Lyngbyastatin 1 and Ibu-epilyngbyastatin 1: Synthesis, stereochemistry, and NMR line broadening

Article abstract of DOI:10.1021/np020117w

The synthesis of a lyngbyastatin 1-Ibu-epilyngbyastatin 1 mixture combined with NMR and molecular modeling studies proved that natural lyngbyastatin 1 was only one Ibu epimer rather than a mixture of both and that the configuration of this epimer in the Ibu unit was R. The substance isolated with lyngbyastatin 1 was Ibu-epidolastatin 12. The extreme broadness in the proton NMR spectra of lyngbyastatin 1 and Ibu-epidolastatin 12 was exchange broadening due to rotation about the Ibu-Ala amide bond. It was a consequence of (1) a small energy difference between the cis and trans forms of this bond, (2) a substantial difference in conformation between these forms, and (3) a lowered barrier between them compared to most amide bonds (due to steric hindrance). The synthetic lyngbyastatin 1-Ibu-epilyngbyastatin 1 mixture had significant activities against cancer cells and in stimulating actin polymerization, but was less active than dolastatin 11 in all assays.

Full text of DOI:10.1021/np020117w

1824
J . Nat. Prod. 2002, 65, 1824-1829
Lyn gbya sta tin 1 a n d Ibu -ep ilyn gbya sta tin 1: Syn th esis, Ster eoch em istr y, a n d
NMR Lin e Br oa d en in g
Ruoli Bai,† Robert B. Bates,*,‡ Ernest Hamel,† Richard E. Moore,§ Pichaya Nakkiew,‡ George R. Pettit, and
Bilal A. Sufi‡
Screening Technologies Branch, Developmental Therapeutics Program, Division of Cancer Treatment and Diagnosis,
National Cancer Institute at Frederick, National Institutes of Health, Frederick, Maryland 21702, Department of Chemistry,
University of Arizona, Tucson, Arizona 85721, Department of Chemistry, University of Hawaii, Manoa, Hawaii 96822, and
Cancer Research Institute and Department of Chemistry, Arizona State University, Tempe, Arizona 85287-1604
Received March 20, 2002
The synthesis of a lyngbyastatin 1-Ibu-epilyngbyastatin 1 mixture combined with NMR and molecular
modeling studies proved that natural lyngbyastatin 1 was only one Ibu epimer rather than a mixture of
both and that the configuration of this epimer in the Ibu unit was R. The substance isolated with
lyngbyastatin 1 was Ibu-epidolastatin 12. The extreme broadness in the proton NMR spectra of
lyngbyastatin 1 and Ibu-epidolastatin 12 was exchange broadening due to rotation about the Ibu-Ala
amide bond. It was a consequence of (1) a small energy difference between the cis and trans forms of this
bond, (2) a substantial difference in conformation between these forms, and (3) a lowered barrier between
them compared to most amide bonds (due to steric hindrance). The synthetic lyngbyastatin 1-Ibu-
epilyngbyastatin 1 mixture had significant activities against cancer cells and in stimulating actin
polymerization, but was less active than dolastatin 11 in all assays.
The isolation of the cytotoxic cyclic depsipeptides lyng-
byastatin 1 (4, earlier assigned structure 3), Ibu-epilyng-
byastatin 1 (3, previously assigned structure 4), dolastatin
12 (5),1 and Ibu-epidolastatin 12 (6) was reported from a
Lyngbya majuscula/ Schizothrix calcicola assemblage and
a L. majuscula strain collected near Guam.2,3 The extreme
broadness of the NMR peaks of these substances was
attributed to the presence of both Ibu-epimers in unsepa-
rated mixtures. We now report the synthesis of a mixture
of 3 and 4 and resulting data which show that the
compounds isolated from Lyngbya/ Schizothrix were lyng-
byastatin 1 (4) and Ibu-epidolastatin 12 (6) and that the
NMR line broadening is due to sterically lowered rotation
barriers about the Ibu-Ala amide bonds with appreciable
amounts of conformationally distinct cis and trans forms
present at equilibrium. This steric hindrance, characteristic
of N,N-dialkylamides with quaternary R-carbons, made the
preparation of the Ibu-N-Me-Ala amide unit difficult, as
described below.
