Turnagainolides A and B, cyclic depsipeptides produced in culture by a Bacillus sp.: Isolation, structure elucidation, and synthesis

Article abstract of DOI:10.1021/np200033y

Two new cyclic depsipeptides, turnagainolides A (1) and B (2), have been isolated from laboratory cultures of a marine isolate of Bacillus sp. The structures of 1 and 2, which are simply epimers at the site of macrolactonization, were elucidated by analysis of NMR data and chemical degradation. A total synthesis of the turnagainolides confirmed their structures. Turnagainolide B (2) showed activity in a SHIP1 activation assay. (Chemical Equation Presented).

Full text of DOI:10.1021/np200033y

ARTICLE
pubs.acs.org/jnp
Turnagainolides A and B, Cyclic Depsipeptides Produced in Culture by
a Bacillus sp.: Isolation, Structure Elucidation, and Synthesis
Dehai Li,† Gavin Carr,† Yonghong Zhang,† David E. Williams,† Ashraf Amlani,‡ Helen Bottriell,†
Alice L.-F. Mui,‡ and Raymond J. Andersen*,†
†Departments of Chemistry and Earth & Ocean Sciences, University of British Columbia, Vancouver, British Columbia, Canada V6T 1Z1
‡Department of Surgery, Jack Bell Research Centre, 2660 Oak Street, Vancouver, British Columbia, Canada, V6H 3Z6
S Supporting Information
b
ABSTRACT: Two new cyclic depsipeptides, turnagainolides A (1)
and B (2), have been isolated from laboratory cultures of a marine
isolate of Bacillus sp. The structures of 1 and 2, which are simply
epimers at the site of macrolactonization, were elucidated by analysis
of NMR data and chemical degradation. A total synthesis of the
turnagainolides confirmed their structures. Turnagainolide B (2)
showed activity in a SHIP1 activation assay.
he phosphatidylinositol-3-kinase (PI3K) signaling pathway
is a crucial regulator of many important cellular processes,
edema, and showed in vitro cytotoxicity toward multiple myeloma
cells, providing proof of principle support for the use of SHIP1
activators as anti-inflammatory and anticancer drugs.5,7 More
recently, we reported that the soft coral diterpenoid australin E
was also an in vitro activator of SHIP1.8
T
including growth, proliferation, survival, and motility.1ꢀ3 A key
second messenger in the pathway is phosphatidylinositol-3,4,5-
triphosphate (PI-3,4,5-P3), which is present in low amounts in
unstimulated cells but is rapidly formed from PI-4,5-P2 by phospha-
tidylinositol-3-kinase in response to diverse extracellular stimuli. The
cellular concentration of PI-3,4,5-P3 is tightly controlled in normal
cells by a balance between the activities of PI3K and the lipid
phosphatase PTEN, which converts PI-3,4,5-P3 back to PI-4,5-P2.
Aberrant activation of the phosphoinositide pathway by mutations
leading to upregulated PI3K activity or loss of PTEN function
contributes to inflammatory diseases and cancers.1ꢀ3 The acknowl-
edged role of unregulated PI3K signaling in disease has made the
pathway a high-value target for small-molecule therapeutic interven-
tion. To date, much of the drug discovery effort aimed at modulating
PI3K signaling has focused on finding potent and isoform-selective
PI3K inhibitors that would prevent the formation of the second
messenger PI-3,4,5-P3.4 An alternate approach to dampening aber-
rant PI3K signaling would be to activate the inositol 5-phosphatase
SHIP1, a negative regulator of the pathway found only in hemato-
poietic cells, which turns the signal off by converting the second
messenger PI-3,4,5-P3 to PI-3,4-P2.5 If this alternate approach is
viable, small-molecule activators of SHIP1 should have potential as
drugs for the treatment of inflammatory disorders and cancers of
the blood.
Several years ago we discovered that the sponge meroterpe-
noid pelorol was an in vitro activator of SHIP1.6 Pelorol was the
first small molecule, other than the endogenous activator PI-3,4-
P2, known to activate SHIP1. Synthetic pelorol analogues inhibited
PI3K signaling in stimulated mast cells and macrophages, showed
in vivo anti-inflammatory activity in standard mouse models of ear
As part of an ongoing program aimed at discovering new
marine natural products that modulate PI3K signaling,6,8,9 we
Received: January 11, 2011
Published: May 03, 2011
r
Copyright
2011 American Chemical Society and
1093
dx.doi.org/10.1021/np200033y J. Nat. Prod. 2011, 74, 1093–1099
American Society of Pharmacognosy
|

Journal of Natural Products
ARTICLE
Figure 1. Selected 2D NMR correlations observed for turnagainolides A (1) and B (2).
found that extracts of a laboratory culture of a marine isolate of
the bacterium Bacillus sp. were active in a kinetic enzyme assay
for SHIP1 activation. Assay-guided fractionation of the extract
led to the identification of the two epimeric cyclic depsipeptides
turnagainolides A (1) and B (2). Turnagainolide B (2) was active
in the SHIP1 activation assay, while the epimer turnagainolide A
(1) was inactive. The turnagainolides have a rare β-hydroxy
carboxylic acid-containing moiety of mixed biogenetic origin, and
they are notable because this residue is found with opposite
configurations in the two naturally occurring cyclic depsipep-
tides. Details of the isolation, structure elucidation, and synthesis
of turnagainolides A (1) and B (2) are described below.
