Complete stereochemistry and preliminary structure-activity relationship of rakicidin A, a hypoxia-selective cytotoxin from micromonospora sp.

Article abstract of DOI:10.1021/np500276c

The complete stereochemistry of rakicidin A, a hypoxia-selective cytotoxin produced by Micromonospora sp., was unambiguously established by extensive chemical degradation and derivatization studies. During the PGME derivatization-based configurational analysis of 3-hydroxy-2,4,16-trimethylheptadecanoic acid, an irregular Δδ distribution was observed, which necessitated further acylation of the 3-hydroxy group to resolve the inconsistency. A hydrogenated derivative of rakicidin A, its ring-opened product, and two congeners with different alkyl chain lengths were tested for hypoxia-selective cytotoxicity. The results indicated that both the conjugated diene unit and appropriate chain length are essential for the unique activity of rakicidin A.

Full text of DOI:10.1021/np500276c

Note
pubs.acs.org/jnp
Complete Stereochemistry and Preliminary Structure−Activity
Relationship of Rakicidin A, a Hypoxia-Selective Cytotoxin from
Micromonospora sp.
†
†
‡
†
†,§
Naoya Oku, Shouhei Matoba, Yohko Momose Yamazaki, Ryoko Shimasaki, Satoshi Miyanaga,
,†
and Yasuhiro Igarashi*
†Biotechnology Research Center and Department of Biotechnology, Toyama Prefectural University, 5180 Kurokawa, Imizu, Toyama
9
39-0398, Japan
‡
Institute of Microbial Chemistry, 3-14-23 Kamiosaki, Shinagawa-ku, Tokyo 141-0021, Japan
S Supporting Information
*
ABSTRACT: The complete stereochemistry of rakicidin A, a
hypoxia-selective cytotoxin produced by Micromonospora sp.,
was unambiguously established by extensive chemical degra-
dation and derivatization studies. During the PGME
derivatization-based configurational analysis of 3-hydroxy-
2
,4,16-trimethylheptadecanoic acid, an irregular Δδ distribu-
tion was observed, which necessitated further acylation of the
-hydroxy group to resolve the inconsistency. A hydrogenated
3
derivative of rakicidin A, its ring-opened product, and two
congeners with different alkyl chain lengths were tested for
hypoxia-selective cytotoxicity. The results indicated that both the conjugated diene unit and appropriate chain length are essential
for the unique activity of rakicidin A.
ocally advanced solid tumors usually contain hypoxic
pound was originally isolated from a culture broth of Micro-
7
L
regions (oxygen tension ≤2.5 mmHg) due to unorganized1
monospora sp. as a moderate cytotoxin; later, rakicidin A was
8
vasculature and extremely high metabolic requirements.
Classically, neoplastic cells in hypoxic loci resist radiotherapy,
shown to have antiallergy activity in a mouse asthma model.
9
,10
Apart from rakicidin congeners, two groups of actinomycetes-
derived depsitripeptides, the vinylamycin-microtermolide A
2
which utilizes oxygen as the sensitizer. However, evidence
1
1,12
13
shows that these cells activate adaptive responses to withstand
the adverse condition and undergo mutation caused by defective
DNA repair and subsequent clonal evolution, resulting in the
development of drug resistance, acquisition of aggressive
class
and the BE-43547 complexes, contain the vDha
residue (Figure 1). Although 15 years have passed since their
discovery, the absolute stereochemistry of this class of
3
phenotype, and poor prognostic outcome. Therefore, hypoxia
has great potential as a target for the definitive therapy of cancer,
and attempts to develop hypoxia-activated prodrugs or inhibitors
of hypoxia-associated proteins have identified several promising
4
drug candidates.
Rakicidin A is a hypoxia-selective cytotoxin identified through
a hypoxia/normoxia differential cytotoxicity screen of 20 000
5
microbial broths. Although the activity of most anticancer
agents significantly decreases under hypoxic conditions, rakicidin
A exhibited enhanced activity against many cancer cell lines, with
the highest hypoxia/normoxia selectivity indices being achieved
with PANC-1 human pancreatic carcinoma (14.9) and HCT-8
5
human ileocecal adenocarcinoma cells (17.5). Moreover, the
hypoxia-selectivity was valid for myelogenous leukemic cells,
6
thus adding to the list of potential applications.
Figure 1. Rakicidin A and related metabolites.
Received: March 27, 2014
Rakicidin A is a depsitripeptide composed of three unusual
amino acids, sarcosine, β-hydroxyasparagine (βOHAsn), and
vinylogous dehydroalanine (vDha), and a β-hydroxy fatty acid, 3-
hydroxy-2,4,16-trimethylheptadecanoic acid (Htha). This com-
©
XXXX American Chemical Society and
American Society of Pharmacognosy
A
dx.doi.org/10.1021/np500276c | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Note
compounds is available only for the two common amino acid
residues in vinylamycin. This is probably because of the very poor
solubility of this metabolite class in most common organic
solvents. In fact, pure rakicidin A is practically insoluble in
Scheme 1. Preparation of a Solubility-Improved Derivative,
Tetrahydrorakicidin A (2), and Attempt at Chemical
Conversion to Establish the Absolute Configuration of Two
Chiral Centers in 2
EtOAc, Et O, and THF and poorly soluble in alcohols, aqueous
2
MeCN, and chloroform. DMSO was used as the solvent for the
NMR experiments of rakicidin A; however, once the solvent is
evaporated, flaky residues are formed, which can be partially
dissolved in the same solvent. The vDha residue may be
responsible for this solubility problem because the replacement
of other residues in the related metabolites did not remarkably
improve the solubility.
