Synthesis of [Gly-1]RA-VII, [Gly-2]RA-VII, and [Gly-4]RA-VII. Glycine-Containing Analogues of RA-VII, an Antitumor Bicyclic Hexapeptide from Rubia Plants

Article abstract of DOI:10.1021/jo030293p

Three analogues of RA-VII (1), an antitumor bicyclic hexapeptide from Rubia plants, were synthesized. Three analogues, [Gly-1]RA-VII (4), [Gly-2]RA-VII (5), and [Gly-4]RA-VII (6), in which one of the three alanine residues in 1 was replaced by a glycine r

Full text of DOI:10.1021/jo030293p

Syn th esis of [Gly-1]RA-VII, [Gly-2]RA-VII, a n d [Gly-4]RA-VII.
Glycin e-Con ta in in g An a logu es of RA-VII, a n An titu m or Bicyclic
Hexa p ep tid e fr om Ru bia P la n ts
Yukio Hitotsuyanagi,† Tomoyo Hasuda,† Takayuki Aihara,† Hiroshi Ishikawa,†
Kentaro Yamaguchi,‡ Hideji Itokawa,† and Koichi Takeya*,†
School of Pharmacy, Tokyo University of Pharmacy and Life Science, 1432-1 Horinouchi, Hachioji,
Tokyo 192-0392, J apan, and Chemical Analysis Center, Chiba University, Yayoi-cho, Inage-ku,
Chiba 263-8522, J apan
takeyak@ps.toyaku.ac.jp
Received September 23, 2003
Three analogues of RA-VII (1), an antitumor bicyclic hexapeptide from Rubia plants, were
synthesized. Three analogues, [Gly-1]RA-VII (4), [Gly-2]RA-VII (5), and [Gly-4]RA-VII (6), in which
one of the three alanine residues in 1 was replaced by a glycine residue, were prepared by linking
of the cycloisodityrosine unit, obtained by degradation of 1, to three different glycine-containing
tetrapeptides followed by macrocyclization. Of these three analogues, analogue 4 showed the highest
cytotoxic activity. The NMR study revealed that in solution the conformer structures and their
ratios of analogue 4 were very similar to those of natural peptide 1, suggesting that the methyl
groups at Ala-2 and Ala-4 should be essential for producing the bioactive conformation, whereas
that at D-Ala-1 is not essential.
In tr od u ction
also be vital because some cycloisodityrosine-containing
RA-VII analogues, in which the tetrapeptide fragment
of the 18-membered ring is modified by reduction,8 by
cleavage of the 18-membered ring,9 or by replacement
with a polyglycine unit,10 show a very weak or no
cytotoxic activity. The configuration of certain amino acid
residues in cyclic peptides takes part in deciding the
conformation of peptide rings11 and thus should affect the
conformation of the whole molecule. In the present study,
we studied the effect of the three alanyl methyls of the
18-membered ring in 1 upon its conformation and cyto-
toxic activity by syntheses of RA-VII analogues in which
one of the three alanines was replaced by glycine.
Information obtained by this study may provide a useful
knowledge for designing more simple and more active
analogues of RA-VII with a reduced number of chiral
centers.
An antitumor bicyclic hexapeptide RA-VII (1) and its
16 congeners have been isolated from Rubia akane and
Rubia cordifolia (Rubiaceae),1-3 and two related peptides,
bouvardin (NSC 259968) (2) and deoxybouvardin (3),
were isolated from another plant of this family, Bou-
vardia ternifolia.4 The structures of the peptides of this
series are characterized by two macrocyclic rings, i.e., an
18-membered cyclohexapeptide ring and a 14-membered
cycloisodityrosine ring, fused together to form a 26-
membered ring. The antitumor action of 1 is considered
to be due to the inhibition of protein synthesis through
its interaction with eukaryotic 80S ribosomes,5,6 and the
cycloisodityrosine ring is proposed to be the pharmaco-
phore in this type of peptides because cycloisodityrosine
itself shows a potent cytotoxic activity.7 However, the role
of the 18-membered ring moiety in the activity should
(7) (a) Boger, D. L.; Myers, J . B.; Yohannes, D.; Kitos, P. A.;
Suntornwat, O.; Kitos, J . C. Bioorg Med. Chem. Lett. 1991, 1, 313-
316. (b) Boger, D. L.; Yohannes, D.; Myers, J . B. J . Org. Chem. 1992,
57, 1319-1321. (c) Boger, D. L.; Yohannes, D.; Zhou, J .; Patane, M. A.
