Structure-activity relationship studies on a novel class of antiproliferative agents derived from Lavendustin A. Part I: Ring A modifications

Article abstract of DOI:10.1016/j.bmc.2008.07.039

The potent antiproliferative agent SDZ LAP 977, which has shown efficacy in a clinical proof of concept study in actinic keratosis patients, has been previously demonstrated to block the cell cycle in mitosis. In the present study, we further explored the mode of action: SDZ LAP 977 binds to the "colchicine binding site" on tubulin and, thus, inhibits tubulin polymerization in vitro. Moreover, we established structure-activity relationships for the effect of modifications in the 2,5-dimethoxyphenyl moiety ("ring A") of the molecule on in vitro antiproliferative activity.

Full text of DOI:10.1016/j.bmc.2008.07.039

Bioorganic & Medicinal Chemistry 16 (2008) 7552–7560
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Structure–activity relationship studies on a novel class of antiproliferative
agents derived from Lavendustin A. Part I: Ring A modifications
*
Peter Nussbaumer , Anthony P. Winiski
Novartis Institutes for BioMedical Research, Brunnerstrasse 59, A-1230 Vienna, Austria
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 5 February 2008
Revised 4 July 2008
Accepted 16 July 2008
Available online 20 July 2008
The potent antiproliferative agent SDZ LAP 977, which has shown efficacy in a clinical proof of concept
study in actinic keratosis patients, has been previously demonstrated to block the cell cycle in mitosis.
In the present study, we further explored the mode of action: SDZ LAP 977 binds to the ‘‘colchicine bind-
ing site” on tubulin and, thus, inhibits tubulin polymerization in vitro. Moreover, we established struc-
ture–activity relationships for the effect of modifications in the 2,5-dimethoxyphenyl moiety (‘‘ring A”)
of the molecule on in vitro antiproliferative activity.
Keywords:
Antiproliferative
Inhibitors
Ó 2008 Elsevier Ltd. All rights reserved.
Structure–activity relationship
Tubulin
1. Introduction
OH
OH
OH
OMe
A
HO
HO
MeO
During a medicinal chemistry program on the natural product
Lavendustin A, we discovered the novel antiproliferative agent 5-
[2-(2,5-dimethoxyphenyl)ethyl]-2-hydroxybenzoic acid methyl
ester (1, SDZ LAP 977; Fig. 1).1 Compound 1 potently inhibits the
proliferation of the human keratinocyte cell line HaCaT and several
tumor cell lines in vitro; it was shown to be effective in vivo in a
model of epidermal hyperplasia after topical application2 and in
selected tumor models after oral and intravenous administration3;
furthermore, 1 showed efficacy in a clinical proof-of-concept study
in actinic keratosis patients.4 In contrast to Lavendustin A and the
fully demethylated SDZ LAP 977 analog 2 (Fig. 1), which are both
potent inhibitors of the epidermal growth factor receptor-associ-
ated tyrosine kinase,1,5,6 compound 1 acts by blocking the cell cycle
in mitosis.1,3,7 Investigations on the antimitotic effect revealed that
the block of the cell cycle in mitosis is due to interference of 1 with
the microtubules of the mitotic spindle apparatus. New results on
the mode of action are presented and discussed in this manuscript.
Next, we established detailed structure–activity relationships
(SAR) based on the inhibitory potencies of the compounds on the
proliferation of the human keratinocyte cell line HaCaT and spot-
checked with selected compounds for the antimitotic mode of ac-
tion. For easier interpretation of SAR results, the lead molecule 1
was dissected into three parts (Fig. 1): the two aromatic ring sys-
N
B
COOH
COOH
COOMe
OH
OH
OH
1 (SDZ LAP 977)
2
Lavendustin A
TK inhibitor
antimitotic
TK inhibitor
Figure 1. Structures and mode of action of Lavendustin A compared to derivatives 1
and 2.
tems (rings A and B) and the spacer connecting them. In general,
we found good correlation between series of analogs with the same
modification in one part of the molecule, but differing at other sites
of the molecule. The present manuscript describes the synthesis
and the SAR of compounds with modifications in ring A.
2. Chemistry
Several methods for the synthesis of 1 and analogs were devel-
oped (Scheme 1; for substitution pattern and compound number
assignment of test compounds see also Table 1). The most gener-
ally applicable procedure was coupling of the two aromatic ring
systems (rings A and B) by a Wittig reaction, followed by catalytic
hydrogenation of the stilbene intermediates (Method A, Scheme 1).
This route was compatible with various substituents, and, as start-
ing materials, A- and B-ring derivatives could optionally be used
either as phosphonium (3) or as benzaldehyde components (4).
* Corresponding author. Current Address: Lead Discovery Center GmbH, Emil-
Figge-Strasse 76a, D-44227 Dortmund. Tel.: +49 231 9742 7006; fax: +49 231 9742
7039.
E-mail address: nussbaumer@lead-discovery.de (P. Nussbaumer).
0968-0896/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.07.039

P. Nussbaumer, A. P. Winiski / Bioorg. Med. Chem. 16 (2008) 7552–7560
7553
4
Method A:
R3
5
-
Br
P+Ph3
2
6
4
CHO
4
R3
5
R3
5
6
a - c
a, b
2
6
2
+
+
COOMe
+
COOMe
OAc
CHO
OH
-
P Ph3
Br
COOMe
OH
3a
4b
4c
4d
4a
R = 2-OMe
R = 3-OMe
R = 2,3-diOMe
R = 2,6-diOMe
R = 3,4-OCH2O-
R = 2,5-diOEt
R = 2-OMe, 5-Cl
R = 2-OMe, 5-NO2
R = 2,5-diEt
3n
R = 2,4-diOMe
R = 2-Br, 5-OMe
R = 2-OMe, 5-SMe
R = 2-OMe
R = 3-OMe
3b
3c
3d
3e
3f
3g
3h
3i
5a
5b
5c
5d
5e
5f
5m
5n
5o
R = 2,3-diOMe
R = 2,4-diOMe
R = 2,6-diOMe
R = 3,4-OCH2O-
R = 2,5-diOEt
4e R = 2-NMe2, 5-OMe
R = 2-OMe, 5-Cl
R = 2-Br, 5-OMe
3j
R = 2-Et, 6-OMe
3k R = 2,3,6-triOMe
5p R = 2-OMe, 5-SMe
R = 3,4,5-triOMe
R = 2-NMe2, 5-OMe
R = 2,5-diEt
R = 2-Et, 6-OMe
R = 2-OMe, 5-NH2
R = 3-OMe, 4-t-butyl
3l
3m
5q
5w
5x
5y
5z
R = 3-OMe, 4- -butyl
t
Method F
d
5r
R = 2-OMe, 5-NMe2
6a R = 2,3,6-triOMe
6b R = 3,4,5-triOMe
Method B:
HO
Method D:
O
OH
O Ac
OMe
e
g
Ac O
O
O
5j
5i
5a
5s
Method C:
O
O
OMe
O
O
f
g
HO
EtO
+
O
5b
5t
5g 2-OH, 5-OMe
5h 2-OMe, 5-OH
5k 2-OEt, 5-OMe
5l 2-OMe, 5-OEt
O
O
O
Method E:
O
Et
g
O
O
+
b
MeO
MeO
COOMe
COOMe
COOMe
OH
OH
OH
5s 2-OMe, 5-COCH3
5t 2-COCH3, 5-OMe
5v 2-OMe, 5-Et
5u 2-Et, 5-OMe
5z
6c
5t
Scheme 1. Methods A–F used for the synthesis of test compounds 5 and 6. Reagents: (a) LDA/THF; (b) H2/Pd/C; (c) H+ or OHÀ; (d) CH2O/NaH2PO3; (e) Ac2O/Et3N; (f) K2CO3/
EtI; (g) AcCl/AlCl3.
