Synthesis, anticancer activity, and inhibition of tubulin polymerization by conformationally restricted analogues of lavendustin A

Article abstract of DOI:10.1021/jm020292+

Compounds in the lavendustin A series have been shown to inhibit both protein-tyrosine kinases (PTKs) and tubulin polymerization. Since certain lavendustin A derivatives can exist in conformations that resemble both the trans-stilbene structure of the PTK

Full text of DOI:10.1021/jm020292+

1670
J . Med. Chem. 2003, 46, 1670-1682
Syn th esis, An tica n cer Activity, a n d In h ibition of Tu bu lin P olym er iza tion by
Con for m a tion a lly Restr icted An a logu es of La ven d u stin A
Fanrong Mu,† Ernest Hamel,‡ Debbie J . Lee,‡ Donald E. Pryor,‡ and Mark Cushman*,†
Department of Medicinal Chemistry and Molecular Pharmacology, School of Pharmacy and Pharmacal Sciences,
Purdue University, West Lafayette, Indiana 47907, and Screening Technologies Branch, Developmental Therapeutics Program,
Division of Cancer Treatment and Diagnosis, National Cancer Institute at Frederick, National Institutes of Health,
Frederick, Maryland 21702
Received J uly 8, 2002
Compounds in the lavendustin A series have been shown to inhibit both protein-tyrosine
kinases (PTKs) and tubulin polymerization. Since certain lavendustin A derivatives can exist
in conformations that resemble both the trans-stilbene structure of the PTK inhibitor
piceatannol and the cis-stilbene structure of the tubulin polymerization inhibitor combretastatin
A-4, the possibility exists that the ratio of the two types of activities of the lavendustins could
be influenced through the synthesis of conformationally restricted analogues. Accordingly, the
benzylaniline structure of a series of pharmacologically active lavendustin A fragments was
replaced by either their cis- or their trans-stilbene relatives, and effects on both inhibition of
tubulin polymerization and cytotoxicity in cancer cell cultures were monitored. Both dihydro-
stilbene and 1,2-diphenylalkyne congeners were also prepared and evaluated biologically.
Surprisingly, conformational restriction of the bridge between the two aromatic rings of the
lavendustins had no significant effect on biological activity. On the other hand, conversion of
the three phenolic hydroxyl groups of the lavendustin A derivatives to their corresponding
methyl ethers consistently abolished their ability to inhibit tubulin polymerization and usually
decreased cytotoxicity in cancer cell cultures as well, indicating the importance of at least one
of the phenolic hydroxyl groups. Further investigation suggested that the phenolic hydroxyl
group in the salicylamide ring was required for activity, while the two phenol moieties in the
hydroquinone ring could be methylated with retention of activity. Two of the lavendustin A
derivatives displayed IC50 values of 1.4 µM for inhibition of tubulin polymerization, which ranks
them among the most potent of the known tubulin polymerization inhibitors.
In tr od u ction
antiproliferative activity in HaCaT cells but was inac-
tive as a PTK inhibitor when tested on the EGFR
tyrosine kinase.9 The antiproliferative activity of 6 was
subsequently found to be due to its antimitotic activity,
and it was also determined to be comparable in anti-
cancer activity to mitomycin C when tested in nude mice
bearing A431 human vulvar carcinomas.10 Earlier work
with the lavendustin C6 analogue 4 had also suggested
that its antiproliferative effects might be related to
actions on cellular targets that are unrelated to protein-
tyrosine kinases.4
The Streptomyces griseolavendus metabolite laven-
dustin A (1) was originally isolated and characterized
as a protein-tyrosine kinase (PTK) inhibitor when
tested vs the epidermal growth factor receptor (EGFR)
in an A431 cell membrane fraction.1 Although initially
thought to be strictly competitive at the ATP binding
site and noncompetitive at the substrate binding site,