J . Am. Chem. Soc. 1993, 115, 3420-3430. (d) Boger, D. L.; Patane, M.
A.; J in, Q.; Kitos, P. A. Bioorg. Med. Chem. 1994, 2, 85-100.
(8) Hitotsuyanagi, Y.; Matsumoto, Y.; Sasaki, S.; Suzuki, J .; Takeya,
K.; Yamaguchi, K.; Itokawa, H. J . Chem. Soc., Perkin Trans. 1 1996,
1749-1755.
* To whom correspondence should be addressed.
† Tokyo University of Pharmacy and Life Science.
‡ Chiba University.
(1) (a) Itokawa, H.; Takeya, K.; Mori, N.; Hamanaka, T.; Sonobe,
T.; Mihara, K. Chem. Pharm. Bull. 1983, 31, 1424-1427. (b) Itokawa,
H.; Takeya, K.; Mihara, K.; Mori, N.; Hamanaka, T.; Sonobe, T.; Iitaka,
Y. Chem. Pharm. Bull. 1984, 32, 284-290.
(2) Itokawa, H.; Takeya, K.; Hitotsuyanagi, Y.; Morita, H. In The
Alkaloids; Cordell, G. A., Ed.; Academic Press: New York, 1997; Vol.
49, pp 301-387.
(9) Hitotsuyanagi, Y.; Kondo, K.; Takeya, K.; Itokawa, H. Tetrahe-
dron Lett. 1994, 35, 2191-2194.
(3) Hitotsuyanagi, Y.; Aihara, T.; Takeya, K. Tetrahedron Lett. 2000,
41, 6127-6130.
(10) Boger, D. L.; Zhou, J .; Winter, B.; Kitos, P. A. Bioorg. Med.
Chem. 1995, 3, 1579-1593.
(4) J olad, S. D.; Hoffmann, J . J .; Torrance, S. J .; Wiedhopf, R. M.;
Cole, J . R.; Arora, S. K.; Bates, R. B.; Gargiulo, R. L.; Kriek, G. R. J .
Am. Chem. Soc. 1977, 99, 8040-8044.
(11) (a) Monahan, M. W.; Amoss, M. S.; Anderson, H. A.; Vale, W.
Biochemistry 1973, 12, 4616-4620. (b) Rich, D. H.; Bhatnagar, P. K.;
J asensky, R. D.; Steele, J . A.; Uchytil, T. F.; Durbin, R. D. Bioorg.
Chem. 1978, 7, 207-214. (c) Mu¨ller, G.; Gurrath, M.; Kessler, H.;
Timpl, R. Angew. Chem., Int. Ed. Engl. 1992, 31, 326-328. (d) Wenger,
R. M.; France, J .; Bovermann, G.; Walliser, L.; Widmer, A.; Widmer,
H. FEBS Lett. 1994, 340, 255-259.
(5) Zalaca´ın, M.; Zaera, E.; Va´zquez, D.; J ime´nez, A. FEBS Lett.
1982, 148, 95-97.
(6) Sirdeshpande, B. V.; Toogood, P. L. Bioorg. Chem. 1995, 23, 460-
470.