As reported for the synthesis of compound 1,1 unprotected 5-form-
ylsalicylic acid methylester (4a) could be used successfully in the
Wittig reaction, when lithium diisopropylamide (LDA) in tetrahy-
drofuran prepared from a highly concentrated solution of n-butyl-
lithium (10 M solution in hexane, Sigma–Aldrich) was employed as
base to deprotonate the phosphonium salts 3a–m. The use of other
bases or a higher percentage of hexane in the solvent mixture (e.g.,
resulting from the preparation of LDA using less concentrated solu-
tions of n-butyllithium) gave substantially lower yields of prod-
ucts, probably due to insufficient solubility of the phenolate
O-acetyl protecting group resulting from 3n was either removed
by acidic (2 N HCl) or mild basic hydrolysis (in some experiments
the O-acetyl group was already partially cleaved during the Wittig
reaction and work-up). Different E/Z-ratios as well as varying
yields of stilbenes were obtained from the Wittig reaction, depend-
ing on the substitution pattern. For instance, the E-isomer was al-
most exclusively formed in 2,6-disubstituted derivatives. In some
cases, the pure, single E- and Z-isomers could be isolated by chro-
matography on silica gel, but usually they were only separated
from the triphenylphosphine oxide and other reaction by-products
and used as mixture in the subsequent hydrogenation step. n-Buty-
lesters of the E/Z-stilbenes were identified as by-products in sev-
eral experiments, in particular, when LDA was prepared from
10 M of n-butyllithium. The formation of the n-butylesters was
probably due to transesterification of the corresponding methy-
lesters by lithium butoxide, an impurity in the butyllithium
intermediate. Alternatively,
a protection/deprotection protocol
for the salicylic hydroxy group in the starting material could be ap-
plied, as it was exercised when the phosphonium component 3n
was the starting material in the Wittig reaction (Scheme 1). In
these cases only a slight excess of LDA (1.2 equiv) was used, in
contrast to the reactions starting from 4a (3 equiv of LDA). The

7554
P. Nussbaumer, A. P. Winiski / Bioorg. Med. Chem. 16 (2008) 7552–7560
O
N+
+
b,c
d, e
a
3j
O
OH
+
OH
O
Br
OH
COOEt
COOEt
COOEt
Scheme 2. Synthetic route toward 3j, intermediate in the synthesis of 5x. Reagents: (a) Et3N; (b) NaH/MeI; (c) DiBAH/toluene; (d) DiPhos/Br2; (e) Ph3P/toluene.
reagent generated by air oxidation. Novartis Process Research and
Development developed a reproducible plant process for the pro-
duction of 1 using a Horner–Wadsworth–Emmons-type olefination
as the key step and without the need for a column chromatogra-
phy.8 When the stilbene product resulting from the Wittig reaction
of 3 h and 4a was hydrogenated, the double bond and the nitro
group were reduced concomitantly to give 5y which was not tested
but further converted to 5r.
of phosphonium salt 3j, a procedure reported for the synthesis
of alkylaryl disubstituted ethyl salicylates9 was adapted. Thus,
1-(3-ethoxycarbonyl-2-oxopropyl)-pyridinium bromide was re-
acted with ethyl vinyl ketone in the presence of triethylamine
to give a mixture of the regioisomeric 4-ethyl- and 6-ethylsalicy-
lates (Scheme 2), which were separated by chromatography on
silica gel. The 6-ethyl product was then transformed into 3j by
four conventional steps.
Several derivatives were prepared from other test compounds
by straightforward functional group transformations. This in-
cludes O-acetylation of 5i to produce compound 5j (Method B,
Scheme 1), selective O-alkylation in 5g,h to give the ethoxy ana-
logs 5k,l (Method C, Scheme 1), catalytic hydrogenation of ace-
tophenones 5s,t over palladium on charcoal in the preparation
of analogs 5u,v (Method E, Scheme 1), and N,N-dimethylation
of the amino function in 5y to yield substance 5r (Method F,
Scheme 1).
Some characteristic physicochemical properties of the new Lav-
endustin derivatives 5 and 6 are listed in Table 2.
3. Results and discussion of biological activities
A previous study focusing on the structural requirements for
the switch in mode of action from the tyrosine kinase inhibitor
Lavendustin A to 1 revealed that the presence of one methoxy
group in the 2,5-disubstituted ring A is sufficient for inducing mi-
totic arrest of keratinocytes.7 Already from this limited study, a
trend for the essential elements needed for potent antiproliferative
activity of 1-related analogs was perceivable. For deeper insight
and the establishment of detailed SAR, series of additional analogs
were synthesized and tested for antiproliferative activity in the
keratinocyte cell line HaCaT. In addition, their antimitotic potential
was determined to confirm the antimitotic mode of action. The
present manuscript deals with the SAR obtained by modifications
in ring A.
The first series of derivatives (5a–f) was designed in order to
probe the importance of the position of the methoxy substituents
in 1 for antiproliferative activity. Both analogs with a single meth-
oxy group in positions 2 and 3 (numbering with respect to the po-
sition of the spacer), respectively, show drastically reduced
antiproliferative activity (5a and 5b, Table 1). Despite their low
antiproliferative activity, both monomethoxy compounds are also
antimitotic. The results obtained for the 2,3-dimethoxy (5c), the
2,4-dimethoxy (5d), the 2,6-dimethoxy (5e), and the 3,4-methyl-
enedioxy analog (5f)—all possible variations of a dialkoxy substitu-
tion pattern of 1—demonstrate that, in addition to the 2,5-, only
the 2,6-disubstitution pattern is tolerated for high activity. With
an IC50 value of 110 nM, the 2,6-dimethoxy analog 5e is about 2-
fold less active than 1. Interestingly, those derivatives, which have
the two (alk)oxy substituents in a neighboring position, that is, 5c
Methoxyacetophenones 5s,t were generally synthesized by
Friedel–Crafts acylation (Method D, Scheme 1). Whereas the
reaction of 5a with acetyl chloride/aluminum chloride afforded
the respective acetylated analogs 5s in good yield, the same
procedure gave a mixture of two isomeric products (2-acetyl-
5-methoxy and 4-acetyl-3-methoxy) when starting from 5b.
The isomers could not be easily separated, and the desired
para-substituted product 5t could only be isolated in small
amount after selective Lewis acid-mediated transformation of
the o-methoxyacetophenone into the corresponding o-hydroxy-
acetophenone by chromatographic separation from this new
product. For the synthesis of a larger sample of this biologically
interesting substance, a new method was developed using the
t-butyl group as protecting group (Scheme 1). The t-butyl group
was chosen to block the o-position against acetylation, thus,
avoiding the formation of the unwanted regioisomer. Further-
more, we anticipated it to be cleaved afterwards by a retro-Fri-
edel–Crafts alkylation reaction. Thus, compound 5z, prepared
via Method A, was treated with 3 equiv of acetyl chloride
and aluminum chloride under usual Friedel–Crafts acylation
conditions. Two products were isolated: compound 6c from
the acetylation reaction and the desired analog 5t formed by
subsequent dealkylation of 6c. The formation of 5t under the
reaction conditions confirmed the envisaged strategy and indi-
cated the feasibility of a one-pot procedure for its synthesis.
Several additional experiments were run at various reaction
temperatures and with different molar equivalents of aluminum
chloride. The optimum reaction conditions for the one-pot con-
version of 5z into 5t were found to be rapid addition of 4 equiv
of aluminum chloride to a solution of 1 equiv of 5z and 2.2
equiv of acetyl chloride in dichloroethane without cooling, fol-
lowed by maintaining the reaction temperature at about 35 °C
for 1 h by external heating. Thus, 5t was obtained directly from
5z in 61% isolated yield, which was even superior to the yield
obtained for the corresponding two-step procedure, that is, in
synthesis of 6c under milder conditions (85% yield) and subse-
quent transformation into 5t by treatment with aluminum chlo-
ride (68% yield).
and 5f, do not exert an antimitotic activity up to 100 lM. This find-
ing is probably not a result of their poor antiproliferative potency
or a matter of the detection limit of the assay, since with other
derivatives induction of mitotic arrest is usually observed at con-
centrations near their antiproliferative IC50 values. For example,
compound 5b shows a similarly low antiproliferative activity as
5c and 5f, but, in contrast, a clear antimitotic activity below
100 lM. Obviously, compounds 5c and 5f exert their antiprolifera-
tive effects via another mechanism of action.
Analogs 5g–m represent a subseries of compounds in which the
2,5- and 2,6-dioxy substitution pattern, respectively, is kept con-
stant. As previously reported,7 the monodemethyl derivatives of
1 (i.e., substances 5g,h) are substantially less active than the parent
compound but still antimitotic. When both methoxy groups are
demethylated (5i), the antiproliferative potency is further reduced
and the ability to block HaCaT cells in mitosis is lost. Thus,
The key starting materials for all test compounds were either
commercially available or synthesized in analogy to standard
procedures (see Section 4). For the preparation of a precursor

P. Nussbaumer, A. P. Winiski / Bioorg. Med. Chem. 16 (2008) 7552–7560
7555
Table 1
In vitro antiproliferative (IC50) and antimitotic activity of Lavendustin derivatives 1, 5, and 6
Compound
Substitution pattern
Pos. 4
HaCaT cells
Pos. 2
Pos. 3
Pos. 5
Pos. 6
Antiproliferative IC50 (nM)
Antimitotic activity
Qualitativea
Quantitative (
l
M)b
1
5a
5b
5c
5d
5e
5f
5g
5h
5i
5j
5k
5l
5m
5n
5o
5p
5q
5r
OMe
OMe
H
OMe
OMe
OMe
H
OH
OMe
OH
OAc
OEt
OMe
OEt
OMe
Br
OMe
NMe2
OMe
OMe
COCH3
Et
OMe
Et
OMe
OMe
H
H
H
H
OMe
H
H
H
H
H
OMe
H
OMe
H
OMe
H
H
H
H
OMe
OH
H
H
H
H
H
OMe
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Et
OMe
H
47
20,000
37,000
33,000
19,000
110
36,000
1100
8700
21,000
24,000
120
540
380
6900
210
860
650
920
660
41
130
430
650
200
+
+
+
-
+
+
36% @ 0.2
16% @ 30
36% @ 100
3.7% @ 100
7.2% @ 100
24% @ 0.4
2.9% @ 100
27% @ 2
57% @ 20
2.2% @ 50
n.d.