A series of preliminary mode of action studies revealed that the
unique activity of rakicidin A is not the result of inhibition of
hypoxia-adaptive responses mediated by hypoxia-inducible
factor-1 (HIF-1), inhibition of the transcriptional activation of
5
this pivotal protein, or the production of oxygen radicals. Thus, a
novel mechanism is indicated. Although rakicidin A was found to
be inactive in vivo against both the murine M109 lung carcinoma
7
and P388 leukemia tumor models, the assignment of its absolute
Scheme 2. Determination of the Relative Stereochemistry of
βOHAsp (6) and Htha (7)
a
stereochemistry is still of great benefit in identifying new
therapeutic targets and developing new anticancer drugs.
Rakicidin A has five chiral centers, and they are present in the
βOHAsn and Htha residues. Attempts to derivatize rakicidin A
by common derivatization procedures, including saponification
(
aqueous LiOH/DMF), methanolysis (NaOMe/MeOH), or
double-bond cleavage (RuCl /NaIO or O /NaBH ), to obtain
3
4
3
4
the derivatives or fragments suitable for chiral analysis failed,
leaving rakicidin A virtually unreacted. Because semitranslucent
lumps, supposedly generated from the flakes of rakicidin A,
remained persistently insoluble during the reactions, we
attributed the chemical inertness of this peptide to its poor
solubility in the reaction solvents. The only exception was the
acid hydrolysis at elevated temperature, affording substantial
amounts of products in the lipophilic extract. However, the
expected product, Htha (7), was contaminated by impurities,
preventing its reliable characterization by NMR.
Obviously, a better solvent for rakicidin A was needed, and we
found that pyridine followed by DMF satisfied this requirement.
An ideal derivatization would be to modify the vDha residue
while using these solvents. Two options were possible:
hydrogenation and cleavage of the olefin moieties. The former
would be preferable because it conserves the cyclic structure and
should facilitate a structure−activity relationship (SAR) study,
which may elucidate the role of vDha.
a
Data reproduced from ref 20.
The hydrogenation of rakicidin A using Pd/C-catalyst in
1
4
pyridine/DMF afforded tetrahydrorakicidin A (2) as a single
peak during HPLC purification (Scheme 1, 52%). As expected,
this derivative was readily soluble in previously incompatible
solvents and thus opened a way for further transformation.
Methanolysis of 2 with NaOMe afforded carboxylic acid 3, which
was methylated with TMS-diazomethane to give methyl ester 4.
Derivatization of the two carbinol carbons with (S)- or (R)-α-
interaction liquid chromatography mode. A sample from the HCl
layer was spotted on a TLC plate and developed with i-PrOH/
25% aqueous NH /2 mM CuSO (4:2:1). Upon visualization
3
4
with ninhydrin, a yellow spot appeared at R 0.35, which was
f
identical to that for a threo isomer (R 0.35 for L-threo; 0.25 for L-
f
erythro). Thus, the original βOHAsn residue in rakicidin A
18
should have a threo-configuration. To establish its absolute
configuration, the HCl layer was dried and reacted with a chiral
derivatizing reagent, 5-fluoro-2,4-dinitrophenyl-5-L-leucine
15
methoxyphenylacetic acid (MPA) was not successful, presum-
ably due to the steric hindrance exerted by the flanking methyl
1
6
19
groups.
amide (FDLA). A complete set of authentic standards were
Then, we applied the previously successful acid hydrolysis to 2.
prepared by derivatizing L-erythro- and L-threo-βOHAsp with
both the enantiomers of FDLA. The reversed-phase LC-MS
analysis exhibited the coelution of the L-DLA-derivatized 6 with
D-DLA-L-threo-βOHAsp. Because the latter was a chromato-
graphic equivalent of L-DLA-D-threo-βOHAsp, the configuration
of the βOHAsn residue in rakicidin A was concluded to be D-
threo.
Extraction of the hydrolysate solution with Et O separated
2
βOHAsp (6) in the 6 N HCl layer and Htha (7) in the Et O layer,
2
the yield of the latter being almost quantitative (97%, Scheme
1
7
2
). The assignment of NMR signals was possible without
purifying the products. The relative stereochemistry of 6 was
determined by silica gel TLC operating in the hydrophilic
B
dx.doi.org/10.1021/np500276c | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Note
To establish the relative stereochemistry of 7, the chemical
shift data for this fatty acid were compared with those available
for the four diastereomers of 3-hydroxy-2,4,6-trimethylheptanoic
2
0
acid. However, none of the profiles matched that of 7 (Figure
S6). The lack of diagnostic validity in this protocol is probably
because of the difference in aliphatic chain lengths between the
three contiguous chiral centers and the terminal isopropyl group.
