Versatile synthesis of new cytotoxic agents structurally related to hemiasterlins

Article abstract of DOI:10.1016/j.bmcl.2010.03.098

A representative series of structural analogues of the antimitotic tripeptides hemiasterlins have been synthesized. The key-step of this synthetic strategy consists of an Ag₂O-promoted nucleophilic substitution on a common precursor, a chiral non-racemic 2-bromoacyl derivative. Simple variation of nucleophile substituents allows a rapid and stereocontrolled development of new series of derivatives. Some reported compounds showed potent biological activity as growth inhibitors of cancer cell lines and tubulin polymerization inhibitors.

Full text of DOI:10.1016/j.bmcl.2010.03.098

Bioorganic & Medicinal Chemistry Letters 20 (2010) 3431–3435
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Versatile synthesis of new cytotoxic agents structurally related to hemiasterlins
Daniele Simoni a, Ray M. Lee b, David E. Durrant b, Nai-Wen Chi c, Riccardo Baruchello a, Riccardo Rondanin a,
Cinzia Rullo a, Paolo Marchetti a,
*
a Department of Pharmaceutical Sciences, University of Ferrara, Via Fossato di Mortara 17/19, 44100, Ferrara, Italy
b Lestoni Corp., 800 E. Leigh St., Richmond, VA 23219, USA
c Department of Medicine, Endocrine Division, University of California, San Diego, La Jolla, CA 92093, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A representative series of structural analogues of the antimitotic tripeptides hemiasterlins have been syn-
thesized. The key-step of this synthetic strategy consists of an Ag2O-promoted nucleophilic substitution
on a common precursor, a chiral non-racemic 2-bromoacyl derivative. Simple variation of nucleophile
substituents allows a rapid and stereocontrolled development of new series of derivatives. Some reported
compounds showed potent biological activity as growth inhibitors of cancer cell lines and tubulin poly-
merization inhibitors.
Received 15 February 2010
Revised 26 March 2010
Accepted 27 March 2010
Available online 1 April 2010
Keywords:
Ó 2010 Elsevier Ltd. All rights reserved.
Hemiasterlin
Taltobulin
Ag2O
Anticancer drugs
Hemiasterlins are members of the family of natural tripeptides,
discovered and isolated from marine sponges some years ago,1
Hemiasterlins contain three highly modified amino acids that are
responsible for their stability and in vivo activity (Fig. 1). They
are highly potent in suppression of microtubule depolymerization
presumably by binding to the vinca alkaloid site of tubulin and
cause mitotic arrest and cell death, thus making them very attrac-
tive molecules for new anticancer drugs.2 In addition, a recent syn-
thetic analogue of hemiasterlin, taltobulin (HTI-286, SPA-110),
wherein a phenyl group replaces the 3-substituted indole ring,
showed more potent in vivo cytotoxicity and antimitotic activity
not only than the natural occurring tripeptides, but also in compar-
ison with vincristine and paclitaxel.3 Moreover, taltobulin ap-
peared to be unaffected by resistance from P-glycoprotein drug
transporter, and it has advanced into clinical trials.4
The synthesis of hemiasterlin was first reported in 1997 by
Andersen et al. who also synthesized its potent derivative taltobu-
lin in 2003, achieved by condensation of the three non-natural
aminoacids A, B and C (Fig. 2).5
Among the three components, the B segment or (L)-tert-leucine
is commercially available, whereas the A and C components must
be synthesized from precursors. In particular, the enantioselective
R2
O
O
N
CO2H
N
CO2H
N
H
N
H
NH
O
NH
O
N
R1
R1 = R2 = Me: Hemiasterlin
Taltobulin (HTI-286, SPA-110)
R1 = H, R2= Me: Hemiasterlin A
R1 = R2= H: Hemiasterlin B
R1 = Me, R2 = H: Hemiasterlin C
Figure 1. Natural hemiasterlins and synthetic derivative taltobulin.
* Corresponding author. Tel.: +39 0532 455920; fax: +39 0532 455953.
