Synthesis and antiproliferative activity of thiazolidine analogs for melanoma

Article abstract of DOI:10.1016/j.bmcl.2007.05.059

We have previously described 2-aryl-thiazolidine-4-carboxylic acid amides as a novel class of antiproliferative agents for prostate cancer. Screening these compounds with melanoma cell lines revealed that several of them have potent antiproliferative activity and selectivity against melanoma. To further improve the potency and selectivity, we synthesized a new series of analogs and tested them in two melanoma cell lines and fibroblast cells (negative controls). Comparison of anticancer effects of these compounds with a standard chemotherapeutic agent, sorafenib, showed that they are very effective in killing melanoma cells with low micromolar to nanomolar antiproliferative activity and provide us a new lead for developing potential drugs for melanoma.

Full text of DOI:10.1016/j.bmcl.2007.05.059

Bioorganic & Medicinal Chemistry Letters 17 (2007) 4113–4117
Synthesis and antiproliferative activity of thiazolidine
analogs for melanoma
Wei Li,a Yan Lu,a Zhao Wang,a James T. Daltonb and Duane D. Millera,*
aDepartment of Pharmaceutical Sciences, College of Pharmacy, University of Tennessee Health Science Center, Memphis,
TN 38163, USA
bDivision of Pharmaceutics, College of Pharmacy, The Ohio State University, Columbus, OH 43210, USA
Received 23 April 2007; revised 16 May 2007; accepted 17 May 2007
Available online 23 May 2007
Abstract—We have previously described 2-aryl-thiazolidine-4-carboxylic acid amides as a novel class of antiproliferative agents
for prostate cancer. Screening these compounds with melanoma cell lines revealed that several of them have potent antiprolif-
erative activity and selectivity against melanoma. To further improve the potency and selectivity, we synthesized a new series of
analogs and tested them in two melanoma cell lines and fibroblast cells (negative controls). Comparison of anticancer effects of
these compounds with a standard chemotherapeutic agent, sorafenib, showed that they are very effective in killing melanoma
cells with low micromolar to nanomolar antiproliferative activity and provide us a new lead for developing potential drugs for
melanoma.
Ó 2007 Elsevier Ltd. All rights reserved.
Melanoma is the most aggressive form of skin cancer
and is the fastest growing cancer in the United States.1,2
Early stage melanoma can be cured surgically. However,
melanoma metastasizing to major organs (stage IV) is
virtually incurable.2 Patients with advanced melanoma
have a median survival time of less than a year, and
the estimated 5-year survival rate is less than 15%.2,3
Dacarbazine (DTIC) is the only FDA-approved drug
to treat advanced melanoma, but complete remission
after DTIC treatment rarely exceeds 5% in patients.4,5
Although additional adjuvant treatment such as the
use of high-dose interferon alpha-2b (IFN-a2b) for
patients at high risk of recurrence of melanoma has also
been approved by FDA, extensive clinical trials have not
detected a survival advantage with the addition of IFN
to DTIC.6–9 Metastatic melanoma is invariably resistant
to existing agents including triazenes (DTIC or Tem-
ozolomide),10,11 nitrosoureas (BCNU, CCNU, or Fote-
mustine),12,13 as well as combination therapies such as
cisplatin and etoposide.14–16
tion of cancer drugs, including DTIC combined with
cisplatin, vinblastine, or carmustine,17,18 but they all
have failed to demonstrate a clear effect against ad-
vanced melanoma.4,6,16,19 Therefore, DTIC remains
the gold standard for advanced melanoma despite its
very limited efficacy.20,21 With the rapidly rising inci-
dence reported for melanoma in the United States,
clearly there is an urgent need to develop more effective
therapeutic agents to combat advanced melanoma.
Recently, novel classes of lipid compounds have been
synthesized and have shown strong activity toward
prostate cancer cells.22,23 Although this class of com-
pounds was initially designed as inhibitors of lysophos-
phatidic acid (LPA) receptors, we discovered that they
did not bind to LPA. They are unlikely to be DNA
alkylating agents such as DTIC. Encouraged by the
results with prostate cancer, we decided to test a library
of such compounds against metastatic melanoma
in vitro. The initial results provided us with three classes
of highly potent lead compounds for metastatic mela-
noma. The most potent lead compound has an IC50
value in the submicromole range with 10-fold selectivity
against cancer cells.24 To further improve potency and
selectivity, we performed extensive synthesis and biolog-
ical testing of additional compounds in this series. Here
we report the synthesis and in vitro antiproliferative
activity of these new compounds against two human
Several extensive clinical trials have been conducted in
recent years with a variety of cancer drugs or combina-
Keywords: Melanoma; Thiazolidine; Antiproliferative; Structure–activ-
ity relationship.
*
Corresponding author. Tel.: +1 901 448 6026; fax: +1 901 448
3446; e-mail: dmiller@utmem.edu
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.05.059