10.1021/jo030293p CCC: $27.50 © 2004 American Chemical Society
Published on Web 02/03/2004
J . Org. Chem. 2004, 69, 1481-1486
1481

Hitotsuyanagi et al.
SCHEME 1
Synthesis of such analogues inevitably demands ef-
ficient preparation of a cycloisodityrosine unit of definite
stereochemistry. However, all the methods reported for
preparation of cycloisodityrosines are not satisfactory in
this regard. They involve a multistep sequence of reac-
tions with elaboration of unusual amino acids,7b,c,12-15 and
the reaction conditions usable for the formation of the
diphenyl ether linkage and further transformations are
strictly limited because of the easily epimerizable proper-
ties of the C-terminal residue of cycloisodityrosines. Thus,
preparation of cycloisodityrosines in quantity has been
an obstacle to the synthesis of RA-VII analogues with
modified 18-membered rings.10,16
In the present study, first, we developed an easier and
more efficient means for preparation of the protected
cycloisodityrosines via partial degradation of natural RA-
VII (1) and then by using this unit prepared three
analogues of 1 with modified 18-membered rings, i.e.,
[Gly-1]RA-VII (4), [Gly-2]RA-VII (5), and [Gly-4]RA-VII
(6). Their conformational properties in solution were
studied by NMR analysis, and the relationship between
the conformation and their cytotoxic activity on P-388
cells was discussed.
retrosynthetically in Scheme 1. Cyclization is generally
the limiting step in the synthesis of cyclic peptides. The
yield of cyclized products is known to depend strongly
upon the sites involved in the cyclization. In all three
reported RA-VII syntheses, the macrocyclization is per-
formed between Tyr-6 and D-Ala-1 residues.12a,13a,14c Thus,
in our present synthesis of these analogues, following
their strategy we linked the cycloisodityrosine unit to the
modified tetrapeptide fragments and subsequently per-
formed macrocyclization by linking the Tyr-6 and D-Ala-1
(Gly-1) residues.17
Cycloisod ityr osin e fr om RA-VII. The procedure is
shown in Scheme 2. The most abundant congeners of
bicyclic hexapeptides of this series in R. akane and R.
cordifolia are RA-VII (1) and deoxybouvardin (3), and 3
is easily converted into 1 by O-methylation. Therefore,
the most readily available peptide 1 was used for the
degradation study. To obtain cycloisodityrosines from
peptide 1, the peptide bonds between Ala-4 and Tyr-5
and between Tyr-6 and D-Ala-1 must be selectively
Resu lts a n d Discu ssion
(17) To evaluate the efficacy of the macrocyclization at different
locations, we also attempted the cyclization of peptide I by linking the
Tyr-3 and Ala-4 residues in DMF by using the reagents A-D of Table
1. Reagents A, B, and D did not afford cyclized products, whereas
reagent C [HBTU (4 equiv), HOBt (4 equiv), and DMAP (8 equiv)] gave
1 in 13% yield. By the same method, peptide II afforded 6 in 17% yield.
The results indicated that cyclization at this position is feasible, but
we did not consider that this strategy is more advantageous over the
scheme described in the text.
Syn th etic P la n for th e Ma cr ocycle. The synthetic
approach to the present analogues 4-6 is illustrated
(12) (a) Inaba, T.; Umezawa, I.; Yuasa, M.; Inoue, T.; Mihashi, S.;
Itokawa, H.; Ogura, K. J . Org. Chem. 1987, 52, 2957-2958. (b) Inoue,
T.; Inaba, T.; Umezawa, I.; Yuasa, M.; Itokawa, H.; Ogura, K.;
Komatsu, K.; Hara H.; Hoshino, O. Chem. Pharm. Bull. 1995, 43,
1325-1335. (c) Inoue, T.; Sasaki, T.; Takayanagi, H.; Harigaya, Y.;
Hoshino, O.; Hara, H.; Inaba, T. J . Org. Chem. 1996, 61, 3936-3937.
