Ulongapeptin, a cytotoxic cyclic depsipeptide from a Palauan marine cyanobacterium Lyngbya sp.

Article abstract of DOI:10.1021/np030050s

Ulongapeptin (1), a cyclic depsipeptide, was isolated from a Palauan marine cyanobacterium Lyngbya sp. The gross structure was elucidated through one-dimensional TOCSY experiments and other spectroscopic techniques. The absolute and relative stereochemistry of the β-amino acid, 3-amino-2-methyl-7-octynoic acid (AMO), in 1 was determined by synthesis of the saturated α-alkyl-β-amino acid and Marfey's analysis of the acid hydrolysate of tetrahydro-1. Ulongapeptin (1) was cytotoxic against KB cells at an IC₅₀ value of 0.63 μM.

Full text of DOI:10.1021/np030050s

J . Nat. Prod. 2003, 66, 651-654
651
Ulon ga p ep tin , a Cytotoxic Cyclic Dep sip ep tid e fr om a P a la u a n Ma r in e
Cya n oba cter iu m Lyn gbya sp .
Philip G. Williams,† Wesley Y. Yoshida,† Michael K. Quon,† Richard E. Moore,*,† and Valerie J . Paul‡,§
Department of Chemistry, University of Hawaii at Manoa, Honolulu, Hawaii 96822, and University of Guam Marine
Laboratory, UOG Station, Mangilao, Guam 96923
Received J anuary 31, 2003
Ulongapeptin (1), a cyclic depsipeptide, was isolated from a Palauan marine cyanobacterium Lyngbya
sp. The gross structure was elucidated through one-dimensional TOCSY experiments and other
spectroscopic techniques. The absolute and relative stereochemistry of the â-amino acid, 3-amino-2-methyl-
7-octynoic acid (AMO), in 1 was determined by synthesis of the saturated R-alkyl-â-amino acid and Marfey’s
analysis of the acid hydrolysate of tetrahydro-1. Ulongapeptin (1) was cytotoxic against KB cells at an
IC50 value of 0.63 µM.
Cyanobacteria are photosynthetic microorganisms that
inhabit diverse habitats ranging from marine to terrestrial.
These ancient prokaryotes are believed to have played a
key role in the evolution of the modern ecosystem since
they were presumably the first organisms to produce
molecular oxygen.1 Filamentous cyanobacteria of the genus
Lyngbya are the most frequently encountered cyanobacte-
ria in tropical areas and have the ability to fix nitrogen
through heterocysts. With few exceptions most secondary
metabolites from Lyngbya arise by the incorporation of
nitrogen into a combination of polyketide and peptide
biosynthetic pathways.2 Here we describe the isolation of
a new nitrogen-containing compound, ulongapeptin (1),
from a Palauan collection of Lyngbya sp.
at m/z 809, 831, and 847 for MH+, MNa+, and MK+, re-
spectively. HR-MALDI, in conjunction with proton and
carbon spectral data, established an elemental composition
for the amorphous powder as C44H68N6O8. From the proton
and carbon spectral data, 13 sp2 carbons in the form of
seven carbonyls and three carbon-carbon double bonds
constituted 10 of the 14 degrees of unsaturation implied
by the molecular formula of 1. The six nitrogens could be
accounted for by three secondary amide (δH 6.19, 8.18, and
8.24) and three tertiary N-methylamide groups (δH 2.75,
1
2.90, and 3.40) according to the H NMR data. In addition
to a broad band at 1654 cm-1 for amide carbonyls, the IR
spectrum showed a strong vibration at 1727 cm-1 charac-
teristic of an ester moiety. The eight oxygens were therefore
located in one ester and six amide carbonyl groups, sug-
gesting that 1 was a depsipeptide consisting of one hydroxy
and six amino acid units.
