N.-H. Nam et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1955–1958
Table 1. Cytotoxicity of synthesized compounds in tumor cell linesa
1957
depleted under continuous exposure of 8e confirms
normal movement of cells through these phases even in
the presence of a drug. Thus, 8e appears to be a rapid
and effective G2–M phase blocker that does not affect
cell cycle progression through the G1 or S phases or
G1–S transitional point. These results strongly suggest
that 8e may interfere with a dynamic engine of cellular
mitosis, whose critical components include microtubule
bundles. Thus, compound 8e and its close analogues
likely act by the same mechanism of combretastatins by
binding to tubulins.
Compd
Cytotoxicity (IC50,
b ng/mL)
B16
1521
HCT116
A431
8a
8b
8c
8e
9a
10a
21
31
12
8
13
45nt
29
13
8
11
34
9
9
ntc
39
12a
15
13
Adriamycin
>1000
>1000
13
>1000
>1000
16
nt
nt
nt
117
90
110
In the anti-cancer drugs development arena, it has been
widely recognized that those compounds that possess
multiple mechanisms of action may be more effective in
killing tumor cells and prevention of multi-drug resis-
tance. Dimmock et al.13 in a recent publication also
proposed that interference at multiple points in biologi-
cal cascades or successive attacks of cellular constituents
may be highly deleterious to malignant cells in pref-
erence to normal cells. In this regard, we attempted to
introduce a second a,b-olefinic moiety in 8a at the 4,5-
position. The resulting 9a possesses a newly formed
conjugation between the 4,5-olefine and 1-one group.
This a,b-unsaturated ketone is sterically unhindered and
therefore may act as a Michael acceptor, and thus,
compound 9a could be expected to be an alkylating
agent also. Indeed, the Michael acceptor is a moiety
often employed as a powerful tool in anti-cancer drugs
design.14 Interestingly, compound 9a was shown to be
about 1.5-fold more potent than 8a. Though this dis-
crepancy was somewhat narrow, it did support the per-
spectives presented above, and further exploration in
this direction in anti-cancer drugs design may prove to
be beneficial.
aB16, murine melanoma; HCT116, human colon carcinoma; A431,
human epidermal carcinoma.
bThe concentration required to reduce the cell growth by 50%.
cNot tested.
lactone oxygen atom in 13 seemed to be not very
important for its cytotoxicity. In contrast, removal of
the carbonyl group, as seen in compound 15, proved to
be detrimental for the bioactivity of this compound;
compound 15 were found inactive up to 1000 ng/mL in
both cell lines. Furthermore, reduction of this carbonyl
moiety to hydroxyl group led to a racemic mixture 10a
with substantially reduced cytotoxicity. Therefore, no
attempt to separate the two enantiomers was further
elaborated. Scission of both lactone oxygen and car-
bonyl group in compound 13 resulted in compound 12a
with lost bioactivity. Thus, it appears that the carbonyl
group might play an important role in the binding of
these compounds (8a–e, 13) with appropriate receptors,
likely via hydrogen bonding.
The combretastatins have been widely known as very
potent anti-mitotic anti-tumor agents. In this study,
compounds 8a–e have been demonstrated to share
common features of the combretastatin class in terms of
structure–activity relationships, and it is expected that
those compounds in this study also act by binding to
tubulins, similar to combretastatins. Though the tubulin
binding experiments are pending, we have alternately
performed cell cycle analyses to examine the effect of
one representative compound, compound 8e, on the
progress of cell cycle events to glean some insights into
mechanism(s) mediating the profound cytotoxicity of
this series. Cell cycle analysis12 of CEM (leukemic) cells
was performed after 0–12 h incubation with 30 ng/mL of
8e. It was found that treatment with 8e for as little as 2 h
led to the increase in the number of G2-M phase cells. It
was also observed that G2-M phase cells accumulated at
the expense of G1 cells and the number of G1 cells was
almost depleted after 8 h. Progressive depletion of S
phase then ensued beginning at 10 h because of cessa-
tion of new cells entering S phase from G1 phase.
Depletion of S phase cells was almost complete by 12 h,
at which time almost all of the liable cells were blocked
in G2-M phase. In addition, increase in the number of
hypodiploid events was seen after 8–10 h, indicating
apoptosis of cells after a prolonged G2-M blockage. The
fact that both G1 and S phases became progressively
Overall, among eight compounds (8a–8e, 9a, 10a, 12a,
15) of four new types of combretastatin analogues
investigated in this work we found six compounds (8a–
8e, 9a, 10a) with potent bioactivities (IC50 values of 8–
45ng/mL). However, these compounds seem to be less
potent than those combretoxazolones 2 and 3 reported
previously.7 For example, the IC50 values of 8a, 1
(R=4-OCH3), and 2 (R=4-OCH3) were 15, 2.8 and
5.4 ng/mL, respectively. One possible reason could be
that combretoxazolones 1 and 2 possess more flexible
conformations, which may allow the correct positioning
of the molecules for a facile binding at the active site
residues of receptors.
Preliminary in vivo evaluation of a representative com-
pound 8e, which offered the best water solubility
(2.1 mg/mL, the hydrochloride salt form), showed that
this compound, when administered at 40 mg/kg/day
(maximum injectable dose) into BDF1 mice inoculated
with Lewis lung carcinoma cells,15 inhibited the growth
of tumor mass by 59%, compared to 78% of etoposide
(36 mg/kg/day, used as a positive control). Of note
however, this compound showed little toxicity com-
pared to etoposide, as evidenced by lower body weight
loss in mice treated with 8e. Thus, elevation of dosage is
possible given that its water solubility is improved. A
series of prodrugs of this compound is being pursued in
Nam, Nguyen-Hai
