Synthesis and biological evaluation of a spongistatin AB-spiroketal analogue.

Article abstract of DOI:10.1016/S0960-894X(02)00305-0

The synthesis of a simplified analogue of the potent, cytotoxic tubulin-depolymerizing agent spongistatin 1, based on the AB spiroketal framework, is presented. The new structural analogue is an extension of a recently described spongistatin congener reported to disrupt microtubules in breast cancer cells in vitro and to alter the microtubule assembly reaction. Cytotoxicity data on the new structural analogue, as well as the parent congener, are reported. We found no significant cytotoxic or antitubulin activity with either compound.

Full text of DOI:10.1016/S0960-894X(02)00305-0

Bioorganic & Medicinal Chemistry Letters 12 (2002) 2039–2042
Synthesis and Biological Evaluation of a Spongistatin
AB-Spiroketal Analogue
Amos B. Smith, III,a,* R. Michael Corbett,a George R. Pettit,b Jean-Charles Chapuis,b
Jean M. Schmidt,b Ernest Hamelc and M. Katherine Jungd
aDepartment of Chemistry, University of Pennsylvania Philadelphia, PA 19104, USA
bCancer Research Institute, Arizona State University, Tempe, AZ 85287, USA
cScreening Technologies Branch, Developmental Therapeutics Program, Division of Cancer Treatment and Diagnosis, National Cancer
Institute at Frederick, National Institutes of Health, Frederick, MD 21702, USA
dScience Applications International Corporation-Frederick, National Cancer Institute at Frederick, National Institutes of Health,
Frederick, MD 21702, USA
Received 10 October 2001; accepted 11 March 2002
Abstract—The synthesis of a simplified analogue of the potent, cytotoxic tubulin-depolymerizing agent spongistatin 1, based on the
AB spiroketal framework, is presented. The new structural analogue is an extension of a recently described spongistatin congener
reported to disrupt microtubules in breast cancer cells in vitro and to alter the microtubule assembly reaction. Cytotoxicity data on
the new structural analogue, as well as the parent congener, are reported. We found no significant cytotoxic or antitubulin activity
with either compound. # 2002 Elsevier Science Ltd. All rights reserved.
Subsequent to the discovery and structural elucidation
of the potent, tubulin-binding antitumor compounds
comprising the spongipyran family,1 intense effort has
been directed at discerning the mode of action2 of the
spongistatins, as well as achieving total syntheses3 of
members of this exciting class of anticancer compounds.
Owing to the impressive structural complexity of the
spongistatins, however, in conjunction with their
extreme scarcity, there is strong impetus for identifica-
tion of simpler structural analogues which retain the
remarkable cytotoxic profile common to members of the
spongipyran family.
favorable for spongistatin binding on the basis of the
potential for stabilizing van der Waals interactions
between the hydrophobic residues in the binding pocket
and the spiroketal moieties embedded in spongistatin 1
(1) (Fig. 1). Molecular modeling using atomic coordi-
nates for the a,b-tubulin hetereodimer derived from the
electron crystallographic data suggested spiroketal-diol
3 would be a suitable target for synthesis and biological
screening. The efficacy of the design rationale was tested
via synthesis of 3, subjection of 3 to in vitro assays to
probe cytotoxicity against human breast cancer cells,
and subsequent screening as a tubulin depolymerizing
agent.4 Spiroketal-diol 3 reportedly disrupted mitotic
spindle formation in human breast cancer cells (BT-20
cells) at the 10 nanomolar level, disrupted microtubule
assembly (MDA-MB-231 cells) at the 500 nanomolar
level, and affected GTP-induced assembly of bovine
brain microtubule protein.
Recently, Uckun and co-workers reported the synthesis
and biological evaluation of a spongistatin surrogate
whose structure was conceived originally via rational
design.4 The design strategy entailed examination of the
electron crystallographic structure of the zinc-induced
tubulin sheets,5 leading to identification of a putative
spongistatin binding site on the a,b-tubulin heterodimer
surface characterized by an array of hydrophobic amino
acid residues. The proposed binding site was deemed
We found the notable biological results for spiroketal-diol
3 intriguing, owing to the extreme structural simplicity.