Then, we carefully examined the NMR data of the model
compounds and found that the spin multiplicities of the 3-
oxymethine protons significantly discriminate the three config-
uration classes. In particular, a pair of small (∼3 Hz) and large
(
∼8 Hz) coupling constants indicates a 2,4-anti-dimethyl
relationship, whereas two middle values (∼6 Hz) indicate a
2
Figure 2. Different conformations proposed for 8b (upper) and 9b
(lower). Note that the presence (8b) or absence (9b) of a hydrogen
bond (dotted lines) on the acyl group defines the orientation of the
PGME group along the C1−C2 axis (yellow bond). As a consequence,
H3 and H4 (yellow atoms) in these derivatives are coplaced in the
opposite compartments divided by the chiral anisotropy group.
,4-syn-dimethyl relationship. The related proton H3 in 7 (δ
H
3
.65 ppm) exhibited a combination of small (3.7 Hz) and large
(
7.9 Hz) values (Scheme 2), indicating a 2,4-anti configuration.
An alternative decision between (2R*,3R*,4S*) and
(
2S*,3R*,4R*) was made based on a pair of homonuclear
decoupling experiments to identify coupled partners associated
12
the structure of microtermolide A tentatively assigned the
configuration of its fatty acid unit as (2R*,3R*,4R*), based on
the computation of the minimum energy conformation.
However, the JH2,H3 and JH3,H4 values (10.0 and 1.0 Hz,
respectively) are very similar to those reported for the Htha
with these values. Upon selective irradiation of H2 (δ 2.68
ppm), the H3 resonance was modulated into a doublet with a 3.7
Hz splitting, while irradiation of H4 (δ 1.66 ppm) resulted in a
residual splitting of 7.9 Hz (Figure S22). This allowed the
assignment of J
relative stereochemistry of 7 as (2R*,3R*,4S*).
H
H
3
3
= 3.7 Hz and J
= 7.9 Hz, and hence the
H3,H4
H2,H3
7
unit in rakicidin A (10.2 Hz, ∼0 Hz), and considering the global
resemblance between the two metabolites, the proposed
stereochemistry should be examined by chemical methods.
The absolute stereochemistry of Htha was established by chiral
21
anisotropy analysis using PGME (phenylglycine methyl ester).
Crude 7 was reacted with either (S)- or (R)-PGME, and the
The hypoxia-selective cytotoxicities of tetrahydrorakicidin A
2), its acyclic derivative 3, rakicidin B, and a new congener,
7
(
chemical shift differences of the products 8a and 8b (Δδ = δ (S)
H
rakicidin E (1: Supporting Information), were evaluated along
with rakicidin A (Table 1). A significant decrease in the
cytotoxicity against PANC-1 and HCT-8 cells under both the
oxygenation conditions was observed for 2 and 3, indicating that
the conjugated diene functionality in the vDHA residue is
essential for cytotoxicity. Moreover, the hypoxia selectivity
gradually decreased with the extension of the aliphatic chain.
These functionalities should be conserved when exploiting the
structure of rakicidin A for drug development and designing
bioprobes to study its mode of action.
−
δ (R)) were calculated to deduce the absolute configuration at
H
the C2. However, the signs for Δδ values around the same chiral
center were irregularly distributed (Scheme 3), preventing the
Scheme 3. Determination of the Absolute Stereochemistry of
Htha (7)
EXPERIMENTAL SECTION
General Experimental Procedures. Cosmosil 75C -PREP
Nacalai Tesque Inc., 75 μm) was used for ODS flash chromatography.
■
(
1
8
Semipreparative HPLC was performed on an Agilent Technologies
260 Infinity LC-DAD system with an automated fraction collector
1
using a Cosmosil Cholester column (Nacalai Tesque Inc., 1 × 25 cm) for
purification of rakicidins A, B, and E (1) and a Shimadzu LC-10AT
pump equipped with a Shimadzu SPD-10A UV−vis detector using a
Cosmosil AR-II column (Nacalai Tesque Inc., 1 × 25 cm) for
purification of 2, 8a/8b, and 9a/9b. NMR spectra were obtained on a
1
Bruker AVANCE 500 spectrometer at 500 MHz for H. Residual solvent
peaks at δH 3.31, 7.27, and 2.50 ppm and HSQC cross-peaks δ /δ
H
C
3
.31/49.2, 7.27/77.0, and 2.50/39.5 ppm in CD OD, CDCl , and
3 3
assignment of its absolute configuration. This anomalous result
was attributed to the conformational restriction caused by a
hydrogen bond formed between the carboxylate oxygen and 3-
hydroxy group (Figure 2). To remove this constraint, 8a and 8b
were acetylated to afford esters 9a and 9b. As expected, a
systematic distribution of the signs for Δδ values was obtained,
which supported a 2S-configuration. By correlating this assign-
ment to the already established relative stereochemistry, the
absolute configuration of 7 was concluded to be (2S,3S,4R).