One-dimensional TOCSY experiments enabled us to
identify the amino acids that constituted 1. Excitation of
the 2° amide proton signals at δΗ 8.18 and 8.24 produced
spectra with signals for protons that were either directly
or relay coupled to these NH groups and suggested the
presence of two valine units. Likewise, excitation of the
doublets at δΗ 3.95 (H-37) and 4.90 (H-26) generated
spectra consistent with two more valine units whose
nitrogens were part of tertiary amide groups. Analysis of
3
the HMBC spectra showed J CH cross-peaks from the
methyl singlets at δH 3.40 (H-30) and 2.75 (H-41) to C-26
and C-37, respectively, verifying this conclusion and thereby
expanded these fragments into two N-Me-Val units. An-
other unit was generated starting from the R-proton signal
3
at 4.99 ppm (H-16) that showed a strong J HH coupling to
a pair of geminal doublet of doublets at 3.90 and 2.53 ppm
(H-17). This fragment was expanded into a N-methylphen-
ylalanine unit via HMBC correlations to C-18 from H-17,
to C-18 from H-20, and from the methylamide signal at δH
2.90 (H-24) to the R-carbon (C-16).
Resu lts a n d Discu ssion .
Lyophilized VP755 was extracted with 1:1 methanol/
ethyl acetate and the concentrated extract fractionated by
normal and reversed-phase chromatography. After re-
peated RP-HPLC 1.1 mg of ulongapeptin (1), which had
an IC50 of 0.63 µM against KB cells,3 was isolated in a yield
of 0.10%.
The two remaining isolated spin systems were assigned
as follows. Irradiation of the methine quartet at δΗ 5.45
(H-43) in a 1D TOCSY experiment showed magnetization
transfer to a methyl doublet at δΗ 1.50 (H-44) to form a
Ulongapeptin (1) was found to have a molecular weight
of 808 Da based on FABMS pseudo-molecular ion peaks
1
two-carbon unit. A J CH cross-peak obtained from the
HSQC spectra connected this downfield methine to a
carbon signal at δC 66.4 (C-43), a chemical shift suggestive
of a carbinol and hence a lactic acid moiety. An initial one-
dimensional TOCSY experiment on the final secondary
* To whom correspondence should be addressed. (R.E.M.) Tel: (808) 956-
7232. Fax: (808) 956-5908. E-mail: moore@gold.chem.hawaii.edu.
† University of Hawaii at Manoa.
‡ University of Guam Marine Laboratory.
§ Present address: Smithsonian Marine Station, Fort Pierce, FL.
10.1021/np030050s CCC: $25.00
© 2003 American Chemical Society and American Society of Pharmacognosy
Published on Web 05/06/2003

652 J ournal of Natural Products, 2003, Vol. 66, No. 5
Williams et al.
Sch em e 1. Synthesis of (2R,3R)-AMO
in Scheme 1. The Horner-Wadsworth-Emmons elonga-
tion of hexanal produced a 1:1 mixture of E:Z methyl oct-
2-enoate, which was easily separated by flash chromatog-
raphy. Michael addition of N-benzyl-R-methylbenzylamine
to trans-3 afforded 4 after purification.6 Subsequent meth-
ylation produced a mixture of diastereomers7 that were
separated after deprotection (5) as their 1-fluoro-2,4-di-
nitrophenyl-leucinamide derivatives (FDLA). Comparison
with the derivatized hydrogenated hydrolysate established
the absolute configuration of the 3-amino-2-methyloct-7-
ynoic acid as 2S, 3S using standard Marfey’s conditions.
The trivial name of 1 has been assigned after the col-
lection site of the cyanobacterium. Ulong Channel has
proven to be a rich source of interesting secondary me-
tabolites including the lyngbyabellins8 and the apramides,9
both of which possess C8 units similar to 1. The structure
of ulongapeptin is an excellent example of the metabolic
themes of cyanobacteria. Four of the five amino acid-
derived units are valines, which is the most commonly
encountered amino acid in cyanobacterial isolates. Two of
the four valine units are N-methylated, a percentage that
mirrors that found in the literature.2 The presence of two
amino acids that have been epimerized to a D-configuration
is unusual though since over 90% of the amino acids
incorporated into cyanobacterial metabolites have an L-
configuration.2 The AMO unit found in 1 has been previ-
ously identified in the mollusk metabolite onchidin.10 Such
â-amino acid units appear to be ubiquitous to cyanobacte-
rial metabolites, and their appearance in compounds iso-
lated from other marine organisms has often been sug-
gested as indicative of either a dietary or symbiotic rela-
tionship between the two organisms.2
F igu r e 1. Key ROESY correlations that suggest the D-configuration
at C-26.