Resu lts a n d Discu ssion
Two unsuccessful strategies for making the critical Ibu-
N-Me-L-Ala unit are shown in Figure 1. The first was to
substitute L-Ala-OBn‚HCl with N-Me-L-Ala-OBn‚HCl (7)
ester 12, followed by coupling with amine 7. Attempts to
dimethylate keto ester 8 using MeI/NaH did not give 12.
When DBU was used as the base, the first methyl group
went to the right place but the second went on nitrogen,
giving 13. Again, the desired product was apparently not
obtained for steric reasons.
In the successful route to the Ibu-N-Me-L-Ala unit
(Figure 2), di-Boc-L-Ala (14)6 was used instead of Boc-L-
Ala to prevent N-methylation. Acid 14 was activated with
CDI and allowed to react with LiCH2COOBn, giving a 66%
yield of â-keto ester 15. Ester 15 was gem-dimethylated to
ester 16 in 78% yield, using MeI/DBU in acetonitrile.
Hydrogenolysis of benzyl ester 16, followed immediately
by coupling with amine salt 7 using BOP-Cl, gave a 93%
yield of Boc2-Ibu-N-Me-L-Ala-OBn (17). The Boc units of
in the synthetic sequence used to make Boc-Ibu-L-Ala-OBn
in the dolastatin 11 (1) synthesis,4 hoping to obtain Boc-
Ibu-N-Me-L-Ala-OBn (11). Benzyl ester 75 was coupled to
the â-keto acid from the hydrogenation of the â-keto ester
8,4 giving dipeptide 9 in 63% yield. However, attempts to
dimethylate 9 with MeI/NaH failed, presumably for steric
reasons; only monomethylation product 10 was obtained.
A second approach to the Ibu-N-Me-L-Ala unit was by
dimethylation of â-keto ester 8 to the gem-dimethylated
* To whom correspondence should be addressed. Tel: (520)621-6317.
Fax: (520)621-8407. E-mail: batesr@u.arizona.edu.
† National Cancer Institute.
‡ University of Arizona.
§ University of Hawaii.
Arizona State University.
10.1021/np020117w CCC: $22.00
© 2002 American Chemical Society and American Society of Pharmacognosy
Published on Web 10/25/2002

Lyngbyastatin 1 and Ibu-epilyngbyastatin 1
J ournal of Natural Products, 2002, Vol. 65, No. 12 1825
F igu r e 1. Unsuccessful strategies for synthesis of the Ibu-N-Me-L-
Ala unit.
F igu r e 3. Synthesis of Ibu-epilyngbyastatin 1 (3) and lyngbyastatin
1 (4).
those of dolastatin 12 (5), except for the peaks due to the
tyrosine side chain; these showed the presence of Ibu-
epilyngbyastatin 1 (now 3). The NMR spectral parameters
of dolastatin 12 (5), previously unreported, were obtained
with the aid of COSY and HETCOR spectra. The peaks of
Ibu-epilyngbyastatin 1 (3) were then assigned by compari-
son with those of dolastatin 12 (5). The peaks of lyngby-
astatin 1 (4) and Ibu-epidolastatin 12 (6)2 remain largely
unassigned due to their extreme broadness.
F igu r e 2. Successful route to the Ibu-N-Me-L-Ala unit.
Ibu-N-Me-L-Ala derivative 17 were removed with TFA to
give TFA salt 18, which was used in the synthesis of 3 and
4.
Except for the use of di-Boc protection and the presence
of the N-methyl group on the alanine, these procedures
follow those used in the dolastatin 11 (1) synthesis, in
which the S configuration was maintained in the Ibu unit
throughout.4 However, in the current synthesis, epimer-
ization in the Ibu unit occurred, probably in the step in
which DBU was used,7 leading to equal amounts of Ibu
epimers 3 and 4.
When the route from dipeptide 18 to depsipeptide 4
analogous to that used for dolastatin 11 (1)4 failed for
unknown reasons, a synthesis was achieved by reversing
the order of protecting group removal to that shown in
Figure 3. The benzyl group of heptadepsipeptide 19 was
removed by hydrogenolysis, yielding carboxylic acid 20
quantitatively. Nonadepsipeptide 21 was made in 79%
yield by coupling TFA salt 18 with acid 20 using TBTU.
The protecting groups were removed by hydrogenolysis and
TFA. The final coupling using HBTU gave a 1:1 ratio of
Ibu epimers 3 and 4 in 40% total yield after HPLC.