Therefore, via adaptation of our previously disclosed3d
successful synthetic route culminating in the total syn-
thesis of spongistatin 2 (2), we reasoned we could
rapidly arrive at a potent, more-highly functionalized
congener of 3 and effectively build upon the privileged
*Corresponding author. Fax: +1-215-898-5129; e-mail: smithab@sas.
upenn.edu
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00305-0

2040
A. B. Smith, III et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2039–2042
Our synthesis of the spongistatin AB-spiroketal mimic 4
began (Scheme 1) with dithiane-penta-ol (+)-5,7 derived
via our one-pot unsymmetrical linchpin dithiane cou-
pling of epoxides.8 Dithiane removal9 under buffered
conditions followed by acid treatment afforded spiro-
ketal (À)-6 (72%). Selective silylation of the free primary
hydroxyl group (86%, based on recovered starting
material) and acetylation of the secondary axial hydroxyl
group (92%) furnished bis-silyl ether (À)-7. The spongi-
statin AB-spiroketal mimic (À)-4 was obtained in 91%
yield upon exposure of (À)-7 to tetrabutylammonium
fluoride (TBAF) in THF. The structure of (À)-4 was
confirmed by single crystal X-ray analysis. Interestingly,
the unit cell of (À)-4 contained four different C(1)–C(2)
rotational isomers of (À)-4 (only one is depicted in
Scheme 1).
With the AB-spiroketal spongistatin analogue (À)-4 in
hand, our first objective was to define the cytotoxicity
via screening against a representative array of known
human tumor lines. Since the design of (À)-4 was
inspired by spiroketal structure (À)-3, we deemed it also
prudent to synthesize (À)-3 as a control and subject
both analogues to the same series of tests. Thus, (À)-3
was prepared using the same route described pre-
viously;4,10 (À)-3 and (À)-4 were then tested side-by-side
against various human tumor cell lines. The results of
the biological assay are presented in Table 1.
Table 1. Results of biological screening of spiroketals (À)-3 and (À)-4
against representative human cancer cell lines (GI50 values in mg/mL)
Figure 1. Structures of the potent, cytotoxic tubulin binders spongi-
statins 1 and 2, and structural analogues.
Cell type (cell line)
(À)-3
(À)-4
Pancreas-a (BXPC-3)
Breast adn (MCF-7)
CNS Gliobl (SF268)
Lung-NSC (NCI-H460)
Colon (KM20L2)
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
structural motif. We thus sought to obtain the spongis-
tatin AB-spiroketal analogue 4 using material made
available through our studies directed toward the
synthesis of the spongistatins. In doing so, we intended
to augment our ongoing program for identification of
effective spongistatin structural analogues.6
Prostate (DU-145)
To our surprise, neither (À)-3 nor (À)-4 displayed
detectable growth inhibitory effects against any of the
cell lines tested. We were particularly concerned by the
complete ineffectiveness of (À)-3 and (À)-4 to inhibit
growth in the breast cancer cell line MCF-7 since (À)-3
was previously reported4a to disrupt the spindle micro-
tubules of BT-20 breast cancer cells at the 10 nM level
and overall microtubule organization in MDA-MB-231
breast cancer cells at a 500 nM concentration. We also
treated the latter cell line with 500 nM (À)-3, but we
observed abundant microtubules. The untreated cells
and the treated cells were indistinguishable. Unfortu-
nately, our results also indicate that (À)-3 possesses no
cytotoxicity against several of the cell lines commonly
used by the National Cancer Institute as standards for
detecting cancer cell growth inhibitory activity.
We also examined (À)-3 and (À)-4 in our standard
tubulin polymerization assay (Fig. 2A). In accord with our
cell data, both compounds failed to affect tubulin assembly
at concentrations up to 200 mM. This assay employs a
highly purified bovine brain tubulin preparation,11 and
the compounds were added at concentrations ranging
Scheme 1. (a) (i) HgClO4 Á4H2O, 2,6-lutidine, MeCN/H2O (10:1);
(ii) HClO4 (3.5% aq), CH2Cl2, 72%; (b) TBSCl (1.5 equiv), ImH
(5 equiv), DMF, 67% (86% borsm); (c) Ac2O (40 equiv), pyridine
(80 equiv), DMAP (1equiv), THF, 92%; (d) TBAF (3 equiv), THF, 93%.