Table 1. NMR Data for Turnagainolides A (1) and B (2) (600
MHz 1H; 150 MHz 13C, DMSO-d6)
1
1
2
2
residue position
δC
δH (J in Hz)
δC
δH (J in Hz)
Val-1
1
168.7
169.0
2
3
58.2 4.23, dd (9.6, 6.3)
28.4 2.24, m
59.7 4.05, dd (9.6, 8.4)
29.0 2.42, m
4
18.9 0.89, d (7.2)
19.6 0.89, d (7.2)
7.59, d (9.6)
19.6 1.95, d (6.6)
19.4 0.94, d (7.2)
7.86, d (8.4)
40
NH
1
Ile
170.3
170.9
2
3
57.1 4.26, m
35.5 2.03, m
57.5 4.22, dd (8.4, 3.6)
35.4 2.01, m
4
23.5 1.24, m
23.6 1.30, m
1.24, m
1.30, m
’ RESULTS AND DISCUSSION
5
6
NH
1
2
3
NH
1
2
3
11.9 0.80, t (7.2)
15.5 0.81, d (6.8)
8.10, d (9.6)
12.1 0.82, t (7.2)
15.7 0.85, d (6.0)
8.22, d (8.4)
Microbial strain RJA2194 was isolated from a sediment sample
collected at a depth of 100 m using a grab sampler near Turnagain
Island in Howe Sound, British Columbia, and it was identified as
a Bacillus sp. by 16S RNA analysis. Laboratory cultures of
RJA2194 were grown as lawns on solid agar that were harvested
by lyophilizing the entire culture, cells and agar, and extracting
the dry residue exhaustively with MeOH. The combined MeOH
extracts were concentrated in vacuo to give a gum, which was
partitioned between EtOAc and H2O. Fractionation of the
EtOAc-soluble portion by sequential application of HP20 chro-
matography and normal-phase Si gel flash chromatography gave
pure samples of turnagainolides A (1) and B (2).
Ala
172.9
173.7
48.7 4.32, m
16.3 1.18, d (6.6)
8.56, d (5.4)
49.3 4.36, m
16.1 1.23, d (7.2)
8.87, d (5.4)
Val-2
172.2
172.6
57.4 4.12, dd (8.4, 6.8)
29.8 1.95, m
18.4 0.87, d (7.3)
19.4 0.87, d (7.3)
7.74, d (8.4)
57.0 4.19, t (8.3)
30.6 1.80, m
18.8 0.87, d (6.6)
19.1 0.88,d (7.2)
7.08, d (8.3)
4
40
NH
1
Hppa
168.6
167.7
2
40.1 2.87, dd (14.2, 11.4)
2.40, dd (14.2, 2.2)
73.0 5.49, m
126.7 6.27, dd (16.1, 7.1)
132.5 6.67, d (16.0)
135.7
126.5 7.44, d (7.4)
128.7 7.34, t (7.4)
128.2 7.27, t (7.4)
41.2 2.79, dd (13.8, 4.8)
2.31, dd (13.8, 3.6)
71.0 5.63, dd (8.4, 4.2)
126.5 6.20, dd (16.2, 6.0)
129.3 6.32, d (16.2)
136.4
126.6 7.46, d (7.4)
128.4 7.29, t (7.4)
127.5 7.21, t (7.4)
Turnagainolide A (1) was isolated as an optically active, white
powder that gave a [M þ Na]þ ion at m/z 579.3152 in the
HRESIMS spectrum, consistent with a molecular formula of
C30H44N4O6, requiring 11 sites of unsaturation. The 13C NMR
spectrum of 1 contained only 28 resolved resonances, suggesting
that there were elements of symmetry in the molecule. COSY,
HSQC, and HMBC data showed that 13C NMR resonances at δ
135.7 (Hppa-C-6), 126.5 (Hppa-C-7/11), 128.7 (Hppa-C8/10),
and 128.2 (Hppa-C9) could be assigned to a monosubstituted
phenyl ring, which accounted for the symmetry required by the
13C NMR resonance count and four of the sites of unsaturation. A
series of 13C resonances at δ 168.6 (Hppa-C-1), 168.7 (Val1-C-1),
170.3 (Ile-C-1), 172.2 (Val2-C-1), and 172.9 (Ala-C-1) were
assigned to amide or ester carbonyls, and two resonances at δ
126.7 (Hppa-C-4) and 132.5 (Hppa-C-5) were assigned to
olefinic methines, accounting for a total of six additional sites
of unsaturation. The lack of 13C NMR evidence for a further
unsaturated functionality required that turnagainolide A (1)
must contain one more ring.