E-mail address: mcp@unife.it (P. Marchetti).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.03.098

3432
D. Simoni et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3431–3435
We explored the possibility of finding a rapid way to build up a
CO2H
H2N
CO2H
series of novel derivatives for SAR studies by exploiting common
precursors to avoid de novo synthesis of particular fragment A syn-
thons every time.
N
H
CO2H
HN
With this aim, a versatile enantioselective approach to synthe-
size novel classes of hemiasterlins that are structurally modified at
fragment A is proposed here. This is based on a stereoselective sub-
stitution reaction in 2-haloamides promoted by Ag2O that we have
previously developed.8 The same synthetic approach was also used
to introduce different radicals at the N-terminal position of the
tripeptide.9
A
C
B
Figure 2. Synthetic building blocks for the achievement of taltobulin.
synthesis of A is very difficult, because the amine group must be
introduced through a reductive amination with induction of chiral-
ity on C-a exploiting a transient Evans’ oxazolidinone. On the other
hand, SAR studies on compounds related to taltobulin and hemias-
Based on this approach, several novel pseudo-peptides bearing
inverted position at the gem-dimethylbenzyl- and the methyl-
groups on
a-carbon and nitrogen with respect to N-terminus
terlins suggest that their activity depends strongly on the presence
aminoacid of HTI-286 were designed and synthesized, and their
preliminary biological activities were reported.
Alternatively, we expanded the procedure to different precur-
sors, starting from both (L)-valine and (L)-tert-leucine as fragment
B, as they represent a known variation allowing a substantial
bioequivalence in literature. After linking the two intermediates
2-bromoacyl compounds with all three or only first two defined
of a a,a-dimethylbenzylic group on C-a of fragment A as well as a
basic NHMe moiety located in the N-terminal position.5c,6
In this context, in order to expand our synthetic and biological
studies on new anticancer tubulin polymerization inhibitors7 and
to explore new structural alteration of hemiasterlins, we have
investigated the consequences of further modifications of frag-
ments A and C.
R1
R1
O
CO2H
a
N
CO2Et
R1
N
CO2Et
+
H2N
N
H
Br
(S)-1
O
Br
O
(S)(S)-2a,b
(S)(S)(S)-3a,b
R1
a: R1 = H; b: R1 = Me
O
O
N
CO2Et
N
CO2H
b
c
N
H
N
H
(S)(S)(S)-3a,b
NH
O
NH
O
(S)(S)(S)-5a,b
(S)(S)(S)-4a,b
1. b (using R2-NH2,80-83%)
2. c
O
(R,S)-1
a
N
O
CO2H
(R,S)(S)(S)-3b
N
+
H
NH
R2
(S)(S)-2b
(R,S)(S)(S)-6: R2 =
(R,S)(S)(S)-7: R2 =
H
N
H2N
CO2Et
a (68%)
CO2Et
Boc
N
H
+
Boc
N
H
CO2H
O
O
N
1. d, 95%
H
N
CO2H
2. a [using (R,S)-1], 71%
H
NH
O
3. b, 88%
4. c, 90%
(R,S)(S)-8
Scheme 1. Reagents and conditions: (a) TMAC, DIPEA, THF, ꢀ78 °C (75–85%); (b) PhC(Me)2–NH2, Ag2O, toluene (sonication, 90–92%); (c) LiOH/MeOH/water, then H+ (85–
90%); (d) trifluoroacetic acid (95%).

D. Simoni et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3431–3435
3433
Figure 3. (A) UCI-101 cells were treated with various concentrations of compounds for 2 days. The tumor growth suppression was determined by Alamar blue staining and
plotted against concentration of the treated drugs. Each concentration was repeated in triplicates. (B) Same studies were done with N1–S1 rat hepatoma cells. (C) Comparison
of tubulin polymerization inhibition curves for compounds (R)(S)(S)-5a and 5b, with those of colchicine (negative control for polymerization), cis-3,40,5-trimethoxy-30-
aminostilbene (st5c) and paclitaxel (positive control for polymerization). (R)(S)(S)-5a and 5b appear to be the most potent tubulin depolymerizers. All compounds are used at
5 lM concentration. Tubulin polymerization study was performed with the polymerization assay kit from Cytoskeleton (Denver, CO). (D) Cell cycle analysis was performed in
UCI-101 cells after treatment with (R)(S)(S)-5a for 16 h. Cells were stained with ethidium bromide before FACScan analysis. Similar result was seen in (R)(S)(S)-5b but not
shown.