38% @ 0.2
31% @ 0.5
20% @ 3
27% @ 30
20% @ 1
37% @ 2.5
42% @ 3
28% @ 2.5
26% @ 2
46% @ 1.5
47% @ 0.4
44% @ 1
H
OCH2O
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
À
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
+
+
OH
À
n.d.
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
OAc
OMe
OEt
OEt
Cl
OMe
SMe
OMe
NMe2
COCH3
OMe
OMe
Et
5s
5t
5u
5v
5w
5x
6a
6b
6c
H
H
OMe
OMe
H
H
H
H
OMe
t-Bu
Et
H
30% @ 2
50% @ 1
25% @ 10
18% @ 10
4.4% @ 100
3500
3500
8200
OMe
OMe
+
À
COCH3
H
a
Qualitative evaluation: +, antimitotic; À, not antimitotic up to 100
l
M; n.d., not determined.
b
Quantitative evaluation: % mitotic cells after 20 h incubation with indicated test concentration; controls of untreated cells were in the range of 1.4–4.7%.
compound 5i does not share the same mode of action with 1 and
its potent antiproliferative analogs. In contrast to dimethylation,
diacetylation of the hydroquinone system of 5i generating deriva-
tive 5j does not lead to an increase in antiproliferative activity, and
5j was found not to be antimitotic. Considering that compounds
5g,h,i are potential metabolites of 1, the in vitro results suggest
that they do not contribute to the observed in vivo efficacy of
SDZ LAP 977 and that substantial deactivation of 1 can be expected
by such a metabolic degradation pathway.
Compounds 5k–m with one or two of the methoxy groups in 1
replaced by the ethoxy substituent retain high antiproliferative
and clear antimitotic activity. Nevertheless, with IC50 values rang-
ing from 120 nM to 540 nM, these analogs are less potent than 1.
Consequently, sterically more demanding alkoxy substituents
were not investigated. Within the subset of 2,5-di(alk)oxy analogs,
antiproliferative activity seems to be more pronouncedly affected
by changes at the 5-O versus the 2-O substituent, as can be seen
by comparing the IC50 values of derivatives 5k (2-OEt, 5-OMe;
120 nM) and 5g (2-OH, 5-OMe; 1100 nM) with those of analogs
5l (2-OMe, 5-OEt; 540 nM) and 5 h (2-OMe, 5-OH; 8700 nM). Addi-
tional pairs of analogs confirm this trend as a general feature of the
SAR (vide infra).
Next, the influence of various other substituents—including
halogens, methylthio, dimethylamino, acetyl, and alkyl resi-
dues—was thoroughly investigated by synthesizing and testing
compounds 5n–5x. The 2-bromo derivative 5o shows high
potency with an IC50 value of 210 nM, whereas the 5-chloro
compound 5n is only weakly active. Replacement of the 5-meth-
oxy group in 1 by the methylthio (5p) or the dimethylamino
substituent (5r), as well as the analogous modification at posi-
tion 2 (5q), results in decreased antiproliferative potency by a
similar factor of about 14–20. Interestingly, highest activity in
the whole series is achieved by introduction of the acetyl substi-
tuent in position 2 (compound 5t, IC50 value = 41 nM). Despite
the obviously different electronic effects of the acetyl relative
to the original methoxy group, compounds 5t and 1 have very
similar activity and the same mode of action. Analogous substi-
tution by acetyl in position 5 of ring A yielding analog 5s, how-
ever, caused a 15-fold reduction in potency. Derivatives 5u–x,
where ethyl groups replace one or two of the methoxy groups
in 1 and its 2,6-isomer 5e, exhibit again altered electronic prop-
erties and show high to moderate activity with IC50 values be-
tween 130 and 650 nM. The 5-ethyl compound 5v is about 3-
fold less potent than the 2-ethyl analog 5u, and replacement
of both methoxy groups in 1 by ethyl results in an additive loss
of activity (5w). The 2,6-disubstituted compound 5x exhibits
strong antiproliferative activity, but is slightly less effective than
the 2,5-disubstituted analog 5u. Although compounds 5n–x cov-
er a broad range of antiproliferative activity (IC50 values ranging
from 41 nM to 6.9 lM), they all exert the antimitotic mode of
action as observed for 1. The results obtained with the described
disubstituted compounds suggest that the antiproliferative activ-
ity of Lavendustin derivatives is generally more sensitive to ste-
ric rather than electronic effects of the substituents in ring A.
Finally, a series of derivatives with three A-ring substituents
were investigated (6a–c), in particular, to compare our compound
class with another class of antimitotic agents, the combretastatins.
The natural product combretastatin A4,10 one of the most potent
representatives, features a 3,4,5-trimethoxyphenyl and a 3-hydro-
xy-4-methoxyphenyl residue connected via a (Z)-ethenylene lin-
ker. Although this compound has some structural similarity to 1
and its derivatives, SAR differs substantially, as demonstrated, for
example, by the weak activity (IC50 = 3.5 lM) of our 3,4,5-trimeth-
oxy-substituted analog 6b. As discussed above, both the 2,5-dime-
thoxy (1) and the 2,6-dimethoxy (5e) substitution patterns provide
high antiproliferative potency. However, the 2,3,6-trimethoxy

Table 2
Physicochemical properties of new Lavendustin derivatives 5 and 6
Compound
Morph.,a mp (°C)
NMR (CDCl3) d
Methodb/% yieldc
A/68
5a
Oil
10.59 (s, 1H), 7.67 (d, J = 2.3 Hz, 1H), 7.29 (dd, J = 2.3 + 8.5 Hz, 1H), 7.20 (dt, J = 1.8 + 7.8 Hz, 1H), 7.08 (dd, J = 1.8 + 7.8 Hz, 1H), 6.84-6.93 (m, 3H), 3.96 (s, 3H), 3.84
(s, 3H), 2.75-2.94 (m, 4H)
5b
5c
5d
5e
5f
Oil
10.60 (s, 1H), 7.66 (d, J = 2.3 Hz, 1H), 7.26 (dd, J = 2.3 + 8.5 Hz, 1H), 7.21 (t, J = 7.9 Hz, 1H), 6.91 (d, J = 8.5 Hz, 1H), 6.7–6.8 (m, 3H), 3.94 (s, 3H), 3.78 (s, 3H), 2.85 (br
s, 4H)