3
4
5
6
7, 11
8, 10
9
Information). Consistent with this observation, detailed analysis
of the COSY, HSQC, and HMBC data led to the routine
identification of alanine, isoleucine, and two valine residues
(Figure 1, Table 1). Subtraction of the atoms attributed to the
four amino acids residues (C19H34N4O4) from the molecular
formula of 1 (C30H44N4O6) showed that the remaining fragment
of the molecule had an elemental composition of C11H10O2.
Elucidation of this fragment started with the observation of
COSY correlations between a carbinol methine resonance at δ
5.49 (Hppa-H-3) and an olefinic methine resonance at δ 6.27
(Hppa-H-4) and a pair of geminal methylene proton resonances
at δ 2.40 (Hppa-H-2) and 2.87 (Hppa-H-20). These correlations
demonstrated that the carbinol methine was flanked by olefinic
and aliphatic methylene functionalities. The downfield chemical
In addition to the carbonyl resonances, the 1H and 13C NMR
spectra of turnagainolide A contained resonances characteristic
of amino acid R-methine and side-chain functionalities, suggest-
ing the molecule contained a peptide component (Supporting
1094
dx.doi.org/10.1021/np200033y |J. Nat. Prod. 2011, 74, 1093–1099

Journal of Natural Products
ARTICLE
1
Scheme 1
Table 2. H NMR Data for Hppa Residue of Compounds 3,
5a, and 5b (600 MHz, CDCl3)
3
5a
5b
position
δH (J in Hz)
δH (J in Hz)
δH (J in Hz)
Hppa
2
3
4
2.47, m
2.84, m
2.80, m
4.74, m
6.08, m
6.08, m
6.24, dd
(16.2, 6.0)
6.09, dd
(14.4, 6.6)
6.26, dd (14.4,6.6)
5
6.69, d (16.2) 6.65, d (14.4) 6.80, d, (14.4)
shift of the carbinol methine proton resonance (δ 5.49) and the
absence of COSY correlations to an OH resonance suggested
that the carbinol methine was attached to an acylated alcohol. An
additional COSY correlation was observed between the olefinic
Figure 2. Mosher ester analysis of the Hppa residue in 5.
1
Scheme 2
methine at δ 6.27 and the second olefinic methine in the H
NMR spectrum at δ 6.67 (Hppa-H-5), which was in turn
correlated in the HMBC spectrum to the aromatic carbon
resonances at δ 135.7 (Hppa-C-6) and 126.5 (Hppa-C7/11),
establishing a connection between the disubstituted olefin and
the phenyl ring. HMBC correlations between the Hppa-H2
(δ 2.40) and Hppa-H-20 (δ 2.87) resonances and a 13C reso-
nance at δ 168.6 (Hppa-C-1) indicated that the methylene was
attached to a carbonyl. The 16 Hz scalar coupling observed
between the two olefinic methine resonances established the
E configuration for the olefin. Thus, the final residue in turn-
againolide A (1) was identified as a derivative of (E)-3-hydroxy-
5-phenylpent-4-enoic acid (Hppa).
The linear sequence of turnagainolide A (1) was established
from ROESY and HMBC data. ROESY correlations between the
Ala-NH (δ 8.56) and Val2-RH (δ 4.12) resonances, between the
Ile-NH (δ 8.12) and Ala-RH (δ 4.32) resonances, and between
the Val2-NH (δ 7.74) and Hppa-H-20 (2.87) resonances identi-
fied the Hppa-CO/Val2-NH, Val2-CO/Ala-NH, and Ala-CO/
Ile-NH amide bonds. An HMBC correlation between the Hppa-
H-3 resonance at δ 5.49 and the Val1-CO resonance at δ 168.7
established an ester linkage between the C-3 carbinol methine in
Hppa and Val1. Finally, the Ile residue had to be inserted betwe-
en Val1 and Ala to give the macrocylic ring needed to satisfy the
unsaturation number required by the molecular formula of 1.
Hydrolysis of 1 in 6 N HCl followed by Marfey’s analysis
established the configurations of the amino acids in turnagaino-
lide A as (S)-Val, (R)-Ala, and (2S,3S)-Ile.10 NaOMe-catalyzed
methanolysis of the ester linkage in the macrocycle gave the
turnagainolide A seco-acid methyl ester 3 (Scheme 1), which was
converted to the S and R Mosher esters 5a and 5b. Mosher ester
analysis (Table 2 and Figure 2) established that the configuration
at Hppa-C-3 in 3 was R, and therefore, the complete structure of
turnagainolide A is as shown in 1.11
A (1). Analysis of the 1D and 2D NMR data collected for 2
(Figure 1; Table 1) confirmed that it had the same constitution as
1, and therefore, it had to be a stereoisomer of turnagainolide A.
Comparison of the 1H NMR chemical shifts observed for 2 with
those observed for 1 (Table 1) showed that the biggest differ-
ences were in the Hppa, Val1, and Val2 residues, suggesting that
the difference in configuration in 1 and 2 was likely located in
these residues. Hydrolysis of 2 with 6 N HCl followed by
Marfey’s analysis10 established the configurations of the amino
acids in turnagainolide B as (S)-Val, (R)Ala, and (2S,3S)-Ile, the
same as in turnagainolide A (1). Therefore, the difference in
configuration had to be located in the Hppa residue, which led to
the assignment of the absolute configuration at C-3 of the Hppa
residue in 2 as S. This assignment was confirmed by total syn-
thesis (vide infra).
