Synthesis and biological activity of naphthalene analogues of phenstatins: Naphthylphenstatins

Article abstract of DOI:10.1016/j.bmcl.2007.03.082

Novel phenstatin analogues with a 2-naphthyl moiety combined with either a 2,3,4- or a 3,4,5-trimethoxyphenyl ring have been synthesized, and their tubulin polymerization inhibiting and cytotoxic activities have been evaluated. The 2-naphthyl ring is a be

Full text of DOI:10.1016/j.bmcl.2007.03.082

Bioorganic & Medicinal Chemistry Letters 17 (2007) 3417–3420
Synthesis and biological activity of naphthalene analogues of
phenstatins: Naphthylphenstatins
a
a
b
´
´
´
´
Concepcion Alvarez, Raquel Alvarez, Purificacion Corchete,
a
a,
a,
*
*
´
´
´
Concepcion Perez-Melero, Rafael Pelaez and Manuel Medarde
aDepartament de Quı´mica Farmace´utica, Facultad de Farmacia, Universidad de Salamanca, Campus Miguel de Unamuno,
37007 Salamanca, Spain
bDepartamento de Fisiologı´a Vegetal, Facultad de Farmacia, Universidad de Salamanca, Campus Miguel de Unamuno,
37007 Salamanca, Spain
Received 27 February 2007; revised 27 March 2007; accepted 27 March 2007
Available online 30 March 2007
Abstract—Novel phenstatin analogues with a 2-naphthyl moiety combined with either a 2,3,4- or a 3,4,5-trimethoxyphenyl ring have
been synthesized, and their tubulin polymerization inhibiting and cytotoxic activities have been evaluated. The 2-naphthyl ring
is a better replacement for the 3-hydroxy-4-methoxyphenyl ring in the phenstatin series than in the combretastatin series. For
the naphthylphenstatins, the carbonyl is required, and the preferred orientation of the trimethoxyphenyl ring is the one found in
combretastatins.
Ó 2007 Elsevier Ltd. All rights reserved.
Ligands binding at the colchicine site of tubulin, such as
colchicine itself, podophyllotoxin, the combretastatins
and the phenstatins, are receiving much attention due
to their interesting antitumour, antiangiogenic and anti-
parasitic activities. Several such ligands are now under-
going clinical trials, for example, the combretastatin A4
phosphate prodrug.1,2 Most of these ligands share com-
mon structural features, often summarized in the struc-
ture of combretastatin A4, such as two non-coplanar
oxygenated aromatic rings (typically one trimethoxy-
phenyl and a 3-hydroxy-4-methoxyphenyl or related
ring). Certain variability is allowed for the size of the
bridge between the two phenyl rings, which ranges from
zero to more than four atoms (Fig. 1).3
analogues, we have also explored the consequences of
the introduction of nitrogen atoms on the 2-naphthyl
ring of naphthylcombretastatins and found that in quin-
oline and isoquinoline analogues of combretastatins the
heteroatom has an effect on the activity which is depen-
dent on its position.7 The moderate solubility increase
OH
MeO
O
O
MeO
MeO
NHAc
O
O
MeO
MeO
O
MeO
OH
OMe
MeO
OMe
OMe
OMe
Colchicine
Podophyllotoxin
Combretastatin A4
We have previously shown that the 3-hydroxy-4-meth-
oxyphenyl ring of combretastatin A4 can be replaced
by a 2-naphthyl ring (Fig. 2) with no substantial loss
of potency. On the other hand, replacement of the
3,4,5-trimethoxyphenyl ring or the introduction of a
1-naphthyl ring is detrimental for the activity.4–6 In an
attempt to increase the aqueous solubility of the naphthyl
Figure 1. Representative colchicine site ligands with different bridge
sizes.
MeO
O
X
Y
O
MeO
MeO
OH
MeO
MeO
MeO
MeO
OMe
MeO
Naphthylcombretastatin
Phenstatin
Naphthylphenstatins
X or Y = MeO
Keywords: Antitumour; Tubulin; Combretastatins; Phenstatins; Cyto-
toxicity; Synthesis.
*
Figure 2. Structures of naphthylcombretastatin, phenstatin and
designed naphthylphenstatins.
Corresponding authors. Tel.: +34 923 294528; fax: +34 923 294515
(R.P.); e-mail addresses: pelaez@usal.es; medarde@usal.es
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.03.082