10.61 (s, 1H), 7.69 (d, J = 2.2 Hz, 1H), 7.30 (dd, J = 2.2 + 8.5 Hz, 1H), 6.95 (t, J = 8 Hz, 1H), 6.91 (d, J = 8.5 Hz, 1H), 6.80 (dd, J = 1 + 8 Hz, 1H), 6.74 (dd, J = 1 + 8 Hz, 1H),
3.95 (s, 3H), 3.88 (s, 3H), 3.84 (s, 3H), 2.78–2.97 (m, 4H)
10.59 (s, 1H), 7.64 (d, J = 2.2 Hz, 1H), 7.26 (dd, J = 2.2 + 8.5 Hz, 1H), 6.96 (d, J = 8.2 Hz, 1H), 6.90 (d, J = 8.5 Hz, 1H), 6.46 (d, J = 2.3 Hz, 1H), 6.40 (dd, J = 2.3 + 8.2 Hz,
1H), 3.96 (s, 3H), 3.81 (s, 6H), 2.80 (br s, 4H)
10.59 (s, 1H), 7.69 (d, J = 2.3 Hz, 1H), 7.32 (dd, J = 2.3 + 8.5 Hz, 1H), 7.15 (t, J = 8.3 Hz, 1H), 6.90 (d, J = 8.5 Hz, 1H), 6.54 (d, J = 8.3 Hz, 2H), 3.96 (s, 3H), 3.79 (s, 6H),
2.87–2.96 (m, 2H), 2.65–2.76 (m, 2H)
10.60 (s, 1H), 7.63 (d, J = 2.3 Hz, 1H), 7.24 (dd, J = 2.3 + 8.5 Hz, 1H), 6.90 (d, J = 8.5 Hz, 1H), 6.72 (d, J = 7.8 Hz, 1H), 6.66 (d, J = 1.6 Hz, 1H), 6.59 (dd, J = 1.6 + 7.8 Hz,
1H), 5.93 (s, 2H), 3.95 (s, 3H), 2.80 (br s, 4H)
10.61 (s, 1H), 7.62 (d, J = 2.3 Hz, 1H), 7.22 (dd, J = 2.3 + 8.5 Hz, 1H), 7.06 (d, J = 8.5 Hz, 1H), 6.97 (dd, J = 2.5 + 8.5 Hz, 1H), 6.94 (d, J = 2.5 Hz, 1H), 6.91 (d, J = 8.5 Hz,
1H), 3.94 (s, 3H), 2.78 (br.s, 4H), 2.31 (s, 3H), 2.28 (s, 3H)
10.59 (s, 1H), 7.67 (d, J = 2.3 Hz, 1H), 7.28 (dd, J = 2.3 + 8.5 Hz, 1H), 6.90 (d, J = 8.5 Hz, 1H), 6.66–6.81 (m, 3H), 3.98 (qua, J = 7 Hz, 2H), 3.94 (s, 3H), 3.74 (s, 3H), 2.84
(br s, 4H), 1.38 (tr, J = 7 Hz, 3H)
10.60 (s, 1H), 7.67 (d, J = 2.3 Hz, 1H), 7.29 (dd, J = 2.3 + 8.5 Hz, 1H), 6.91 (d, J = 8.5 Hz, 1H), 6.68–6.82 (m, 3H), 3.96 (qua, J = 7 Hz, 2H), 3.95 (s, 3H), 3.78 (s, 3H), 2.82
(br s, 4H), 1.41 (tr, J = 7 Hz, 3H)
10.59 (s, 1H), 7.68 (d, J = 2.3 Hz, 1H), 7.29 (dd, J = 2.3 + 8.5 Hz, 1H), 6.92 (d, J = 8.5 Hz, 1H), 6.65-6.81 (m, 3H), 3.99 (qua, J = 7 Hz, 2H), 3.96 (s, 3H), 3.94 (qua, J = 7
Hz, 2H), 2.76–2.93 (m, 4H), 1.42 (tr, J = 7 Hz, 3H), 1.38 (tr, J = 7 Hz, 3H)
10.60 (s, 1H), 7.65 (d, J = 2.3 Hz, 1H), 7.28 (dd, J = 2.3 + 8.5 Hz, 1H), 7.15 (dd, J = 2.6 + 8.6 Hz, 1H), 7.07 (d, J = 2.6 Hz, 1H), 6.91 (d, J = 8.5 Hz, 1H), 6.77 (d, J = 8.6 Hz,
1H), 3.95 (s, 3H), 3.80 (s, 3H), 2.74–2.90 (m, 4H)
10.63 (s, 1H), 7.67 (d, J = 2.3 Hz, 1H), 7.43 (d, J = 8.5 Hz, 1H), 7.30 (dd, J = 2.3 + 8.5 Hz, 1H), 6.92 (d, J = 8.5 Hz, 1H), 6.69 (d, J = 3 Hz, 1H), 6.65 (dd, J = 3 + 8.5 Hz,
1H), 3.95 (s, 3H), 3.74 (s, 3H), 2.78–2.99 (m, 4H)
10.59 (s, 1H), 7.65 (d, J = 2.3 Hz, 1H), 7.27 (dd, J = 2.3 + 8.5 Hz, 1H), 7.18 (dd, J = 2.4 + 8.5 Hz, 1H), 7.07 (d, J = 2.3 Hz, 1H), 6.9 (d, J = 8.5 Hz, 1H), 6.81 (d, J = 8.5 Hz,
1H), 3.96 (s, 3H), 3.81 (s, 3H), 2.74–2.94 (m, 4H), 2.41 (s, 3H)
10.60 (s, 1H), 7.69 (d, J = 2.3 Hz, 1H), 7.33 (dd, J = 2.3 + 8.5 Hz, 1H), 7.07–7.12 (m, 1H), 6.91 (d, J = 8.5 Hz, 1H), 6.72–6.76 (m, 2H), 3.95 (s, 3H), 3.76 (s, 3H), 2.82–3.02
(m, 4H), 2.62 (s, 6H)
10.59 (s, 1H), 7.69 (d, J = 2.3 Hz, 1H), 7.32 (dd, J = 2.3 + 8.5 Hz, 1H), 6.92 (d, J = 8.5 Hz, 1H), 6.81 (d, J = 8.5 Hz, 1H), 6.63 (dd, J = 3 + 8.5 Hz, 1H), 6.60 (d, J = 3 Hz, 1H),
3.96 (s, 3H), 3.79 (s, 3H), 2.79–2.91 (m, 10H)
10.60 (s, 1H), 7.86 (dd, J = 2.3 + 8.6 Hz, 1H), 7.74 (d, J = 2.3 Hz, 1H), 7.66 (d, J = 2.3 Hz, 1H), 7.28 (dd, J = 2.3 + 8.5 Hz, 1H), 6.91 (d, J = 8.5 Hz, 1H), 6.89 (d, J = 8.6 Hz,
1H), 3.96 (s, 3H), 3.90 (s, 3H), 2.77–2.95 (m, 4H), 2.54 (s, 3H)
10.62 (s, 1H), 7.79 (d, J = 8.7 Hz, 1H), 7.71 (d, J = 2.3 Hz, 1H), 7.38 (dd, J = 2.3 + 8.5 Hz, 1H), 6.91 (d, J = 8.5 Hz, 1H), 6.79 (dd, J = 2.7 + 8.7 Hz, 1H), 6.69 (d, J = 2.7 Hz,
1H), 3.94 (s, 3H), 3.82 (s, 3H), 3.11–3.19 (m, 2H), 2.76–2.84 (m, 2H), 2.56 (s, 3H)
A/56
A/78
A/76
A/79
A/83
B/39
C/36
C/49
A/78
A/66
Ad/58
A/48
A/59
F/53
D/64
D/61e
E/85
E/93
A/77
A/63
A/55
A/61
A/73
75–77
64–66
95–97
89
5j
Oil
5k
5l
Oil
Oil
5m
5n
5o
5p
5q
5r
Oil
59–62
67–68
Oil
Oil
Oil
5s
5t
107–108
74–75
5u
5v
5w
5x
5y
5z
6a
Oil
10.65 (s, 1H), 7.69 (d, J = 2.3 Hz, 1H), 7.30 (dd, J = 2.3 + 8.5 Hz, 1H), 7.13 (d, J = 8.3 Hz, 1H), 6.95 (d, J = 8.5 Hz, 1H), 6.75 (dd, J = 2.8 + 8.3 Hz, 1H), 6.73 (d, J = 2.8 Hz,
1H), 3.97 (s, 3H), 3.79 (s, 3H), 2.82–2.91 (m, 4H), 2.62 (qua, J = 7.5 Hz, 2H), 1.22 (t, J = 7.5 Hz, 3H)
10.58 (s, 1H), 7.67 (d, J = 2.3 Hz, 1H), 7.29 (dd, J = 2.3 + 8.5 Hz, 1H), 7.02 (dd, J = 2.3 + 8.2 Hz, 1H), 6.91 (d, J = 2.3 Hz, 1H), 6.90 (d, J = 8.5 Hz, 1H), 6.79 (d, J = 8.2 Hz,
1H), 3.96 (s, 3H), 3.82 (s, 3H), 2.76–2.91 (m, 4H), 2.56 (qua, J = 7.6 Hz, 2H), 1.19 (t, J = 7.6 Hz, 3H)
10.61 (s, 1H), 7.67 (d, J = 2.2 Hz, 1H), 7.29 (dd, J = 2.2 + 8.5 Hz, 1H), 7.13 (d, J = 7.8 Hz, 1H), 7.02 (dd, J = 1.8 + 7.8 Hz, 1H), 6.97 (d, J = 1.8 Hz, 1H), 6.93 (d, J = 8.5 Hz,
1H), 3.97 (s, 3H), 2.76–2.93 (m, 4H), 2.65 (qua, J = 7.5 Hz, 2H), 2.60 (qua, J = 7.6 Hz, 2H), 1.25 (t, J = 7.5 Hz, 3H), 1.23 (t, J = 7.6 Hz, 3H)
10.62 (s, 1H), 7.7 (d, J = 2.3 Hz, 1H), 7.34 (dd, J = 2.3 + 8.5 Hz, 1H), 7.17 (t, J = 7.9 Hz, 1H), 6.94 (d, J = 8.5 Hz, 1H), 6.84 (d, J = 7.9 Hz, 1H), 6.76 (d, J = 7.9 Hz, 1H),
3.98 (s, 3H), 3.84 (s, 3H), 2.88–2.96 (m, 2H), 2.69–2.78 (m, 2H), 2.63 (qua, J = 7.5 Hz, 2H), 1.22 (t, J = 7.5 Hz, 3H)
10.59 (s, 1H), 7.67 (d, J = 2.3 Hz, 1H), 7.28 (dd, J = 2.3 + 8.5 Hz, 1H), 6.89 (d, J = 8.5 Hz, 1H), 6.70 (d, J = 8.3 Hz, 1H), 6.53 (dd, J = 2.8 + 8.3 Hz, 1H), 6.49 (d, J = 2.8 Hz,
1H), 3.95 (s, 3H), 3.75 (s, 3H), 2.79 (s, 4H)
83–86
47
Oil
Oil, HCl: 172–175
Oil
10.61 (s, 1H), 7.65 (d, J = 2.3 Hz, 1H), 7.30 (dd, J = 2.3 + 8.5 Hz, 1H), 7.19 (d, J = 7.8 Hz, 1H), 6.92 (d, J = 8.5 Hz, 1H), 6.73 (dd, J = 1.8 + 7.8 Hz, 1H), 6.66 (d, J = 1.8 Hz,
1H), 3.95 (s, 3H), 3.81 (s, 3H), 2.86 (br.s, 4H), 1.36 (s, 9H)
10.60 (s, 1H), 7.70 (d, J = 2.3 Hz, 1H), 7.34 (dd, J = 2.3 + 8.5 Hz, 1H), 6.91 (d, J = 8.5 Hz, 1H), 6.74 (d, J = 8.9 Hz, 1H), 6.56 (d, J = 8.9 Hz, 1H), 3.96 (s, 3H), 3.83 (s, 3H),