We undertook the total synthesis of turnagainolides A (1) and
B (2) in order to confirm their structures, particularly their
epimeric relationship at Hppa-C-3. The synthesis started with
preparation of the protected Hppa residue 9 as shown in
Scheme 2. Base-catalyzed condensation of EtOAc with (E)-
cinnamaldehyde gave the β-hydroxy ester 7 in high yield as a
racemic mixture. Reaction of the alcohol 7 with TBDMS-Cl gave
Turnagainolide B (2) was also isolated as an optically active,
white powder that gave a [M þ Na]þ ion at m/z 579.3151 in the
HRESIMS spectrum, consistent with a molecular formula of
C30H44N4O6, identical to the molecular formula of turnagainolide
1095
dx.doi.org/10.1021/np200033y |J. Nat. Prod. 2011, 74, 1093–1099

Journal of Natural Products
ARTICLE
Scheme 3
the TBDMS derivative 8, which was hydrolyzed with aqueous
NaOH to give the desired racemic acid 9.
NMR data recorded for the synthetic turnagainolides matched
the data collected for the natural products (Supporting In-
formation), confirming the structures assigned to the natural
products. The synthetic turnagainolides 1 and 2 were each se-
parately hydrolyzed using NaOMe to give the methylated seco
acids 3 and 4, respectively, which were subjected to Mosher’s
ester11 analysis to confirm the assignments of the configurations
at Hppa-C-3 as R for 1 and S for 2.
Natural and synthetic samples of turnagainolide B (2) were
active in the SHIP1 activation assay, while turnagainolide A (1)
was inactive. The observed efficacy of turnagainolide B (2) was
comparable to that found for the previously reported SHIP
activator MN100 (17) (Figure 3).5
Standard solid-phase peptide synthetic methodology using a
2-chlorotrityl resin was employed to prepare the polymer-bound
tetrapeptide resin-O-Val1-Ile-Ala-Val2-NH2 (14) as shown in
Scheme 3. PyBOP- and HOBt-catalyzed coupling of the pro-
tected Hppa fragment 9 with the resin-bound tetrapeptide 14,
followed by removal of the silyl protecting group with TBAF,
gave the resin-bound turnagainolide seco acids 15. Treatment of
the resin-bound seco acids 15 with TFA gave the free seco acids
16. Macrocyclization of the mixture of epimeric seco acids 16 was
carried out in refluxing CH2Cl2 with DMAP and DIC catalysis to
give a mixture of turnagainolide A (1) and turnagainolide B (2) in
15% and 33% yield, respectively, which was separated by Si gel
chromatography and reversed-phase HPLC to give pure compounds
(Scheme 4). It is interesting to note that the chemical macro-
cyclization gave turnagainolide B (2) as the major product
(≈2:1), whereas turnagainolide A (1) was the major product
’ CONCLUSIONS
The two new cyclic depsipeptides turnagainolides A (1) and B
(2) contain a mixture of R and S amino acids and the (E)-3-
hydroxy-5-phenylpent-4-enoic acid fragment Hppa, which are
1
isolated from the microbial cultures (≈8:1). The H and 13C
1096
dx.doi.org/10.1021/np200033y |J. Nat. Prod. 2011, 74, 1093–1099

Journal of Natural Products
ARTICLE
Scheme 4
the Hppa residue were not assigned. Comparison of the NMR
data for EGM-556 (18) with the NMR data recorded for
turnagainolide A (1), and a reversed-phase HPLC comparison
of authentic samples of both compounds, showed that they are
identical, requiring that the structure of EGM-556 should be
revised to 1.15 It is also interesting that EGM-556 is produced by
a fungal culture and the turnagainolides are produced by a
Bacillus sp. culture. We previously found that rhizoxins are
produced by a marine Pseudomonad, even though all previous
reports of their isolation were from fungal cultures.16 Later it was
shown that rhizoxin production by cultures of fungi in the genus
Rhizopus was due to the presence of a symbiotic bacterium iden-
tified as a Burkholderia sp.17 Therefore, it is possible that the
production of EGM-556 by Microascus cultures is also related to
the presence of a symbiotic/associated bacterium.
’ EXPERIMENTAL SECTION
General Experimental Procedures. Optical rotations were
measured using a Jasco P-1010 polarimeter with sodium light
1
(589 nm). The H and 13C NMR spectra were recorded on a Bruker
AV-600 spectrometer with a 5 mm CPTCI cryoprobe. 1H chemical shifts
are referenced to the residual DMSO-d6, CDCl3, or acetone-d6 signal
(δ 2.49, 7.24, or 2.04 ppm, respectively), and 13C chemical shifts are
referenced to the DMSO-d6, CDCl3, or acetone-d6 solvent peak (δ 39.5,
77.0, or 206.0 ppm, respectively). Low- and high-resolution ESI-QIT-
MS data were recorded on a Bruker-Hewlett-Packard 1100 EsquireꢀLC
system mass spectrometer. Reversed-phase HPLC purifications were
performed on a Waters 600E System Controller liquid chromatography
system attached to a Waters 996 photodiode array detector with a CSC-
Inertsil 150A/ODS2 column. All solvents used for HPLC were Fisher
HPLC grade.