The stereochemistry established for rakicidin A is readily
applicable to rakicidin B and rakicidin D because of the similarity
in the NMR data and optical properties. A recent publication on
DMSO-d , respectively, were used as chemical shift reference signals.
6
TLC silica gel 60 F254 (Merck KGaA) was used for thin-layer
chromatography. Analysis of βOHAsp derivatized with FDLA was
performed on an Agilent 1100 LC-DAD system using a Cosmosil 5PE-
MS column (Nacalai Tesque Inc., 4.6 × 250 mm). HR-ESITOFMS were
recorded on a Bruker micrOTOF focus mass spectrometer. Optical
rotation and UV and CD spectra were recorded on a JASCO DIP-3000
polarimeter, a Hitachi U-3210 spectrophotometer, and a JASCO J-
7
20W spectropolarimeter, respectively.
Extraction and Isolation. Fermentation of the rakicidins producer
strain Micromonospora sp. TP-A0860 was described in our previous
article. The 1-butanol extract (59.9 g) from the fermentation broth
(17.6 L) was chromatographed on silica gel, eluting with a step gradient
2
2
C
dx.doi.org/10.1021/np500276c | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Note
Table 1. Cytotoxicity (IC : μM) of Rakicidins A, B, and E (1), Tetrahydrorakicidin A (2), and Carboxylic Acid (3) under Two
5
0
Oxygenation Conditions and Indices of Hypoxia Selectivity
a
cell
condition
rakicidin A
rakicidin B
1
2
3
Dox
2.4
HCT-8
normoxia
hypoxia
3.9
2.1
9.4
1.8
5.2
17.9
4.7
3.8
27.7
6.5
33.9
63.4
0.54
27.4
>80
0.30
12.8
2.3
0.21
10.0
3.1
19.6
0.12
1.1
b
SI
4.3
PANC-1
normoxia
hypoxia
19.3
75.4
0.26
0.37
6.3
0.71
4.3
5.4
b
SI
0.21
a
b
Doxorubicin. Index of hypoxia selectivity calculated from the equation IC (normoxia)/IC (hypoxia).
50
50
1
of CHCl /MeOH from 1:0 to 0:1 (v/v). Fractions eluted with CHCl /
TFA, 4.0 mL/min, 210 nm) to afford 3 (0.5 mg, 0.8 μmol): H NMR
3
3
MeOH = 1:1 and 0:1 were combined (total 7.028 g) and then separated
on ODS by a stepwise elution with aqueous MeCN. A fraction eluted
between 70% and 80% aqueous MeCN (628.1 mg) was purified by
reversed-phase HPLC on a Cosmosil Cholester column using 70%
aqueous MeCN as a solvent (4.0 mL/min, UV = 280 nm) to afford
rakicidin A (10.8 mg, t = 12.7 min), rakicidin B (7.3 mg, t = 17.9 min),
(500 MHz, CD OD) δ 5.01 (d, J = 14.6 Hz, 1H, βOHAsn α-CH), 4.63
3
(d, J = 18.5 Hz, 1H, βOHAsn β-CH), 4.14 (s, 2H, Sar α-CH ), 3.92 (m,
2
1H, Apa H-4), 3.58 (m, 1H, Htha H-3), 3.04 (s, 3H, Sar N-CH ), 2.44
3
(m, 1H, Htha H-2), 1.55 (m, 1H, Htha H-4), 1.50 (m, 1H, Htha H-16),
1.30 (brs, 22H, Htha H-5−H-15), 1.17 (m, 2H, Apa, H-3), 1.15 (m, 3H,
Apa H-5), 1.08 (m, 3H, Htha H-19), 0.87 (d, J = 7.0 Hz, 6H, Htha H-17,
R
R
13
and a new congener, rakicidin E (1) (1.8 mg, t = 25.2 min). The
H-18), 0.86 (m, 3H, Htha H-20); C NMR (data assigned from HSQC,
R
physicochemical properties of rakicidins A and B were essentially the
CD OD) δ 76.6 (Htha C-3), 72.7 (βOHAsn β-CH), 56.3 (βOHAsn α-
3
7,10
same as those previously reported.
CH), 51.5 (Sar α-CH ), 46.0 (Apa C-4), 45.4 (Htha C-2), 40.1 (Apa, C-
2
Rakicidin E (1): white floc; [α]20 −4.3 (c 0.1, benzene); UV
3), 36.8 (Sar N-CH ), 29.5 (Htha C-5 - C-15), 23.1 (Htha C-17, C18),
D
3
(
2
EtOH) λmax nm (log ε) 260 (4.8), 219 (4.7); CD (EtOH) λ nm (Δε)
21.1 (Apa C-5), 15.7 (Htha C-19), 13.1 (Htha C-20). Apa CH -2 was
2
−
59 (+0.2), 219 (−0.5); HRESITOFMS 633.4223 [M − H] ,
not observed.