Attempts to separate these epimers by HPLC failed since
their retention times were so similar. However, it was clear
from the NMR spectra (Figure 4) what compounds were
present in the mixture, as described below.
Three arguments that the configurations in the Ibu units
of depsipeptides 3-6 are assigned correctly and not re-
versed are as follows: (a) dolastatins 11 (1) and 12 (5),
found in the same organism, are more likely to share the
same configuration at the Ibu center than not, and the
configuration of dolastatin 11 (1) was shown to be S at the
1
Ibu center by synthesis from S-alanine;4 (b) the H NMR
spectrum (Table 1) of dolastatin 12 (5) is closer to that of
dolastatin 11 (1) than to that of Ibu-epidolastatin 11 (2),
especially noticeable in the valine and Hmp units;8 (c) the
NMR line-broadening considerations discussed next.
The severe line broadening in the proton NMR spectrum
of lyngbyastatin 1 (4) and much less noticeable broadening
for Ibu-epilyngbyastatin 1 (3) are related to the lowered
rotation barrier about the amide bond in the Ibu-N-Me-
Ala unit compared to most amides. For example, in N,N-
dimethylamides, the energy barrier to rotation about the
amide bond drops from 88 kJ /mol in N,N-dimethylforma-
mide to 71 kJ /mol in N,N-dimethylacetamide, 67 kJ /mol
in N,N-dimethylisobutyramide, and only 50 kJ /mol in N,N-
dimethylpivalamide.9 When the R-carbon is quaternary, as
in N,N-dimethylpivalamide and the Ibu-N-Me-Ala units of
depsipeptides 3-6, the interconversion of cis and trans
amides occurs at about the same rate as the NMR mea-
surement and can result in broadness. In strong support
of this interpretation, the broadened lines in the carbon
One set of mostly very broad peaks matched those of
natural lyngbyastatin 1, which we now assign as 4. (The
extensive degradative work on lyngbyastatin 12 had es-
tablished the structure except for the configuration in the
Ibu unit.) A second set of much sharper peaks matched

1826 J ournal of Natural Products, 2002, Vol. 65, No. 12
Bai et al.
F igu r e 4. 1H NMR spectra (CDCl3) of N-methyl-L-alanine derivatives related to lyngbyastatin 1 (4). (A) Natural lyngbyastatin 1 (4, R configuration
in Ibu unit, 500 MHz). (B) Synthetic mixture of lyngbyastatin 1 (4, R in Ibu unit) and Ibu-epilyngbyastatin 1 (3, S in Ibu unit, 500 MHz). (C)
Natural dolastatin 12 (5, S configuration in Ibu unit, 400 MHz).
spectrum of dolastatin 12 (5), due to a lowered barrier
about its Ibu-N-Me-L-Ala amide bond, are observed for
carbons in its Map, Ala, Ibu, and Tyr units (Table 2).
That extreme broadness would be observed with the
compounds that have R configurations in the Ibu unit
(lyngbyastatin 1 (4) and Ibu-epidolastatin 12 (6)) and much
less with compounds that have S configurations in the Ibu
unit (Ibu-epilyngbyastatin 1 (3) and dolastatin 12 (5)) is
supported by the molecular modeling studies described
next, which suggest that at probe temperature depsipep-

Lyngbyastatin 1 and Ibu-epilyngbyastatin 1
J ournal of Natural Products, 2002, Vol. 65, No. 12 1827
Ta ble 1. 1H NMR Assignments of Dolastatin 11 (1), Ibu-epidolastatin 11 (2), Dolastatin 12 (5), Ibu-epidolastatin 12 (6),
