Structure-activity relationships of side-chain modified didemnins

Article abstract of DOI:10.1016/S0960-894X(00)00638-7

The synthesis and antitumor activity of a novel didemnin B analogue containing a ψ[CH₂NH] amide bond surrogate between N-Me-D-Leu⁷ and Pro⁸ are reported. The analogue shows activity (GI₅₀ = 4 nM) comparable to that of didemnin B (GI₅₀ = 13 nM) in the NCI-60 tumor cell screen. This result, along with new data from previously reported synthetic didemnin analogues, is discussed within the context of the side-chain SAR for didemnins.

Full text of DOI:10.1016/S0960-894X(00)00638-7

Bioorganic & Medicinal Chemistry Letters 11 (2001) 231±234
Structure±Activity Relationships of Side-Chain
Modi®ed Didemnins
Xiaobin Ding, Matthew D. Vera, Bo Liang, Yuming Zhao, Michael S. Leonard
and Madeleine M. Joullie*
Department of Chemistry, University of Pennsylvania, Philadelphia, PA 19104-6323, USA
Received 28 August 2000; accepted 7 November 2000
AbstractÐThe synthesis and antitumor activity of a novel didemnin B analogue containing a É[CH2NH] amide bond surrogate
between N-Me-d-Leu7 and Pro8 are reported. The analogue shows activity (GI50=4 nM) comparable to that of didemnin B (GI50=
13 nM) in the NCI-60 tumor cell screen. This result, along with new data from previously reported synthetic didemnin analogues, is
discussed within the context of the side-chain SAR for didemnins. # 2001 Elsevier Science Ltd. All rights reserved.
The didemnin class of cyclic depsipeptides has attracted
considerable interest because of its wide spectrum of
biological activities. Didemnin B (Fig. 1), the initial lead
congener, was isolated from the didemnidae family of
tunicates.1,2 It has shown potent antitumor, antiviral,
and immunosuppressive activities, as well as the ability to
inhibit in vitro protein biosynthesis3 and to induce rapid
apoptosis.4 The antitumor action of these compounds has
attracted substantial attention, with didemnin B entering
Phase I and II clinical trials against cancer in the late
1980s.5,6 Although didemnin B showed poor ecacy and
high toxicity in the clinic, the molecular basis of its multi-
faceted activities is incompletely understood and still under
investigation. In an attempt to improve the biological
pro®le, we investigated the use of amide bond surrogates.
Although the reduced peptide bond is only one of many
possible alternatives for the amide bond linkage, it has
the advantage of being easily incorporated by reductive
alkylation.
As a continuation of our investigation of the structure±
activity relationships (SARs) of side-chain modi®ed di-
demnin B analogues (1),14 we have prepared a compound
(2) containing a É[CH2NH] surrogate for the amide bond
between the N-Me-d-leucine7 and l-proline8 residues.
Analogue 2 contains a lactyl residue at the terminal posi-
tion. Since, aplidine ([pyruvyl9]didemnin B), isolated from
the polyclinidae family of tunicates,15 is reported to be a
highly potent cytotoxic agent,16 we attempted to synthesize
the É[CH2NH] surrogate analogue with an oxidized
side chain (3). Unfortunately, compound 3 was not
stable over extended periods of time.
The amide bond of a peptide is a polar and enzymatically
fragile moiety. These attributes cause poor rates of trans-
fer across cell membranes and intestinal mucosa and result
in rapid degradation by various proteases.7 The use of
amide surrogates is one approach to modulate the bio-
activity of peptides. An appropriate amide surrogate may
make a peptide analogue more resistant toward enzymatic
degradation, more stable, and even orally active. The
É[CH2NH] surrogate is one of the simplest isosteres of the
amide bond. It has been used in the design of potent ACE
inhibitors8 and in the generation of antagonists against
bombesin,9 gastrin,10 substance P,11 and secretin12 recep-
tors. The reduced moiety É[CH2NH] acts as a non-
hydrolyzable analogue of the tetrahedral transition state
formed during hydrolysis of the peptide bond.13
The tolerance for changes in the residue beyond the N-
Me-d-Leu7 is precedented but limited. The introduction
of unsaturation at positions 3 and 4 of the proline ring
([3,4-dehydroPro8]didemnin B) and the elimination of
the lactyl group in the didemnin B side chain ([Pro8]di-
demnin A) were shown previously to produce two very
potent analogues.14 For the purposes of comparison to
these previously tested analogues, we prepared the con-
gener (5) that is truncated with 3,4-dehydroproline,8
thus containing both modi®cations. To further probe
the tolerance to side-chain modi®cations a ¯uorescent
didemnin was prepared in order to study its cellular
localization by ¯uorescence microscopy.17
The aforementioned analogues were prepared by coup-
ling the modi®ed side chains to the macrocycle. The
*Corresponding author. Tel.: +1-215-898-3158; fax: +1-215-898-5129;
e-mail: mjoullie@sas.upenn.edu
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00638-7