a more detailed kinetic analysis indicated that laven-
dustin A is a hyperbolic mixed-type inhibitor with
respect to both ATP and the substrate.2 The first
synthesis of lavendustin A passed through the inter-
mediate 2, which was found to be as active as 1 as a
PTK inhibitor.1 Although the high polarity of 1 pre-
cludes it from penetrating cellular membranes, several
of its more lipophilic analogues, including the methyl
ester 33 and the N-n-hexyl derivative 4,4,5 are effective
PTK inhibitors in cellular systems. In addition, a large
number of ester and amide derivatives of the active
lavendustin A fragment 2 have been synthesized and
found to inhibit EGF-stimulated DNA synthesis in ER
22 cells, with the 3-phenyl-n-propyl ester 5 being among
the best compounds in the series.6-8 These results led
to the synthesis of compound 6, which displayed strong
Recently, a series of amide derivatives of the phar-
macologically active lavendustin A fragment 2 were
synthesized for evaluation as EGFR and Syk tyrosine
kinase inhibitors.11 The most cytotoxic compound in the
series was the N-(4-fluorophenethyl)salicylamide de-
rivative 7, which yielded a GI50 mean-graph midpoint
(MGM) value of 0.3 µM when tested in the National
Cancer Institute (NCI) cytotoxicity screen of approxi-
mately 55 human cancer cell lines. However, a COM-
PARE analysis12 of the cytotoxicity profile of 7 revealed
that it was likely to be acting on tubulin.11 When tested
for inhibition of tubulin polymerization in a cell-free
system, the amide 7 and related amides were in fact
determined to be active.11 Compound 7 and many of the
other derivatives were also found to be active against
both the nonreceptor PTK Syk and the receptor PTK
EGFR. Therefore, both tubulin polymerization inhibi-
* To whom correspondence should be addressed. Phone: 765-494-
1465. Fax: 765-494-1414. E-mail: cushman@pharmacy.purdue.edu.
† Purdue University.
‡ National Institutes of Health.
10.1021/jm020292+ CCC: $25.00 © 2003 American Chemical Society
Published on Web 04/02/2003

Analogues of Lavendustin A
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 9 1671
lated receptors.4 (4) The potencies of amide 7 and its
analogues as inhibitors of both the EGF receptor and
the nonreceptor protein-tyrosine kinase Syk did not
correlate particularly well with their cytotoxic activities
against cancer cell cultures.11
The activities of the lavendustin A amide derivatives
7 and related compounds as inhibitors of both PTKs and
tubulin polymerization are intriguing. Compound 7 does
bear some structural resemblance to the PTK inhibitors
piceatannol (8)14 and structurally related polyhydroxy-
lated trans-stilbenes.15,16 On the other hand, the struc-
tural relationship between the lavendustin A analogues
and the tubulin polymerization inhibitor combretastatin
A-4 (9),17 as well as the structurally allied polymethoxy-
lated cis-stilbene, dihydrostilbene, and benzylaniline
tubulin polymerization inhibitors, is apparent.9,18,19
Whereas combretastatin A-4 (9) and a large number of
polymethoxylated cis-stilbene analogues are potent tu-
bulin polymerization inhibitors that are highly cytotoxic
in human cancer cell cultures, their corresponding trans
isomers, including trans-combretastatin A-4 (10), are
inactive as tubulin polymerization inhibitors and are
less cytotoxic.18 This suggests that the PTK inhibitory
activity of the lavendustin A amide analogues could
reside in the anti conformation of the linkage between
the two aromatic rings, as displayed in structure 7a ,
while the syn conformation, as shown in structure 7b,
might be responsible for their antitubulin effects. Since
the cytotoxicity of the lavendustin A amide analogues,
including 7, seems to be due to inhibition of tubulin
polymerization and not PTK inhibition, we decided to
synthesize a series of conformationally restricted lav-
endustin A amide analogues and monitor their effects
on tubulin polymerization.
