Substituted ajoenes as novel anti-cancer agents

Article abstract of DOI:10.1016/j.bmcl.2008.08.056

A new synthesis of the ajoene pharmacophore core is presented involving the regioselective radical addition of a thiyl radical to a terminal alkyne as the key step. The synthesis allows structural variation of the two end groups on sulfur, and a range of novel derivatives varying the R¹ group (sulfoxide end) has been prepared and tested against CT-1 transformed fibroblast cells for anti-cancer activity. The results indicate comparable or even improved activity compared to the parent natural product ajoene isomers. This opens up the way to systematically studying the biology of the ajoene core.

Full text of DOI:10.1016/j.bmcl.2008.08.056

Bioorganic & Medicinal Chemistry Letters 18 (2008) 5277–5279
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Substituted ajoenes as novel anti-cancer agents
a,
b
a
a
a
Roger Hunter a, , Catherine H. Kaschula , Iqbal M. Parker , Mino R. Caira , Philip Richards , Susan Travis ,
*
*
Francois Taute a, Thozama Qwebani a
a Department of Chemistry, University of Cape Town, Rondebosch 7701, South Africa
b Department of Medical Biochemistry, Institute of Infectious Disease and Molecular Medicine, University of Cape Town, Observatory 7925, South Africa
a r t i c l e i n f o
a b s t r a c t
Article history:
A new synthesis of the ajoene pharmacophore core is presented involving the regioselective radical addi-
tion of a thiyl radical to a terminal alkyne as the key step. The synthesis allows structural variation of the
two end groups on sulfur, and a range of novel derivatives varying the R1 group (sulfoxide end) has been
prepared and tested against CT-1 transformed fibroblast cells for anti-cancer activity. The results indicate
comparable or even improved activity compared to the parent natural product ajoene isomers. This opens
up the way to systematically studying the biology of the ajoene core.
Received 11 April 2008
Revised 15 August 2008
Accepted 18 August 2008
Available online 22 August 2008
Keywords:
Cancer
Ó 2008 Elsevier Ltd. All rights reserved.
Garlic
Synthesis
Ajoene
Radical addition
Ajoene is a stable sulfoxide rearrangement product of allicin, a
natural product found in freshly crushed garlic (Fig. 1), whose
structure was established in the mid 1980s following seminal work
by Block and Apitz-Castro.1 It possesses an intriguing allyl vinyl
disulfide functional grouping, which is likely to account for its
range of biological activities via acting as a sulfenylating agent to-
wards protein sulfhydryl groups.2 Originally, the focus of its bio-
logical activity centered around its anti-thrombotic activity,3 but
in subsequent years it has been demonstrated to possess a range
of other biological activities to include antimicrobial,4 anti-obes-
ity,5 antifungal,6 and anti-cancer7 activities.
Regarding the latter, while garlic dietary supplements and ex-
tracts have been reported to reduce the risk of cancer8 as well as
alter the activation of several carcinogens9 and to cause growth
inhibition and/or death of tumor cells,10 the overall verdict on gar-
lic is controversial. The fact that crude extracts of garlic contain
numerous organosulfur compounds with varying stability and bio-
logical activity has made it impossible to reach any firm conclu-
sions from clinical trials11 about the chemopreventative effect of
garlic.