(13) (a) Boger, D. L.; Yohannes, D. J . Am. Chem. Soc. 1991, 113,
1427-1429. (b) Boger, D. L.; Patane, M. A.; Zhou, J . J . Am. Chem.
Soc. 1994, 116, 8544-8556. (c) Boger, D. L.; Zhou, J .; Borzilleri, R.
M.; Nukui, S. Bioorg. Med. Chem. Lett. 1996, 6, 1089-1092. (d) Boger
D. L.; Zhou, J . J . Org. Chem. 1996, 61, 3938-3939. (e) Boger, D. L.;
Zhou, J .; Borzilleri, R. M.; Nukui, S.; Castle, S. L. J . Org. Chem. 1997,
62, 2054-2069. (f) Krenitsky, P. J .; Boger, D. L. Tetrahedron Lett. 2002,
43, 407-410.
(14) (a) Beugelmans, R.; Bigot, A.; Bois-Choussy, M.; Zhu, J . J . Org.
Chem. 1996, 61, 771-774. (b) Bigot, A.; Beugelmans, R.; Zhu, J .
Tetrahedron 1997, 53, 10753-10764. (c) Bigot, A.; Dau, M. E. T. H.;
Zhu, J . J . Org. Chem. 1999, 64, 6283-6296.
(15) Poupardin, O.; Ferreira, F.; Genet, J . P.; Greck, C. Tetrahedron
Lett. 2001, 42, 1523-1526.
(16) Boger, D. L.; Zhou, J . J . Am. Chem. Soc. 1995, 117, 7364-7378.
1482 J . Org. Chem., Vol. 69, No. 5, 2004

Synthesis of [Gly-1]RA-VII, [Gly-2]RA-VII, and [Gly-4]RA-VII
SCHEME 2a
a
Reagents and conditions: (a) see ref 18; (b) 7, dioxane, 91%; (c) MeI, K2CO3, acetone; (d) 6 M HCl, MeCN; neutralized with 1 M
K2CO3; phenyl isothiocyanate; 6 M HCl, MeCN; neutralized with 1 M K2CO3; Boc2O, 78% from 10; (e) LiOH, H2O2, THF-H2O; (f)
(trimethylsilyl)diazomethane, MeCN-MeOH, 97% from 14; (g) benzyl alcohol, DEAD, Ph3P, THF, 90% from 14.
SCHEME 3a
a
Reagents and conditions: (a) H-Ala-OMe‚HCl, EDC, HOBt, Et3N, CH2Cl2, 71%; (b) H-Gly-OMe‚HCl, EDC, HOBt, Et3N, CH2Cl2,
79%; (c) CH2N2, Et2O-MeOH, 100%; (d) H2, Pd/C, HCl, MeOH; Boc-Ala-OH, PyBOP, i-Pr2NEt, CH2Cl2, 88% for 20 and 73% for 28; (e) H2,
Pd/C, HCl, MeOH; Boc-Gly-OH, PyBOP, i-Pr2NEt, CH2Cl2, 96%; (f) TFA; Boc-Gly-OH, EDC, HOBt, CH2Cl2, 95%; (g) TFA; Boc-D-Ala-OH,
EDC, HOBt, CH2Cl2, 84% for 24 and 79% for 29; (h) LiOH, THF-MeOH-H2O, 99% for 22, 98% for 25 and 93% for 30.