Design of Con for m a tion a lly Restr icted
An a logu es
To restrict the conformation of the linker connecting
the two aromatic rings in the pharmacologically active
lavendustin A derivative 7 in an anti conformation to
create a series of analogues that would more closely
resemble piceatannol (8), a series of trans-stilbenes 11
was designed. A number of dihydro analogues 12, which
would have more conformational freedom, were also
planned, since this would provide insight into the
question of whether any difference in activity between
7 and 11 was due to a conformational effect or simply
to the absence or presence of the amino group. Meth-
ylation of the phenolic hydroxyl groups of 7 and restric-
tion of the linker connecting the aromatic rings in a cis
alkene resulted conceptually in the series of analogues
represented by the general structure 13, which bears a
closer resemblance to combretastatin A-4 (9). Demeth-
ylation of the three methoxyl groups of 13 led to series
14, which would more closely resemble 7 but would
approximate the syn conformation 7b. Finally, replace-
ment of the cis alkene linker of 13 with a linear alkyne
linker would result in series 15, and a similar structural
transformation of 14 would provide series 16. A com-
parison of the pharmacological activities of the phenolic
compounds 11, 12, 14, and 16 should provide a rigorous
probe of the optimal spatial orientation of the two
aromatic rings, and the methyl ethers 13 and 14 should
tory activity and PTK inhibitory activity were found to
reside in the same molecules. However, it seems un-
likely for the following four reasons that the cytotoxici-
ties of these lavendustin A amide analogues are due to
protein-tyrosine kinase inhibition: (1) Whereas MCF-
10A cells are EGF-dependent while MCF-7 cells are not,
lavendustin A amides were equally effective as inhibi-
tors of DNA synthesis in these two breast cancer cell
lines.11 (2) Since almost all of the EGFR tyrosine kinase
activity must be inhibited before effects are seen on cell
growth,13 it is unlikely that the potencies of amide 7
and its congeners as EGFR inhibitors could possibly be
responsible for the effects seen on cancer cell growth
because the IC50 values for EGF PTK inhibition are
close to their IC50 values for cytotoxicity.11 (3) The fact
that lavendustin A (1) did not inhibit the PTK activity
of the mutant protein pp60F527, but nevertheless did
exhibit antiproliferative activity, suggests that the
antiproliferative effects of 1 could be due to actions on
cellular targets downstream from pp60F527 or on unre-

1672 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 9
Mu et al.
Sch em e 1a
a
Reagents and conditions: (a) RNH2, DCC, pyridine, 70-80 °C
(5 h).
Sch em e 2a
a
Reagents and conditions: (a) BnBr, Me2CO, K2CO3, reflux (12
h); (b) (1) MePPh3Br, n-BuLi, THF 23 °C (1.5 h), (2) 20, 23 °C (1
h), 70 °C (6 h); (c) 18a -c, Pd(OAc)2, Et3N, tri-o-tolyphosphine,
CH3CN, 100 °C (30 h); (d) BBr3, CH2Cl2, -78 to 23 °C (4 h); (e)
H2, Pd/C, EtOAc (12 h).
provide insight into the question of whether the tubulin
polymerization inhibitory activity of the lavendustins
could be maximized by converting them to compounds
that more closely resemble combretastatin A-4 (9).
reaction. Heck coupling of 21 with the three iodosali-
cylamides 18a -c was performed at high temperature
in acetonitrile in the presence of triethylamine, using
Pd(OAc)2 and tri-o-tolylphosphine as the catalyst. As
expected from the Heck reaction, under these reaction
conditions, the trans isomers 22a -c were obtained
exclusively. Cleavage of the benzyl groups from 22a -c
with boron tribromide at low temperature in methylene
chloride afforded the desired trans-styrenes 11a -c. The
formation of the trans isomers 11a -c was confirmed
by the characteristic 1H NMR coupling constant of 16.5-
18.1 Hz observed for the two alkene protons. Catalytic
hydrogenation of 22a -c provided the corresponding
compounds 12a -c resulting from cleavage of the two
benzyl groups and reduction of the double bond.
Ch em istr y
In the design of the piceatannol-lavendustin A hybrid
molecules 11, the aminomethylene linkage between the
two aromatic rings of the pharmacologically active
lavendustin A fragment 7 is replaced by a trans alkene
linker, which could possibly be accomplished by a Heck
coupling reaction. Aryl iodide coupling partners were
constructed as shown in Scheme 1. Treatment of com-
mercially available 5-iodosalicylic acid (17) with phen-
ethylamine, 4-fluorophenethylamide, and hexylamine,
in the presence of DCC, provided the required synthetic
intermediates 18a -c.