Isolation and Cultivation of Strain RJA2194. Strain RJA2194,
identified as a Bacillus sp., was isolated from a sediment sample collected
on Oct 11, 2007, at a depth of 104 m near Turnagain Island in Howe
Sound, British Columbia, at N 49°,31.473°; W 123°,59.014° and grown
on ISP2 at 15 °C. NCBI blastn analysis of the partial 16S rRNA sequence
of Bacillus sp. RJA2194 (deposited in GenBank with an accession no.
JF827033) is 99% identical to Bacillus atrophaeus strain EHFS2_S04Ha
(GenBank accession no. EU071547).
Production cultures of strain RJA2194 were grown as lawns on solid
ISP2 medium (per liter: yeast extract, 4 g; malt exract, 10 g; glucose, 4 g;
agar, 15 g) in sixty 8 ꢁ 12 ꢁ 1 inch pans at room temperature (rt) for 14
days. The mature cultures, cells and media together, were lyophilized in
the growth pans.
Figure 3. AQX-MN100 (17),5 turnagainolide B (2), or EtOH control
(10 μM each) was incubated with SHIP1 enzyme at 23 °C for 10 min.
Then 80 μM IP4 was added, and the reaction was incubated at 37 °C for
10 min prior to the addition of malachite green containing stop solution.
The wells were read at 650 nM, and enzyme activity is expressed as fold
increase over the activity seen in the EtOH control wells.5
hallmarks of nonribosomal peptide biosynthesis.12 Hppa has
been previously reported as an acyl substituent in marinoid E, an
iridoid glucoside isolated from the mangrove plant Avicennia
marina.13 Biogenetically, Hppa appears to be derived from a (E)-
cinnamoylCoA residue that has been extended with an acetate
unit to give a β-ketoester. Non-stereoselective reduction of the
ketone would lead to the opposite configurations observed at the
Hppa-C-3 site of macrolactonization in turnagainolides A (1)
and B (2). We have not been able to find other examples of
natural cyclic depsipeptides co-occurring as epimers at the ma-
crocyclic lactone carbinol carbon.
Vervoort et al. have recently reported the isolation of EGM-
556 (18) from cultures of the marine fungal isolate Microascus sp.14
EGM-556 (18) has the same constitution as turnagainolides A
(1) and B (2), but the configuration of one of the two valine
residues has been assigned as R.14 The positions of the R and S
valine residues in EGM-556 (18) and the configuration at C-3 of
Extraction and Purification. The lyophilized solid agar (8 L) and
cells were exhaustively extracted with MeOH (3 ꢁ 4 L). The combined
MeOH extract was concentrated in vacuo and then partitioned between
EtOAc and water. The EtOAc-soluble portion was fractionated on HP20
using gradient elution from H2O to acetone. The SHIP active fraction
(300 mg, 80% acetone) was chromatographed on Si gel eluting with
CH2Cl2/MeOH (20:1) to yield turnagainolide A (1) (40 mg) and
turnagainolide B (2) (5 mg).
Turnagainolide A (1): white powder; [R]20 ꢀ54 (c 0.05, 1:1
D
CH2Cl2/MeOH); UV (1:1 CH2Cl2/MeOH) λmax (log ε) 202 (4.23),
1
253 (4.36) nm; H and 13C NMR see Table 1; (þ)-HRESIMS m/z
[M þ Na]þ 579.3152 (calcd for C30H44N4O6Na, 579.3159).
Turnagainolide B (2): white powder; [R]20 ꢀ90 (c 0.05, 1:1
D
CH2Cl2/MeOH); UV (1:1 CH2Cl2/MeOH) λmax (log ε) 209 (4.14),
1
252 (4.27) nm; H and 13C NMR see Table 1; (þ)-HRESIMS m/z
[M þ Na]þ 579.3151 (calcd for C30H44N4O6Na, 579.3159).
Acid Hydrolysis of Compounds 1 and 2. Purified compounds
1 (0.2 mg) and 2 (0.2 mg) were hydrolyzed separately in 0.5 mL of 6 M
1097
dx.doi.org/10.1021/np200033y |J. Nat. Prod. 2011, 74, 1093–1099

Journal of Natural Products
ARTICLE
HCl at 108 °C for 16 h in screw-top vials. After cooling, the solvent was
removed under N2 and the residue was dissolved in 100 μL of H2O.
Derivatization of Amino Acids with Marfey’s Reagent and
HPLC Analysis. To a 0.5 mL vial containing 2.0 μmol of the pure
amino acid standard in 40 μL of H2O was added 2.8 μmol of FDAA in
80 μL of acetone followed by 20 μL of 1 M NaHCO3. The mixture was
heated for 1 h at 40 °C. After cooling to rt, 10 μL of 2 M HCl was added,
and the resulting solution was filtered through a 4.5 μm filter and stored
in the dark until HPLC analysis.