1
13
C H N O ; calcd 633.4223; H and C NMR data, see Supporting
Methyl Ester Formation of 3. To a solution of 3 (0.3 mg, 0.48
μmol) in MeOH (150 μL) was added a 10% n-hexane solution of TMS-
diazomethane (150 μL), and the reaction mixture was stirred for 30 min
at room temperature. The solution was then quenched by AcOH (150
3
4
58
4
7
Information.
Evaluation of Hypoxia-Selective Cytotoxicity. Human color-
ectal adenocarcinoma cell line HCT-8 or human pancreatic
adenocarcinoma cell line PANC-1 was seeded in duplicated 96-well
1
μL) and concentrated in vacuo to afford 4 (0.3 mg, 0.47 μmol): H NMR
4
microplates at a density of 2 × 10 /100 μL medium per well and
(500 MHz, CD OD) δ 4.15 (s, 1H, Apa Sar α-CH b), 4.08 (s, 1H, Sar α-
3
2
preincubated for 24 h to ensure complete adherence to the substratum.
To these microcultures were added serially diluted sample solutions, and
test cultures thus prepared were separately incubated in either a
normoxic (5% CO ) or hypoxic condition (1% O ). Hypoxic conditions
CH a), 3.92 (m, 1H, Apa H-4), 3.72 (s, 3H, βOHAsn α-CO CH ), 3.58
2
2
3
(m, 1H, Htha H-3), 3.09 (s, 3H, Sar N-CH ), 2.47 (m, 2H, Apa H-2),
3
2.44 (m, 1H, Htha H-2), 1.56 (m, 1H, Htha H-4), 1.31 (brs, 22H, Htha
H-5−H-15), 1.17 (m, 3H, Apa H-5), 1.17 (m, 2H, Apa H-3), 1.08 (d, J =
7.1 Hz, 3H, Htha H-19), 0.87 (d, J = 6.7 Hz, 6H, Htha H-17, H-18), 0.86
2
2
were achieved using a MIC-101 modular incubator chamber (Billups-
Rosenberg, Del Mar, CA, USA) equipped with an oxygen indicator.
After 2 d of incubation, cytotoxicity at both the oxygenation conditions
was compared by the crystal violet method under normoxic
13
(m, 3H, Htha H-20); C NMR (data assigned from HSQC, CD OD) δ
3
75.5 (Htha C-3), 51.3 (βOHAsn α-CO CH ), 49.2 (Sar α-CH ), 44.7
2
3
2
(Apa C-4), 44.3 (Htha C-2), 39.0 (Apa C-3), 36.0 (Sar N-CH ), 35.0
3
23
conditions.
(Htha C-4), 29.6 (Apa C-2), 21.9 (Htha C-17, C-18), 19.9 (Apa C-5),
14.5 (Htha C-19), 11.9 (Htha C-20). Signals for α- and β-methines
βOHAsn and Htha CH-16 were not observed.
Hydrogenation of Rakicidin A. To a solution of rakicidin A (2.5
mg, 4.12 μmol) in a mixture of pyridine (150 μL) and DMF (200 μL)
was added Pd/C (2.5 mg). The reaction mixture was stirred for 2 h
Acid Hydrolysis of 2. Compound 2 (3.4 mg, 5.57 μmol) was placed
in a hydrolysis tube and dissolved in 6 N HCl (600 μL). The tube was
degassed, sealed, and heated at 110 °C for 18 h. After leaving the
under a H atmosphere at room temperature and then filtered through
2
Celite to remove Pd/C. The filtrate and an ethanolic wash of the
retentate were combined, and the resulting concentrate was purfied by
HPLC (Cosmosil Cholester column, 60% aqueous MeCN for 5 min and
hydrolysis tube until cool, the solution was extracted with Et O. The
2
aqueous layer was concentrated in vacuo to give a residue containing
amino acid 6, while the combined organic extracts were concentrated in
1
%/min concentration ramp to 85%, 4.0 mL/min, 210 nm) to give
1
1
tetrahydrorakicidin A (2) (1.3 mg, 2.14 μmol): H NMR (500 MHz,
vacuo to afford crude fatty acid 7 (1.8 mg, 5.49 μmol) as a yellow oil: H
CD OD) δ 5.25 (d, J = 1.9 Hz, 1H, βOHAsn α-CH), 4.97 (dd, J = 3.2,
NMR for 7 (500 MHz, CDCl ) δ 3.65 (dd, J = 2.8, 8.3 Hz, 1H, H-3),
3
3
8.7 Hz, 1H, Htha H-3), 4.77 (d, J = 1.9 Hz, 1H, βOHAsn β-CH), 4.40 (d,
2.68 (m, 1H, H-2), 1.63 (m, 1H, H-4), 1.53 (m, 1H, H-16), 1.27 (brs,
J = 18.6 Hz, 1H, Sar α-CH b), 3.93 (m, 1H, Apa H-4), 3.88 (d, J = 18.6
20H, H-5−H14), 1.22 (m, 3H, H-19), 1.16 (m, 2H, H-15), 0.90 (d, J =
2
1
3
Hz, 1H, Sar α-CH a), 2.96 (s, 3H, Sar N-CH ), 2.84 (m, 1H, Htha H-2),
7.1 Hz, 3H, H-20), 0.87 (d, J = 6.7 Hz, 6H, H-17, H-18); C NMR for 7
2
3
2
4
.22 (m, 1H, Apa H-2b), 2.10 (m, 1H, Apa H-2a), 1.98 (m, 1H, Htha H-
), 1.53 (m, 1H, Htha H-16), 1.30 (brs, 20H, Htha H-5−H-14), 1.18
(data assigned from HSQC, CDCl ) δ 75.6 (C-3), 42.2 (C-2), 38.7 (C-
3
15), 34.7 (C-4), 29.7 (C-5−C-14), 27.6 (C-16), 22.3 (C-17, C-18), 13.8
(C-19), 12.7 (C-20).