ꢀ
3418
C. Alvarez et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3417–3420
achieved and the more cumbersome handling make such
a modification less attractive.
NBS
Br
MeO
OMe
OMe
CCl4
68%
MeO
OMe
OMe
Recently, much synthetic effort has been devoted to the
bisarylketones named phenstatins (Fig. 2) and related
analogues, which show substantial differences in their
structure–activity relationships when compared to the
combretastatins.8–10 In this paper, we disclose our preli-
minary results on the synthesis and biological activity of
2-naphthophenones (Fig. 2) bearing a 2-naphthyl ring
(in substitution of the 3-hydroxy-4-methoxyphenyl or
related ring) combined with either a 3,4,5-trimethoxy-
phenyl ring (such as the one present in combretastatins
and podophyllotoxin) or a 2,3,4-trimethoxyphenyl ring
(such as that found in colchicine). We have termed these
compounds naphthylphenstatins, in analogy to the
naphthylcombretastatins.
nBuLi
THF
- 40ºC - r.t.
MeO
X
MeO
X
X
+
MeO
OMe
OMe
MeO
ox.
1 X = OH, H
2 X = O
3 X = OH, H
4 X = O
ox.
Scheme 2. Formation of mono- and dialkylated products.
Acetylation of the resulting alcohols with acetic anhy-
dride and pyridine occurred in lower than usual yields
(around 70%), probably due to the ease of formation
of the corresponding bisbenzilic cation and its transfor-
mation into byproducts during chromatography. Oxida-
tion of the alcohols was attempted with PCC, according
to literature procedures, but low yields of the desired ke-
tones were obtained. When the crude reaction mixture
was treated with PCC in CH2Cl2, the only oxidized
products were 9 and 10. In our hands, the best condi-
tions found for the oxidation were: KMnO4 and a phase
transfer catalyst (tetrabutylammonium bromide) with
wet CH2Cl2 as solvent. Under these conditions, the ke-
tones could be readily isolated after oxidation of the
unpurified alcohols in 70% or better yields.
We planned the synthesis of the naphthylphenstatins
based on the reaction of an organometallic (formed by
transmetallation of the corresponding bromoderivative)
with an aldehyde (Scheme 1). Of the two possible pairs
of starting materials, we set out to start the synthesis
with the halogenated trimethoxyphenyl rings and com-
mercial 2-naphthaldehyde, a strategy we thought might
be more convenient for the later preparation of related
analogues, such as the quinolines, and which had previ-
ously shown itself successful.
Accordingly (Scheme 2), we brominated 1,2,3-trimeth-
oxybenzene with NBS in carbon tetrachloride. The bro-
mide was transmetallated with n-BuLi and treated with
2-naphthaldehyde. We obtained mixtures of mono- (1)
and dialkylated (3) products even in the presence of ex-
cess of bromide. These byproducts could not be chro-
matographically resolved, not even after oxidation to
the corresponding ketones (2 and 4), rendering this
route inconvenient for this and related synthesis.
The synthesized compounds were assayed for tubulin
polymerization inhibition6 at a single concentration of
20 or 40 lM (Table 1). Compound 7 was the only active,
showing an IC50 one order of magnitude lower than its
naphthylcombretastatin analogue (Table 1) and three
times lower than CA-4. Literature values for tubulin
polymerization inhibition for phenstatin show that it is
only one to two times more potent than CA-4.8–10 Cross
comparison of these data suggests that 7 is at least equi-
potent to phenstatin. In combretastatins, replacement of
the 3-hydroxy-4-methoxyphenyl ring by the 2-naphthyl
one decreases somewhat the potency. In phenstatins,
no reduction is observed, indicating that the 2-naphthyl
is an even better replacement in these structures. Neither
the alcohols (1 and 6), nor the acetates (5 and 8) were
active. The active orientation for the trimethoxyphenyl
ring corresponds to that of combretastatins and podo-
phyllotoxin. The lack of activity for the one-carbon
bridged naphthylphenstatins with a 2,3,4-trimethoxy-
phenyl ring agrees with our earlier observation on the
same ring combination (2-naphthyl and 2,3,4-trimeth-
oxyphenyl) in zero-carbon bridged analogues6 and indi-
cates that it is not a good replacement in colchicine-like
compounds.
The alternative approach, starting from 2-bromonaph-
thalene and the isomeric trimethoxybenzaldehydes,
was then attempted (Scheme 3). Initially, treatment of
the 2-naphthyl bromide with Mg in THF was selected
for the formation of the organometallic species, in order
to avoid direct addition of the organolithic reagent to
the aldehyde. However, even under inert atmosphere,
mixtures of alcohols and ketones were formed. There-
fore, we chose n-BuLi for the transmetallation reaction.
With this reagent, the major byproducts (9 and 10) were
formed by direct addition of n-BuLi to the aldehyde, but
they were easily removed by chromatography.
Br
X
X
Y
Z
CHO
Br
MeO
MeO
Y
OMe
OMe
CHO
X
Z = O
Z = H, OH
The synthesized compounds were assayed following a
described procedure6 against several cancer cell lines,
including human cervix epitheloid carcinoma HeLa,
human lung carcinoma A-549 and human colon adeno-
Naphthylphenstatins
Y
OMe
OMe
X or Y = MeO
Scheme 1. Retrosynthetic analysis for naphthylphenstatins.