cleaved. Previously, we reported that, when treated with
2,4-bis(methylthio)-1,3-dithia-2,4-diphosphetane-2,4-dis-
ulfide (Davy-Reagent-Methyl) (7) at room temperature,
peptide 1 gives bis(thioamide) 8 in 38% yield, with
accompanying thioamides and other bis(thioamides).8
Bis(thioamide) 8 should be a suitable substrate for the
present degradation. However, even when the reaction
conditions were modified, the yield of 8 was not success-
fully increased beyond 54%. On the other hand, when
[N-methyl-Ala-2]RA-VII (9), which was readily prepared
from peptide 1 in 97% yield,18 was thionated by using
the same procedure, bis(thioamide) 10 possessing thio-
amide bonds at the same positions as in 8 was produced
in 91% yield. This marked enhancement of the yield of
the bis(thioamide) may be attributed to the fact that the
introduction of a methyl group into the Ala-2 amide
nitrogen makes the D-Ala-1/Ala-2 peptide bond a less
reactive tertiary amide and thus stabilizes the structure
of 9. As a result, the two reactive secondary amide
carbonyls at Tyr-3 and Tyr-6 were selectively thionated
to afford 10 in a high yield. Bis(thioamide) 10 was then
converted into bis(thioimidate) 11 by using iodomethane
and potassium carbonate, and the subsequent treatment
of 11 with 6 M HCl in acetonitrile caused the cleavage
(18) Hitotsuyanagi, Y.; Suzuki, J .; Takeya, K. Itokawa, H. Bioorg.
Med. Chem. Lett. 1994, 4, 1633-1636.
J . Org. Chem, Vol. 69, No. 5, 2004 1483

Hitotsuyanagi et al.
SCHEME 4a
of the two thioimidate linkages to produce a mixture of
two tripeptide segments 12 and 13. Without separation,
the mixture of 12 and 13 was subjected to Edman
degradation with phenyl isothiocyanate. Treatment of the
resultant crude thiourea with 6 M HCl and successive
neutralization and N-protection using di-tert-butyl di-
carbonate afforded cycloisodityrosine thioester 14 in 78%
yield from bis(thioamide) 10. In regard to the stereo-
chemistry at the R-carbon of the C-terminal amino acid
residue, the natural (9S,12S)-cycloisodityrosine esters are
known to epimerize readily to become thermodynamically
more stable (9R,12S)-cycloisodityrosine esters.13c-e,14b
However, in thioester 14, both of the stereocenters at 9
and 12 were confirmed to be S by X-ray crystallography.19
When thioester 14 was hydrolyzed with lithium hydrox-
ide in THF-H2O, a considerable amount of epimerized
acid contaminated the product to demonstrate facile
epimerization of the compound. Accordingly, hydrolysis
of 14 was performed with more nucleophilic and less basic
lithium hydroperoxide prepared in situ. Subsequent
treatment with (trimethylsilyl)diazomethane or benzy-
lation using Mitsunobu conditions20 afforded methyl ester
15 (97%) or benzyl ester 16 (90%), respectively, in an
isomerically pure form. The spectroscopic data of 15 were
in good agreement with those of the synthetic 15
reported.13d,14b The series of chemical conversions de-
scribed in Scheme 2 proceeded efficiently, and the overall
yields of cycloisodityrosines 15 and 16 from peptide 1
were 67% and 62%, respectively. In the subsequent
reactions for the analogue syntheses, benzyl ester 16 was
used.
Syn th esis of Tetr a p ep tid es. The tetrapeptide frag-
ments of 4, 5, and 6, each consisting of residues 1-4, were
synthesized by the stepwise extension of Cbz-N-methyl-
L-tyrosine (17)12a,b as shown in Scheme 3. For the
preparation of the fragment for 4, 17 was first coupled
to L-alanine methyl ester and subsequently O-methylated
with diazomethane to afford dipeptide 19 (Scheme 3). The
Cbz protecting group was removed to give a less reactive
dipeptide ester having an N-methyl group at the N-
terminus, which was then coupled to Boc-L-alanine by
using 1H-benzotriazol-1-yloxytripyrrolidinophosphonium
hexafluorophosphate (PyBOP) to afford tripeptide 20 in
good yield. Further elongation of the peptide chain of 20
at the N-terminus with Boc-glycine gave tetrapeptide 21,
which on treatment with lithium hydroxide provided acid
22. The tetrapeptide fragments 25 and 30 for the
synthesis of analogues 5 and 6, respectively, were
prepared analogously following the processes illustrated
in Scheme 3 via tripeptides 23 and 28, respectively.