A. B. Smith, III et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2039–2042
2041
from 200 nM to 200 mM. In all cases the reactions with
both (À)-3 and (À)-4 could be superimposed on the
control reaction without drug, as shown in Figure 2A
for the 200 mM reaction mixtures. In contrast, as
described previously,2a spongistatin 1 (1) caused pro-
gressive inhibition of assembly as the concentration was
increased from 2 to 10 mM.
possible that the major effect of (À)-3 was on micro-
tubule disassembly as opposed to inhibition of poly-
merization, since Uckun et al.4a reported an unusual
loss of turbidity in a reaction mixture of microtubule
protein containing 0.2 mM (À)-3 held at 37 ꢀC upon
addition of GTP. To evaluate this possibility, we did
two types of experiments. First, microtubule protein
was assembled with GTP, and (À)-3 or spongistatin 1
was added. We found no effect with (À)-3, whereas slow
disassembly, as manifested by decline in turbidity,
occurred with spongistanin 1 (data not presented). Sec-
ond, we set up experiments similar to those described by
Uckun et al.4a as shown in Figure 3. Reaction mixtures
containing all components except GTP were placed in
cuvettes at 0 ꢀC, and at zero time the temperature was
set at 37 ꢀC, reaching that temperature in about 75 s.
The increasing temperature caused a decrease in tur-
bidity in all reaction mixtures. This had been attributed
to disassembly of pre-existing ring oligomers composed
of tubulin+microtubule-associated proteins.12 Within
3 min, turbidity in the reaction mixtures began to rise at
an almost linear rate (curve 0, no GTP), probably
resulting from slow tubulin aggregation. Addition of
GTP and mixing it into the reaction mixture (this took
about 10 s), both without drug (curve G) or with differ-
ent concentrations of (À)-3 (curves 1, 2 and 3), caused a
small drop in turbidity, which was followed by a rapid
rise in turbidity, representing microtubule assembly.
The data obtained form the reaction mixtures contain-
ing (À)-3 are indistinguishable from the control. We
thus do not observe any effect of (À)-3 in this reaction
system. In contrast, with spongistatin 1 (curve S), in the
absence of GTP, there was a small rise in the turbidity
relative to the control. Addition and mixing of GTP
resulted in a similar loss in turbidity, but the rapid
microtubule assembly reaction did not occur. Thus,
taken together, our cytotoxicity and tubulin studies call
Figure 2. Inability of compounds (À)-3 and (À)-4 to inhibit either the
glutamate-induced assembly of purified tubulin (panel A) or glycerol-
enhanced assembly of microtubule protein (panel B). Tubulin and
microtubule protein were prepared as described previously.11 Tubulin
assembly was followed turbidimetrically at 350 nm in either Gilford
250 (A) or Beckman DU7400/7500 (B) spectrometers equipped with
electronic temperature controllers. For the panel A experiments, reac-
tion mixtures contained tubulin (1.0 mg/mL=10 mM), 0.8 M mono-
sodium glutamate (pH 6.6, adjusted with HCl in 2 M stock solution),
0.4 mM GTP, 4% (v/v) dimethyl sulfoxide (the drug solvent), and
drug as follows: none (curve 1), spongistatin 1 (1) at 2, 5, or 10 mM
(curves 2, 3, and 4, respectively); compound (À)-3 at 200 mM (curve 5);
compound (À)-4 at 200 mM (curve 6). Following a 15-min tubulin-
drug pre-incubation at 30 ꢀC without GTP, the reaction mixtures were
placed on ice and GTP was added. They were transferred to cuvettes
at 0 ꢀC, and baselines were established. Incubation was at 30 ꢀC.
Curves 5 and 6 were essentially identical to curve 1, and the three
curves are shown superimposed. For the panel B experiments, the
reaction conditions of Uckun et al.4a were used, except that leupeptin
and aprotinin were not included, and the microtubule protein (1.0 mg/
mL) was prepared in the laboratory rather than purchased. All com-
ponents were mixed at 0 ꢀC. Reaction mixtures were placed in cuvettes
at 5 ꢀC, and baselines were established. Samples were held at 5 ꢀC for
3 min and then incubated for 20 min at 37 ꢀC (four readings per min).