4114
W. Li et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4113–4117
Scheme 1. Reagents: (a) C2H5OH, H2O (70–99%); (b) Boc2O, 1 N NaOH, 1,4-dioxane, H2O (89–97%); (c) EDCI, HOBt, Et3N, amine, CH2Cl2; (d)
TFA (42–68% two steps); (e) HCHO, NaBH3CN, HOAc (67%).
melanoma cell lines and fibroblast cells to determine
their selectivity.
of 2-aryl-thiazolidine-4-carboxylic acid with EDC/
HOBt.
The general synthesis of 2-aryl-thiazolidine-4-carboxylic
acid amides is shown in Scheme 1.
Derivatives 26 and 27 with a 4-amino-phenyl group were
synthesized by deacetylation of compounds 14 and 17,
which was accomplished by acid hydrolysis in methanol
(Scheme 2).
L- or D-Cysteine was reacted with appropriate benzalde-
hydes in ethanol and water at ambient temperature to
give cyclized 2-aryl-thiazolidine-4-carboxylic acid, which
was converted to the corresponding Boc derivatives.
Reaction of Boc-protected carboxylic acids with differ-
ent amines using EDC/HOBt gave corresponding
amides, which were treated with TFA to form the target
compounds 1–25.22 Reductive alkylation with formalde-
hyde and sodium cyanoborohydride of the amino group
in compound 17 gave methylation derivative 30.25
Dimer 31 was obtained by intramolecular condensation
L-Cysteine and appropriate benzonitriles were dissolved
in a 1:1 (v/v) mixture of phosphate buffer (pH 6.4) and
methanol and stirred at 50 °C to give cyclized 2-aryl-
4,5-dihydro-thiazole-4-carboxylic acid, which was
reacted with tetradecylamine using EDC/HOBt to give
corresponding compounds 28 and 29 as shown in
Scheme 3.26
We characterized each compound with NMR, mass
spectroscopy, and elemental analysis.
We examined the antiproliferative activity of these new-
ly synthesized compounds in two human melanoma cell
lines (SK-MEL-188 and WM-164) and in a fibroblast
cell line. We used activity on fibroblast cells as a control
to determine the selectivity of these compounds against
melanoma. Standard sulforhodamine B (SRB) assay
was used. Cells were exposed to a wide range of concen-
trations for 48 h in round-bottomed 96-well plates. Cells
were fixed with 10% trichloroacetic acid and washed five
times with water. After cells were air-dried overnight
and stained with SRB solution, total proteins were mea-
sured at 560 nm with a plate reader. IC50 (i.e., concen-
Scheme 2.
Scheme 3. Reagents: (a) MeOH/pH 6.4 phosphate buffer (58–75%); (b) EDCI, HOBt, Et3N, n-C14H29NH2, CH2Cl2 (58–87%).

W. Li et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4113–4117
4115
Table 1. Antiproliferative activity of thiazolidine analogs and their comparison with that of sorafenib and DTIC (ND, not detected)
tration which inhibited cell growth by 50% of DMSO-
treated controls) values were obtained by nonlinear
regression analysis with GraphPad Prism (GraphPad
Software, San Diego, CA).
this early stage, all compounds were used as a diastereo-
meric mixture if they contain chiral centers in order to
select the most promising compounds for further
development.
The ability of thiazolidine derivatives to inhibit the
growth of two melanoma cancer cell lines and fibroblast
cells is summarized in Table 1. Sorafenib (Velcade) has
been used extensively in clinical trials for melanoma,
hence we selected this compound and DTIC as reference
standards to assess the activity of our compounds. At
Examination of antiproliferative effects for a variety of
substitutions on the phenyl ring revealed the chain-
length dependence for these compounds (1–5, 6–9, 10–
13, 14–19). Short chain length such as a C10 chain
(for example, compounds 2, 6, 10) displayed low
potency for both cancer cells and fibroblast cells. As