a
Reagents and conditions: (a) TFA; 22, EDC, HOObt, THF,
97% for 31; (b) TFA; 25, EDC, HOObt, THF, 79% for 32; (c) TFA;
30, EDC, HOAt, THF, 92% for 33; (d) TFA, 95% for 34, 98% for
35 and 94% for 36; (e) H2, Pd/C, EtOH.
of isomeric hexapeptides were produced, which were
probably the epimers at the Ala-4 residue. This epimer-
ization during the coupling process was effectively sup-
pressed by using 3-hydroxy-4-oxo-3,4-dihydro-1,2,3-
benzotriazine (HOObt) as the additive to afford desired
hexapeptides 31 and 32 in good yields.
Cycliza t ion of H exa p ep t id es. The macrocycliza-
tion of the cycloisodityrosine-containing linear peptides
37, 38, and 39, obtained by debenzylation of 34, 35, and
36, respectively, was performed in the presence of a
coupling reagent (Scheme 4). The effects of the reagents
diphenylphosphoryl azide (DPPA), pentafluorophenyl di-
phenylphosphinate (FDPP), O-benzotriazol-1-yl-N,N,N′,N′-
tetramethyluronium hexafluorophosphate (HBTU), and
EDC on the macrocyclization were examined. All the
Cou p lin g Rea ction of Tetr a p ep tid e to Cycloisod i-
tyr osin e Un it. The coupling reaction and subsequent
macrocyclization are shown in Scheme 4. Tetrapeptides
22, 25, and 30, obtained as described above, were then
subjected to the coupling reaction with N-deprotected
cycloisodityrosine derived from 16 by using 1-(3-dimethyl-
aminopropyl)-3-ethylcarbodiimide hydrochloride (EDC)
and 1-hydroxybenzotriazole (HOBt) or 1-hydroxy-7-aza-
benzotriazole (HOAt) (Scheme 4). When the tetrapeptides
employed were 22 and 25, a significant amount (28-34%)
(19) Hitotsuyanagi, Y.; Hasuda, T.; Matsumoto, Y.; Yamaguchi, K.;
Itokawa, H.; Takeya, K. Chem. Commun. 2000, 1633-1634.
(20) Mitsunobu, O. Synthesis 1981, 1-28.
1484 J . Org. Chem., Vol. 69, No. 5, 2004

Synthesis of [Gly-1]RA-VII, [Gly-2]RA-VII, and [Gly-4]RA-VII
TABLE 1. Cycliza tion of 37-39a
entry substrate reagentb
solvent
product
yieldc (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
37
37
37
37
37
38
38
38
38
38
39
39
39
39
39
A
B
C
D
D
A
B
C
D
D
A
B
C
D
D
DMF
DMF
DMF
DMF
dioxane
DMF
DMF
DMF
DMF
dioxane
DMF
DMF
DMF
DMF
4
4
4
4
4
5
5
5
5
5
6
6
6
6
6
0
32
0
29
18
0
0
0
10
0
0
18
12
54
0
dioxane
a
All reactions were carried out for 4 days at room temperature
b
at a concentration of 0.0013 M. Conditions: (A) DPPA (2 equiv),
Et3N (10 equiv); (B) FDPP (3 equiv), i-Pr2NEt (3 equiv); (C) HBTU
(4 equiv), HOBt (4 equiv), DMAP (8 equiv); (D) EDC (8 equiv),
HOObt (8 equiv). c Two-step yield from benzyl ester 34, 35, or 36.