and to strongly inhibit metastasis to lung in the B16/BL6 mela-
noma tumor model in C57BL/6 mice.13 Topical application of ajo-
ene to the tumors of 21 human patients with either nodular or
superficial basal cell carcinoma caused a reduction in tumor size
in 17/21 patients.14 Ajoene has been shown to induce apoptosis
and arrest HL60 leukemia cells in the G2/M phase of the cell cycle
in a dose-dependent manner,12c,12d and ajoene-treated leukemia
cells have been shown to undergo a time-dependent reduction in
the anti-apoptotic bcl-2 protein that results in release of cyto-
chrome c and the activation of caspase-3.1,2 These results support
the hypothesis that ajoene-induced apoptosis in leukemia cells
proceeds via the mitochondria-dependent caspase cascade path-
way rather than the triggering of cell-surface death receptors. Ajo-
ene has also been shown to decrease the expression of
a4b1
integrin in murine melanoma cells,12e and to induce complete dis-
assembly of the microtubule network in HL60 cells.12c
The only synthesis published in the literature to date for ajoene
is the biomimetic thermal rearrangement of allicin in aqueous ace-
tone due to Block in his original work.1 Although the synthesis is a
one-pot conversion, the synthesis suffers from being low-yielding
as well as not being realistically amenable to producing structural
By comparison, the fact that ajoene is a relatively stable com-
pound has allowed more accurate data to be collected, and it has
been demonstrated that it is able to induce apoptosis in a number
of tumor cell-lines.12 Ajoene has been shown to offer strong pro-
tection against TPA-promoted carcinogenesis on the mouse skin,
O
S
O
S
S
S
S
Allicin
Ajoene
* Corresponding authors. Tel.: +27 21 650 2544; fax: +27 21 689 7499.
E-mail address: Roger.Hunter@uct.ac.za (R. Hunter).
Figure 1. Structures of garlic-derived allicin and ajoene.
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.08.056

5278
R. Hunter et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5277–5279
utilizes a regioselective radical addition of thiolacetic acid to a ter-
O
S
minal alkyne to generate a vinyl thioacetate based on the known
addition16 of thiolacetic acid to 1-hexyne as a model reaction.
Scheme 1 summarizes the overall reaction sequence.
S
R1
S
Figure 2. General structure of substituted ajoenes available from the present
Thus, propargylation of thiol R1SH available commercially or
synthesized via an isothiouronium salt formed from the bromide
or tosylate (R1X), followed by regioselective radical addition of thi-
olacetic acid to the terminus of the triple bond generated vinyl-
thioacetate A. Optimal conditions for the key step were identified
in order to maximize both alkyne conversion as well as the yield
of the mono-addition product. These involved dropping thiolacetic
acid (1.1 equiv) into the alkyne and the initiator in deoxygenated
toluene at 85 °C over 1 h and continuing heating (1–2 h) until
TLC indicated maximum conversion. Yields for the radical addition
step ranged routinely in around the 60–70% region after chroma-
tography, and the vinylthioacetates A were obtained as a mixture
of E/Z stereoisomers, with the Z-isomer predominating (ꢀ2:1) as
reported for the free-radical addition of thiolacetic acid to 1-hex-
yne.16 E/Z-Stereochemistry was assigned on the basis of the vinyl
coupling constants as ꢀ15 Hz for the E-isomer and ꢀ10 Hz for
the Z-isomer. The stereoselectivity is consistent with the quench-
ing of the intermediate vinyl radicals (E and Z) by thiolacetic acid
to be kinetically favoured for the Z-radical intermediate in view
of minimized steric interaction between the attached thioacetate
synthesis.
(i)
(ii)
R1S
S
S
(iii)
R1S
S
R1SH
A
O
S
O
S
(iv)
R1S
R1
S
B
C
Scheme 1. Reagents and conditions: (i) KOH, MeOH, propargyl bromide; (ii)
CH3COSH (1.1 equiv), AIBN (2 mol%), Tol, 85 °C; (iii) (a) KOH (1.05 equiv), MeOH,
À78 °C; (b) p-TolSO2Sallyl (1.1 equiv); (iv) m-CPBA (1.1 equiv), CH2Cl2, À78 °C to rt.
variants. In this communication, we report on the first synthetic
sequence15 that can access a range of ajoene derivatives containing
the central vinyl disulfide/sulfoxide core while varying the end
group R1 (Fig. 2).
For the purpose of preliminary biological investigation, it was
elected to change only the R1 group and maintain the other end-
group as allyl as in the parent ajoene. The key step in the synthesis
Table 1
IC50 of ajoene derivatives of C against CT-1 fibroblast cells
O
S
S
R1
S
C
No.