4116
W. Li et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4113–4117
chain length increased, potency increased, as well as tox-
icity as measured on fibroblast cells except when the
acetyl amino group was substituted on the phenyl ring
(compounds 14–17). Both C15 and C16 chains with this
substitution displayed both high potency and high selec-
tivity against cancer cells, with an IC50 for melanoma
cells as low as 600 nM (compound 17). Further chain
length increases, however, reduced potency and selectiv-
ity. At a chain length of C18 (compound 19), the IC50
value was higher than 10 lM for all three cell lines.
Interestingly, adding either a cis- or trans-double bond
in the C18 side chain restored potency dramatically
(compounds 20Z and 20E), demonstrating that both
length and composition of the side chain are critical
for their activity. There is no significant difference in
their activity between the cis- and trans-isomers.
compounds are identified, pure optical isomers will be
separated by preparative HPLC for both in vitro and
in vivo animal testing.
Acknowledgments
This research is supported by funds from Department of
Pharmaceutical Sciences, College of Pharmacy, Univer-
sity of Tennessee Health Science Center (W.L.), and the
Van Vleet Endowed Professorship (D.D.M.). We thank
Dr. David L. Armbruster for editorial assistance.
References and notes
1. Jemal, A.; Tiwari, R. C.; Murray, T.; Ghafoor, A.;
Samuels, A.; Ward, E.; Feuer, E. J.; Thun, M. J. CA
Cancer J. Clin. 2004, 54, 8.
2. Carlson, J. A.; Ross, J. S.; Slominski, A.; Linette, G.;
Mysliborski, J.; Hill, J.; Mihm, M., Jr. J. Am. Acad.
Dermatol. 2005, 52, 743.
3. Buzaid, A. C.; Anderson, C. M. Curr. Oncol. Rep. 2000, 2,
322.
4. Anderson, C. M.; Buzaid, A. C.; Legha, S. S. Oncology
(Williston Park) 1995, 9, 1149.
Removing the acetyl amino group on the phenyl ring
(compounds 21 and 22) resulted in the loss of selectiv-
ity, although potency was similar to those with this
substitution (compounds 15 and 17). Replacing the al-
kyl chain with a methoxyl group completely abolished
potency (compound 23). Changing the chirality from
an R to S configuration at the C4 position on the
thiazolidine ring did not substantially affect either
potency or selectivity (compound 4 vs 24 and com-
pound 17 vs 25). Selectivity has a strong dependence
on the substitutions in the phenyl ring. For example,
with a C12 chain, potency is similar for all the substi-
tutions we studied (compounds 3, 7, 11, 14, and 26).
However, selectivity improves dramatically when prop-
er substitutions are present (compound 17 vs com-
pounds 5, 9, 13, 22, and 27).
5. Serrone, L.; Zeuli, M.; Sega, F. M.; Cognetti, F. J. Exp.
Clin. Cancer Res. 2000, 19, 21.
6. Lawson, D. H. Cancer Control 2005, 12, 236.
7. Bajetta, E.; Di Leo, A.; Zampino, M. G.; Sertoli, M. R.;
Comella, G.; Barduagni, M.; Giannotti, B.; Queirolo, P.;
Tribbia, G.; Bernengo, M. G. J. Clin. Oncol. 1994, 12,
806.
8. Thomson, D. B.; Adena, M.; McLeod, G. R.; Hersey, P.;
Gill, P. G.; Coates, A. S.; Olver, I. N.; Kefford, R. F.;
Lowenthal, R. M.; Beadle, G. F. Melanoma Res. 1993, 3,
133.
9. Young, A. M.; Marsden, J.; Goodman, A.; Burton, A.;
Dunn, J. A. Clin. Oncol. (R. Coll. Radiol.) 2001, 13,
458.
10. Chang, J.; Atkinson, H.; A’Hern, R.; Lorentzos, A.; Gore,
M. E. Eur. J. Cancer 1994, A, 2093.
When the amino group in the thiazolidine ring is
substituted (compound 30) or a double bond is intro-
duced (compounds 28 and 29), the resulting com-
pounds are largely inactive with IC50 values above
20 lM. We also tested the intermediate compound in
which the amino group is protected by a Boc group,
and that compound is inactive also (data not shown).
Furthermore, when we removed the aliphatic chain
and the amino group by synthesizing a dimer, we
obtained an inactive compound (compound 31). These
results clearly demonstrate the importance of the
amino group in the thiazolidine ring.
11. Fletcher, W. S.; Daniels, D. S.; Sondak, V. K.; Dana, B.;
Townsend, R.; Hynes, H. E.; Hutchins, L. F.; Pancoast, J.
R. Am. J. Clin. Oncol. 1993, 16, 359.
12. Madajewicz, S.; West, C. R.; Park, H. C.; Ghoorah, J.;
Avellanosa, A. M.; Takita, H.; Karakousis, C.; Vincent,
R.; Caracandas, J.; Jennings, E. Cancer 1981, 47, 653.
13. Avril, M. F.; Aamdal, S.; Grob, J. J.; Hauschild, A.;
Mohr, P.; Bonerandi, J. J.; Weichenthal, M.; Neuber, K.;
Bieber, T.; Gilde, K.; Guillem Porta, V.; Fra, J.; Bonnet-
erre, J.; Saiag, P.; Kamanabrou, D.; Pehamberger, H.;
Sufliarsky, J.; Gonzalez Larriba, J. L.; Scherrer, A.; Menu,
Y. J. Clin. Oncol. 2004, 22, 1118.
14. Franciosi, V.; Cocconi, G.; Michiara, M.; Di Costanzo,
F.; Fosser, V.; Tonato, M.; Carlini, P.; Boni, C.; Di Sarra,
S. Cancer 1999, 85, 1599.
15. Middleton, M. R.; Lorigan, P.; Owen, J.; Ashcroft, L.;
Lee, S. M.; Harper, P.; Thatcher, N. Br. J. Cancer 2000,
82, 1158.
Not surprising, DTIC was inactive (IC50 > 100 lM) in
our in vitro assay due to lack of bioactivation.27 Recent
clinical trials indicated that sorafenib has promising ef-
fect against melanoma, and it has very low toxicity.28
Our in vitro assay indicated that compound 17 was more
potent and selective against melanoma cells than sorafe-
nib, as indicated by the ratio of its IC50 values for fibro-
blast cells over melanoma cells (10–20 for 17 vs 3–4 for
sorafenib).
16. Mandara, M.; Nortilli, R.; Sava, T.; Cetto, G. L. Expert
Rev. Anticancer Ther. 2006, 6, 121.
In conclusion, we have synthesized novel analogs of
thiazolidine compounds based on initial studies. When
compared with existing anticancer drugs, our com-
pounds were more potent and selective. Further optimi-
zation of the structure to improve their activities is
currently in progress. Once highly potent and selective
17. Chapman, P. B.; Einhorn, L. H.; Meyers, M. L.; Saxman,
S.; Destro, A. N.; Panageas, K. S.; Begg, C. B.; Agarwala,
S. S.; Schuchter, L. M.; Ernstoff, M. S.; Houghton, A. N.;
Kirkwood, J. M. J. Clin. Oncol. 1999, 17, 2745.
18. Eton, O.; Legha, S. S.; Bedikian, A. Y.; Lee, J. J.; Buzaid,
A. C.; Hodges, C.; Ring, S. E.; Papadopoulos, N. E.;