232
X. Ding et al. / Bioorg. Med. Chem. Lett. 11 (2001) 231±234
altered to employ a benzyl ester, which then could be
removed under non-aqueous conditions. The Boc-pro-
tected d-leucine benzyl ester (8) was methylated to yield 9.
The Boc group was then removed with tri¯uoroacetic
acid to a€ord 10, which was subsequently condensed
with protected prolinal 13 by reductive amination. The
secondary amine was liberated by treatment with HCl in
dioxane to produce 15. Coupling of the amine hydro-
chloride salt with the appropriate acid (lactic or pyruvic
acid) was accomplished with BOP and NMM in methyl-
ene chloride. Hydrogenolysis of the benzyl esters (16a,b)
gave the desired unprotected side chains, and the ®nal
coupling to the didemnin macrocycle salt was performed
in methylene chloride using HATU and DIEA.
Figure 1. Didemnin B.
synthesis of the macrocycle was based on the approach
previously developed in our group,14,18 with some mod-
i®cations including cyclization with HATU.19
The 3,4-dehydroproline side chain was synthesized
according to a previously reported procedure14 and was
attached to the macrocycle as described above. The syn-
thesis of the ¯uorescent analogue 7 has been published.17
All analogues were submitted to the NCI-60 tumor cell
screen.20 The results are collected in Table 1.
The synthesis of the two É[CH2NH] surrogate side chains
is described in Scheme 1. Although the Z-d-leucine
methyl ester was initially used, its saponi®cation at a
later stage proved dicult and required large amounts
of base. As the resultant acid was very hydrophilic, its
puri®cation was dicult. Therefore, the scheme was
Results from several groups have demonstrated that in
vitro antitumor activity is closely related to the structure
of the side-chain peptide attached to the amino group of
Thr.6 As shown in Table 1, replacement of the amide bond
with the É[CH2NH] surrogate led to a compound (2)
having antitumor activity similar to that of didemnin B.
Antitumor activity is quite intolerant of modi®cations to
the absolute con®guration21 of the N-Me-d-Leu.7 How-
ever, N-acylation of this residue, whether by Pro8Lac9 or
short (C2±C6) aliphatic residues, enhances cytotoxicity.16
This e€ect has been attributed to improved cell perme-
ability resulting from either increased lipophilicity or the
elimination of basic nitrogen.16 Since the É[CH2NH] sur-
rogate (2)24 exhibited activity comparable to that of
didemnin B, the N-acylation of N-Me-d-Leu7 is appar-
ently not a strict requirement for intact biological activity,
and a basic nitrogen is also tolerated.
Compounds 5 and 6 address the importance of the Lac9
residue. The enhanced cytotoxicity of aplidine, having a
ketone rather than a secondary hydroxyl at position 9,
seems to imply a signi®cant role for residue 9 as a deter-
minant of cytotoxicity. However, our results demonstrate
that the Lac9 residue can be deleted without adverse
e€ects on cytotoxicity. In fact, compound 6 exhibits
modestly enhanced growth inhibition and cytotoxic
activities.14 Rinehart has independently demonstrated
that several didemnins having O-acyl substituents on the
Lac9 hydroxyl group also show enhanced cytotoxicity.16
These apparently con¯icting results could be reconciled
with the observation that analogues 5 and 6, as well as
aplidine and the O-acylated analogues, contain no
hydrogen bond donor at position 9. Kessler has repor-
ted a solution NMR study of didemnin B showing evi-
dence of a hydrogen bond between the Lac9 hydroxyl
group and the carbonyl of the macrocyclic Me2Tyr5
residue.22 It is possible to speculate that the absence of
this hydrogen bond may alter the conformation of the
molecule in such a way as to enhance antitumor activity.
The possibility that potency enhancements result from
Scheme 1. (a) MeI, NaHMDS, CH2Cl2 (78%); (b) TFA/CH2Cl2
(90%); (c) MeOH, SOCl2 (95%); (d) Boc2O, Et3N, CH2Cl2 (85%); (e)
.
NaBH4, LiCl, THF/EtOH (85%); (f) SO3 pyridine complex, DMSO,
Et3N, CH2Cl2 (81%); (g) Na(AcO)3BH, AcOH, CH2Cl2 (88%); (h)
HCl in dioxane (90%); (i) lactic acid or pyruvic acid, BOP, NMM,
CH2Cl2 (70%); (j) H2, Pd/C; (99%); (k) didemnin macrocycle salt,
DIEA, HATU, CH2Cl2 (72%).