The other coupling partner, 2,5-dibenzyloxystyrene
(21), was prepared in two steps from 2,5-dihydroxy-
benzaldehyde as outlined in Scheme 2. First, the two
phenolic hydroxyl groups were protected as benzyl
ethers by treatment with benzyl bromide in acetone,
with potassium carbonate as the base, to yield 20,20
which was then converted to 21 through the Wittig
The alkyne 27 (Scheme 3) was envisaged as an
intermediate in the synthesis of the cis alkene systems
13 and 14 as well as the alkynes 15 and 16, and the
Sonogashira coupling was investigated as the key step
in its assembly. The required monosubstituted alkyne
intermediate 25 was obtained by Sonogashira coupling

Analogues of Lavendustin A
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 9 1673
Sch em e 3a
Sch em e 4a
a
Reagents and conditions: (a) HCtC-Si(CH3)3, PdCl2, PPh3,
CuI, piperidine, reflux (24 h); (b) n-Bu4NF, THF, H2O, 23 °C
(3 h); (c) (CH3)2SO4, K2CO3, Me2CO, reflux (12 h); (d) (PPh3)2PdCl2,
CuI, Et2NH, 50-60 °C (8 h).
of trimethylsilylacetylene with 1-bromo-2,5-dimethoxy-
benzene (23) in the presence of PdCl2, PPh3, and CuI
in refluxing piperidine, followed by desilylation of
intermediate 24 with tetra-n-butylammonium fluoride
in THF. Meanwhile, 5-iodosalicylic acid (17) was con-
verted to methyl 5-iodo-2-methoxybenzoate (26) with
dimethyl sulfate under basic conditions. A second So-
nogashira coupling of 25 and 26 under the same reaction
conditions afforded the alkyne intermediate 27.
As shown in Scheme 4, in an iodobenzene system
containing a free hydroxyl group, the attempt to use an
analogous Sonogashira approach to the synthesis of the
alkyne 28 led to an unexpected result. Instead of
obtaining the alkyne 28, another compound was isolated
that was assigned the ketone structure 29 on the basis
of the 1H and 13C NMR data, along with chemical
ionization mass spectrometry and elemental analysis.
The position of the carbonyl group in 29 was subse-
quently verified by X-ray crystallography, and the
resulting ORTEP diagram is displayed in Figure 1.
a
Reagents and conditions: (a) PdCl2, PPh3, CuI, piperidine,
reflux (24 h); (b) H2, Pd/C, EtOAc, 23 °C (4 h).
Hydrogenation of 29 in the presence of palladium on
carbon effectively removed the carbonyl group to afford
the reduction product 30.
The formation of 29 most likely involves initial
formation of the alkyne, followed by a regioselective
palladium-catalyzed hydration. Oxypalladation reac-
tions of alkynes are known,21,22 including the regiose-
lective hydration of alkynones by palladium catalysis
F igu r e 1. ORTEP diagram resulting from X-ray crystallographic analysis of ketone 29.

1674 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 9
Mu et al.
Sch em e 5
Sch em e 6a
a
Reagents and conditions: (a) NaOH, MeOH, 60-70 °C (3 h);
(b) H2, Pd/CaCO3 (poisoned with Pb), EtOH, EtOAc, 23 °C (2 h);
(c) RNH2, Et3N, EDCI, HOBT, DMF, 23 °C (48 h); (d) BBr3,
CH2Cl2, 0 °C (3 h); (e) BBr3, CH2Cl2, -20 °C (2 h).
Sch em e 7a
to form ketones.23 In the present case, the regioselec-
tivity of the reaction is influenced by the substitution
pattern on both aromatic rings. As proposed in Scheme
5, this could possibly be explained by the conversion of
the initially formed alkyne 28 to the palladium-
substituted allene 31, followed by attack of water on the
quinone methide moiety to form 32.22 Protonation of 32
and loss of palladium would then generate the enol,
which would tautomerize to the observed ketone 29. The
regioselectivity of the reaction is then explained by the
participation of the phenol in the formation of the
palladium-substituted allene 31.