3.80 (s, 3H), 3.76 (s, 3H), 2.70–2.95 (m, 4H)
87–90
6b
6c
90
114–117
10.61 (s, 1H), 7.64 (d, J = 2.3 Hz, 1H), 7.27 (dd, J = 2.3 + 8.5 Hz, 1H), 6.92 (d, J = 8.5 Hz, 1H), 6.36 (s, 2H), 3.95 (s, 3H), 3.83 (s, 6H), 2.78–2.89 (m, 4H)
10.62 (s, 1H), 7.74 (s, 1H), 7.69 (d, J = 2.3 Hz, 1H), 7.41 (dd, J = 2.3 + 8.5 Hz, 1H), 6.92 (d, J = 8.5 Hz, 1H), 6.59 (s, 1H), 3.94 (s, 3H), 3.83 (s, 3H), 3.09–3.17 (m, 2H),
2.78–2.86 (m, 2H), 2.58 (s, 3H), 1.39 (s, 9H)
A/82
D/85
a
Morph., morphology.
b
c
Method A consists of 2 steps (Wittig reaction and subsequent hydrogenation) and overall yields are given.
Yields (not optimized) of isolated, analytically pure products.
Rhodium (10% on charcoal) instead of Pd was used in the hydrogenation step for the synthesis of 5o.
Yield for two steps starting from 5z.
d
e

P. Nussbaumer, A. P. Winiski / Bioorg. Med. Chem. 16 (2008) 7552–7560
7557
analog 6a, where both substitution patterns are merged into one
structure, shows strongly reduced activity (IC50 = 3.5 M). Addi-
are the 2,5-dimethoxy derivative 1 and the 2-acetyl-5-methoxy
analog 5t with IC50 values of 47 and 41 nM, respectively. Based
on these results, the 2,5-dimethoxy substitution pattern has been
used throughout the subsequent SAR studies, that is, derivation
of the spacer and ring B in this compound class. We demonstrated
that compound 1 binds to tubulin at the promiscuous ‘‘colchicine
binding site” and, consequently, inhibits tubulin polymerization.
This interference with the spindle apparatus results in inhibition
of cell proliferation by blocking the cell cycle in mitosis. The com-
pound class is of high clinical interest since it was demonstrated
for compound 1 that it affects interphase microtubules only at
much higher concentrations and that it exhibits selective antipro-
liferative activity against certain tumor cells including multidrug
resistance phenotypes without suppressing hematopoiesis.3
l
tion of the bulky t-butyl residue to 5t to yield 6c, the most potent
compound of the whole study, results in loss of the antimitotic
mode of action. Due to the consistently weak activity of com-
pounds 6a–c, no further trisubstituted analogs were investigated.
Searching for the molecular basis of the antimitotic activity
exhibited by this compound class, we initiated detailed investiga-
tions with 1. In initial studies with HaCaT keratinocytes, we ob-
served that mitotic arrest of cells occurred within a few hours
resulting in large numbers of cells containing micronuclei after 1
day. The microtubules of the spindle apparatus were perturbed
at the low, antiproliferative concentrations as visualized with indi-
rect immunofluorescence using antibodies against a- and b-tubu-
lin,3 strongly suggesting either tubulin or microtubule-associated
proteins as the molecular target. Dose-response curves revealed
an excellent correlation between the antiproliferative and antimi-
totic effects. Effects on interphase microtubules were seen at much
4. Experimental
4.1. Chemistry
higher concentrations, that is, in the
then tested in a tubulin polymerization assay in vitro and exhibited
an IC50 of 1.0 M (Fig. 2; podophyllotoxin was used as reference
compound and gave an IC50 of 0.56 M). Thus, with this activity,
l
M range.11 Compound 1 was
4.1.1. Materials and methods
l
1,4-Diethylbenzene and 2,4-dimethoxybenzaldehyde were pur-
chased from Fluka AG. 2-Methoxy-, 2,3-dimethoxy-, 3,4-(methyl-
enedioxy)-, 2,5-diethoxy-, 3,4,5-trimethoxybenzyl alcohol, 3-
methoxy-, 2-methoxy-5-nitrobenzyl bromide, and 2-t-butyl-5-
methylanisol, were purchased from Sigma–Aldrich. 2,6-Dime-
thoxybenzyl alcohol and 2,3,6-trimethoxybenzaldehyde were
purchased from Apin Chemicals and Oakwood Products, Inc.,
respectively. 2-Bromo-5-methoxybenzaldehyde was purchased
from ABCR GmbH and 5-chloro-2-methoxybenzaldehyde and
5-formyl-2-hydroxybenzoic acid methylester from 3B Scientific
Corporation. 2-Methoxy-5-methylthio-,18 2-dimethylamino-5-
methoxybenzaldehyde,19 and 2-acetoxy-5-(bromomethyl)benzoic
acid methylester20 were prepared according to published
procedures.
In general, benzyltriphenylphosphonium bromides 3 were pre-
pared by heating an equimolar solution of either the corresponding
benzyl bromides and PPh3 or the corresponding benzyl alcohols
and PPh3 Â HBr (Sigma–Aldrich) in toluene at 80 °C for about 5 h.
The solution was cooled, and the solid precipitate was filtered,
dried, and used directly without further purification. Some benzyl
alcohols were synthesized from the corresponding benzaldehydes
by treatment with sodium borohydride in methanol.
l
1 is among the most potent known tubulin polymerization inhibi-
tors. There are several defined binding sites on tubulin for antimi-
totic agents. Based on some structural similarity to known binders,
we investigated whether 1 binds to the promiscuous ‘‘colchicine
binding site”.12 Indeed, 1 inhibited the binding to tubulin of a com-
mercially available fluorescent colchicine derivative with a similar
efficiency as unlabeled colchicine (43% vs 72% at 10 lM), thus, con-
firming binding to the ‘‘colchicine binding site” on its molecular
target tubulin.
Several additional compound classes derived from Lavendustin
A that inhibit tubulin polymerization have been discovered and re-
ported.13–15 Moreover, the structure of 1 also resembles another
known class of antimitotic agents, the combretastatins10,16,17 How-
ever, the results we obtained for analogs of 1 demonstrate that our
substances have to be considered as a new compound class with
clear-cut different SARs.