Lyngbyastatin 1 (4), and Ibu-epilyngbyastatin 1 (3)
compound
unit
Ma p
R-CH
â-CH3
â-CH
γ-CH2
δ-CH3
NH
1
2
3a
2.87m
1.08m
4.10m
4b
5a
6b
2.79qd(7,2.5)
1.10d(7)
2.75m
2.87m
1.12d(7)
1.09d(7)
4.13m
1.35,1.59m
0.92t(7)
6.95br s
4.47m
1.46,1.56m
0.93t(7.5)
7.09d(10.5)
0.95t(7)
6.91d(10)
0.85-0.95
Ala
R-CH
â-CH3
NH/NCH3
Ibu
4.44p(7)
1.07d(7)
7.78d(8.5)
4.61p(7)
1.08d(7)
7.52d(8)
5.23m
1.30d(6.5)
2.58br s
5.26q(7)
1.30d(7)
2.59br s
1.21d(7)
1.21d(7)
â-CH3
γ-CH
δ-CH3
NH
1.44,1.49s
4.91p(7)
1.13d(7)
7.15d(9)
1.52,1.52s
4.98p(6.5)
1.22d(6.5)
7.31d(8)
1.40,1.54s
4.89m
1.38d(7)
1.32s
1.42,1.56s
4.91p(7.5)
1.39d(7)
1.32s
1.21d(7)
1.19d(7)
6.80d(7.5)
Tyr
R-CH
â-CH2
5.11dd(8.5,6.5)
2.82dd(14,8.5),
3.25dd(14,6.5)
7.14d(8.5)
5.04t(8)
2.78m, 3.27dd
(13.5,8)
7.14d(9)
6.83d(9)
4.94m
4.98m
2.95m 3.17dd
(13,4)
7.23d(7.5)
7.34t(7.5)
7.27t(7.5)
δ-CH
ꢀ-CH
ú-CH
OCH3
NCH3
Va l
7.12d(8)
6.83d(8)
7.14d(8)
6.84d(8)
7.25m
7.25m
7.25m
6.81d(8.5)
3.75s
2.96s
3.77s
2.99s
3.77s
2.92br s
3.76br s
2.75-3.10
2.94s
2.75-3.10
R-CH
â-CH
γ-CH3
4.78d(10.5)
2.23m
0.37d(6.5),
0.74d(6.5)
2.95s
4.89d(10.5)
2.28m
0.57d(6.5),
0.81d(6.5)
2.93s
4.69d(10)
4.72d(10.5)
2.16m
0.24d(6.5),
0.69d(6.5)
2.92s
0.29br s,
0.69br s
2.89br s
0.40br s,
0.70br s
2.75-3.10
0.34br s,
0.70br s
2.75-3.10
NCH3
Gly-1
R-CH2
3.60dd(18,2)
4.42dd(18,7.5)
7.41dd(7.5,2)
3.65dd(18,2),
4.44dd(18,7.5)
3.94m,
4.40m
3.96m
4.44dd(18,7)
7.30m
NH
Leu
R-CH
â-CH2
5.37dd(11,5)
1.62m,1.87ddd
(13.5,11,4)
1.56m
0.92d(6.5),
0.98d(6.5)
3.14s
5.20dd(10,5.5)
1.81ddd
(13,10.5,4)
5.19m
1.83ddd
(14,10,4)
5.20dd(10.5,5)
1.57m,1.85ddd
(14,10,4)
1.45m
0.90d(7),
0.97d(6.5)
3.06s
γ-CH
δ-CH3
0.89-0.94m
0.85-0.95
3.04br s
0.85-0.95
2.75-3.10
0.85-0.95
2.75-3.10
NCH3
Gly-2
R-CH2
3.10s
3.58dd(16,4.5),
4.46dd(16,7)
7.35dd(7,4.5)
3.71dd(16,4.5),
4.48dd(16,7)
7.34dd(7,4.5)
3.92m, 4.27m
6.55br s
3.94m,
4.31dd(19,6)
6.59m
NH
Hm p
R-CH
â-CH
γ-CH3
γ-CH2
δ-CH3
5.19d(3.5)
2.07m
0.89d(7)
1.23,1.47m
0.88t(7.5)
5.27d(3)
2.07m
0.89d(7.5)
5.10d(5)
5.12d(5.5)
2.13m
0.95d(6.5)
1.23,1.53m
0.92t(7)
0.85-0.95
0.85-0.95
0.85-0.95
0.85-0.95
0.85-0.95
0.85-0.95
0.88t(7.5)
a
b
Some broadening of spectral lines in Ibu-Ala region. All spectral lines very broad.
tides 4 and 6 have appreciable amounts of trans- and cis-
Ibu-Ala-amide forms with very different conformations
(and thus very different NMR shifts), while depsipeptides
3 and 5 have a predominant trans form accompanied by a
cis form, which does not differ much in conformation (and
thus NMR shifts) from the trans form except in the vicinity
of the Ibu and Ala units. These calculations were done by
first locking the Ibu-Ala amide bond in the trans configu-
ration and then in the cis configuration. Table 3 shows the
calculated energies of the lowest-energy cis and trans forms
of lyngbyastatin 1 (4, R Ibu configuration, 4.4 kJ /mol
difference) to be closer in energy than those of Ibu-
epilyngbyastatin 1 (3, S Ibu configuration, 8.0 kJ /mol
difference). The smaller energy difference between cis and
trans forms for lyngbyastatin 1 (3) and resulting larger
amounts of the minor form is one factor that should
contribute to greater broadness in its NMR spectra.