After the alkyne 27 was obtained in reasonable
quantities, trihydroxylated and trimethoxylated alkynes
and cis alkenes were synthesized as outlined in Scheme
6. The methyl ester present in 27 was first hydrolyzed
to the acid 34, which was then converted to the corre-
sponding amide analogues 15a -c by reaction with the
required amines in the presence of 1-(3-dimethylami-
nopropyl)-3-ethylcarbodiimide hydrochloride (EDCI) and
1-hydroxybenzotriazole hydrate (HOBT). Demethylation
of 15a with boron tribromide in methylene chloride then
afforded the triphenol 16a . On the other hand, hydro-
genation of the alkyne 27 using Lindlar catalyst (5%
Pd on CaCO3, poisoned with Pb) led to the cis alkene
35, which was demethylated as before to afford the
triphenol 36.
a
Reagents and conditions: (a) H2, Pd/CaCO3 (poisoned with Pb),
EtOAc, EtOH 23 °C (2 h); (b) BBr3, CH2Cl2, -20 °C (2 h).
substituted alkenes 13a -c, which were demethylated
to provide the correspinding triphenols 14a -c.
Biologica l Resu lts a n d Discu ssion
The lavendustin A analogues were examined for
antiproliferative activity against the human cancer cell
lines in the National Cancer Institute cytotoxicity
screen, in which the activity of each compound was
evaluated using approximately 55 different cancer cell
As outlined in Scheme 7, hydrogenation of the alkynes
15a -c using the Lindlar catalyst yielded the three cis

Analogues of Lavendustin A
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 9 1675
Ta ble 1. Cytotoxicities and Antitubulin Activities of Lavendustin A Analogues
cytotoxicity (GI50 in µΜ)a
inhibition of
tubulin
leukemia
lung
colon
CNS
melanoma ovarian
renal prostate
breast
polymerizationc
compd CCRF-CEM HOP-62 HCT-116 SF-539 UACC-62 OVCAR-3 SN12C DU-145 MDA-MB-435 MGMb (IC50, µM ( SD)
7
0.29
2.1
1.8
12
9.3
8.8
9.2
1.8
1.9
9.2
2.4
2.2
7.8
2.0
2.1
11
2.9
2.7
7.7
5.1
5.6
18
10.1
11.0
13
4.0
3.8
0.35
3.8
3.5
6.4 ( 0.7
4.0 ( 0.6
2.9 ( 0.1
5.0 ( 0.8
5.5 ( 1
9.7 ( 2
6.2 ( 0.07
>40
11a
11b
11c
12a
12b
12c
13a
13b
13c
14a
14b
14c
15a
15b
15c
16a
27
13.9
2.6
7.4
12.5
1.0
3.6
47.3
3.8
21.1
3.5
3.8
3.9
12.0
4.7
6.9
10.5
37.7
4.0
5.3
32.3
20.1
11.0
2.8
3.3
16.4
3.5
25.0
2.1
2.3
5.8
24.4
6.0
56.8
1.9
4.4
1.0
2.4
5.0
4.2
16.1
1.0
2.1
1.7
6.7
3.0
49.8
2.1
2.2
14.5
8.0
7.4
3.9
2.3
3.2
3.0
2.1
10.3
1.7
12.9
4.5
2.5
1.2
6.4
2.8
4.4
0.4
5.1
12.3
3.4
24.7
0.2
1.8
2.2
>100
>100
>100
6.4 >100
63.1 45.6
>100
38.9
8.1
31.9
3.0
6.3
8.3
5.9
23.6
1.6
3.8
3.3
4.5
16.6
1.4
3.5
4.1
40.7
11.5
31.6
3.8
30.9
3.6
6.9
18.6
15.5
>40
>40
6.8
4.0
11.6
1.8
15.2
10.5
5.4 ( 0.6
4.2 ( 1
4.2 ( 0.4
>40
1.3
1.2
>100
>100
>100
>100
>100
>100
>100
>100
4.7
7.9
0.3
36.8
2.4
3.2
4.9
11.8
10.3
>40
>40
3.7 ( 0.08
>40
>100
>100
>100
9.6
81.1
2.8
13.6
1.4
3.2
24.4
4.9
5.7
37.9
4.3
15.8
38.2
30.2
12.7
43.6
1.7
4.2
26.9
18.1
2.9
29.3
4.0
6.6
22.6
14.8
>100
29
30
34
35
1.5
15.0
25.6
32.9
0.6
0.5
8.8
1.4 ( 0.08
1.4 ( 0.4
>40
3.2
38.5
10.6
55.0
10.9
>40
a
The cytotoxicity GI50 values are the concentrations corresponding to 50% growth inhibition. Compounds producing a mean-graph
midpoint (MGM) of 10 µM or less on the first determination were tested a second time, and the standard deviations were calculated.