ꢀ
C. Alvarez et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3417–3420
3419
CHO
HO
X
Y
OH
MeO
OMe
OMe
X
Br
MeO
MeO
nBuLi
- 40ºC - r.t.
Y
OMe
OMe
9 X = OMe, Y = H
10 X = H, Y = OMe
1 X = OMe, Y = H
6 X = H, Y = OMe
O
O
Ac2O
pyr
X
Y
KMnO4
Bu4NBr
MeO
MeO
X
Y
O
MeO
MeO
5 X = OMe, Y = H
8 X = H, Y = OMe
2 X = OMe, Y = H
7 X = H, Y = OMe
Scheme 3. Synthesis of naphthylphenstatins.
Table 1. Tubulin polymerization inhibition assay results for com-
pounds 1, 2, and 5–8
tubulin inhibitory activity but lower cytotoxicity than
their combretastatin counterparts.11 This apparent con-
tradiction might be explained by the different concentra-
tion ranges at which the two measured effects are
produced and/or by differences in permeability. In such
a context, the small difference in cytotoxicity observed
between 7 and its naphthylcombretastatin analogue
again suggests a more favourable substitution of the 3-
hydroxy-4-methoxyphenyl ring by the 2-naphthyl in this
family of compounds than in the combretastatins. As
different trends of structure–activity relationships for
both families have been previously found,8–10 these
results suggest that the 2-naphthyl ring represents a
structural intersection between the structure activity
relationships of both families of compounds.
b
Compound
% inhibition
of tubulin
polymerization
(lM)a
IPT IC50 (lM)
1
0 (20)
4 (20)
26 (20)
0 (40)
100 (40)
5 (20)
100 (20)
—
>20
>20
2
5
>20
>20
6
7
1.1
>20
8
CA-4
3 (1–3)
10
(0.4–1.5)
Naphthylcombretastatin
Phenstatin
—
a Values are means of at least three experiments (concentration
assayed).
b Concentration inhibiting 50% of 1.0 mg/mL microtubular protein
polymerization (literature values).
In summary, we have synthesized new naphthylphensta-
tins with two trimethoxyphenyl ring orientations. The
most potent compound has a ketone on the bridge and
a 3,4,5-trimethoxyphenyl, equivalent to that found in
combretastatins and podophyllotoxin. On the other
hand, a 2,3,4-trimethoxyphenyl, equivalent to that
found in colchicine, turned out to be suboptimal. These
results suggest that the 3D structure of podophyllotoxin
in complex with tubulin might be a better model in
which to perform docking experiments than the complex
with colchicine. When compared to this family of com-
pounds, the 2-naphthyl ring seems to be an even better
substitution for the 3-hydroxy-4-methoxyphenyl ring
than in the combretastatin family.
carcinoma HT-29, and compared with naphthylcombre-
tastatin and CA-4 (Table 2). The most sensitive cell line
was HeLa, in which all the compounds except 1 showed
micromolar to submicromolar (compound 7) potencies.
The other two cell lines were more resistant to the com-
pounds, again being 7 the only one with remarkable
activity. Opposite to the tubulin polymerization inhibi-
tory activity, 7 is less cytotoxic than CA-4 and even than
naphthylcombretastatin. These observations agree well
with previous work on phenstatins: they display higher
Acknowledgments
Table 2. Cytotoxicity results (IC50/lM) for compounds 1, 2, 5, 7, and 8
Compound
HeLaa
A549a
HT29a
We thank the MEC (Ref CTQ2004-00369/BQU), the
´
Junta de Castilla y Leon (Refs. SA090A06 and
SA009A06) and the EU (Structural Funds) for the
1
>10
1.7
>10
>10
>10
0.25
>10
0.003
0.02
>10
>10
>10
—
2
5
2.7
0.042
3.8
´
financial support. R.A. thanks the Fundacion Ramon
Areces and the Spanish Ministerio de Educacion y Cul-
´
7
´
8
CA-4
Naphthylcombretastatin
>10
0.003
0.02
tura for predoctoral Fellowships. The authors are grate-
ful to CIDTA for equipment facilities and to Janis de la
Iglesia, Consuelo Gajate and Faustino Mollinedo for
assistance with the biological assays.
0.003
0.02
a Values are means of three experiments.