amide proton signal at δΗ 6.19 gave correlations to a
nitrogen-bearing methine at δΗ 4.27 (H-3), a methyl doublet
(H-9), a quartet of doublets at δΗ 2.82 (H-2), and the proton
signals for H-4. These diagnostic resonances implied an
R-methyl-â-amino acid unit common to cyanobacterial
metabolites.2 A second TOCSY experiment with a longer
mixing time showed the secondary amide proton (3-NH)
was relay-coupled to the terminal alkyne proton at δH 1.98
(H-8). Considering all the fragments previously identified
and the constraints imposed by the molecular formula, this
had to be a C8 unit, viz., a 3-amino-2-methyloct-7-ynoic acid
unit.
With the molecular formula satisfied, the gross structure
was assembled via HMBC correlations. Cross-peaks from
the amide proton signals to C-10, C-15, and C-36 estab-
lished the sequences (AMO)-(Val)-(N-Me-Phe) [C-1 to C-24]
and (Val)-(N-Me-Val) [C-31 to C-41]. These two fragments
3
could be linked by J CH correlations from the N-methyl-
amide signals to the remaining units. Specifically, cross-
peaks to C-25 from H-24, to C-31 from H-30, and to C-42
from H-41 connected all of these fragments into a linear
chain, i.e., (AMO)-(Val)-(N-Me-Phe)-(N-Me-Val)-(Val)-(N-
Me-Val)-(lactic acid). HMBC4 and ROESY experiments
failed to show any correlations that supported a connection
between C-1 of the AMO unit and the lactic acid oxygen,
but given the ester carbonyl vibration in the IR at 1727
cm-1 and the degrees of unsaturation required by the
molecular formula, ulongapeptin had to be the cyclic
depsipeptide 1.
The absolute configuration of 1 was established by analy-
sis of the degradation products. A small sample of 1 was
hydrogenated to reduce the terminal alkyne and then
hydrolyzed with 6 N HCl to liberate the amino acids. These
were analyzed by chiral HPLC and the retention times
compared with authentic standards. The proteogenic amino
acids and the hydroxy acid were shown to have an L-con-
figuration. Of the N-methylated amino acids, the phen-
ylalanine-derived unit was clearly D, but both enantiomers
of N-methylvaline were encountered in an equal amount.
Hydrolysis of 1 at a lower temperature (90 °C) gave the
same result and confirmed the presence of both enanti-
omers of N-Me-Val in 1. Analysis of the ROESY data,
specifically a cross-peak from H-16 to H-26 and a correla-
tion between H-17b and H-28, suggested a D-configuration
around C-26 (Figure 1). Molecular modeling of this dipep-
tide fragment with the stereochemistry of C-26 reversed
[(N-Me-D-Phe)-(N-Me-L-Val)] did not yield a low-energy
conformer consistent with the ROESY data. Correlations
between the lactic acid moiety and the adjacent N-Me-Val
unit were in line with this stereochemical conclusion and
further suggested an L-configuration around C-37.5
Exp er im en ta l Section
Gen er a l Exp er im en ta l P r oced u r es. The optical rotation
was measured on a J asco-DIP-700 polarimeter at the sodium
D line (589 nm). The UV spectrum was determined on a
Hewlett-Packard 8453 spectrophotometer, and the IR spec-
trum was recorded on a Perkin-Elmer 1600 FTIR instrument
as a film on a NaCl disk. The NMR spectra of 1 were recorded
in CDCl3 on a Varian 500 operating at 500 and 125 MHz using
the residual solvent signal as an internal reference. NMR
analyses of the synthetic products were carried out at 300 and
75 MHz using a Varian spectrometer. FABMS and HR-MALDI
were recorded in the positive mode on a VG ZAB2SE and a
DE-STR spectrometer, respectively. HPLC separations were
performed on a Beckman 110B apparatus coupled to an
Applied Biosystems 759A absorbance detector. All synthetic
reagents and amino acids were purchased from Aldrich, and
all synthetic procedures were not optimized.