b
Compounds producing a mean-graph midpoint (MGM) greater than 10 µM on the first determination were not evaluated further. Mean-
graph midpoint for growth inhibition of all human cancer cell lines successfully tested. c Minimum of two independent determinations.
Tubulin concentration was 1.2 mg/mL (12 µM).
lines of diverse tumor origins. The mean-graph midpoint
values (MGMs) listed in Table 1 are based on a calcula-
tion of the average GI50 values for all of the cancer cell
lines tested (approximately 55) in which GI50 values
below and above the test range (10-4-10-8 M) are taken
as the minimum (10-8 M) and maximum (10-4 M) drug
concentrations used in the screening test.24 The cyto-
toxicity GI50 values are also listed in Table 1 for nine
representative cell lines, along with the IC50 values for
inhibition of tubulin polymerization.
The evaluation of the compounds as inhibitors of
tubulin polymerization led to some unexpected results.
On the basis of the assumption that our lead compound
7 would adopt a syn conformation 7b in order to more
closely resemble the cis-stilbene structure of combret-
astatin A-4 (9), one would expect the cis-stilbene lav-
endustin A analogues (14a -c) to be more potent as
tubulin polymerization inhibitors than the trans-stil-
bene lavendustin A analogues (11a -c). However, the
results clearly show that the alkene geometry makes
no difference in activity, since all six compounds were
active and had IC50 values in the 2.9-5.4 µM range. The
unimportance of the geometry of the linker connecting
the two aromatic rings was corroborated further by the
activity of the alkyne 16a as a tubulin polymerization
inhibitor (IC50 ) 3.7 µΜ), as well as the activities of the
three compounds having a saturated ethylene linker:
12a (IC50 ) 5.5 µΜ), 12b (IC50 ) 9.7 µΜ), and 12c
(IC50 ) 6.2 µΜ).
ketone 29 is 4- to 5-fold more active than the lead
compound 7.
By examination of the relative potencies of similar
compounds that only vary by having different substit-
uents on the amide nitrogen, it is possible to draw
conclusions about how the phenethyl (substituent “a”),
4-fluorophenethyl (substituent “b”), and n-hexyl (sub-
stituent “c”) groups on the amide influence activity. The
results in Table 1 reveal that the amide substituent has
very little, if any, real effect on the abilities of these
compounds to inhibit tubulin polymerization. Also, with
the exception of 13b, the amide substituent has no
significant effect on cytotoxicity in human cancer cell
cultures.
Methylation of the three phenolic hydroxyl groups in
14a -c might be expected to make these compounds
closer in structure to combretastatin A-4 (9) and con-
ceivably increase their potencies as tubulin polymeri-
zation inhibitors, but the results show that the tri-
methoxylated intermediates 13a -c were completely
inactive as tubulin polymerization inhibitors. These
results indicate the importance of the phenolic hydroxyl
groups for activity. In agreement with this conclusion,
while the triphenolic alkyne 16a is active, the corre-
sponding trimethoxylated derivative 15a is inactive, as
are 15b and 15c. The additional inactive methyl ethers
include alkynes 27 and 34. However, it should be noted
that compounds 29 and 30, which bear two methoxyl
groups in the benzyl aromatic ring and one phenolic
hydroxyl group in the salicylamide aromatic ring, are
the two most potent compounds in the series as inhibi-
tors of tubulin polymerization, which indicates that only
the free phenolic hydroxyl group in the “lower” ring is
critical for activity. The IC50 values for 29 and 30 as
tubulin polymerization inhibitors (both 1.4 µΜ) rank
them with combretastatin A-4 (IC50 ) 1.9 µΜ), podo-
Comparison of the tubulin polymerization inhibitory
activities of the amine 7 (IC50 ) 6.4 µΜ), the hydrocar-
bon 12b (IC50 ) 9.7 µΜ), and the ketone 29 (IC50 ) 1.4
µΜ) provides some insight into the importance of the
amine nitrogen in the benzylamine linker chain of the
lavendustins. The evidence indicates that the basic
nitrogen atom is not necessary for activity, and the

1676 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 9
Mu et al.