X. Ding et al. / Bioorg. Med. Chem. Lett. 11 (2001) 231±234
Table 1. NCI-60 mean data from NCI-60 tumor cell screen
233
Reference
1,2
Compound
Side-chain structure
GI50 (nM)
TGI (nM)
LC50 (mM)
1
13
66
3.8
2
3
4
5
6
4
76
3.0
24
2.8
3
20
39
12
0.083
14
14
14
7
0.62
0.21
7
11
20.5
7.2
17
increased lipophilicity cannot be ruled out. However,
this hypothesis would be at odds with a result from
Schmidt, showing that the more polar hydroxylactic
acid at position 9 results in equal or enhanced growth
inhibition potency.23
the Pro8Lac9 is still unclear. The overall trends seem to
indicate a quite stringent requirement for the intact N-
Me-d-Leu7 residue, preferably bearing an N-acyl group.
Additionally, signi®cant but ill-de®ned attenuation of
cytotoxic activity results from modi®cations to the rest
of the side chain. Nevertheless, there is notable toler-
ance to modi®cations at the side-chain terminus, as
demonstrated by the successful replacement of Pro8Lac9
with the Gly-dimethylaminocoumarin moiety.
Compound 4 illustrates that installation of an unsatu-
ration in the ring of Pro8 leads to an incremental
increase in growth inhibition potency and a larger
increase in cytotoxicity. Comparison with 5 implies that
the e€ects of the unsaturated proline and the deletion of
Lac9 do not appear to be additive.
Acknowledgements
The tolerance of side-chain modi®cations led us to
attempt the incorporation of a glycine-linked coumarin
¯uorophore. The synthesis of this analogue (7), envi-
sioned as a molecular probe to study cellular and tissue
localization by ¯uorescence microscopy, was recently
reported.17 The data in Table 1 now establish 7 as
essentially equipotent with didemnin B in the NCI-60
tumor cell screen. This result suggests that 7 could serve
as a useful tool to study the fate of didemnins in cells.
Work in this area is currently underway.
We thank NIH (CA-40081) and NSF (CHE 99-01449)
for ®nancial support.
References and Notes
1. Rinehart, K. L.; Gloer, J. B., Jr.; Hughes, R. G., Jr.; Renis,
H. E.; McGovren, J. P.; Swynenberg, E. B.; Stringfellow,
D. A.; Kuentzel, S. L.; Li, L. H. Science 1981, 212, 933.
2. Rinehart, K. L., Jr.; Gloer, J. B.; Cook, J. C., Jr.; Mizsak,
S. A.; Scahill, T. A. J. Am. Chem. Soc. 1981, 103, 1857.
3. SirDeshpande, B. V.; Toogood, P. L. Biochemistry 1995,
34, 9177.
In summation, the acylation of N-Me-d-Leu7 is not
required for activity, and the presence of basic nitrogen
in the side chain is tolerated in the case of compounds 2,
5, and 6. However, the structure±activity relationship of
4. Grubb, D. R.; Wolvetang, E. J.; Lawen, A. Biochem. Bio-
phys. Res. Commun. 1995, 215, 1130.