Biologica l Ma ter ia l. The dark reddish-black clumps of
cyanobacterium, designated VP755, were collected at Ulong
Channel in Palau. The sample was identified by V. J . Paul
and a voucher is maintained at the Smithsonian Marine
Station, Fort Pierce, FL.
Extr a ction a n d Isola tion of Ulon ga p ep tin (1). VP755
was extracted with 1:1 EtOAc/MeOH to yield 1.11 g of
lipophilic extract that was partitioned between hexane and
The R-methyl-â-amino acid was synthesized as a 5:2
mixture of C-2 diastereomers (2R, 3R and 2S, 3R) as shown

Cytotoxic Cyclic Depsipeptide from Lyngbya
J ournal of Natural Products, 2003, Vol. 66, No. 5 653
Ta ble 1. NMR Spectral Data for Ulongapeptin in CDCl3
C/H no.
δH (J in Hz)
δC
1H-1H COSY
HMBC
TOCSY
ROESY
AMO
1
2
3
177.7, s
41.8, d
49.4, d
2, 9
9
9
2.82, qd (7.0, 3.6)
4.27, m
3, 9
2, 3-NH, 4
3, 9
2, 9
3-NH
4
6.19, d (10.1)
1.59, m
1.30, m
1.28, m
2.18, m
3
2, 3, 4, 5, 6, 8, 9
4, 11-NH
2, 3-NH, 5, 6, 19
33.0, t
3, 4b
4a
2, 3
3
5
6
24.1, t
17.7, t
4a
5, 8
5, 8
2.02, m
7
8
9
83.9, s
68.9, d
14.7, q
1.98, t (2.6)
1.13, d (7.0)
6
2
2, 3, 13
Val
10
11
11-NH
12
13
14
171.4, s
61.2, d
3-NH, 11
4.28, t (10.2)
8.24, d (10.2)
1.62, m
0.97, d (6.7)
0.78, d (6.6)
11-NH, 12
11
11, 13, 14
12
12
13, 14
11, 12, 13, 14
3-NH, 16
13, 14
31.8, d
19.5, q
20.3, q
11, 13, 14
12, 14
11, 12, 13
9, 11, 12, 14
11, 12, 13, 24
N-Me Phe
15
16
17
167.5, s
62.0, d
35.8, t
11, 16, 17b, 11-NH
17a, 17b, 24
4.99, dd (11.5, 1.9)
3.90, dd (-13.2, 11.5)
2.53, dd (-13.2, 1.9)
17
17b
17a
17a, 17b
11-NH, 17b, 26
17b, 24
17a, 19, 28
18
137.6, s
129.4, d
128.6, d
126.7, d
29.2, q
17, 20
17, 19, 21
19/23
20/22
21
7.26, d (7.4)
7.30, dd (7.4, 6.0)
7.25, t (6.0)
2.90, s
20
19, 21
4, 17b, 28
14, 17a
19
24
N-Me-Val
25
26
27
28
170.9, s
58.4, d
28.5, d
20.1, q
19.6, q
31.6, q
16, 24, 26
28, 29, 30
28, 29
27, 29
26, 27, 28
26
4.90, d (10.9)
2.39, m
1.07, d (6.5)
0.94, d (6.4)
3.40, s
27
27, 28, 29
16, 28, 29
30
26, 28, 29
27
27
17b, 19, 27
26, 30, 35
27, 29, 32, 35
29
30
Val
31
32
32-NH
33
34
175.7, s
54.6, d
26, 30, 32
34, 35
4.76, t (9.4)
8.18, d (9.4)
2.06, m
0.87, d (6.7)
0.84, d (6.8)
32-NH, 33
30, 34, 35
37
32
32
33
33
32, 33, 34, 35
31.2, d
18.6, q
19.7, q
32, 34, 35
32, 35
34
34, 35
32, 33
29, 35
35
N-Me-Val
36
37
38
39
167.6, s
66.5, d
26.1, d
20.7, q
19.2, q
29.3, q
32, 32-NH, 37
39, 40, 41
37, 40
37, 38, 40
37, 38, 39
37
3.95, d (10.5)
2.46, m
1.03, d (6.3)
0.87, d (6.7)
2.75, s
38
38, 39, 40
32-NH, 39, 40, 43
38, 40, 41
37
37, 39, 40
38
38
40
41
38, 44
34, 38, 40
Lac
42
43
170.7, s
66.4, d
17.7, q
37, 41, 44
44
43
5.45, q (6.6)
1.50, d (6.6)
44
43
44
37, 44
40, 43
44
80% aqueous MeOH. After drying, the aqueous methanol
residue was partitioned between water and n-butanol. The
residue from the organic layer was subjected to normal-phase
flash chromatography eluting with increasing amounts of
methanol in dichloromethane. This resulted in the cytotoxicity
concentrated primarily in the 5% methanol fraction. Subse-
quent separation on a C18 column with increasing amounts of
MeCN in H2O resulted in the activity being concentrated
primarily in the 60% MeCN in H2O fraction. This sample was
purified twice by RP-HPLC [Ultracarb 5 ODS 30, 10 × 250
mm, 3 mL/min, detection at 220 nm], first with 70% MeCN in
H2O (tR 21.3 min) and then with 80% MeOH in H2O to yield
1.1 mg of 1 (tR 25.4 min).