The hydrolysate mixture of 1 (40 μL) and 2 (40 μL) was treated with
FDAA as described above, respectively. The resulting solutions were
analyzed by employing an Alltech Econosil C18 HPLC column eluting
with 60% mixed buffer in MeCN (flow rate: 1 mL/min). The mixed
buffer was prepared by mixing an aqueous solution of sodium borate,
potassium chloride, citric acid, and tris(hydroxymethyl)aminomethane
(each 25 mM) and 0.05 M HCl in a 3:2 ratio (pH 2ꢀ3).
(1H, d, J = 15.9 Hz, H-5), 6.34 (1H, dd, J = 15.9, 6.7 Hz, H-4), 4.83 (1H,
m, H-3), 2.57 (2H, m, H-2), 0.91 (9H, s, H-13/14/15), 0.12 (3H, s,
H-16), 0.09 (3H, s, H-17); 13C NMR (100 MHz, acetone-d6) δ 172.2
(C, C-1), 137.8 (C, C-6), 132.8 (CH, C-5), 130.7 (CH, C-9), 129.5
(CH, C-7/11), 128.5 (CH, C-4), 127.4 (CH, C-8/10), 71.7 (CH, C-3),
44.1 (CH2, C-2), 26.3 (CH3, C-13/14/15), 18.8 (C, C-12), ꢀ4.0 (CH3,
C-16), ꢀ4.7 (CH3, C-17).
Compound 16. To CH2Cl2-washed polymer-bound 2-chlorotrityl
chloride (572 mg) under N2 was added Fmoc-(S)-Val-OH (330 mg,
0.97 mmol) in CH2Cl2 (10 mL), followed by DIPEA (0.3 mL, 1.72
mmol). The reaction mixture was stirred for 1 h at rt before the solid-
phase resin was washed with CH2Cl2. The solid-phase resin was capped
by adding a solution of CH2Cl2/MeOH/DIPEA (ratio 17:2:1, 10 mL)
and stirring for 1 h at rt. The solid-phase resin was then washed with
CH2Cl2, followed by DMF. The Fmoc protecting group of (S)-Val was
removed by treatment with 20% piperidine in DMF (5 mL) for 5 min at rt.
This reaction was repeated twice more with 20% piperidine in DMF
(5 mL), followed by washing the solid-phase resin with DMF, CH2Cl2,
and DMF again. The coupling of the next amino acid ((2S,3S)-Ile) was
performed by adding a solution of Fmoc-(2S,3S)-Ile-OH (330 mg, 0.93
mmol), PyBOP (560 mg, 1.08 mmol), and HOBt (2.0 mL of a 0.5 M
solution, 1.0 mmol) in DMF (10 mL), followed by the addition of
DIPEA (0.2 mL, 1.15 mmol), and stirring for 1 h at rt.
Retention times in minutes for the derivatized amino acid standards
were as follows: (S)-Ala 7.94; (R)-Ala 9.95; (S)-Val 14.83; (R)-Val
19.89; (2S,3S)-Ile 23.90; (2R,3R)-Ile 41.86. The retention times for
compound 1 were 9.94, (R)-Ala; 14.75, (S)-Val; 23.43 (2S,3S)-Ile. For
compound 2 they were 9.99, (R)-Ala; 14.73 (S)-Val; 23.57, (2S,3S)-Ile.
Methanolysis of 1 to Ester 3. Compound 1 (3 mg) was dissolved
in 5% NaOMe/MeOH (2 mL) and stirred for 1.5 h at rt. The reaction
mixture was neutralized with 1 M aqueous HCl, and then the aqueous
phase was extracted three times with EtOAc. The residue, after solvent
removal, was purified by C18 reversed-phase HPLC using 11:9 MeCN/
H2O (tR = 12.58 min) to yield compound 3 (1.8 mg).
The above deprotection/coupling procedure was repeated for (R)-
Ala and (S)-Val using approximately 1 mmol of each reagent as des-
cribed above. The final coupling step was performed by adding a solution
of 9 (306 mg, 1.0 mmol), PyBOP (560 mg, 1.08 mmol), and HOBt
(2.0 mL of a 0.5 M solution, 1.0 mmol) in DMF (10 mL), followed by
the addition of DIPEA (0.2 mL, 1.15 mmol), and stirring for 1 h at rt.
The solid-phase resin was then washed with DMF, followed by THF.
Deprotection of the silyl protecting group was accomplished by treat-
ment with a solution of TBAF in THF (10 mL of a 0.2 M solution, 2.0
mmol) for 1 h at rt. This reaction was repeated once more for an
additional hour, followed by washing the solid-phase resin with THF and
then CH2Cl2.
Cleavage from the solid-phase resin was accomplished by treatment
with CH2Cl2/PDT/TFA (ratio 98:1:1, 10 mL) for 1 h at rt under N2. This
procedure was repeated for an additional hour at rt, followed by washing
the solid-phase resin with CH2Cl2 and MeOH. The washings were
combined and dried in vacuo to give the crude linear peptide (16). The
peptide was washed with CH2Cl2 and purified by Si gel chromatography
(eluent: gradient from 1:9 MeOH/EtOAc to 1:3 MeOH/EtOAc) to give
the linear peptide 16 (387 mg) as a mixture of diastereomers.