Derivatization of 8 with Modified Marfey’s Reagents. To a
(
m, 2H, Apa H-3), 1.18 (m, 2H, Htha H-15), 1.16 (m, 3H, Apa H-5),
1
=
.15 (m, 3H, Htha H-19), 0.91 (d, J = 6.7 Hz, 3H, Htha H-20), 0.88 (d, J
6.6 Hz, 6H, Htha H-17, H-18); C NMR (data assigned from HSQC,
13
solution of 6 in 0.1 N NaHCO (100 μL) was added a solution of L-
3
CD OD) δ 81.3 (Htha C-3), 71.0 (βOHAsn β-CH), 55.2 (βOHAsn α-
FDLA (50 μg, 0.16 mol) in acetone (50 μL) at room temperature. The
reaction mixture was heated at 60 °C for 40 min and then quenched with
0.2 N HCl (50 μL) to afford 6a.
Derivatization of L-threo β-OHAsp Standards with Modified
Marfey’s Reagents. To a solution of L-threo β-OHAsp (1.0 mg) in 0.1
3
CH), 45.4 (Apa C-4), 43.0 (Htha C-2), 38.9 (Htha C-15), 35.4 (Htha C-
4
), 35.1 (Sar N-CH ), 32.8 (Htha H-15), 29.5 (Htha C-5−C-14), 28.8
3
(
Apa C-2), 28.0 (Htha C-16), 21.8 (Htha C-17, C-18), 19.7 (Apa H-5),
14.6 (Htha C-20), 14.5 (Htha C-19).
Saponification of 2. Compound 2 (0.8 mg, 1.31 μmol) in a mixture
N NaHCO (100 μL) was added a solution of L-FDLA (50 μg, 0.16 mol)
3
of MeOH (220 μL) and 1 N LiOH (693 μL) was stirred overnight at
room temperature. To quench and further acidify the reaction, 1 N HCl
in acetone (50 μL). The reaction mixture was stirred for 30 min at 50 °C
and then quenched with 0.1 N HCl (50 μL). In an entirely analogous
manner, D-DLA-L-threo β-OHAsp was prepared using D-FDLA.
HPLC Analysis of FDLA Derivatives. The quenched reaction
solutions were diluted with 50% aqueous MeCN containing 0.05% TFA
(
1.5 mL) was added to the solution, and this was extracted with Et O.
2
The combined organic extracts were concentrated and then purified by
HPLC (Cosmosil AR-II column, 60% aqueous MeCN containing 0.05%
D
dx.doi.org/10.1021/np500276c | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Note
(
100 μL), and 5 μL aliquots were analyzed by HPLC (Cosmosil 5PE-
AUTHOR INFORMATION
■
MS column, 9% aqueous MeCN containing 50 mM AcONH , 1.0 mL/
4
Corresponding Author
E-mail (Y. Igarashi): yas@pu-toyama.ac.jp.
Present Address
Faculty of Health Sciences, Hokkaido University, Kita-12,
Nishi-5, Sapporo, Hokkaido 060-0812, Japan.
min, 340 nm). The retention times for the D-FDLA and L-FDLA
derivatives of L-threo β-OHAsp were 18.8 and 21.3 min, respectively,
whereas that for 6a was 18.8 min.
*
§
Preparation of Htha (R)-PGME Amide (8a). To a solution of 7
(
0.4 mg, 1.22 μmol) in DMF (10 μL) was added (R)-PGME·HCl (0.6
mg, 2.99 μmol), PyBOP (1.2 mg), HOBt (0.5 mg), and TEA (200 μL) at
room temperature. The reaction mixture was stirred for 1 h and then
concentrated in vacuo. The dried residue was directly purified by HPLC
Notes
The authors declare no competing financial interest.