+ H]+ 809.5226 (calcd for C44H69N6O8 809.5171, 5.5 mDa
error).
Absolu te Ster eoch em istr y of th e Am in o Acid -Der ived
Un its. A 0.3 mg sample of 1 dissolved in 0.5 mL of methanol
was placed in a 1 mL vial along with 1 mg of 30% Pd/C. The
solution was stirred under an atmosphere of H2 for 18 h at
room temperature. The reaction mixture was filtered through
a pad of Celite, which was then flushed with methanol. After
removal of the solvent, the residue was taken up in 0.3 mL of
6 N HCl and heated to 90 °C for 18 h. The acid was removed
under a stream of nitrogen and the hydrolysate analyzed by
chiral HPLC, comparing the retention times of the components
of the hydrolysate with those of authentic standards [Chirex
Phase 3126 (D), 4.6 × 250 mm, Phenomenex; flow rate 1
mL/min, detection at 254 nm, solvent 2 mM CuSO4 for the
valines; 2 mM CuSO4/MeCN (95:5) for lactic acid and 85:15
for N-Me-Phe]. The retention times (tR, min) of the standards
were L-Val (30.3), D-Val (54.5), N-Me-L-Val (23.4), N-Me-D-Val
Ulon ga p ep tin (1): amorphous powder: [R]21 -16° (c 0.4,
D
MeOH); UV (MeOH) λmax (log ꢀ) 204 (3.84) nm; IR (film) νmax
3336, 1727, 1654 (br), 1508, 1458, 1259, 1078 cm-1; 1H NMR,
13C NMR, 1H-1H COSY, HMBC, TOCSY, and ROESY data,
see Table 1; FABMS m/z 809.4 [M + H]+; HR-MALDI m/z [M

654 J ournal of Natural Products, 2003, Vol. 66, No. 5
Williams et al.
(41.8), L-lactic acid (18.5), D-lactic acid (31.5), N-Me-L-Phe
(31.2), and N-Me-D-Phe (33.5). The retention times of the
components in the hydrolysate were L-lactic acid (18.5), L-Val
(30.3), N-Me-L-Val (23.4), N-Me-D-Val (41.8), and N-Me-D-Phe
(33.5). The identities of the peaks were also confirmed by co-
injection.
Hz) 9.11 (1H, s), 8.93 (1H, d, 9.2), 8.55 (1H, d, 6.4), 7.55 (1H,
br s), 6.63 (1H, br s), 6.11 (1H, s), 4.30 (1H, m), 4.11 (1H, m),
2.85 (1H, m), 1.86 (2H, m), 1.70 (1H, m), 1.47 (2H, m), 1.27
(6H, m), 1.22 (3H, d, 7.1), 1.04 (3H, d, 5.7), 0.95 (3H, d, 5.9),
0.85 (3H, t, 8.0).