3434
D. Simoni et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3431–3435
chiral carbons, we were then able to produce a wide series of deriv-
atives just by varying the nucleophile in the last reaction. In the
present report, a limited representative set of derivatives are
shown.
To obtain more information, we were also interested to expand
modifications to fragment C. It is possible to synthesize a new ser-
ies of more rigid derivatives replacing the alkene functional group
with a cyclic aromatic ring, which can be easily achieved with an
appropriate amino benzoic acid precursor.
Finally, for the purpose of exploiting the full potential of the
stereoselective synthetic procedure, compounds 5a,b were synthe-
sized as single diastereomers but, for other derivatives, preliminary
indications on activity were considered our main goal, and the syn-
thesis of diastereomeric mixtures was preferred as being more
informative.
Scheme 1 shows the synthesis of some target compounds via
the key intermediates (S)(S)(S)-3a,b, which was conducted by
adapting the literature procedure described for the fragments
BC.5 For clarity, only (S)(S)(S)-configuration are illustrated for
5a,b but, in the same way, (R)(S)(S)-isomers were obtained using
(R)-1 as starting material.10 For compound 8, which bears the more
rigid aromatic ring in place of fragment C, the intermediate BC was
pounds 5a,b, the rigid modification at fragment C as in (R,S)(S)-8
lead to loss of activity.
Similarly to other hemiasterlin derivatives, the most active
compounds (R)(S)(S)-5a and 5b were also tested for their effects
on tubulin polymerization and cell cycle inhibition. In these stud-
ies, they demonstrated more potent activity to inhibit tubulin
polymerization than the known natural tubulin inhibitors such as
colchicine. Their activity is similar to the synthetic tubulin inhibi-
tor cis-3,40,5-trimethoxy-30-aminostilbene (named st5c),12 as com-
parison (Fig. 3C). In cell cycle analysis, moreover, both (R)(S)(S)-5a
and 5b induced cell cycle arrest at G2 phase, consistent with the
fact that they are tubulin inhibitors. (R)(S)(S)-5a is shown in Figure
3D, with similar result in (R)(S)(S)-5b.
In summary, we here describe an efficient synthesis of novel
hemiasterlins congeners. This very flexible procedure allows us
to open a way to synthesize series of compounds where a number
of A fragments may be synthesized from an
sor, in turn available from any diazotizable
a
-bromoacid precur-
a-aminoacid, and dif-
ferent nucleophiles. Further efforts to optimize the overall
synthetic scheme and to generate a library of novel hemiasterlins
are currently ongoing in the laboratory.
easily obtained through condensation of Boc-(
L
)-tert-leucine and
Acknowledgment
3-aminobenzoic acid ethyl ester.
More in detail, fragments BC [(S)(S)-2a,b] were condensed, via
trimethylacetyl chloride (TMAC) at ꢀ78 °C, with (S)-2-bromoprop-
This work was financially supported in part by Ministero
dell’Università e della Ricerca Scientifica e Tecnologica (PRIN
2006).