In summary, the most important structural requirement in ring
A for high antiproliferative potency of 1 and analogs is a 2,5 or 2,6
substitution pattern with a preference for 2,5-disubstituted ana-
logs. Highest activity is achieved when at least one of the two sub-
stituents is the methoxy group. Structural modifications at position
5 have a more pronounced negative effect on antiproliferative
activity relative to the same changes at position 2, whereby the
activity is generally more sensitive to steric rather than electronic
effects of the A-ring substituents. All potent compounds share the
same antimitotic mode of action. The most potent representatives
Melting points were determined on a Reichert Thermovar
microscope, and are not corrected. The temperature is given in
Celsius units. Thin-layer chromatography was performed using sil-
ica gel F254 plates (Merck) visualizing with UV or potassium per-
manganate. Column chromatography was performed using silica
gel 60 (0.040–0.063 mm, Merck), pressure 3–5 bar. 1H NMR spectra
were recorded at 250 MHz (Bruker WM 250) usually in CDCl3 with
(CH3)4Si as internal standard. Chemical shifts are given as d units.
4.1.2. General procedure for the synthesis of test compounds 5
and 6
4.1.2.1. Method A. Wittig reaction + hydrogenation. Synthesis of
5-[2-(3,4,5-trimethoxyphenyl)ethyl]-2-hydroxybenzoic acid methyl-
ester (6b).
(E)- and (Z)-5-[2-(3,4,5-Trimethoxyphenyl)ethenyl]-
2-hydroxybenzoic acid methylester. At À30 °C, n-butyllithium
(0.57 mL, 5.7 mmol, 10 M solution in hexane, Aldrich Chemical
Co.) was added to
a solution of diisopropylamine (0.57 g,
5.7 mmol) in dry tetrahydrofuran (25 mL). After stirring for
30 min, the mixture was cooled to À50 °C, and phosphonium bro-
mide 3l21 (1 g, 1.9 mmol) was added as solid. The orange suspen-
sion was stirred for 1 h at À50 °C, cooled to À70 °C, and treated
with a solution of 4a (345 mg, 1.9 mmol) in dry tetrahydrofuran
(10 mL). The mixture was stirred for 1 h at À70 °C and for 16 h
Figure 2. Dose-dependent effect of compound 1 on tubulin polymerization in vitro.

7558
P. Nussbaumer, A. P. Winiski / Bioorg. Med. Chem. 16 (2008) 7552–7560
after removal of the cooling bath, and then was poured into aque-
ous ammonium chloride solution. Extraction with ethyl acetate
yielded a product mixture, which was chromatographed on silica
gel (hexane/ethyl acetate = 9:1) to give pure (Z)-5-[2-(3,4,5-trime-
methoxyphenyl)ethyl]-2-hydroxybenzoic acid methylester. Pure
5t was obtained using the following procedure.
4.1.2.5. Synthesis of 5-[2-(2-acetyl-5-methoxyphenyl)ethyl]-2-
hydroxybenzoic acid methylester (5t) and 5-[2-(2-acetyl-4-t-
butyl-5-methoxyphenyl)ethyl]-2-hydroxybenzoic acid methyl-
thoxyphenyl)ethenyl]-2-hydroxybenzoic
acid
methylester
[168 mg, 26%; colorless crystals: mp 103 °C, NMR d 10.74 (s, 1H),
7.81 (d, J = 2.2 Hz, 1H), 7.42 (dd, J = .2 + 8.6 Hz, 1H), 6.86 (d,
J = 8.6 Hz, 1H), 6.50 (s, 2H), 6.48 (s, 2H), 3.92 (s, 3H), 3.85 (s, 3H),
3.70 (s, 6H)], followed by a mixed fraction (125 mg, 19%) and its
pure E-isomer [157 mg, 24%; colorless crystals: mp 155 °C, NMR
d 10.78 (s, 1H), 7.97 (d, J = 2.3 Hz, 1H), 7.66 (dd, J = 2.3 + 8.7 Hz,
1H), 7.00 (d, J = 8.7 Hz, 1H), 6.97 (d, J = 16.5 Hz, 1H), 6.91 (d,
J = 16.5 Hz, 1H), 6.73 (s, 2H), 3.99 (s, 3H), 3.92 (s, 6H), 3.88 (s, 3H)].
Only 1.2 equiv of lithium diisopropylamide (preferably pre-
pared from 1.6 M solution of n-butyllithium in hexane) was used
in Wittig reactions with phosphonium bromide 3n and benzalde-
hydes 4b–e as starting materials. In most cases, the stilbene iso-
mers were not separated, but isolated and used as mixture for
the next step.
ester (6c).
Aluminum chloride (1.8 g, 13.6 mmol) was added
in one portion without cooling to a solution of 5z (1.16 g,
3.4 mmol) and acetyl chloride (588 mg, 7.5 mmol) in dichloroeth-
ane (40 mL), which resulted in a raise of the temperature of the
mixture to 35 °C. This temperature was maintained for about 1 h
using a water bath. The mixture was then poured into 2 N aqueous
hydrochloric acid and extracted with ethyl acetate. The combined
organic extracts were washed with brine, dried over magnesium
sulfate, and concentrated in vacuo. The residue was chromato-
graphed on silica gel (cyclohexane/ethyl acetate = 4:1) to obtain a
small fraction consisting of 6c (66 mg, mp 114–117 °C), followed
by pure 5t (683 mg, 61 %) as colorless crystals, mp 74–75 °C.
5-[2-(3,4,5-Trimethoxyphenyl)ethyl]-2-hydroxybenzoic acid meth-
ylester (6b). (Z)- or (E)-5-[2-(3,4,5-Trimethoxyphenyl)ethenyl]-2-
hydroxybenzoic acid methylester or a mixture thereof (160 mg,
0.46 mmol) was dissolved in ethyl acetate (20 mL) and hydroge-
nated over palladium (20 mg, 10% on charcoal) for 4 h at atmo-
spheric pressure. The mixture was filtered over Celite, and the
filtrate was concentrated in vacuo. The residue was crystallized
from ethanol or chromatographed on silica gel to afford 6b
(154 mg, 96%) as colorless crystals: mp 90 °C.
4.1.2.6. Method E: Hydrogenation. Synthesis of 5-[2-(2-ethyl-5-
methoxyphenyl)ethyl]-2-hydroxybenzoic
acid methylester
(5u). A solution of 5t (105 mg, 0.31 mmol) in dry ethyl acetate
(10 mL) was hydrogenated over palladium (20 mg, 10% on char-
coal) for 3 h at atmospheric pressure. The mixture was filtered over
Celite, and the filtrate was concentrated in vacuo. The residue was
chromatographed on silica gel to afford 5u (85 mg, 85%) as color-
less oil.
4.1.2.7. Method F: Synthesis of 5-[2-(5-dimethylamino-2-
methoxyphenyl)ethyl]-2-hydroxybenzoic
(5r). A mixture of 5y (103 mg, 0.34 mmol; obtained by stan-
acid methylester
4.1.2.2. Method B: O-acetylation. Synthesis of 5-[2-(2,5-diacet-
oxyphenyl)ethyl]-2-hydroxybenzoic acid methylester (5j).
A
dard hydrogenation of the stilbene resulting from reacting 3 h
and 4a), 1 N aqueous sodium dihydrogenphosphite solution
(3.4 ml, 3.4 mmol), formaldehyde (0.3 ml, 37% solution in water),
and dioxane (12 ml) was heated to reflux for 2 h. Then the mixture
was poured into water and extracted with ethyl acetate. The com-
bined organic layers were washed with brine, dried over magne-
sium sulfate, and concentrated in vacuo. The residue was
chromatographed on silica gel (toluene/ethyl acetate = 95:5) to
give pure 5r (59 mg, 53%) as yellowish oil.
solution of 5i (200 mg, 0.7 mmol) in dichloromethane/tetrahydro-
furane (2:1) was treated successively with triethylamine (280 mg,
2.8 mmol) and acetic anhydride (230 mg, 2.2 mmol). The mixture
was stirred for 2 h at room temperature, poured into water, and
stirred for an additional 30 min. Extraction with ethyl acetate gave
after drying and evaporation of the solvent a crude product, which
was purified by chromatography (silica gel, hexane/ethyl ace-
tate = 4:1) yielding 5j (102 mg, 39%) as colorless oil.
4.1.2.3. Method C: O-alkylation. Synthesis of 5-[2-(2-ethoxy-5-
methoxyphenyl)ethyl]-2-hydroxybenzoic acid methylester
4.1.3. Starting materials
4.1.3.1. General procedure for the synthesis of benzyltriphen-
ylphosphonium bromides 3. 4.1.3.1.1. (2,5-Diethylbenzyl)tri-
(5k).