(
Cosmosil 5C18-AR-II column, 90% MeCN to 100% MeCN) to afford
1
ACKNOWLEDGMENTS
This work was supported by a fund from Research Supporters
Association for Toyama Prefectural University FY2011.
compound 8a as a colorless oil (0.4 mg, 0.89 μmol): H NMR (500
MHz, CD OD) δ 7.37 (m, 6H, PGME Phe), 5.48 (s, 1H, PGME β-CH),
3
1
■
3
.70 (s, 3H, PGME OCH ), 3.56 (dd, J = 4.1, 8.3 Hz, 1H, H-3), 2.61 (m,
3
H, H-2), 1.56 (m, 1H, H-4), 1.52 (m, 1H, H-16), 1.28 (brs, 20H, H-5 -
H-14), 1.17 (m, 2H, H-15), 1.10 (m, 3H, H-19), 0.87 (d, J = 6.6 Hz, 6H,
REFERENCES
H-17, H-18), 0.84 (d, J = 6.9 Hz, 3H, H-20); 13C NMR (data assigned
■
(
(
1) Vaupel, P.; Mayer, A. Cancer Metastasis Rev. 2007, 26, 225−239.
2) Moeller, B. J.; Richardson, R. A.; Dewhirst, M. W. Cancer Metastasis
Rev. 2007, 26, 241−248.
3) Luoto, K. R.; Kumareswaran, R.; Bristow, R. G. Genome Integr.
013, 4, 5.
from HSQC, CD OD) δ 129.0 (PGME Phe), 76.7 (C-3), 58.0 (PGME
3
β-CH), 52.8 (PGME OCH ), 44.8 (C-2), 40.1 (C-15), 36.1 (C-4), 30.6
3
(
C-5−C-14), 22.9 (C-17, C-18), 13.6 (C-20). C-16 and C-19 were not
observed.
Preparation of Htha (S)-PGME Amide (8b). Compound 7 (0.4
(
2
(
(
4) Wilson, W. R.; Hay, M. P. Nat. Rev. Cancer 2011, 11, 393−410.
5) Yamazaki, Y.; Kunimoto, S.; Ikeda, D. Biol. Pharm. Bull. 2007, 30,
mg, 1.22 μmol) was derivatized with (S)-PGME·HCl (2.4 mg, 12.0
μmol) by essentially the same procedure using PyBOP (4.8 mg), HOBt
1
261−265.
(
2.2 mg), and TEA (160 μL) to give 8b (0.3 mg, 0.67 μmol): H NMR
(
6) Takeuchi, M.; Ashihara, E.; Yamazaki, Y.; Kimura, S.; Nakagawa, Y.;
(
500 MHz, CD OD) δ 7.36 (m, 6H, PGME Phe), 5.48 (s, 1H, PGME β-
3
Tanaka, R.; Yao, H.; Nagao, R.; Hayashi, Y.; Hirai, H.; Maekawa, T.
Cancer Sci. 2011, 102, 591−596.
7) McBrien, K. D.; Berry, R. L.; Lowe, S. E.; Keddermann, K. M.;
Bursuker, I.; Huang, S.; Klohr, S. E. J. Antibiot. 1995, 48, 1446−1452.
8) Kajiura, T.; Suzuki, H.;Sato, S.; Ishizaki, S.; Shinozaki, J.; Shiozaki,
M. WO01/68121A1, 2001.
9) Hu, J.-F.; Wunderlich, D.; Sattler, I.; Feng, X.-Z.; Grabley, S.;
Thiericke, R. Eur. J. Org. Chem. 2000, 3353−3356.
10) Igarashi, Y.; Shimasaki, R.; Miyanaga, S.; Oku, N.; Onaka, H.;
Sakurai, H.; Saiki, I.; Kitani, S.; Nihira, T.; Wimonsiravude, W.;
Panbangred, W. J. Antibiot. 2010, 63, 563−565.
11) Igarashi, M.; Shida, T.; Sasaki, Y.; Kinoshita, N.; Nagasawa, H.;
Hamada, M.; Takeuchi, T. J. Antibiot. 1999, 52, 873−879.
12) Carr, G.; Poulsen, M.; Klassen, J. L.; Hou, Y.; Wyche, T. P.; Bugni,
T. S.; Currie, C. R.; Clardy, J. Org. Lett. 2012, 14, 2822−2825.
13) Nishioka, H.; Nakajima, S.; Nagashima, M.; Kojiri, K.; Suda, H. JP
Patent 10147594, 1998.
CH), 3.69 (s, 3H, PGME OCH ), 3.54 (m, 1H, H-3), 2.62 (m, 1H, H-2),
3
1
.56 (m, 1H, H-4), 1.52 (m, 1H, H-16), 1.28 (brs, 20H, H-5−H-14),
(
1
.17 (m, 2H, H-15), 1.06 (m, 3H, H-19), 0.87 (m, 6H, H-17, H-18), 0.87
13
(
(
m, 3H, H-20); C NMR (data assigned from HSQC, CD OD) δ 130.0
3
(
PGME Phe), 77.0 (C-3), 58.0 (PGME β-CH), 52.9 (PGME OCH3),
4
4.7 (C-2), 40.1 (C-15), 36.0 (C-4), 30.5 (C-5−C14), 27.8 (C-16), 22.7
(
(
C-17, C-18), 14.1 (C-19), 13.4 (C-20).