anoic acid [(S)-1], in turn easily obtainable from (L)-alanine
through a diazotization–bromuration reaction, to give the bromoa-
cyl-peptides (S)(S)(S)-3a,b. In the next step, solid Ag2O-promoted
bromine displacement in key intermediates (S)(S)(S)-3a,b by
appropriate nucleophiles, with full control of diastereoselectivity,
giving N-alkyl peptides 4a,b with retention of configuration. Inter-
estingly, when the reaction was conducted with hindered nucleo-
philes as dimethyl benzyl amine, no substitution products were
detected without the presence of Ag2O, which becomes essential
not only to obtain the desired stereochemistry. Finally, hydrolysis
of esters 4a,b yielded the expected derivates 5a,b without configu-
rational losses. Diastereomeric excesses of compounds 5a,b were
checked by NMR and HPLC analyses in comparison with diastereo-
meric mixtures obtained when starting from racemic (R,S)-1
instead of (R) or (S)-1.11 Using racemic (R,S)-1, compounds
(R,S)(S)(S)-6,7 and (R,S)(S)-8 were similarly obtained as diastereo-
meric mixtures.
The cytotoxicity of each compound was examined in UCI-101
human ovarian cancer cells and N1–S1 rat hepatoma cells (Figs.
3A and B). The two most active compounds are (R)(S)(S)-5a and
5b with IC50 of 20 nM, whereas (R,S)(S)(S)-7 and (R,S)(S)-8 have
no cytotoxic activity even at 1 mM. The other two compounds
(S)(S)(S)-5b and (R,S)(S)(S)-6 have IC50 at 200 nM. The fact that
(R)(S)(S)-5a and 5b are the two most potent compounds compared
with their stereoisomers (S)(S)(S)-5a and 5b suggests the essential
role of the first (R) configuration. With respect to taltobulin deriv-
atives, in which (S)(S)(S) stereochemistry is reported to have po-
tent activity, there is some discrepancy. However, the bulky
dimethyl benzyl group in the most active diastereomers occupies
the same place in both series. It seems, therefore, that the correct
placement of dimethyl benzyl group is more important than the
secondary amine to give functional interaction with binding site.
The poor activity found with the aromatic 2-naphthyl-2-propyl
group of 6, suggests the presence of a large pocket that could be
occupied by an aromatic group, similarly to the indole ring in nat-
ural parent compound hemiasterlin. Among the other substituents
at nitrogen of fragment A, cyclohexyl did not lead to an active com-
pound. Lack of the aryl portion also seems to be incompatible with
growth inhibition, which is consistent with the results described in
taltobulin series. Moreover, maintaining the fragment A as in com-
References and notes
1. (a) Talpir, R.; Benayahu, Y.; Kashman, Y.; Pannell, L.; Schleyer, M. Tetrahedron
Lett. 1994, 35, 4453; (b) Coleman, J. E.; de Silva, E. D.; Kong, F.; Andersen, R. J.;
Allen, T. M. Tetrahedron 1995, 51, 10653; (c) Gamble, W. R.; Durso, N. A.; Fuller,
R. W.; Westergaard, C. K.; Johnson, T. R.; Sackett, D. L.; Hamel, E.; Cardellina, J.
H., II; Boyd, M. R. Bioorg. Med. Chem. 1999, 7, 1611.
2. Anderson, H. J.; Coleman, J. E.; Andersen, R. J.; Roberge, M. Cancer Chemother.
Pharmacol. 1997, 39, 223.
3. (a) Zask, A.; Kaplan, J.; Musto, S.; Loganzo, F. J. Am. Chem. Soc. 2005, 127, 17667;
(b) Ayral-Kaloustian, S.; Zask, A. Drugs Future 2005, 30, 254.
4. (a) Hadaschik, B. A.; Ettinger, S.; Sowery, R.; Richard, D.; Zoubeidi, A.; Andersen,
R. J.; Roberge, M.; Gleave, M. E. Int. J. Cancer 2008, 122, 2368; (b) Hadaschik, B.
A.; Adomat, H.; Fazli, L.; Fradet, Y.; Andersen, R. J.; Gleave, M. E.; So, A. I. Clin.
Cancer Res. 2008, 14, 1510.
5. (a) Andersen, R.; Coleman, J. Tetrahedron Lett. 1997, 38, 317; (b) Andersen, R.;
Piers, E.; Nieman, J.; Coleman, J.; Roberge, M. Hemiasterlin analogs, 1999, WO99/
32509.; (c) Nieman, A.; Coleman, J. E.; Wallace, D. J.; Piers, E.; Lim, L. Y.;
Roberge, M.; Andersen, R. J. Nat. Prod. 2003, 66, 183.