W. Li et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4113–4117
4117
Plager, C.; East, M. J.; Zhan, F.; Benjamin, R. S. J. Clin.
Oncol. 2002, 20, 2045.
24. Li, W.; Wang, Z.; Gududuru, V.; Zbytek, B.; Slominski,
A. T.; Dalton, J. T.; Miller, D. D. Anticancer Res. 2007,
27, 883.
25. Borch, R. F.; Hassid, A. I. J. Org. Chem. 1972, 37,
1673.
26. Zamri, A.; Abdallah, M. A. Tetrahedron 2000, 56, 249.
27. Daidone, G.; Maggio, B.; Raffa, D.; Plescia, S.;
Schillaci, D.; Valeria Raimondi, M. Farmaco 2004, 59,
413.
28. Eisen, T.; Ahmad, T.; Flaherty, K. T.; Gore, M.; Kaye, S.;
Marais, R.; Gibbens, I.; Hackett, S.; James, M.; Schuch-
ter, L. M.; Nathanson, K. L.; Xia, C.; Simantov, R.;
Schwartz, B.; Poulin-Costello, M.; O’Dwyer, P. J.; Ratain,
M. J. Br. J. Cancer 2006, 95, 581.
19. Kaufmann, R.; Spieth, K.; Leiter, U.; Mauch, C.; von den
Driesch, P.; Vogt, T.; Linse, R.; Tilgen, W.; Schadendorf,
D.; Becker, J. C.; Sebastian, G.; Krengel, S.; Kretschmer,
L.; Garbe, C.; Dummer, R. J. Clin. Oncol. 2005, 23, 9001.
20. Eggermont, A. M.; Kirkwood, J. M. Eur. J. Cancer 2004,
40, 1825.
21. Atallah, E.; Flaherty, L. Curr. Treat Options Oncol. 2005,
6, 185.
22. Gududuru, V.; Hurh, E.; Dalton, J. T.; Miller, D. D. J.
Med. Chem. 2005, 48, 2584.
23. Gududuru, V.; Hurh, E.; Sullivan, J.; Dalton, J. T.; Miller,
D. D. Bioorg. Med. Chem. Lett. 2005, 15, 4010.