Another factor that should make the spectra of lyngby-
astatin 1 (4) much broader is that its lowest-energy cis and
trans forms (4c and 4t, respectively) have very different
overall shapes, and thus they should have greater chemical
shift differences between the cis and trans forms through-
out the molecule. It should be noted that (except for
conformation 5 of 4t) all of the conformations in Table 3
share essentially the same conformation of the 30-
membered ring as the other conformations in that column
and differ only in side-chain rotations. These calculations
of energy and shape differences between cis and trans
forms thus both support the view that lyngbyastatin 1, with
much broader NMR peaks, is 4, and Ibu-epilyngbyastatin
1 is 3.
This mixture of synthetic lyngbyastatin 1 (4) and its Ibu
epimer (3) showed significant cytotoxicities against several
human tumor cells and stimulated actin polymerization,10

1828 J ournal of Natural Products, 2002, Vol. 65, No. 12
Bai et al.
Ta ble 2. 13C NMR Chemical Shifts for Dolastatins 11 (1) and
12 (5) in CDCl3
Cl ) bis(2-oxo-3-oxazolidinyl)phosphinic chloride, CDI ) 1,1′-
carbonyldiimidazole, DBU ) 1,8-diazabicyclo[5.4.0]undec-7-
ene, EDC ) 1-[3-(dimethylamino)propyl]-3-ethylcarbodiimide
hydrochloride, HBTU ) O-benzotriazol-1-yl-N,N,N′N′-tetra-
methyluronium hexafluorophosphate, Hmp ) (2S,3S)-2-hy-
droxy-3-methylpentanoic acid, Ibu ) (S)-4-amino-2,2-dimethyl-
3-oxopentanoic acid, Map ) (2S,3R)-3-amino-2-methylpen-
tanoic acid, TBTU ) O-benzotriazol-1-yl-N,N,N′N′-tetrameth-
yluronium tetrafluoroborate, and TFA ) trifluoroacetic acid.
N-Me-L-Ala -OBn ‚HCl (7). Prepared by the method of
Coppola et al.:5 1H NMR (DMSO-d6) δ 1.44 (d, J ) 7 Hz, 3H,
â-Me), 2.55 (s, 3H, N-Me), 4.15 (q, J ) 7 Hz, 1H, a-CH), 5.24
(s, 2H, BnCH2), 7.39 (m, 5H, Ar).
Ben zyl 4-(N,N-Di-Boc-a m in o)-3-oxop en ta n oa te (15). A
solution of 14 (1.0 g, 3.5 mmol) and 1,1′-carbonyldiimidazole
(0.61 g, 3.8 mmol) in THF (5 mL) was stirred at 0 °C for 30
min and 25 °C for 2 h. To prepare LDA, BuLi (1.6 M in hexane,
6.5 mL, 10.4 mmol) was added dropwise to a solution of
diisopropylamine (1.45 mL, 10.4 mmol) in THF (10 mL) at -78
°C. After warming to 0 °C for 15 min and recooling to -78 °C,
benzyl acetate (1.5 mL, 10.4 mmol) was added and stirring
was continued at -78 °C for 1.25 h. The acylimidazole
prepared earlier was cannulated into this solution, stirring was
continued at -78 °C for 15 min, and 1 N HCl (10 mL) was
added. The mixture was warmed to 0 °C, acidified to pH 3
with citric acid, and extracted with EtOAc (3 × 100 mL). The
combined organic phase was washed with 5% NaHCO3 (2 ×
100 mL) and brine (2 × 100 mL). Solvent evaporation and
HPLC gave ester 15 (0.96 g, 66%): 1H NMR δ 1.41 (d, J ) 6.5
Hz, 3H, δ-Me), 1.49 (s, 18H, Boc), 3.52 and 3.59 (d, J ) 16 Hz,
2H, R-CH2), 4.85 (q, J ) 6.5 Hz, 1H, γ-CH), 5.16 (s, 2H,
BnCH2), 7.34 (br s, 5H, Ar).