The 20-min turbidity readings in the presence of the indicated con-
centrations of drugs as a percentage of a simultaneously obtained
control reaction are plotted. No significant drug effects were observed
in the tracing except with spongistatin 1. Drugs are indicated as fol-
lows: !, spongistatin 1 (1); &, compound (À)-3; *, compound (À)-4.
Upon reviewing the report from Uckun and collabor-
ators,4a however, we noted that in their studies on inhi-
bition of tubulin assembly they used reaction conditions
substantially different from our standard condition.
They used a commercial bovine brain microtubule pro-
tein preparation (tubulin unresolved from microtubule-
associated proteins), with further enhancement of
assembly by addition of 2.5 M glycerol. In addition,
they specifically reported inhibition of assembly by
(À)-3 at the relatively low concentration of 200 nM. We
therefore further examined the effects of spongistatin 1
(1), (À)-3, and (À)-4 with microtubule protein under the
reaction conditions described by Uckun and collabor-
ators.4a As shown in Figure 2B, no significant inhibitory
effect was observed with either (À)-3 or (À)-4 across a
broad concentration range (200 nM to 200 mM), while
spongistatin 1 (1) again showed progressive inhibition
that was nearly complete by 5 mM. However, it was
Figure 3. Reaction mixtures contained the components described by
Uckun et al.,4a except for leupeptin and aprotinin, and we used our
own preparation of microtubule protein. All components except GTP
were added on ice, and the reaction mixtures were transferred to cuv-
ettes held at 0 ꢀC in Gilford 250 spectrophotometers. At zero time the
temperature controller was set to 37 ꢀC. Curves G and 0–3 were iden-
tical. At the indicated times GTP was added to the reaction mixtures
represented by curves G, S, and 1–3 (no addition for curve 0). The
reaction mixture represented by 0 and G contained no drug, while S
contained 5 mM spongistatin 1 (1), and 1–3 contained (À)-3 at 0.2, 2,
and 50 mM, respectively. At the end of the experiment the reaction
mixtures were chilled to 0 ꢀC (not shown), and this resulted in sub-
stantial decrease in turbidity in the reaction mixtures represented by
curves G and 1–3 but not in the reaction mixtures represented by
curves 0 and S, indicating no microtubule assembly in the latter two
reaction mixtures.

2042
A. B. Smith, III et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2039–2042
into question the original claims4 that (À)-3 disrupts
breast cancer cell microtubules and interferes with the
assembly of microtubule protein.
70329, contract N01-CO-56000, and a postdoctoral
fellowship, CA-81337 to R.M.C.
References and Notes
Although the work presented here makes it unlikely that
such simple compounds as (À)-3 and (À)-4 represent the
pharmacophore of the spongistatins, it is nonetheless
possible that only a portion of the structure of these
highly active compounds will turn out to be responsible
for their biological activities. For example, the biologi-
cal activity of halichondrin B (8), an antitumor com-
pound with structural and biological similarities to the
spongistatins, can be duplicated in analogues less than
2/3 the size of the parent compound. Compound 9 has
potent antitumor activity in vivo13a and is nearly indis-
tinguishable from halichondrin B (8) in its interactions
with tubulin.13b
1. (a) Pettit, G. R.; Tan, R.; Gao, F.; Williams, M. D.; Dou-
bek, D. L.; Boyd, M. R.; Schmidt, J. M.; Chapuis, J.-C.;
Hamel, E.; Bai, R.; Hooper, J. N. A.; Tacket, L. P. J. Org.
Chem. 1993, 58, 2538. (b) Pettit, G. R.; Cichacz, Z. A.; Gao,
F.; Herald, C. L.; Boyd, M. R. J. Chem. Soc., Chem. Commun.
1993, 1166. (c) Fusetani, N.; Shinoda, K.; Matsunaga, S. J.
Am. Chem. Soc. 1993, 115, 3977. (d) Kobayashi, M.; Aoki, S.;
Sakai, K.; Kawazoe, K.; Kihara, N.; Sakai, T.; Kitigawa, I.
Tetrahedron Lett. 1993, 34, 2795. (e) Kobayashi, M.; Aoki, S.;
Kitigawa, I. Tetrahedron Lett. 1994, 35, 1243.