phyllotoxin (2.1 µM), and thiocolchicine (1.4 µΜ).18 (In
experiments contemporaneous with those in which these
lavendustin analogues were examined, combretastatin
A-4 (9) yielded a lower IC50 value, 0.94 µM, for inhibition
of tubulin polymerization.)
which abolishes tubulin polymerization inhibitory activ-
ity and also makes them much less cytotoxic.18 These
observations lead to the conclusion that the present
lavendustin derivatives may bind to tubulin in a way
that is structurally distinct from their combretastatin
A-4 relatives. This possibility was explored by further
examining the interaction of lavendustin analogues with
tubulin. The most potent inhibitors of assembly de-
scribed here (compounds 11a b, 14bc, 16a , 29, and 30)
were examined for inhibitory effects on the binding of
[3H]colchicine to tubulin, and compound 29 was the
most active. However, like 7,11 29 was a relatively weak
inhibitor in that significant inhibition required at least
a 10-fold excess of 29 over the [3H]colchicine. In contrast,
50% inhibition of colchicine binding by combretastatin
A-4 (9) is readily demonstrated even when the concen-
tration of 9 is less than the colchicine concentration.27
With the exception of analogue 13b, the cytotoxicities
of the compounds appear to correlate fairly well with
their abilities to inhibit tubulin polymerization. In
general, compounds having MGM values of less than
10 µΜ also had IC50 values of less than 10 µΜ. On the
other hand, except for 13b, all of the compounds with
IC50 values greater than 40 µΜ for inhibition of tubulin
polymerization were less inhibitory for cell growth.
These results indicate that the observed cytotoxicities
of the compounds in this series may in fact be related
to inhibition of tubulin polymerization. Furthermore,
the MGM values of the trans-stilbenes 11a (MGM ) 4.0
µΜ) and 11b (MGM ) 2.9 µΜ) vs their cis isomers 14a
(MGM ) 5.4 µΜ) and 14b (MGM ) 4.2 µΜ) agree with
the antitubulin data in indicating a lack of geometrical
effects on the biological activity. These data contrast
with those reported recently by Pettit et al. for resvera-
trol 37 and its cis isomer 38, which demonstrated that
the trans isomer is more cytotoxic than the cis isomer.25
The same study also documented the fact that the cis
trimethyl ether derivative 39 of cis-resveratrol was
significantly more cytotoxic than resveratrol trimethyl
ether (40). The greater potency of cis-resveratrol tri-
methyl ether (39) vs the trans isomer 40 is in line with
what one would expect from the known structure-
activity data for combretastatin A-4 (9) and its trans
isomer 10.18,26 These results contrast with those of the
present series of stilbenes, which showed no effect of
alkene geometry on biological activity. The unusual lack
of a geometrical effect in the present series may relate
to the presence of the salicylamide moiety.
The binding of colchicine, as well as other colchici-
noids such as thiocolchicine, to tubulin has a number
of unusual properties (for a review, see ref 28).28 Two
features of the drug-tubulin interaction are especially
notable. First, binding is negligible at 0 °C, and even
at higher temperatures, depending on the precise reac-
tion conditions, it can take hours to fully saturate
tubulin with the drug. Second, once bound, colchicine
and thiocolchicine dissociate very slowly from tubulin,
with the half-lives of these tubulin-drug complexes
being at least 24 h at 37 °C.29 As a consequence, in
practical terms, formation of the tubulin-colchicine
complex is generally considered an irreversible reaction.
In contrast, in our experience, binding of most colchi-
cine site compounds to tubulin is substantially more
rapid and reversible than that of colchicinoids. Maxi-
mum inhibition of tubulin assembly by colchicine and
its analogues is enhanced (i.e., a substantially lower IC50
value obtained) by a drug-tubulin preincubation prior
to addition of the GTP required for assembly (for
example, see ref 30).30 The IC50 values obtained for 29,
as well as for combretastatin A-4 (9), changed relatively
little if the preincubation step used in the experiments
presented in Table 1 was omitted (data not presented).