ꢀ
3420
C. Alvarez et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3417–3420
7. Perez-Melero, C.; Maya, A. B.; del Rey, B.; Pelaez, R.;
Caballero, E.; Medarde, M. Bioorg. Med. Chem. Lett.
2004, 14, 3771.
8. Liou, J. P.; Chang, J. Y.; Chang, C. W.; Chang, C. Y.;
Mahindroo, N.; Kuo, F. M.; Hsieh, H. P. J. Med. Chem.
2004, 47, 2897.
References and notes
1. Tron, G. C.; Pirali, T.; Sorba, G.; Pagliai, F.; Busacca, S.;
Genazzani, A. A. J. Med. Chem. 2006, 49, 3033.
2. Hsieh, H. P.; Liou, J. P.; Mahindroo, N. Curr. Pharm.
Des. 2005, 11, 1655.
9. Liou, J. P.; Chang, C. W.; Song, J. S.; Yang, Y. N.;
Yeh, C. F.; Tseng, H. Y.; Lo, Y. K.; Chang, Y. L.;
Chang, C. M.; Hsieh, H. P. J. Med. Chem. 2002, 45,
2556.
10. Liou, J. P.; Chang, Y. L.; Kuo, F. M.; Chang, C. W.;
Tseng, H. Y.; Wang, C. C.; Yang, Y. N.; Chang, J.
Y.; Lee, S. J.; Hsieh, H. P. J. Med. Chem. 2004, 47,
4247.
3. Nguyen, T. L.; McGrath, C.; Hermone, A. R.; Burnett, J.
C.; Zaharevitz, D. W.; Day, B. W.; Wipf, P.; Hamel, E.;
Gussio, R. J. Med. Chem. 2005, 48, 6107.
4. Maya, A. B.; del Rey, B.; Lamamie de Clairac, R. P.;
Caballero, E.; Barasoain, I.; Andreu, J. M.; Medarde, M.
Bioorg. Med. Chem. Lett. 2000, 10, 2549.
5. Medarde, M.; Maya, A. B.; Perez-Melero, C. J. Enzyme
Inhib. Med. Chem. 2004, 19, 521.
11. Pettit, G. R.; Toki, B.; Herald, D. L.; Verdier-Pinard, P.;
Boyd, M. R.; Hamel, E.; Pettit, R. K. J. Med. Chem. 1998,
41, 1688.
6. Maya, A. B.; Perez-Melero, C.; Mateo, C.; Alonso, D.;
Fernandez, J. L.; Gajate, C.; Mollinedo, F.; Pelaez, R.;
Caballero, E.; Medarde, M. J. Med. Chem. 2005, 48, 556.