2. (a) Bai, R.; Cichacz, Z.; Herald, C. L.; Pettit, G. R.; Hamel,
E. Mol. Pharmacol. 1993, 44, 757. (b) Luduena, R. F.; Roach,
M. C.; Prasad, V.; Pettit, G. R.; Cichacz, Z. A.; Herald, C. L.
Drug Develop. Res. 1995, 35, 40. (c) Bai, R. L.; Taylor, G. F.;
Cichacz, Z. A.; Herald, C. L.; Kepler, J. A.; Pettit, G. R.;
Hamel, E. Biochemistry 1995, 34, 9714.
3. (a) Evans, D. A.; Trotter, B. W.; Cote, B.; Coleman, P. J.;
Dias, L. C.; Tyler, A. N. Angew. Chem. 1997, 109, 2957,
Angew. Chem. Int. Ed. Engl. 1997, 36, 2744. (b) Evans, D. A.;
Trotter, B. W.; Coleman, P. J.; Cote, B.; Dias, L. C.; Raja-
pakse, H. A. Tetrahedron 1999, 55, 8671. (c) Hayward, M. M.;
Roth, R. M.; Duffy, K. J.; Dalko, P. I.; Stevens, K. L.; Guo,
J.; Kishi, Y. Angew. Chem. 1998, 110, 202, Angew. Chem. Int.
Ed. Engl. 1998, 37, 192. (d) Smith, A. B., III; Lin, Q.;
Doughty, V. A.; Zhuang, L.; McBriar, M. D.; Kerns, J. K.;
Brook, C. S.; Murase, N.; Nakayama, K. Angew. Chem. 2001,
113, 202 , Angew. Chem. Int. Ed. Engl. 2001, 40, 196.
4. (a) Uckun, F. M.; Mao, C.; Vassilev, A. O.; Huang, H.;
Jan, S.-T. Bioorg. Med. Chem. Lett. 2000, 10, 541. (b) Huang,
H.; Mao, C.; Jan, S.-T.; Uckun, F. M. Tetrahedron Lett. 2000,
41, 1699.
5. Nogales, E.; Wolf, S. G.; Downing, K. H. Nature 1998,
391, 199.
6. Smith, A. B., III; Lin, Q.; Pettit, G. R.; Chapuis, J.-C.;
Schmidt, J. M. Bioorg. Med. Chem. Lett. 1998, 8, 567.
7. Lin Q., University of Pennsylvania, unpublished data.
8. Smith, A. B., III; Boldi, A. M. J. Am. Chem. Soc. 1997,
119, 6925.
9. Fujita, E.; Nagao, Y.; Kaneko, K. Chem. Pharm. Bull.
1978, 26, 3743.
10. Spiroketal (À)-3 had spectroscopic properties (1H, 13C
NMR, IR, optical rotation) identical to those reported
previously.4b
In conclusion, we have carried out a short synthesis of a
highly functionalized spongistatin AB-spiroketal mimic
(À)-4 starting from material available from our total
synthesis of spongistatin 2. Neither (À)-4 nor the parent
compound (À)-3 was active against several lines of
human cancer cells regularly utilized in drug cytotoxi-
city trials. Additionally, neither (À)-3 nor (À)-4 inhib-
ited polymerization of tubulin or caused polymer
disassembly using several assays.
11. Hamel, E.; Lin, C. M. Biochemistry 1984, 23, 4173.
12. Spann, U.; Renner, W.; Mandelkow, E.-M.; Bordas, J.;
Mandelkow, E. Biochemistry 1987, 26, 1123.
13. (a) Towle, M. J.; Salvato, K. A.; Budrow, J.; Wels, B. F.;
Kuznetsov, G.; Aalfs, K. K.; Welsh, S.; Zheng, W.; Seletsk,
B. M.; Palme, M. H.; Habgood, G. J.; Singer, L. A.; Dipietro,
L. V.; Wang, Y.; Chen, J. J.; Quincy, D. A.; Davis, A.; Yosh-
imatsu, K.; Kishi, Y.; Yu, M. J.; Littlefield, B. A. Cancer
Res. 2001, 61, 1013. (b) Verdier-Pinard, P.; Bai, R.; Yu, M. J.;
Littlefield, B. A.; Hamel, E. Mol. Biol. Cell 2000, 11, 188a.
Acknowledgements
Support was provided by the National Institutes of
Health (National Cancer Institute) through grant CA-