Synthetic Turnagainolides A (1) and B (2). A solution of
DMAP (208 mg, 1.70 mmol), pTSA monohydrate (162 mg, 0.85
mmol), and DIC (0.75 mL, 4.79 mmol) in CHCl3 (200 mL) was heated
to reflux. To this reaction mixture at reflux was added a solution of the
linear peptide (16) (95.8 mg, 0.167 mmol) in DMF (20 mL) dropwise
over the course of 10 h, and heating was continued at reflux overnight.
The reaction mixture was then cooled, dried in vacuo, and purified by Si
gel flash chromatography (eluent: gradient from CH2Cl2 to 1:19
MeOH/CH2Cl2) and Si gel chromatography (eluent: gradient from
CH2Cl2 to 1:1 acetone/CH2Cl2) to give crude turnagainolide B (2)
(31.5 mg, 33%) and crude turnagainolide A (1) (14.0 mg, 15%)
(combined yield = 48%). Portions of the crude turnagainolides A and
B were further purified separately by C18 reversed-phase HPLC (eluent:
3:2 MeOH/H2O) to give pure turnagainolide A (1) and pure turn-
againolide B (2). Turnagainolide A (1) was converted to the Mosher
esters 5a and 5b as described above for the natural product 1.
Compound 3: white powder; [R]20 ꢀ62.5 (c 0.12, 1:1 CH2Cl2/
D
MeOH); UV (1:1 CH2Cl2/MeOH) λmax (log ε) 210 (4.14), 250
1
(4.22) nm; H and 13C NMR see Supporting Information Table S1;
(þ)-LRESIMS m/z [M þ Na]þ 611.3.
Mosher Esters 5a and 5b. Compound 3 (0.6 mg) was dissolved
in pyridine (1 mL), and dimethylaminopyridine (1 mg) and (R)-MTPA-
Cl (10 μL) were added in sequence. The reaction mixture was stirred for
5 h at rt, and two drops of H2O were then added. The resulting solution
was concentrated under reduced pressure to dryness. The residue was
purified by C18 reversed-phase HPLC eluting with 13:7 MeCN/H2O
(tR = 29.72 min) to yield compound 5a (0.4 mg). In the same fashion, a
solution of 3 (0.6 mg) in pyridine was treated with dimethylaminopyr-
idine (1 mg) and (S)-MTPA-Cl (10 μL) to afford, after isolation as
above (tR = 29.02 min), 0.3 mg of 5b. 1H NMR see Table 2.
Compound 7. A flask containing a stir bar and THF (5 mL) under
N2 was cooled to ꢀ78 °C, and to this flask was added a solution of LDA
in THF (10.0 mL of a 1.8 M solution, 18.0 mmol). EtOAc (1.0 mL, 10.2
mmol) was added dropwise, followed by (E)-cinnamaldehyde (2.0 mL,
15.9 mmol). The reaction mixture was stirred for 45 min at ꢀ78 °C
before being quenched by the addition of saturated aqueous NH4Cl. The
mixture was extracted with CH2Cl2, dried in vacuo, and purified by silica
gel chromatography (eluent: gradient from 1:1 CH2Cl2/hexane to
CH2Cl2) to give compound 7 (2.17 g, 96%).
Compound 8. To a solution of compound 7 (2.17 g, 9.88 mmol) in
THF were added TBDMS-Cl (1.65 g, 10.9 mmol) and imidazole (1.02 g,
15.0 mmol), and the reaction mixture was refluxed for 12 h. The reaction
mixture was dried in vacuo and purified by Si gel chromatography
(eluent: 1:3 CH2Cl2/hexane to 1:1 CH2Cl2/hexane) to give compound
8 (2.56 g, 78%).
Compound 9. To a solution of compound 8 (2.56 g, 7.65 mmol) in
THF (5 mL) and MeOH (5 mL) was added 1 M aqueous NaOH
(5 mL), and the reaction mixture was stirred for 3 h at rt. The reaction
mixture was acidified with 1 M aqueous HCl and extracted with CH2Cl2.
The organic layer was dried in vacuo and purified by Si gel chromatog-
raphy (eluent: gradient from CH2Cl2 to 1:19 MeOH/CH2Cl2) to give 9
(1.64 g, 70%): 1H NMR (400 MHz, acetone-d6) δ 7.44 (2H, d, J = 7.3
Hz, H-7/11), 7.33 (2H, t, J = 7.5 Hz, H-8/10), 7.24 (1H, m, H-9), 6.67
Methanolysis of Synthetic 2 to Ester 4. Synthetic 2 (6.1 mg)
was dissolved in 2 mL of 5% NaOMe/MeOH and stirred overnight at rt.
The reaction was acidified with 1 M aqueous HCl, diluted with H2O, and
extracted with CH2Cl2. The organic layer was purified by C18 reversed-
phase HPLC using 3:2 MeOH/H2O to give compound 4 (3.2 mg).