1
5
1
5
1
5
Ma p
Tyr
CdO
Leu
CdO
a-CH
â-CH3
â-CH
γ-CH2
δ-CH3
172.6 173.5
168.0 167.7 CdO
61.1 61.2 R-CH
34.7 35.3a â-CH2
128.7 136.1 δ-CH
171.7 171.6b
54.7 54.3
38.1 37.6
24.9 25.0
42.4 42.5a R-CH
9.9 11.5a â-CH2
51.4 52.8a γ-C
25.9 25.1a δ-CH
10.9 11.2a ꢀ-CH
ú-CH
130.4 129.1 δ1-CH3 24.4 21.7
114.4 129.2 δ2-CH3 23.2 23.1
158.7 127.4 NCH3
55.3
29.4 29.3 Gly-2
CdO
30.2 29.8
Ala
OCH3
CdO
R-CH
â-CH3
NCH3
172.8 172.2ab NCH3
48.3 52.6a
170.0 169.6b
40.7 41.5
15.5 13.9a Va l
31.9a CdO
R-CH2
170.1 169.7b
58.2 57.8 Hm p
27.1 27.0 CdO
R-CH
â-CH
Ibu
CdO
R-C
170.1 169.8b
78.4 78.4
37.4 36.5
11.6 11.3
23.8 24.3
15.5 15.1
171.9 170.8ab γ1-CH3 18.4 18.3 R-CH
54.9 55.4a γ2-CH3 18.5 18.4 â-CH
â1-CH3 21.6 21.5a NCH3
â2-CH3 22.0 25.2a
29.2 29.0 γ-CH3
γ-CH2
â-CdO 209.7 210.3
Gly-1
δ-CH3
γ-CH
δ-CH3
51.2 50.3
CdO
169.3 169.3
41.1 41.5
19.2 19.6a R-CH2
a
b
Shows broadening. May be interchanged.
Ta ble 3. Calculated Energies (kJ /mol) of the Conformations
within 10 kJ /mol of the Lowest-Energy Form for Lyngbyastatin
1 (4) and Ibu-epilyngbyastatin 1 (3) in Chloroform
3 (S Ibu configuration) 4 (R Ibu configuration)
conformation
trans
cis
trans
cis
Ben zyl 4-(N,N-Di-Boc-a m in o)-2,2-d im eth yl-3-oxop en -
ta n oa te (16). To a solution of ester 15 (0.22 g, 0.52 mmol) in
CH3CN (15 mL) were added DBU (0.155 mL, 1.03 mmol) and
methyl iodide (0.16 mL, 1.03 mmol). The mixture was stirred
at 25 °C for 16 h, the solvent was evaporated, and the residue
was washed with water, giving methyl derivative 16 as a
yellow oil (0.18 g, 78%): 1H NMR δ 1.36 (d, J ) 6.5 Hz, 3H,
δ-Me), 1.40 and 1.44 (s, 6H, â-Me’s), 1.49 (s, 18H, Boc), 5.07
(q, J ) 6.5 Hz, 1H, γ-CH), 5.13 and 5.21 (d, J ) 12.5 Hz, 2H,
CH2), 7.33 (br s, 5H, Ar).
N,N-Di-Boc-Ibu -N-Me-L-Ala -OBn (17). A mixture of ben-
zyl ester 16 (67 mg, 0.15 mmol), Pd/C (15 mg), and CH2Cl2 (5
mL) was stirred under H2 (1 atm) for 3 h at -5 °C. After
quickly filtering the mixture into a solution of amine salt 7
(37 mg, 0.16 mmol) in CH2Cl2 (10 mL) at 0 °C, bis(2-oxo-3-
oxazolidinyl)phosphinic chloride (49 mg, 0.19 mmol) and
diisopropylethylamine (67 µL, 0.38 mmol) were added. After
stirring at 0 °C for 4 h and 25 °C for 16 h, solvent evaporation
and HPLC gave dipeptide 17 as an oil (74 mg, 93%): 1H NMR
(mixture of rotamers) δ 1.38 (d, J ) 7 Hz, 3H, Ibu-δ-Me), 1.44
(d, J ) 6.5, 3H, Ala-Me), 1.41 and 1.45 (s, 6H, Ibu-â-Me’s),
1.49 (s, 18H, Boc), 2.78 and 2.81 (s, 3H, NMe), 4.82 (q, J ) 7
Hz, 1H, Ala-CH), 5.11 (q, J ) 7 Hz, 1H, Ibu-CH), 5.13 and
5.16 (d, J ) 16 Hz, 2H, BnCH2), 7.34 (br s, 5H, Ar); FABMS
m/z 535 [M + 1]+.