Name
–R1
Nb
CT-1
IC50 (M)
95% CIc (M)
1
2
E-ajoene
Z-ajoene
6
6
17.6
15.5
17.4–17.8
15.3–15.6
3
4
E-propyl
Z-propyl
5
5
26.7
17.0
26.5–26.9
16.8–17.2
5
E/Z-tert-butyl
5
33.1
32.5–33.6
6
7
E-OHa
Z-OHa
5
6
23.1
22.8
23.0–23.3
22.5–23.1
OH
8
9
E-OPMB
Z-OPMB
5
6
23.5
21.7
23.2–23.8
21.6–21.9
O
O
O
10
11
E-phthal
Z-phthal
4
4
95.9
34.6
95.0–96.8
34.3–34.9
N
O
12
E/Z-benzyl
5
16.6
11.2
16.5–16.7
11.1–11.3
13
E/Z-PMB
12
O
a
b
c
Prepared via deprotection of its TBDMS ether with HF/CH3CN.
Number of independent experimental observations.
95% Confidence interval in micromolar.

R. Hunter et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5277–5279
5279
group and incoming thiolacetic acid in the hydrogen-quenching
propagation step. Generally, the stereoisomers could not be sepa-
rated by column chromatography and were characterized as a mix-
ture. Subsequent thioacetate deprotection with hydroxide in
methanol at low temperature (À78 °C) followed by sulfenylation
of the enethiolate with S-allyl p-toluenesulfonylthioate17 afforded
a high yield (>90%) of the vinyl disulfide B (R2 = allyl) as the same
E/Z-mixture, the latter indicating sulfenylation to be faster than
isomerization of the intermediate enethiolate. Finally, chemoselec-
tive oxidation of B with m-CPBA (1.1 equiv) afforded the target ajo-
ene derivative C in the same E/Z-ratio as that in B. Yield for the
latter step varied (60–90%) and the optimal temperature range
for the reaction was highly substrate specific. Thus, the overall
yield for the synthesis was usually no less than about 35%. In most
cases (exceptions = entries 5, 12, and 13 in Table 1), the E- and Z-
isomers of C could be separated by slow column chromatography
or preparative TLC. This was considered to be important because
the two stereoisomers of ajoene have been shown1 to have signif-
icantly different biological activities. Unfortunately, the parent E-
and Z-ajoenes could not be prepared via this synthetic sequence
because the intermediate vinyl radical precursor to A presumably
cyclizes onto the allyl group double bond of R1 via either a 5-exo-
or 6-endo-trig process but this aspect was not investigated further.
A preliminary investigation into the use of a protecting group on R1
to circumvent this problem failed to resolve the issue. Thus, the
parent ajoenes (1 and 2 in Table 1) were prepared by the method
of Block.1 The final products C were characterized (see Annexure
Supplementary data) by the normal range of spectroscopic (1H
and 13C NMR) and analytical techniques including high resolution
mass spectrometry in view of their nature as oils. A particularly
diagnostic set of resonances for C pertained to the methylene
and two vinyl protons of the vinyl disulfide grouping in the 1H
NMR spectrum in which a consistent coupling set could be identi-
fied for each derivative. Finally, the synthesis should also be ame-
nable to changing the other allyl end-group that was kept constant
in this particular study.
creased chemical stability of the analogues over the parent ajoene.
In summary, the synthesis described herein18 opens up the way for
a more comprehensive study to be carried out involving a broader
range of derivatives and cell-lines in the hope of identifying syn-
thetic ajoenes with significantly improved anti-cancer activity.
Furthermore, it will allow an appraisal of substituted ajoenes as
chemosensitizing agents for established drugs like cytarabine to-
wards drug-resistant tumors to be carried out.7d
Acknowledgments
We thank the National Research Foundation of South Africa,
Atlantic Philanthropies (bursary to T. Qwebani), the Council for Sci-
entific and Industrial Research (CSIR), the Medical Research Coun-
cil (MRC), and the University of Cape Town for financial support.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2008.08.056.
References and notes
1. (a) Block, E.; Ahmad, S.; Jain, M. K.; Crecely, R. W.; Apitz-Castro, R.; Cruz, M. R. J.