6. (a) Yamashita, A.; Norton, E. B.; Kaplan, J. A.; Niu, C.; Loganzo, F.; Hernandez, R.;
Beyer, C. F.; Annable, T.; Musto, S.; Discafani, C.; Zask, A.; Ayral-Kaloustian, S.
Bioorg. Med. Chem. Lett. 2004, 5317; (b) Zask, A.; Birnberg, G.; Cheung, K.;
Kaplan, J.; Niu, C.; Norton, E.; Suayan, R.; Yamashita, A.; Cole, D.; Tang, Z.;
Krishnamurthy, G.; Williamson, R.; Khafizova, G.; Musto, S.; Hernandez, R.;
Annable, T.; Yang, X.; Discafani, C.; Beyer, C.; Greenberger, L. M.; Loganzo, F.;
Ayral-Kaloustian, S. J. Med. Chem. 2004, 47, 4774; (c) Qie, J.; Zhou, W.; Zhou, X.;
He, J.; Zhang, Y.; Liu, K. Int. J. Pept. Res. Ther. 2009, 15, 187.
7. (a) Simoni, D.; Romagnoli, R.; Baruchello, R.; Rondanin, R.; Grisolia, G.; Eleopra,
M.; Rizzi, M.; Tolomeo, M.; Giannini, G.; Alloatti, D.; Castorina, M.; Marcellini,
M.; Pisano, C. J. Med. Chem. 2008, 19, 6211; (b) Simoni, D.; Invidiata, F. P.;
Eleopra, M.; Marchetti, P.; Rondanin, R.; Baruchello, R.; Grisolia, G.; Tripathi, A.;
Kellogg, G. E.; Durrant, D.; Lee, R. M. Bioorg. Med. Chem. 2009, 17, 512.
8. (a) D’Angeli, F.; Marchetti, P.; Cavicchioni, G.; Maran, F.; Bertolasi, V.
Tetrahedron: Asymmetry 1991, 2, 1111; (b) D’Angeli, F.; Marchetti, P.;
Bertolasi, V. J. Org. Chem. 1995, 60, 4013; (c) Maran, F. J. Am. Chem. Soc. 1993,
115, 6557.
9. (a) Salvadori, S.; Tomatis, R.; Balboni, G.; Marchetti, P.; D’Angeli, F.; Lazarus, L.
H. Peptides 1992, 185; (b) D’Angeli, F.; Marchetti, P.; Salvadori, S.; Balboni, G. J.
Chem. Soc., Chem. Commun. 1993, 3, 304.
10. As sample procedures: To a suspension of Ag2O (464 mg, 2 mmol) in anhydrous
toluene (5 mL), (S)(S)(S)-3b (446 mg, 1 mmol) and 2-phenyl-2-aminopropane
(270 mg, 3 mmol) were added. The mixture was stirred by sonication under
argon atmosphere for 1 h and then filtered over Celite. Evaporation to constant
weight of the organic solution gave the crude product as a yellow oil that was
purified by column chromatography (toluene/AcOEt 3.5:3) to give (S)(S)(S)-4b
as a solid (451 mg, 90%). By comparison with TLC and NMR of diastereomeric
mixture (R,S)(S)(S)-4b obtained starting from (R,S)(S)(S)-3b, we confirmed the
optical purity of (S)(S)(S)-4b. 1H NMR: d 0.79 (d, 3H, J = 6.8 Hz), 0.87 (d, 3H,
J = 6.8 Hz), 0.97 (s, 9H), 1.21 (s, 3H), 1.32 (t, 3H, J = 7.2 Hz), 1.40 (d, 3H,

D. Simoni et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3431–3435