Boc-Ma p -Hm p -Gly-N-Me-L-Leu -Gly-N-Me-L-Va l-O,N-d i-
Me-L-Tyr -Ibu -N-Me-L-Ala -OBn (21). To dipeptide 17 (6.2
mg, 0.0012 mmol) was added TFA (15 mL), and the solvent
was evaporated. To the residue was added EtOAc (10 mL),
and the solvent was evaporated to give TFA salt 18 (6.4 mg,
100%). Heptadepsipeptide 19 (14 mg, 0.014 mmol)4 was stirred
with Pd/C (15 mg) in CH2Cl2 under H2 (1 atm) for 2 h; catalyst
filtration and solvent evaporation gave carboxylic acid 20 (12.4
mg, 100%). A solution of TFA salt 18 (6.4 mg, 0.0012 mmol),
20 (10.3 mg, 0.0012 mmol), diisopropylethylamine (8 µL, 0.046
mmol), and TBTU (22 mg, 0.069 mmol) in DMF (3 mL) was
stirred for 20 h. Solvent evaporation and HPLC gave peptide
21 (11 mg, 79%): 1H NMR δ 0.40 and 0.75 (m, 6H, Val-Me’s),
1.42 (s, 9H, Boc), 2.40, 2.45, 2.68, 2.72, 2.91, 2.94, and 2.99 (s,
12H, NMe’s), 3.72, 3.75, and 3.77 (s, 3H, OMe), 5.14 and 5.20
(d, J ) 15 Hz, 2H, BnCH2), 6.78 and 7.09 (m, 4H, Tyr-Ar),
7.34 (br s, 5H, Bn-Ar); FABMS m/z 1207 [M + 1]+.
1
2
3
4
5
-596.8
-593.8
-588.8
-588.0
-580.3
-579.7
-577.8
-576.2
-575.6
-584.7
-580.6
-579.8
-579.4
but the mixture was less active than dolastatin 11 (1) in
all assays. The human cancer cell growth inhibitions (GI50
µg/mL) of the mixture, probably largely due to the Ibu
epimer 3 since Ibu-epidolastatin 11 (2) was much less
active than dolastatin 11 (1),8 were 0.031 and 0.24 against
the NCI-H460 (lung) and DU-145 (prostate) cell lines,
respectively, compared to 0.0013 and 0.22, respectively, for
dolastatin 11 (1). For induction of actin polymerization, the
mixture had an EC50 value of 40 ( 7 µM; cf. 42 ( 3 µM for
phalloidin and 9.5 ( 0.7 µM for dolastatin 11 (1).10
In summary, a synthesis of a lyngbyastatin 1 (4)-Ibu-
epilyngbyastatin 1 (3) mixture combined with NMR and
molecular modeling studies proved that natural lyngby-
astatin 1 (4) has the R configuration in the Ibu unit and is
not accompanied by the S epimer as suggested earlier.2 The
natural substance isolated along with lyngbyastatin 1 (4)2
was Ibu-epidolastatin 12 (6), unaccompanied by dolastatin
12 (5). The severe broadness in the proton NMR spectra of
lyngbyastatin 1 (4) and Ibu-epidolastatin 12 (6) was
exchange broadening from rotation about the Ibu-Ala
amide bond; for such N,N-dialkylamides with quaternary
R-carbons, when the cis-trans energy difference is small,
considerable broadening should be expected for the parts
of the molecule where their conformations differ signifi-
cantly. The mixture of depsipeptides 3 and 4 displayed
significant activities against human cancer cells and in
stimulating actin hyperassembly, but was less active than
dolastatin 11 (1).
;
Exp er im en ta l Section
Gen er a l Exp er im en ta l P r oced u r es. General procedures
have been described previously.4 Abbreviations used are BOP-

Lyngbyastatin 1 and Ibu-epilyngbyastatin 1
J ournal of Natural Products, 2002, Vol. 65, No. 12 1829
3t, 4c, and 4t. This material is available free of charge via the Internet
at http://pubs.acs.org.