Am. Chem. Soc. 1984, 106, 8295; (b) Block, E.; Ahmad, S.; Catalfamo, J. L.; Jain, M.
K.; Apitz-Castro, R. J. Am. Chem. Soc. 1986, 108, 7045.
2. Gallwitz, H.; Bonse, S.; Martinez-Cruz, A.; Schlichting, I.; Schumacher, K.;
Krauth-Siegel, R. L. J. Med. Chem. 1999, 42, 364.
3. (a) Rendu, F.; Daveloose, D.; Denouzy, J. C.; Bourdeau, N.; Levy-Toledano, S.;
Jain, M. K.; Apitz-Castro, R. Biochem. Pharmacol. 1989, 38, 1321; (b) Apitz-
Castro, R.; Badimon, J.; Badimon, L. Thromb. Res. 1994, 75, 243.
4. (a) Yoshida, S.; Kasuga, S.; Hayashi, N.; Ushiroguchi, T.; Matsuura, H.;
Nakagawa, S. Appl. Environ. Microbiol. 1987, 53, 615; (b) San-Blas, G.; San-
Blas, F.; Gil, F.; Marino, L.; Apitz-Castro, R. Antimicrob. Agents Chemother. 1989,
33, 1642; (c) Urbina, J.; Marchan, E.; Lazardi, K.; Visbal, G.; Apitz-Castro, R.; Gil,
F.; Aguirre, T.; Piras, M. Biochem. Pharmacol. 1993, 45, 2381; (d) Naganawa, R.;
Iwata, N.; Ishikawa, K.; Fukuda, H.; Fujino, T.; Suzuki, A. Appl. Environ. Microbiol.
1996, 62, 4238.
5. (a) Yang, J.-Y.; Della-Fera, M.; Nelson-Dooley, C.; Baile, C. A. Obesity 2006, 14,
388; (b) ang, J.-Y.; Della-Fera, M.; Hausman, D. B.; Baile, C. A. Apoptosis 2007, 12,
1117.
6. (a) San-Blas, G.; San-Blas, F.; Gil, F.; Marino, L.; Apitz-Castro, R. Antimicrob.
Agents Chemother. 1989, 33, 1641; (b) Visbal, G.; San-Blas, G.; Murgich, J.;
Franco, H. Curr. Drug Target Infect. Dis. 2005, 5, 211.
7. (a) Scharfenberg, K.; Wagner, R.; Wagner, K. G. Cancer Lett. 1990, 53, 103; (b)
Li, M.; Min, J. M.; Cui, J. R. Nutr. Cancer 2002, 42, 241; (c) Terrasson, J.; Xu,
B.; Li, M.; Allart, S.; Davignon, J.-L.; Zhang, L.-H.; Wang, K.; Davrinche, C.
Fundam. Clin. Pharmacol. 2007, 21, 281; (d) Hassan, H. T. Leuk. Res. 2004, 28,
667.
8. (a) Pinto, J. T.; Krasnikov, B. F.; Cooper, A. J. L. J. Nutr. 2006, 136, 835S; (b)
Dorant, E.; van den Brandt, P. A.; Goldbohm, R. Carcinogenesis 1996, 17, 477.
9. Sumiyoshi, H.; Wargovich, M. J. Cancer Res. 1990, 50, 5084.
10. (a) Reddy, B. S.; Rao, C. V.; Rivenson, A.; Kelloff, G. Cancer Res. 1993, 53, 3493;
(b) Pinto, J. T.; Qiao, C.; Xing, J.; Rivlin, R. S.; Protomastro, M. L.; Weissler, M. L.;
Tao, Y.; Thaler, H.; Heston, W. D. W. Am. J. Clin. Nutr. 1997, 66, 398.
11. (a) Agarwal, K. Med. Res. Rev. 1996, 16, 111; (b) Lea, M. Adv. Exp. Med. Biol. 1996,
401, 47.