A mixture of 5g (100 mg, 0.33 mmol), potassium car-
bonate (50 mg, 0.36 mmol), iodoethane (52 mg, 0.33 mmol), and
acetone (7 mL) was heated to reflux overnight. Then it was poured
into water and extracted with ethyl acetate. The combined organic
layers were dried over magnesium sulfate and concentrated in va-
cuo. The residue was chromatographed on silica gel (hexane/ethyl
acetate = 5:1) to separate the title compound 5k (39 mg, 36%; col-
orless oil) and recovered starting material 5g (46 mg).
phenylphosphonium bromide (3i). 2,5-Diethylbenzaldehyde.
A
mixture of 1,4-diethylbenzene (1 g, 7.45 mmol), hexamethylene-
tetramine (1.05 g, 7.45 mmol), and trifluoroacetic acid (10 mL)
was heated to reflux overnight. The solvent was distilled off, and
the residue was treated with cold water (40 mL). The mixture
was basified with sodium carbonate, stirred well for 20 min, and
then extracted with ethyl acetate. The combined organic layers
were dried over magnesium sulfate and concentrated in vacuo.
The residue was chromatographed on silica gel (hexane/ethyl ace-
tate = 10:1) to give the pure aldehyde (651 mg, 54%) as an oil: NMR
d 10.26 (s, 1H), 7.67 (d, J = 2 Hz, 1H), 7.36 (dd, J = 2 + 7.8 Hz, 1H),
7.22 (d, J = 7.8 Hz, 1H), 3.04 (qua, J = 7.5 Hz, 2H), 2.70 (qua, J = 7.6
Hz, 2H), 1.21–1.31 (m, 6H). 2,5-Diethylbenzyl alcohol. 2,5-Diethyl-
benzaldehyde (632 mg, 3.9 mmol) was dissolved in methanol
(15 mL), and the solution was treated with sodium borohydride
(147 mg, 3.9 mmol). The mixture was stirred for 15 min at room
temperature and then poured into brine. Extraction with ethyl ace-
tate afforded after drying over magnesium sulfate and evaporation
in vacuo the crude benzyl alcohol (630 mg), which was used in the
next step without purification: NMR d 7.09–7.24 (m, 3H), 4.73 (s,
2H), 2.70 (qua, J = 7.5 Hz, 2H), 2.64 (qua, J = 7.5 Hz, 2H), 1,60 (br.
s, 1H), 1.24 (t, J = 7.5 Hz, 6H). 2,5-Diethylbenzyl bromide. Bromine
4.1.2.4. Method D: C-acetylation. Synthesis of 5-[2-(5-acetyl-2-
methoxyphenyl)ethyl]-2-hydroxybenzoic acid methylester
(5s).
Acetyl chloride (180 mg, 2.3 mmol) was added without
cooling to a stirred mixture of 5a (265 mg, 0.93 mmol), aluminum
chloride (370 mg, 2.8 mmol), and dichloroethane (10 mL). The
mixture was stirred for 3 h at ambient temperature, then poured
into 1 N aqueous hydrochloric acid, and extracted with ethyl ace-
tate. The combined organic extracts were washed with brine, dried
over magnesium sulfate, and concentrated in vacuo. The residue
was chromatographed on silica gel (toluene/ethyl acetate = 93:7)
to obtain pure 5s (198 mg, 65 %) as colorless crystals, mp 107–
108 °C.
Starting with 5b the same procedure afforded an inseparable
o,p-isomeric mixture of 2-acetyl-5-methoxy-(5t) and 4-acetyl-3-

P. Nussbaumer, A. P. Winiski / Bioorg. Med. Chem. 16 (2008) 7552–7560
7559
(583 mg, 3.65 mmol) was added to a solution of 1,2-bis(diphenyl-
phosphino)ethane (727 mg, 1.82 mmol) in dry methylene chloride
at 0 °C under argon atmosphere. The mixture was stirred for
30 min, treated with a solution crude 2,5-diethylbenzyl alcohol
(600 mg, 3.65 mmol), and then stirred for an additional 30 min.
The solvent was evaporated in vacuo, and the residue was well trit-
urated with hexane/diethyl ether (1:1). The mixture was filtered to
remove the crystalline phosphinoxide by-products, and the filtrate
was evaporated to obtain the crude benzyl bromide (760 mg, 92%),
which was used in the next step without further purification: NMR
d 7.09–7.20 (m, 3H), 4.55 (s, 2H), 2.76 (qua, J = 7.5 Hz, 2H), 2.62
(qua, J = 7.5 Hz, 2H), 1.30 (t, J = 7.5 Hz, 3H), 1.24 (t, J = 7.5 Hz,
3H). (2,5-Diethylbenzyl)triphenylphosphonium bromide (3i). A mix-
ture of crude 2,5-diethylbenzyl bromide (750 mg, 3.3 mmol), tri-
phenylphosphine (1.3 g, 5 mmol), and toluene (20 mL) was
heated to 80 °C for 3 h. After cooling, the colorless crystalline prod-
uct was filtered and washed with diethyl ether (1.38 g, 85%): mp
287–290 °C.
tion, suspended in fresh medium, and seeded into 96-well microti-
ter plates at 4000 cells/0.2 mL/well. After 24 h, the medium was
replaced with fresh medium containing graded concentrations of
test compound. After 3 days of incubation, the extent of cellular
proliferation was measured by a colorimetric assay using sulforho-
damine B.23 The results represent the average of at least two inde-
pendent measurements.
4.2.2. Measurement of antimitotic activity
Cells were seeded into 8-well Lab-Tek plastic chamber slides
(Permanox, Nunc, Vienna, Austria) at a final density of 24000
cells/0.4 ml/well. After 24 h, the medium was replaced with fresh
medium containing specified concentrations of the test compound
(Table 1), which were usually 2- to 5-fold above the corresponding
antiproliferative IC50 values. After 20 h of incubation time, the
medium was removed from each slide chamber, and 90% acetone
was added for 5 min at room temperature to fix the cells. DNA
was stained with DAPI in PBS (0.5 lg/ml for 1 h at room tempera-
4.1.3.1.2. (2-Ethyl-6-methoxybenzyl)triphenylphosphonium bro-
ture). Fluorescence microscopy was performed with a Zeiss IM
microscope equipped with epifluorescence optics. Cells arrested
in metaphase were scored for each sample and the mitotic index
(100% Â number of mitotic cells/number of total cells) determined.
At least 300 cells were counted at each drug concentration tested.
mide (3j).
Mp 242 °C; 2-Ethyl-6-hydroxybenzoic acid ethylester.
A mixture of 1-(3-ethoxycarbonyl-2-oxopropyl)-pyridinium bro-
mide9 (3 g, 10.4 mmol), ethyl vinyl ketone (0.9 g, 10.7 mmol), tri-
ethylamine (1.08 g, 10.7 mmol), and ethanol (400 mL) was heated
to reflux for 15 h. The solvent was partly distilled off, and the res-
idue poured into brine. Extraction with ethyl acetate yielded after
drying over magnesium sulfate and evaporation in vacuo an iso-
meric product mixture, which was separated by chromatography
on silica gel (cyclohexane/ethyl acetate = 100:2). The first fraction
consisted of the unwanted isomer 4-ethyl-2-hydroxybenzoic acid
ethylester [NMR d 10.81 (s, 1H), 7.75 (d, J = 8.1 Hz, 1H), 6.82 (d,
J = 1.7 Hz, 1H), 6.72 (dd, J = 1.7 + 8.1 Hz, 1H), 4.39 (qua, J = 7.1 Hz,
2H), 2.63 (qua, J = 7.6 Hz, 2H), 1.40 (t, J = 7.1 Hz, 3H), 1.23 (t,
J = 7.5 Hz, 3H)], followed by pure 2-ethyl-6-hydroxybenzoic acid
ethylester (pure fraction: 355 mg, 18%): NMR d 11.24 (s, 1H),
7.31 (t, J = 8 Hz, 1H), 6.84 (dd, J = 1 + 8 Hz, 1H), 6.74 (dd, J = 1 + 8
Hz, 1H), 4.44 (qua, J = 7.1 Hz, 2H), 2.95 (qua, J = 7.4 Hz, 2H), 1.44
(t, J = 7.1 Hz, 3H9, 1.21 (t, J = 7.4 Hz, 3H). 2-Ethyl-6-methoxybenzoic
acid ethylester. Sodium hydride (37 mg, 1.2 mmol; 80% suspension
in mineral oil) was added to a solution of 2-ethyl-6-hydroxybenzo-
ic acid ethylester (190 mg, 1 mmol) in dry DMF (8 mL). The mix-
ture was stirred for 20 min at room temperature, treated with
iodomethane (0.5 mL), and stirred for an additional hour. The mix-
ture was poured into water and extracted with ethyl acetate. The
organic layers were combined, dried over magnesium sulfate,
and concentrated in vacuo to obtain the oily product (200 mg),
which was used in the next step without further purification:
NMR d 7.28 (t, J = 8 Hz, 1H), 6.83 (d, J = 7.6 Hz, 1H), 6.76 (d,
J = 8.4 Hz, 1H), 4.40 (qua, J = 7.1 Hz, 2H), 3.82 (s, 3H), 2.60 (qua,
J = 7.6 Hz, 2H), 1.38 (t, J = 7.1 Hz, 3H), 1.21 (t, J = 7.6 Hz, 3H). 2-
Ethyl-6-methoxybenzyl alcohol. A solution of crude 2-ethyl-6-meth-
oxybenzoic acid ethylester (200 mg, 1 mmol) in dry toluene
(15 mL) was cooled to À40 °C and treated slowly with diisobutyl
aluminum hydride (3 mL, 3 mmol; 1 M solution in toluene). The
cooling bath was removed, and the mixture was stirred for 1 h at
room temperature. Then it was poured into brine and extracted
with ethyl acetate. The combined organic layers were dried over
magnesium sulfate and evaporated in vacuo to give the crude title
compound (160 mg), which was used in the next step without fur-
ther purification.