Boc-Ma p -Hm p -Gly-N-Me-L-Leu -Gly-N-Me-L-Va l-O,N-d i-
Me-L-Tyr -Ibu -N-Me-L-Ala (22). A solution of benzyl ester 21
(11 mg, 0.0009 mmol) in CH2Cl2 was stirred with Pd/C (15 mg)
under H2 (1 atm) for 2 h. Removal of catalyst by filtration and
solvent evaporation gave carboxylic acid 22 (10 mg, 99%): 1H
NMR δ 0.37 and 0.75 (d, J ) 7 Hz, 6H, Val-Me’s), 1.42 (s, 9H,
Boc), 2.37, 2.68, 2.79, 2.92, 2.94, and 2.97 (s, 12H, NMe’s), 3.76
and 3.77 (s, 3H, OMe), 6.78, 6.82, and 7.12 (m, 4H, Ar).
Lyn gbya sta tin 1 (4) a n d Ibu -ep ilyn gbya sta tin 1 (3). To
acid 22 (10 mg, 0.009 mmol) was added TFA (10 mL) followed
by solvent evaporation. EtOAc (5 mL) was added and evapo-
rated. To the residual TFA salt (10 mg, 0.009 mmol) in DMF
(3.5 mL) were added Et3N (6 µL, 0.045 mmol) and HBTU (34
mg, 0.09 mmol). Stirring for 7 h, solvent evaporation, and
HPLC gave lyngbyastatin 1 (4, 1.8 mg, 20%) admixed with
Ibu-epilyngbyastatin 1 (3, 1.8 mg, 20%): 1H NMR, Table 1 and
Figure 4; FABMS m/z 999 [M + 1]+.
Refer en ces a n d Notes
(1) Pettit, G. R.; Kamano, Y.; Kizu, H.; Dufresne, C.; Herald, C. L.;
Bontems, R. J .; Schmidt, J . M.; Boettner, F. E.; Nieman, R. A.
Heterocycles 1989, 28, 553-558.
(2) Harrigan, G. G.; Yoshida, W. Y.; Moore, R. E.; Nagle, D. G.; Park, P.
U.; Biggs, J .; Paul, V. J .; Mooberry, S. L.; Corbett, T. H.; Valeriote,
F. A. J . Nat. Prod. 1998, 61, 1221-1225.
(3) “Ibu-epi-” is preferred to using a number, since dolastatins1 and
lyngbyastatins2 have been numbered in different ways. Dolastatin
numbering1 is used herein.
(4) Bates, R. B.; Brusoe, K. G.; Burns, J . J .; Caldera, S.; Cui, W.;
Gangwar, S.; Gramme, M. R.; McClure, K. J .; Rouen, G. P.; Schadow,
H.; Stessman, C. C.; Taylor, S. R.; Vu, V. H.; Yarick, G. V.; Zhang,
J .; Pettit, G. R.; Bontems, R. J . Am. Chem. Soc. 1997, 119, 2111-
2113.
(5) Coppola, B. P.; Noe, M. C.; Schwartz, D. J .; Abdon, R. L., II; Trost,
B. M. Tetrahedron 1994, 50, 93-116.
(6) Gunnarsson, K.; Ragnarsson, U. Acta Chem. Scand. 1990, 44, 944-
951.
(7) Evans, P. A.; Roseman, J . D.; Garber, L. T. J . Org. Chem. 1996, 61,
4880-4881.
Ack n ow led gm en t. For financial support, we thank the
NIH (grant R01 CA78750), the Camille and Henry Dreyfus
Foundation (Senior Scientist Mentor Initiative), the Research
Corporation, Bristol-Myers Squibb Company, Aldrich Chemical
Co., and the Government of Thailand (fellowship to P.N.).
(8) Unpublished results from the laboratory of RBB.
(9) J ackman, L. M.; Cotton, F. A. Dynamic Nuclear Magnetic Resonance
Spectroscopy; Academic: New York, 1975; pp 208-210.
(10) Bai, R.; Verdier-Pinard, P.; Gangwar, S.; Stessman, C. C.; McClure,
K. J .; Sausville, E. A.; Pettit, G. R.; Bates, R. B.; Hamel, E. Mol.
Pharmacol. 2001, 59, 462-469.
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra of 7, 15-
17, 21, and 22; table of the 30-membered ring torsion angles for 3c,
NP020117W