FIGURE 1. Conformer population and key sequential NOESY/
ROESY correlations commonly observed in 1, 4, and 6 in
solution.
reactions were performed at room temperature for 4 days,
the concentration of peptides being 0.0013 M. The results
are summarized in Table 1. The results showed that the
macrocyclization was to be governed by the coupling
reagents and the solvent used. EDC generally favored
the macrocyclization of this series, and when used in
DMF it gave a yield of as high as 54% to 6 (entry 14),
29% to 4 (entry 4), and 10% to 5 (entry 9). The effect of
solvent on the reaction may not be ignored. For macro-
cyclization of 38 and 39, EDC in dioxane gave no positive
result, but for the reaction of 37 the combination pro-
duced the desired product (entries 5, 10, and 15). FDPP
in DMF also worked as a reagent for macrocyclization
(entries 2 and 12). The most commonly used coupling
reagent for macrocyclization of linear peptides, DPPA
in DMF, which was used in the synthesis of RA-VII
(1),13a,14c bouvardin (2),13b and some related analogues,7b,10,16
apparently did not work on peptides 37, 38, and 39
(entries 1, 6, and 11). HBTU appeared to be unsuitable
for the cyclization of these peptides either; only 39
cyclized to 6 in 12% yield (entries 3, 8, and 13).
X-ray crystallography, NOESY experiments, and com-
putational methods.21
1
The H NMR spectrum of 4 in CDCl3 was very similar
to that of peptide 1. It demonstrated the presence of two
conformers in a ratio of 89:11 at 300 K, and their
structures were analyzed by the ROESY spectra. The
major conformer showed ROESY correlations between
Ala-2 HR/Tyr-3 NMe, Tyr-3 NMe/Tyr-3 HR, Ala-4 HR/Tyr-5
NMe, Ala-4 Me/Tyr-5 NMe, and Tyr-5 HR/Tyr-6 HR,
whereas the minor one between Ala-2 HR/Tyr-3 HR, Ala-4
HR/Tyr-5 NMe, Ala-4 Me/Tyr-5 NMe, and Tyr-5 HR/Tyr-6
HR (Figure 1). These correlations demonstrated that the
major conformer of analogue 4 possessed t-t-t-t-c-t amide
configurations as the major conformer in peptide 1 and
that the minor one in 4 had t-c-t-t-c-t configurations as
the minor conformer in 1. In addition to the above
ROESY correlations, the very similar chemical shifts of
the proton signals in the spectrum of 1 and that of 4
indicated that the structures of these two conformers in
the two peptides were almost identical.
Solu tion Con for m a tion of An a logu es 4-6. The
biological activity is considered to be related to the
conformation of molecules in solution. Therefore, we
studied the conformer structures of those peptide ana-
logues in solution by NMR spectroscopy to examine the
effect of each alanyl methyl group of the peptide molecule.
The results were compared with those of the parent
active peptide 1, whose solution form is well studied. The
most distinctive features which characterize the confor-
mation of cyclic peptides are their backbone amide
configurations. Peptide 1 possesses three modified N-
methyltyrosines at residues 3, 5, and 6, and the am-
ide bonds between Ala-2/Tyr-3, Ala-4/Tyr-5, and Tyr-5/
Tyr-6 can be either cis or trans. Peptide 1 is known to
adopt two stable conformations in CDCl3 and CD3OD
solutions in ratios of 89:11 and 84:16, respectively. The
major conformer possesses trans-trans-trans-trans-cis-
trans (t-t-t-t-c-t) amide configurations between D-Ala-1/
Ala-2, Ala-2/Tyr-3, Tyr-3/Ala-4, Ala-4/Tyr-5, Tyr-5/Tyr-
6, and Tyr-6/D-Ala-1, respectively, whereas the minor
conformer t-c-t-t-c-t configurations, as characterized by
In the case of analogue 5, the 1H NMR signals in CDCl3
and DMSO-d6 at 300 K and at higher temperatures were
too broad to allow reliable conformational analysis. At a
1
lower temperature (283 K), the H NMR spectrum was
very complex, suggesting that it was a mixture of three
or more conformers. Thus, the conformational analysis
of 5 by NMR studies was not possible in the present
experiment, and we concluded that analogue 5 existed
in a mixture of three or more unstable conformers in
solution.