4.2.3. Tubulin polymerization assay24
GTP and colchicine were from Sigma, and podophyllotoxin was
from Aldrich. Monosodium glutamate was purchased from United
States Biochemical. Microtubule protein was prepared from pig
brain by two cycles of temperature-dependent assembly-disas-
sembly25 and drop-frozen in liquid nitrogen after the final cold
centrifugation, and then stored at À80 °C. Purified tubulin was pre-
pared from the microtubule protein by a total of three cycles of
temperature-dependent assembly–disassembly in 1 M monoso-
dium glutamate, pH 6.6, with 1 mM GTP26 with the following mod-
ification: monosodium glutamate powder was added to the rapidly
thawed microtubule protein solution with stirring at 4 °C (i.e., in
the cold-room) to a final concentration of 1 M followed by the
addition of GTP to 1 mM. The solution was then incubated at
37 °C for 1 h to start the first warm cycle. After the final cold cen-
trifugation, the purified tubulin was stored in aliquots in liquid
nitrogen. Polymerization was followed turbidimetrically at
350 nm in a computer-controlled Cary Model 1E spectrophotome-
ter equipped with electronic temperature controllers. Preincuba-
tions (15 min, 30 °C) of compounds (initially dissolved in DMSO)
with tubulin were performed in 0.8 M monosodium glutamate
(pH 6.6) in the absence of GTP in a 0.48 ml volume, with all con-
centrations referring to a final reaction volume of 0.5 ml. Following
the preincubation, the reaction mixtures were chilled on ice, 20 ll
of 10 mM GTP was added, and the samples were transferred to
cuvettes held at 0 °C (final DMSO conc. = 4% v/v). Polymerization
was initiated by a temperature jump to 30 °C. The IC50 was defined
as the drug concentration (obtained by linear interpolation be-
tween data points at specific drug concentrations) required to inhi-
bit the extent of polymerization by 50% following a 20-min
incubation (i.e., following the temperature jump).
4.2.4. Tubulin binding assay27
Fluorescein Colchicine (FC) was purchased from Molecular
Probes. Purified tubulin (2
lM) was mixed with FC (2 lM) test
substance (10 M) at 37 °C (15 min), cooled on ice (5 min), and
l
then passed through small gel filtration columns (BioSpin 30 chro-
matography columns, BioRad). Tubulin comes through the column
with any FC bound. Samples are placed in a black microtiter plate
and the fluorescence measured in a fluorescence microtiter plate
reader. The degree of binding is determined relative to the control
sample without test substance (i.e. only tubulin + FC).
4.2. Biological experiments
4.2.1. HaCaT keratinocyte proliferation assay
HaCaT cells22 were cultivated in DMEM (Gibco) containing 5%
FCS. For the proliferation assay, cells were detached by trypsiniza-

7560
P. Nussbaumer, A. P. Winiski / Bioorg. Med. Chem. 16 (2008) 7552–7560
11. Winiski, A. P.; Meingassner, J. G. J. Invest. Dermatol. 1994, 102, 565. Abstract 251.
Supplementary data
12. Islam, M. N.; Iskander, M. N. Mini-Rev. Med. Chem. 2004, 4, 1077.
13. Mu, F.; Coffing, S. L.; Riese, D. J.; Geahlen, R. L.; Verdier-Pinard, P.; Hamel, E.;
Johnson, J.; Cushman, M. J. Med. Chem. 2001, 44, 441.
14. Mu, F.; Lee, D. J.; Pryor, D. E.; Hamel, E.; Cushman, M. J. Med. Chem. 2002, 45,
4774.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2008.07.039.
15. Mu, F.; Hamel, E.; Lee, D. J.; Pryor, D. E.; Cushman, M. J. Med. Chem. 2003, 46,
1670.
References
16. Pettit, G. R.; Cragg, G. M.; Herald, D. L.; Schmidt, J. M.; Lohavanijaya, P. Can. J.
Chem. 1982, 60, 1374.
17. Lin, C. M.; Singh, S. B.; Chu, P. S.; Dempcy, R. O.; Schmidt, J. M.; Pettit, G. R.;
Hamel, E. Mol. Pharmacol. 1988, 34, 200.
18. Ando, M.; Emoto, S. Bull. Chem. Soc. Jpn. 1978, 51, 2435.
19. Damhaut, P.; Lemaire, C.; Plenevaux, A.; Brihaye, C.; Christiaens, L.; Comar, D.
Tetrahedron 1997, 53, 5785.
1. Nussbaumer, P.; Winiski, A. P.; Cammisuli, S.; Hiestand, P.; Weckbecker, G.;
Stütz, A. J. Med. Chem. 1994, 37, 4079.
2. Nussbaumer, P.; Winiski, A. P.; Meingassner, J. G.; Cammisuli, S.; Weckbecker,
G.; Stuetz, A. J. Invest. Dermatol. 1994, 102, 565. Abstract 252.
3. Cammisuli, S.; Winiski, A. P.; Nussbaumer, P.; Hiestand, P.; Stütz, A.;
Weckbecker, G. Int. J. Cancer 1996, 65, 351.
20. Regnier, G.; Canevari, R.; Le Douarec, J.-C. Bull. Chem. Soc. Fr. 1966, 9, 2821.
21. Asakawa, Y.; Tanikawa, K.; Aratani, T. Phytochemistry 1976, 15, 1057.
22. Boukamp, P.; Petrussevska, R. T.; Breitkreutz, D.; Hornung, J.; Markham, A.;
Fusenig, N. E. J. Cell Biol. 1988, 106, 761.
23. Skehan, P.; Storeng, R.; Scudiero, D.; Monks, A.; McMahon, J.; Vistica, D.;
Warren, J. T.; Bokesch, H.; Kenney, S.; Boyd, M. R. J. Natl. Cancer Inst. 1990, 82,
1107.
4. Ure, I.; Niederecker, C.; Nussbaumer, P.; Winiski, A. P.; Hegemann, L.; Stütz, A.;
Wolff, K. Skin Pharmacol. 1996, 9, 166.
5. Onoda, T.; Iinuma, H.; Sasaki, Y.; Hamada, M.; Isshiki, K.; Naganawa, H.;
Takeuchi, T.; Tatsuta, K.; Umezawa, K. J. Nat. Prod. 1989, 52, 1252.
6. Smyth, M. S.; Stefanova, I.; Hartmann, F.; Horak, I. D.; Osherov, N.; Levitzky, A.;
Burke, T. R. J. J. Med. Chem. 1993, 36, 3010.
7. Nussbaumer, P.; Winiski, A. P.; Stütz, A. Scient. Pharm. 1996, 64, 601.
8. Kucerovy, A.; Li, T.; Prasad, K.; Repic, O.; Blacklock, T. J. Org. Process Res. Dev.
1997, 1, 287.
24. Hamel, E.; Blokhin, A. V.; Nagle, D. G.; Yoo, H.-D.; Gerwick, W. H. Drug Dev. Res.
1995, 34, 110.
25. Tiwari, S. C.; Suprenant, K. A. Anal. Biochem. 1993, 215, 96.
26. Hamel, E.; Lin, C. M. Arch. Biochem. Biophys. 1981, 209, 29.
27. Clark, J. I.; Garland, D. J. Cell Biol. 1978, 76, 619.
9. Eichinger, K.; Nussbaumer, P.; Balkan, S.; Schulz, G. Synthesis 1987, 1061.
10. Cushman, M.; Nagarathnam, D.; Gopal, D.; Chakraborti, A. K.; Lin, C. M.; Hamel,
E. J. Med. Chem. 1991, 34, 2579.