3435
J = 6.8 Hz), 1.52 (s, 3H), 1.87–1.91 (m, 1H), 1.90 (d, 3H, J = 0.8 Hz), 2.90–3.01
(br, 2H), 3.01 (s, 3H), 4.17–4.22 (m, 3H), 4.79 (d, 1H, J = 9.6 Hz), 5.08 (dd, 1H,
J = 10.2 and 10 Hz), 6.63 (dd, 1H, J = 9.6 and 0.8 Hz), 7.19–7.49 (m, 5H). 13C
NMR d 13.9, 14.3, 18.8, 19.5, 21.3, 26.5, 27.3, 30.0, 31.2, 32.2, 35.4, 54.7, 56.3,
61.0, 68.2, 124.6, 126.6, 128.4, 132.8, 138.2, 150.1, 167.8, 171.8, 174.6. To a
solution of tripeptide ethyl ester (S)(S)(S)-4b (60 mg, 0.12 mmol) in MeOH
(2 mL) and H2O (1 mL), LiOH monohydrate (50 mg, 1.2 mmol) was added and
the reaction mixture stirred for 3 h at rt. The mixture was acidified with TFA
and then extracted with AcOEt (3 ꢁ 5 mL). The combined extracts were dried
over anhydrous Na2SO4 and concentrated in vacuo to give (S)(S)(S)-5b as
trifluoroacetate salt. The crude product was then purified by trituration with
Et2O (3 ꢁ 3 mL). Solid, (70 mg, 86%). 1H NMR (CD3OD) d 0.77 d, 3Y, J = 6.4 Hz,
0.86 d, 3Y, J = 6.4 Hz, 0.90 s, 9Y, 1.30 (d, 3H, J = 7.2 Hz), 1.75 (s, 3H), 1.80 (s, 3H),
1.84 (d, 3H, J = 0.6 Hz), 1.90–2.03 (m, 1H), 2.62 (d, 2H, J = 15.6 Hz), 3.03 (s, 3H),
3.80 (q, 1H, J = 7.2 Hz), 4.58 (s, 1H), 4.95 (dd, 1H, J = 10 and 9.6 Hz), 6.65 (dd,
1H, J = 9.6 and 0.6 Hz), 7.42–7.56 (m, 5H). 13C NMR (CD3OD) d 13.4, 15.1, 18.6,
19.3, 19.5, 19.6, 26.0, 26.2, 28.7, 30.8, 34.7, 42.8, 52.1, 54.4, 55.8, 64.8, 72.1,
125.8, 127.6, 128.5, 131.2, 131.8, 138.3, 168.5, 171.2, 174.8. With a similar
procedure, starting with (R)(S)(S)-3b, compound (R)(S)(S)-5b was obtained: 1H
NMR (CD3OD) d 0.89 (d, 3H, J = 6.8 Hz), 0.92 (d, 3H, J = 6.4 Hz), 0.96 (s, 9H), 1.39
(d, 3H, J = 6.8 Hz), 1.63 (s, 3H), 1.75 (s, 3H), 1.89 (d, 3H, J = 0.8 Hz), 2.01–2.10
(m, 1H), 3.09 (s, 3H), 3.78 (q, 1H, J = 6.8 Hz), 4.73 (s, 1H), 5.01 (dd, 1H, J = 9.8
and 9.4 Hz), 6.90 (dd, 1H, J = 9.0 and 0.8 Hz), 7.42–7.57 (m, 5H).13C NMR d 14.2,
18.9, 19.4, 19.9, 26.2, 26.8, 27.1, 31.1, 31.9, 35.6, 53.9, 57.9, 58.2, 64.4, 127.6,
130.6, 130.7, 134.0, 139.6, 139.9, 168.6, 171.0, 172.9.
11. Single diastereomeric compounds (S)(S)(S)- and (R)(S)(S)-3a were also
achieved by column chromatography separation of the mixture (R,S)(S)(S)-3a,
obtained starting from racemic 1.
12. Cao, T. M.; Durrant, D. E.; Tripathi, A.; Liu, J.; Tsai, S.; Kellogg, G. E.; Simoni, D.;
Lee, R. M. Am. J. Hematol. 2008, 83, 390.