Absolu te Ster eoch em istr y of th e 3-Am in o-2-m eth yloct-
7-yn oic Acid . The hydrogenated hydrolysate was derivatized
with L-FDLA by the standard procedure and compared with
the derivatized synthetic standards.11 The analysis was carried
out by RP-HPLC [YMC-Pack AQ-ODS, 10 × 250 mm, 50%
MeCN in 0.01 N TFA, flow rate 2.5 mL/min, PDA detection].
The retention times (min) of the L-FDLA derivatized standards
were (2R,3S)-5 (25.7), (2S,3S)-5 (26.7), (2R,3R)-5 (45.6), and
(2S,3R)-5 (54.1)12 with the retention times of L-FDLA+-
(2R,3S)-5 and l-FDLA-(2S,3S)-5 being inferred from the reten-
tion times of D-FDLA+(2S,3R)-5 and d-FDLA+(2S,3S)-5 re-
spectively. The retention time of the â-amino acid in the
hydrolysate was 26.7 min (2S, 3S), the identity of which was
confirmed by co-injection of the (2S,3S)-5. The previously
identified amino acids appeared at 12.0 (L-Val), 15.1 (N-Me-
L-Val), and 20.5 min (N-Me-D-Val).
Ack n ow led gm en t. We would like to thank the Republic
of Palau for the marine research permit and the National
Cancer Institute for providing support through R01 grant
CA12623 and NCNPDDG grant CA53001. The upgrades to the
NMR spectrometers used in this research were funded by
grants from the CRIF Program of the National Science
Foundation (CHE9974921), the Air Force Office of Scientific
Research (F49620-01-1-0524), and the Elise Pardee Founda-
tion. The cyanobacterium was collected by J . Starmer and
identified with the assistance of E. Cruz-Rivera. G. Tien at
the University of Hawaii, Department of Chemistry carried
out the bioassays. The UCR Mass Spectrometry Facility,
Department of Chemistry, University of California, performed
the mass spectral analyses at Riverside.
Syn th esis of 3-Am in o-2-m eth ylocta n oic Acid . To 15 mL
of THF in a 100 mL flask under N2 were added trimethyl
phosphonoacetate (2) (5.5 mmol) and n-BuLi (5.6 mmol). After
stirring at -78 °C for 1 h, this mixture was cannulated into
20 mL of THF containing hexanal (5.5 mmol). The reaction
was allowed to warm to room temperature overnight. The
solvent was removed and the residue partitioned between
diethyl ether and water. The residue from the organic layer
was dissolved in a 20:1 mixture of petroleum ether/diethyl
ether and chromatographed on silica to yield pure methyl oct-
2(E)-enoate (3). This R,â-unsaturated ester was then treated
according to Davies’ procedure13 except the diastereomers were
not separated after methylation with KHMDS and iodomethane.
The C-2 diastereomers were separated after deprotection as
their FDLA derivatives in the manner previously described.
D-F DLA + (2R,3R)-3-Am in o-2-m eth ylocta n oic Acid : 1H
NMR (acteone-d6, 500 MHz) δ (integration, multiplicity, J in
Hz) 9.11 (1H, s), 8.93 (1H, d, 9.2), 8.55 (1H, d, 6.4), 7.55 (1H,
br s), 6.63 (1H, br s), 6.11 (1H, s), 4.26 (1H, m), 3.99 (1H, m),
2.96 (1H, m), 1.86 (2H, m), 1.68 (1H, m), 1.40 (2H, m), 1.30
(6H, m), 1.23 (3H, d, 6.3), 1.01 (3H, d, 5.9), 0.93 (3H, d, 5.7),
0.85 (3H, t, 7.2).
Su p p or tin g In for m a tion Ava ila ble: The 1H, 13C, and 2D NMR
data of 1 and the full model for Figure 1 are available free of charge
via the Internet at http://pubs.acs.org.
Refer en ces a n d Notes
(1) Staley, J . T.; Castenholz, R. W.; Colwell, R. R.; Holtz, J . G.; Cane, M.
D.; Pace, N. R.; Salyers, A. A.; Tiedje, J . M. The Microbial World:
Foundation of the Biosphere, a report from The American Acadamey
of Microbiology; 1997.