Acetylation of 8a. To a solution of compound 8a (0.4 mg, 0.89
(
μmol) in pyridine (400 μL) was added Ac O (200 μL) at room
2
temperature. The reaction mixture was stirred for 1 h, lyophilized, and
then purified by HPLC with the same condition adopted for the
(
purification of 8a to afford compound 9a (0.1 mg, 0.20 μmol) as a
1
colorless oil: H NMR (500 MHz, CD OD) δ 7.37 (m, 6H, PGME
3
(
Phe), 5.46 (s, 1H, PGME β-CH), 5.08 (dd, J = 2.2, 10.3 Hz, 1H, H-3),
3
.68 (s, 3H, PGME OCH ), 2.83 (m, 1H, H-2), 1.88 (s, 3H, OAc), 1.73
3
(
(
=
m, 1H, H-4), 1.28 (m, 20H, H-5−H-14), 1.16 (m, 2H, H-15), 1.09 (d, J
13
7.2 Hz, 3H, H-19), 0.89 (m, 3H, H-20), 0.87 (m, 6H, H-17, H-18); C
(
(
14) Sajiki, H.; Hirota, K. Chem. Pharm. Bull. 2003, 51, 320−324.
15) Trost, B. M.; Belletire, J. L.; Godleski, S.; McDougal, P. G.;
Balkovec, J. M.; Baldwin, J. J.; Christy, M. E.; Ponticello, G. S.; Varga, S.
NMR (data assigned from HSQC, CD OD) δ 127.8 (PGME Phe), 76.7
3
(
5
C-3), 56.4 (PGME β-CH), 51.4 (PGME OCH ), 38.6 (C-15), 29.3 (C-
−C-14), 21.6 (C-17, C-18), 12.1 (C-20). C-2, C-3, C-16, and C-19
3
L.; Springer, J. P. J. Org. Chem. 1986, 51, 2370−2374.
were not observed.
(
2
(
16) Exner, C. J.; Turks, M.; Fonquerne, F.; Vogel, P. Chem.Eur. J.
Acetylation of 8b. Compound 9b (0.2 mg, 0.41 μmol) was
prepared by essentially the same procedure described above except for
the amounts of reagents used: 8b (0.3 mg, 0.67 μmol), pyridine (300
011, 17, 4246−4253.
17) Extraction with EtOAc caused formation of βOHAsp ethyl ester.
1
(18) Kato, A.; Nakaya, S.; Ohashi, Y.; Hirata, H.; Fujii, K.; Harada, K. J.
Am. Chem. Soc. 1997, 119, 6680−6681.
19) Harada, K.; Fujii, K.; Hayashi, K.; Suzuki, M. Tetrahedron Lett.
996, 37, 3001−3004.
20) Turk, J. A.; Visbal, G. S.; Lipton, M. A. J. Org. Chem. 2003, 68,
μL), and Ac O (150 μL). H NMR (500 MHz, CD OD) δ 7.36 (m, 6H,
2
3
PGME Phe), 5.41 (s, 1H, PGME β-CH), 5.14 (dd, J = 2.4, 10.0 Hz, 1H,
(
1
(
H-3), 3.67 (s, 3H, PGME OCH ), 2.84 (m, 1H, H-2), 1.88 (s, 3H, OAc),
3
1
.74 (m, 1H, H-4), 1.51 (m, 1H, H-16), 1.28 (m, 20H, H-5−H-14), 1.17
(
m, 2H, H-15), 1.09 (d, J = 7.2 Hz, 3H, H-19), 0.91 (m, 3H, H-20), 0.87
13
7841−7844.
(
m, 6H, H-17, H-18); C NMR (data assigned from HSQC, CD OD) δ
3
(
(
21) Nagai, Y.; Kusumi, T. Tetrahedron Lett. 1995, 36, 1853−1856.
22) Igarashi, Y.; Miyanaga, S.; Onaka, H.; Takeshita, M.; Furumai, T. J.
127.7 (PGME Phe), 76.8 (C-3), 57.0 (PGME β-CH), 52.6 (PGME
OCH ), 42.2 (C-2), 39.0 (C-15), 33.8 (C-4), 29.4 (C-5−C-14), 27.8
3
Antibiot. 2005, 58, 350−352.
23) Kueng, W.; Silber, E.; Eppenberger, U. Anal. Biochem. 1989, 182,
6−19.
(
C-16), 21.8 (C-17, C-18), 13.3 (C-19), 12.2 (C-20).
(
1
ASSOCIATED CONTENT
■
*
S
Supporting Information
H NMR chemical shift differences of 7 relative to literature
1
values, results of hypoxia/normoxia differential cytotoxicity
1
13
testings, comparison of H and C NMR data between 1 and
rakicidin A, NMR spectra for 1−4, 7, 8a/b, and 9a/b. This
material is available free of charge via the Internet at http://pubs.
acs.org.
E
dx.doi.org/10.1021/np500276c | J. Nat. Prod. XXXX, XXX, XXX−XXX