this article can be found in
the online version, at doi:10.1016/j.tet.2011.05.047. These data
include MOL files and InChiKeys of the most important compounds
described in this article.
References and notes
1. Staunton, J.; Weissman, K. J. Nat. Prod. Rep. 2001, 18, 380e416.
2. Heathcote, M. L.; Staunton, J.; Leadlay, P. F. Chem. Biol. 2001, 8, 207e220.
3. Kim, B. S.; Cropp, T. A.; Beck, B. J.; Sherman, D. H.; Reynolds, K. A. J. Biol. Chem.
2002, 277, 48028e48034.
4. Kotowska, M.; Pawlik, K.; Butler, A. R.; Cundliffe, E.; Takano, E.; Kuczek, K.
Microbiology 2002, 148, 1777e1783.
5. Butler, A. R.; Bate, N.; Cundliffe, E. Chem. Biol. 1999, 6, 287e292.
6. Doi-Katayama, Y.; Yoon, Y. J.; Choi, C.-Y.; Yu, T.-W.; Floss, H. G.; Hutchinson, C. R.
J. Antibiot. 2000, 53, 484e495.
7. Hu, Z.; Pfeifer, B. A.; Chao, E.; Murli, S.; Kealey, J.; Carney, J. R.; Ashley, G.; Khosla,
C.; Hutchinson, C. R. Microbiology 2003, 149, 2213e2225.
8. Zhou, Y.; Meng, Q.; You, D.; Li, J.; Chen, S.; Ding, D.; Zhou, X.; Zhou, H.; Bai, L.;
Deng, Z. Appl. Environ. Microbiol. 2008, 74, 7235e7242.
1H NMR (D2O)
m); 13C NMR (D2O)
d
0.90 (2H, m), 0.92 (3H, d, J¼7.0 Hz), 2.26 (1H,
d
20.0 (t, J¼18.5 Hz), 20.3, 37.7, 184.1; 2H NMR
(H2O)
d 0.87; High resolution ESI-MS: observed m/z 88.0516
[MꢁH]ꢁ (calcd for C4H62HO2, 88.0514).
4.7.5. (2S)-2-Methyl-3-(tetrahydro-2H-pyran-2-yloxy)propan-1-ol
(11). Methyl (R)-(ꢁ)-3-hydroxyisobutyrate 10 (1.09 g, 9.26 mmol)
was treated in the same manner as described for the preparation of
7 to give 11 (1.70 g, 97%) as a colorless oil.
9. Chen, S.; Roberts, J. B.; Xue, Y.; Sherman, D. H.; Reynolds, K. A. Gene 2001, 263,
255e264.
10. Harvey, B. M.; Hong, H.; Jones, M. A.; Hughes-Thomas, Z. A.; Goss, R. M.;
Heathcote, M. L.; Bolanos-Garcia, V. M.; Kroutil, W.; Staunton, J.; Leadlay, P. F.;
Spencer, J. B. ChemBioChem 2006, 7, 1435e1442.
Diastereomer mixture: 1H NMR (CDCl3)
d
0.90 (1.5H, d, J¼7.1 Hz),
0.91 (1.5H, d, J¼7.1 Hz), 1.53e1.59 (4H, m), 1.67e1.82 (2H, m),
2.01e2.07 (1H, m), 2.60 (0.5H, br), 2.64 (0.5H, br), 3.35 (0.5H, dd,
J¼7.4 and 10.3 Hz), 3.48e3.70 (4H, m), 3.84e3.87 (1.5H, m), 4.58
11. Liu, T.; You, D.; Valenzano, C.; Sun, Y.; Li, J.; Yu, Q.; Zhou, X.; Cane, D. E.; Deng, Z.
(1H, br); 13C NMR (CDCl3)
d 13.4, 13.5, 19.6, 25.3, 25.3, 30.5, 30.6,
Chem. Biol. 2006, 13, 945e955.
12. Kinashi, H.; Mori, E.; Hatani, A.; Nimi, O. J. Antibiot. 1994, 47, 1447e1455.
35.4, 35.6, 62.5, 62.5, 67.4, 67.5, 72.1, 72.2, 99.1, 99.4; High resolution
ESI-MS: observed m/z 197.1146 [MþNa]þ (calcd for C9H18O3Na,
197.1148).
€
€
€
13. Gaumann, E.; Hutter, R.; Keller-Schierlein, W.; Neipp, L.; Prelog, V.; Zahner, H.
Helv. Chim. Acta 1960, 80, 601e606.
14. Mochizuki, S.; Hiratsu, K.; Suwa, M.; Ishii, T.; Sugino, F.; Yamada, K.; Kinashi, H.
Mol. Microbiol. 2003, 48, 1501e1510.
15. Arakawa, K.; Kodama, K.; Tatsuno, S.; Ide, S.; Kinashi, H. Antimicrob. Agents
4.7.6. 2-[(2S)-[3-2H]Isobutoxy]tetrahydro-2H-pyran (12). Compound
11 (1.14 g, 6.54 mmol) was treated in the same manner as described
for the preparation of 8 to give 12 (770 mg, 78%) as a colorless oil.
Chemother. 2006, 50, 1946e1952.
16. Arakawa, K.; Suzuki, T.; Kinashi, H. Actinomycetologica 2008, 22, 35e41.
17. Arakawa, K.; Mochizuki, S.; Yamada, K.; Noma, T.; Kinashi, H. Microbiology 2007,
153, 1817e1827.
Diastereomer mixture: 1H NMR (CDCl3)
d 0.89e0.92 (2H, m),
18. Ellestad, G. A.; Kunstmann, M. P.; Lancaster, J. E.; Mitscher, L. A.; Morton, G.
0.91 (1.5H, d, J¼6.8 Hz), 0.93 (1.5H, d, J¼6.8 Hz), 1.51e1.63 (4H, m),
1.68e1.74 (1H, m), 1.81e1.90 (2H, m), 3.16 (1H, dd, J¼6.4 and
9.5 Hz), 3.50 (1H, dd, J¼7.1 and 9.5 Hz), 3.50 (1H, m), 3.85 (0.5H, dd,
J¼3.4 and 8.3 Hz), 3.88 (0.5H, dd, J¼2.8 and 7.7 Hz), 4.58 (1H, t,
Tetrahedron 1967, 23, 3893e3902.
19. Schmid, R.; Grisebach, H. Eur. J. Biochem. 1970, 14, 243e252.
20. Paulsen, H.; Redlich, H. Angew. Chem., Int. Ed. Engl. 1972, 11, 1021e1023.
21. Park, J.-S.; Yang, H.-O.; Kwon, H.-C. J. Antibiot. 2009, 62, 171e175.
22. Kieser, T.; Bibb, M. J.; Buttner, M. J.; Chater, K. F.; Hopwood, D. A. Practical
Streptomyces Genetics; The John Innes Foundation: Norwich, United Kingdom,
2000.
23. Zhang, H.; Shinkawa, H.; Ishikawa, J.; Kinashi, H.; Nimi, O. J. Ferment. Bioeng.
1997, 83, 217e221.
24. Ishikawa, J.; Niino, Y.; Hotta, K. FEMS Microbiol. Lett. 1996, 145, 113e116.
J¼4.3 Hz); 13C NMR (CDCl3)
d
19.1 (dt, J¼7.3 and 19.6 Hz), 19.4 (d,
J¼7.3 Hz), 19.6, 25.5, 28.4, 30.7, 62.1, 74.3, 98.9; 2H NMR (CHCl3)
d
0.93; High resolution ESI-MS: observed m/z 182.1260 [MþNa]þ
(calcd for C9H172HO2Na, 182.1262).