(2) Gerwick, W. H.; Tan, L. T.; Sitachitta, N. In The Alkaloids; Cordell,
G. A., Ed.; Academic Press: San Diego, 2001; Vol. 57, pp 75-184.
(3) The IC50 values for in vitro cytotoxicity were determined using the
sulforhodamine B assay. Skehan, P.; Storeng, R.; Scudiero, D.; Monks,
A.; McMahon, J .; Vistica, D.; Warren, J . T.; Bokesch, H.; Kenney, S.;
Boyd, M. R. J . Natl. Cancer Inst. 1990, 82, 1107-1112.
3
(4) Two HMBC experiments with J CH optimized for
7 and 4 Hz,
respectively, failed to show any correlation between C-1 and H-43.
(5) ROESY cross-peaks between H-37 and H-43 indicate a cis amide bond,
while correlations between H-44 and H-40 of the L-lactic acid and
adjacent N-Me-L-Val units indicate the two side chains are on the
same face of the molecule. Analysis of models suggested that H-40
was responsible for the latter correlation rather than the overlapping
H-34. Unfortunately too little material remains to conclusively
establish which methyl group shows a ROESY correlation to H-44.
(6) Agami, C.; Cheramy, S.; Dechoux, L.; Melaimi, M. Tetrahedron 2001,
57, 195-200.
(7) The methylation occurs predominantly from the re face of the enolate
to yield the anti R-methyl-â-amino acid as the major product. Cf.:
Davies 1994.
(8) Luesch, H.; Yoshida, W. Y.; Moore, R. E.; Paul, V. J .; Mooberry, S. L.
J . Nat. Prod. 2000, 63, 611-615.
(9) Luesch, H.; Yoshida, W. Y.; Moore, R. E.; Paul, V. J . J . Nat. Prod.
2000, 63, 1106-1112.
(10) Fernandez-Suarez, M.; Mun˜oz, L.; Fernandez, R.; Riguera, R. Tetra-
hedron: Asymmetry 1997, 8, 1847-1854, and references within.
(11) Fujii, K.; Yoshitomo, I.; Oka, H.; Suzuki, M.; Harada, K. Anal. Chem.
1997, 69, 5146-5151.
(12) The relative integration of the HPLC peaks after derivatization with
L- and DL-FDLA was used to determine the elution order of the AMO
standards. The major peaks were assigned as (2R, 3R) and ent-(2R,
3R).
L-F DLA + (2R,3R)-3-Am in o-2-m eth ylocta n oic Acid : 1H
NMR (acteone-d6, 500 MHz) δ (integration, multiplicity, J in
Hz) 9.11 (1H, s), 8.91 (1H, d, 8.9), 8.61 (1H, d, 6.3), 7.54 (1H,
br s), 6.73 (1H, br s), 6.11 (1H, s), 4.26 (1H, m), 3.98 (1H, m),
2.93 (1H, m), 1.89 (2H, m), 1.68 (1H, m), 1.44 (2H, m), 1.31
(6H, m), 1.20 (3H, d, 7.3), 1.02 (3H, d, 6.2), 0.94 (3H, d, 6.1),
0.85 (3H, t, 7.9).
D-F DLA + (2S,3R)-3-Am in o-2-m eth ylocta n oic Acid : 1H
NMR (acteone-d6, 500 MHz) δ (integration, multiplicity, J in
Hz) 9.10 (1H, s), 8.54 (1H, d, 8.9), 8.52 (1H, d, 5.5), 7.71 (1H,
br s), 6.64 (1H, br s), 6.20 (1H, s), 4.35 (1H, m), 4.25 (1H, m),
2.92 (1H, m), 1.84 (2H, m), 1.68 (1H, m), 1.40 (2H, m), 1.30
(6H, m), 1.23 (3H, d, 7.0), 1.02 (3H, d, 6.0), 0.94 (3H, d, 6.2),
0.83 (3H, t, 7.0).
(13) Davies, S. G.; Walters, I. A. S. J . Chem. Soc., Perkins Trans. 1 1994,
9, 1129-1139.
L-F DLA + (2S,3R)-3-Am in o-2-m eth ylocta n oic Acid : 1H
NMR (acteone-d6, 500 MHz) δ (integration, multiplicity, J in
NP030050S