234
X. Ding et al. / Bioorg. Med. Chem. Lett. 11 (2001) 231±234
5. Maroun, J. A.; Stewart, D.; Verma, S.; Eisenhauer, E.
Invest. New Drugs 1998, 16, 51.
6. Williamson, S. K.; Wolf, M. K.; Eisenberger, M. A.;
O'Rourke, M.; Brannon, W.; Crawford, E. D. Invest. New
Drugs 1995, 13, 167.
19. Ding, X. Total Synthesis of Didemnin B Analogues, Uni-
versity of Pennsylvania, Philadelphia, PA, 2000.
20. Boyd, M. R.; Paull, K. D. Drug Dev. Res. 1995, 34, 91.
21. Jouin, P.; Poncet, J.; Dufour, M.-N.; Aumelas, A.; Pan-
taloni, A. J. Med. Chem. 1991, 34, 486.
7. Spatola, A. F.; Wen, J. J. J. Peptide Res. 1997, 49, 3.
8. Szele, I.; Zollinger, H. J. Am. Chem. Soc. 1978, 100, 2811.
9. Coy, D. H.; Heinz-Erian, P.; Jiang, N. Y.; Sasaki, Y.;
Taylor, J.; Moreau, J. P.; Wolfrey, W. T.; Gardner, J. D.;
Jensen, R. T. J. Biol. Chem. 1988, 263, 5056.
10. Martinez, J.; Bali, J.-P.; Rodriguez, M.; Castro, B.;
Magous, R.; Laur, J.; Lignon, M. F. J. Med. Chem. 1985, 28,
1874.
22. Kessler, H.; Will, M.; Antel, J.; Beck, H.; Sheldrick, G. M.
Helv. Chim. Acta 1989, 72, 530.
23. Schmidt, U.; Griesser, H.; Haas, G.; Kroner, M.; Riedl,
B.; Schumacher, A.; Sutoris, F.; Haupt, A.; Emling, F. J.
Peptide Res. 1999, 54, 146.
24. Compound 2: Rf 0.40 (10% MeOH/CH2Cl2); [a]2D5 +81.2
1
(c 0.15, CHCl3); H NMR (500 MHz, CDCl3) d 0.85±0.97 (m,
24H), 1.17±1.27 (m, 2H), 1.28±1.39 (s, 3H), 1.40 (d, 3H), 1.43
(d, 3H), 1.46±1.50 (m, 1H), 1.61 (t, 1H), 1.68±1.73 (m, 1H),
1.74±1.79 (m, 1H), 1.82±1.88 (m, 2H), 2.02±2.08 (m, 1H) and
2.11±2.17 (m, 2H), 2.34±2.37 (m, 1H), 2.57 (s, 3H), 2.38 (d, 2H),
3.18 and 3.39 (dd, 1H), 2.83 (s, 3H), 3.37±3.57 (m, 2H), 3.60 (d,
1H), 3.70±3.73 (m, 1H), 3.80 (s, 3H), 4.05±4.14 (m, 3H) and 4.36±
4.48 (m, 2H), 4.27 (q, 1H), 4.56 (dd, 1H), 4.65 (dd, 1H), 4.81 (t,
1H), 5.19 (d, 1H), 5.35 (dd, 1H), 5.42 (dd, 1H), 5.61 (dd, 1H),
5.73±5.75 (m, 1H), 6.11 (dd, 1H), 6.85 (d, 2H), 7.08 (d, 2H), 7.20
(d, 1H), 7.54 (d, 1H), 7.78 (d, 1H); 13C NMR (125 MHz, CDCl3)
d 11.7, 14.7, 15.2, 16.3, 16.9, 18.6, 20.1, 20.9, 23.3, 23.7, 24.8,
27.2, 24.9, 25.0, 33.9, 27.9, 31.2, 31.3, 33.9, 36.2, 38.7, 38.8, 41.3,
46.9, 49.46, 49.52, 53.0, 67.9, 55.3, 55.4, 55.5, 57.2, 57.6, 63.9,
65.9, 66.4, 66.5, 70.4, 81.5, 114.1, 124.0, 129.1, 130.0, 130.3,
158.6, 168.6, 169.3, 170.5, 170.6, 171.3, 171.5, 172.4, 173.9,
204.9; IR (neat) 3331 (br), 2956 (s), 2871 (m), 1732 (s), 1638.3
(br, overlap), 1543 (m), 1513 (s), 1448 (m), 1379 (m), 1247 (w),
1167 (w) cm 1; HRMS m/z calcd for C57H91N7O14 (M+H)
1098.6702, found 1098.6726.
11. Quian, J. M.; Coy, D. H.; Jiang, N. Y.; Gardner, J. D.;
Jensen, R. T. J. Biol. Chem. 1989, 264, 1667.
12. Ha€ar, B. M.; Hocart, S. J.; Coy, D. H.; Mantey, S.;
Chiang, V. H. C.; Jensen, R. T. J. Biol. Chem. 1991, 266, 316.
13. Szelke, M.; Leckie, B.; Hallett, A.; Jones, D. M.; Sueiras,
J.; Atrash, B.; Lever, A. F. Nature 1982, 299, 555.
14. Mayer, S. C.; Ramanjulu, J.; Vera, M. D.; P®zenmayer, A.;
Joullie, M. M. J. Org. Chem. 1994, 59, 5192. Mayer, S. C. PhD
Thesis, University of Pennsylvania, Philadelphia, PA, 1994.
15. Rinehart, K. L. Med. Res. Rev. 2000, 20, 1.
16. Sakai, R.; Rinehart, K. L.; Kishore, V.; Kundu, B.; Fair-
cloth, G.; Gloer, J. B.; Carney, J. R.; Namikoshi, M.; Sun, F.;
Hughes, J. R. G.; Gravalos, G.; de Quesada, T. G.; Wilson,
G. R.; Heid, R. M. J. Med. Chem. 1996, 39, 2819.
17. Portonovo, P.; Ding, X.; Leonard, M. S.; Joullie, M. M.
Tetrahedron 2000, 56, 3687.
18. Li, W.-R.; Ewing, W. R.; Harris, B. D.; Joullie, M. M. J.
Am. Chem. Soc. 1990, 112, 7659.