NMR spectra of analogue 6 were studied in CD3OD
because of its poor solubility in CDCl3. In this solvent,
analogue 6 was noted to consist of two conformers in a
ratio of 72:28 at 300 K. The NOESY correlations dem-
onstrated that the major conformer of 6 had t-c-t-t-c-t
amide configurations and was identical to the minor
conformer of 4 or 1 and that the minor conformer of 6
having t-t-t-t-c-t amide configurations was identical to
(21) Morita, H.; Kondo, K.; Hitotsuyanagi, Y.; Takeya, K.; Itokawa,
H.; Tomioka, N.; Itai, A.; Iitaka, Y. Tetrahedron 1991, 47, 2757-2772.
J . Org. Chem, Vol. 69, No. 5, 2004 1485

Hitotsuyanagi et al.
TABLE 2. Cytotoxicity of RA-VII (1) a n d Its An a logu es
of this active conformer. A difference is noted between
the cytotoxic activities of 1 and 4, though their bioactive
conformer structures are about the same. No further
investigation was performed as to the cause of difference
in cytotoxic activity, but the alanyl methyl group at 1
may play some role in the interaction with 80S ribosomes.
4-6 a ga in st P -388 Leu k em ia Cells
compd
IC50 (µg/mL)
RA-VII (1)
0.0038
0.014
0.032
0.16
[Gly-1]RA-VII (4)
[Gly-2]RA-VII (5)
[Gly-4]RA-VII (6)
Con clu sion
We prepared three analogues 4-6 of RA-VII (1), in
which one of the three alanine residues was replaced by
glycine, and demonstrated the possible role of the methyl
groups in the three alanine residues upon their confor-
mational features and accordingly upon the cytotoxic
activity. The methyl group of D-Ala-1 residue apparently
affects neither the backbone structure nor the population
of the conformers in solution, whereas the methyl group
of Ala-2 residue seemed to be essential for maintaining
the stable conformation. The methyl group of Ala-4
residue does not affect the backbone structures of the
conformers, but affects the ratio of those conformers.
Thus, the presence of the methyl groups of both Ala-2
and Ala-4 residues is necessary for producing the con-
formational property of RA-VII (1) responsible for the
activity.
the major conformer of 4 or 1 (Figure 1). The similarity
of the chemical shifts of proton signals for each conformer
positively supported that the major conformer in peptide
6 was almost identical to the minor conformer of 1 and
the minor conformer in 6 was the major conformer of 1.
Cytotoxicity of An a logu es 4-6. Cytotoxicity of
analogues 4-6 was evaluated by using P-388 murine
leukemia cells, and the results are shown in Table 2. RA-
VII (1) was also assayed for comparison (Table 2). Of
those three glycine analogues, analogue 4 was the most
active, which may be due to its having the conformational
features very similar to those of 1. Both of the two
conformers of 6 were found in 1 and 4, yet it was 11 times
less cytotoxic than 4. It may support the suggestion that
the major conformer of 1 and 4 or the minor one of 6
having t-t-t-t-c-t amide configurations is the bioactive
conformer of this peptide series.22 Thus, the reduced
activity of analogue 6 may be due to the low population
Su p p or tin g In for m a tion Ava ila ble: Experimental de-
tails, 1H NMR spectra of 4-6, 10, 14-16, and 18-36, and
X-ray data of 14. This material is available free of charge via
the Internet at http://pubs.acs.org.
(22) Itokawa, H.; Kondo, K.; Hitotsuyanagi, Y.; Isomura, M.; Takeya,
K. Chem. Pharm. Bull. 1993, 41, 1402-1410.
J O030293P
1486 J . Org. Chem., Vol. 69, No. 5, 2004