The R1 group was varied in order to investigate the influence of
lipophilicity on the anti-cancer activity against CT-1 transformed
fibroblast cells with the individual ajoene isomers as the reference
standard. Table 1 summarizes the various derivatives synthesized
and their IC50 values against the CT-1 cell-line.
A number of interesting points emerge from analysis of the IC50
values depicted for the small library listed in Table 1. Firstly, in
general, retention of anti-cancer activity for the derivatives at a le-
vel similar to that of the parent ajoenes was noted, supporting the
idea that the central pharmacophore resides in the vinyl disulfide
grouping. Moreover, there are various literature reports on the
anti-cancer activity of Z-ajoene7,12 but no general data on the activ-
ity of the E-isomer. The literature indicates that Z-ajoene is gener-
ally more potent than its E-isomer for a range of diseases, but
interestingly in the current study, the anti-cancer activity of E-ajo-
ene was found to be only marginally less than that of Z-ajoene
12. (a) Dirsch, V. M.; Gerbes, A. L.; Volmar, A. M. Mol. Pharmacol. 1998, 53, 402; (b)
Dirsch, V. M.; Antlsperger, D. S. M.; Hentze, H.; Vollmar, A. M. Leukemia 2002,
16, 74; (c) Li, M.; Ciu, J.-R.; Ye, Y.; Min, J.-M.; Zhang, L.-H.; Wang, K.; Gares, M.;
Cros, J.; Wright, M.; Leung-Track, J. Carcinogenesis 2002, 23, 573; (d) Xu, B.;
Monsarrat, B.; Gaitin, J. E.; Girbal-Neuhauser, E. Fundam. Clin. Pharmacol. 2004,
18, 171; (e) Ledezma, E.; Apitz-Castro, R.; Cardier, J. Cancer Lett. 2004, 206, 35.
13. Nishikawa, T.; Yamada, N.; Hattori, A.; Fukuda, H.; Fujino, T. Biosci. Biotechnol.
Biochem. 2002, 66, 2221.
14. Tilli, C. M. L. J.; Stavast-Kooy, A. J. W.; Vuerstaek, J. D. D.; Thissen, M. R. T. M.;
Krekels, G. A. M.; Ramaekers, F. C. S.; Neumann, H. A. M. Arch. Dermatol. Res.
2003, 295, 117.
15. Experimental detail for the synthetic sequence depicted in Scheme 1 is given in
the Supplementary data.
(17.6 vs 15.5
of the synthesized ajoene analogues, namely entries 3 and 4, (26.7
vs 17.0 M), 6 and 7 (23.1 vs 22.8 M), and 8 and 9 (23.5 vs
21.7 M). A notable exception were the E- and Z-phthalimidopro-
pyl derivatives in entries 10 and 11 (95.9 vs 34.6 M). The data
lM). This trend was also observed (Table 1) for most
l
l
l
l
suggests that a shape-selectivity at a binding site may be an impor-
tant parameter for anti-cancer activity. Only one of the derivatives
as the E/Z-mixture of entry 13 (R1 = p-methoxybenzyl) returned a
16. Kampmeier, J. A.; Chen, G. J. Am. Chem. Soc. 1965, 87, 2608.
17. (a) Scholz, D. Liebigs Ann. Chem. 1984, 2, 259; (b) Binns, M. R.; Goodridge, R. J.;
Haynes, R. K.; Ridley, D. D. Tetrahedron Lett. 1985, 26, 6381.
higher activity (11.2
lM) than that of Z-ajoene (15.5 lM), suggest-
ing this compound to be a lead for further study.
18. A patent on the synthesis and the derivatives prepared has been filed under the
name of the University of Cape Town entitled ‘Compounds useful for the
inhibition of the growth of tumor cells’. South African Patent Application No
2008/06780. Inventors: Roger Hunter, Mohamed Iqbal Parker, Catherine Hart
Kaschula.
Ajoene itself degrades over time when stored at À20°C neat, as
observed by the formation of a less polar TLC product. Based on TLC
studies, in general, R1 substitution of allyl in ajoene offers an in-