1098
dx.doi.org/10.1021/np200033y |J. Nat. Prod. 2011, 74, 1093–1099

Journal of Natural Products
ARTICLE
Mosher’s Esters 6a and 6b. Compound 4 (0.4 mg) was dissolved
in CH2Cl2 (1 mL), and to this solution were added one crystal of
dimethylaminopyridine and (R)-MTPA-Cl (10 μL). The reaction
mixture was stirred overnight at rt and purified by Si gel chromatography
using a gradient from CH2Cl2 to 1:3 acetone/CH2Cl2 to yield com-
pound 6a. In the same fashion, a solution of 4 (0.4 mg) in CH2Cl2
(1 mL) was treated with one crystal of dimethylaminopyridine and (S)-
MTPA-Cl (10 μL). The reaction mixture was stirred overnight at rt and
purified by Si gel chromatography using a gradient from CH2Cl2 to 1:3
acetone/CH2Cl2 to yield 6b.
(16) Roberge, M.; Cinel, B.; Anderson, H. J.; Lim, L.; Jiang, X. X.;
Xu, L.; Bigg, C. M.; Kelly, M. T.; Andersen, R. J. Cancer Res. 2000,
60, 5052–5058.
(17) Partida-Martinez, L. P.; Hertweck, C. Nature 2005, 437, 884
888.
’ ASSOCIATED CONTENT
S
Supporting Information. NMR spectra and data for
b
compounds 1 to 3 are available free of charge via the Internet
at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*Tel: 604 822 4511. Fax: 604 822 6091. E-mail: randersn@
interchange.ubc.ca.
’ ACKNOWLEDGMENT
Financial support was provided by the Canadian Cancer
Society (R.J.A.) and Aquinox Pharmaceuticals (R.J.A. and A.L.
F.M.). Prof. P. Crews kindly supplied an authentic sample of
EGM-556.
’ REFERENCES
(1) Bunney, T. D.; Katan, M. Nat. Rev. Cancer 2010, 10, 342–352.
(2) Martelli, A. M.; Evangelisti, C.; Chiarini, F.; Grimaldi, C.;
Manzoli, L.; McCubrey, J. A. Expert Opin. Investig. Drugs 2009, 18
1333–1349.
(3) Harris, S. J.; Parry, R. V.; Westwick, J.; Ward, S.G. J. Biol. Chem.
2008, 283, 2465–2469.
(4) Pereira, A. R.; Strangman, W. K.; Marion, F.; Feldberg, L.; Roll,
D.; Mallon, R.; Hollander, I.; Andersen, R. J. J. Med. Chem. 2010,
53, 8523–8533.
(5) Ong, C. J.; Ming-Lum, A.; Nodwell, M.; Ghanipour, A.; Yang, L.;
Williams, D. E.; Kim, J.; Demirjan, L.; Qasimi, P.; Ruschmann, J.; Cao,
L. P.; Ma, K. W.; Chung, S. W.; Duronio, V.; Andersen, R. J.; Krystal, G.;
Mui, A. L. F. Blood 2007, 110, 1942–1949.
(6) Yang, L.; Williams, D. E.; Mui, A.; Ong, C.; Krystal, G.; van Soest,
R.; Andersen, R. J. Org. Lett. 2005, 7, 1073–1076.
(7) Kennah, M.; Yau, T. Y.; Nodwell, M.; Krystal, G.; Andersen, R. J.;
Ong, C. J.; Mui, A. L. F. Exp. Hematol. 2009, 37, 1274–1283.
(8) Williams, D. E.; Amiani, A.; Dewi, A. S.; Patrick, B. O.; van
Ofwegen, L.; Mui, A. L. F.; Andersen, R. J. Aust. J. Chem. 2010,
63, 895–900.
(9) Marion, F.; Williams, D. E.; Patrick, B. O.; Hollander, I.; Mallon,
R.; Kim, S. C.; Roll, D. M.; Feldberg, L.; van Soest, R.; Andersen, R. J.
Org. Lett. 2006, 8, 321–324.
(10) Marfey, P. Carlsberg Res. Commun. 1984, 49, 591–596.
(11) Ohtani, I.; Kusumi, T.; Kashman, Y.; Kakisawa, H. J. Am. Chem.
Soc. 1991, 113, 4092–4096.
(12) Jones, A. C.; Monroe, E. A.; Eisman, E. B.; Gerwick, L.;
Sherman, D. H.; Gerwick, W. H. Nat. Prod. Rep. 2010, 27, 1048–1065.
(13) Sun, Y.; Ouyang, J.; Deng, Z.; Li, Q.; Lin, W. Magn. Reson.
Chem. 2008, 46, 638–642.
(14) Vervoort, H. C.; Draskovic, M.; Crews, P. Org. Lett. 2010,
13, 410–413.
(15) An authentic sample of EGM-556 was provided by Prof.
P. Crews.
1099
dx.doi.org/10.1021/np200033y |J. Nat. Prod. 2011, 74, 1093–1099