The reversibility of binding of 29 at the colchicine site
was evaluated in a more complex experiment. Because
compound 41, as well as colchicine, fluoresces more
intensely upon binding to tubulin,28 its interaction with
the protein has been studied extensively, with its
association and dissociation parameters well-character-
ized.28 Besides binding rapidly to tubulin, compound 41
dissociates rapidly from the protein. One can therefore
obtain good insight into the behavior of a new colchicine
site drug by comparing its inhibitory effect on colchicine
binding as a function of time to the rapidly dissociating
41 and to the stably bound thiocolchicine.
To summarize the tubulin polymerization results,
when the molecules in the present lavendustin series
were constrained in ways that should have made them
more potent on the basis of the assumption that their
SAR would mimic the SAR observed with combretasta-
tin A-4 and its analogues, either no change in activity
or a large decrease in potency occurred. Especially
surprising was the lack of an effect resulting from
structural changes in the hydrocarbon linker connecting
the two aromatic rings. This is in contrast to the
isomerization of the cis double bond in combretastatin
A-4 (9) and related stilbenes to their trans isomers,
Such a study is shown in Figure 2. In this experiment,
inhibitors and tubulin were preincubated prior to ad-
dition of [3H]colchicine, and the extent of inhibition of
colchicine binding was followed over time. If a drug
binds reversibly at the colchicine site, the extent of
inhibition “decays” over time, with the exact pattern
observed reflecting the combined effects of the inability
of [3H]colchicine to dissociate significantly once bound,
the rate of dissociation of the inhibitor, the binding rate
of colchicine, and the rebinding rate of the inhibitor. The

Analogues of Lavendustin A
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 9 1677
Inhibition of tubulin assembly by 9 and 41 does not
differ greatly from inhibition by 29 (IC50 values of 0.94
and 1.5 µM versus 1.4 µM for 29). If we make the
assumption that similar effects on assembly reflect
similar rapid binding rates, then the dissociation rates
may be roughly proportional to the inhibition decay half-
lives. Since the kf for 41 is 0.04-0.07 s-1 (t1/2 ) 15-25
s),32-34 we can estimate the dissociation rate for 29 as
3-fold higher, or about 0.15 s-1 (t1/2 ) 7 s). It would also
be reasonable to conclude that the relatively weak
inhibitory effects of 29 and other lavendustin analogues
on colchicine binding are derived from rapid dissociation
from the colchicine site.
In conclusion, the present series of compounds has
yielded at least two representatives, 29 and 30, that are
potent tubulin polymerization inhibitors, despite an
apparently weak binding interaction with tubulin at the
colchicine site. The observed cytotoxicities of these
compounds in human cancer cell cultures warrant their
further investigation as potential anticancer agents.
Both compounds are structurally related to the antimi-
totic agent 6, which has displayed in vivo tumor growth
inhibition that is comparable to 5-fluorouracil and
mitomycin C in nude mice.10
F igu r e 2. The experiment was performed essentially as
described previously.36 Reaction mixtures (3.0 mL) were
prepared containing tubulin at 0.1 mg/mL (1.0 µM), either no
inhibitor or the indicated inhibitor at 20 µM with 5% (v/v)
dimethyl sulfoxide (final concentration) as solvent for the
inhibitors, 1.0 M monosodium glutamate (pH 6.6 with HCl),
0.1 M glucose 1-phosphate, 1 mM MgCl2, 1 mM GTP, and 0.5
mg/mL bovine serum albumin. The reaction mixtures were
warmed to 37 °C for 45 min and then chilled on ice. Aliquots
(0.1 mL each) of each reaction mixture were distributed into
test tubes, and a total of 10 µL containing 200 pmol of [3H]-
colchicine was added to each tube. The samples were incubated
for the indicated times at 30 °C. Duplicate data points were
taken for each sample at each time point. For each time point,
reaction mixtures containing each inhibitory drug were com-
pared with the control reaction mixture without inhibitor, with
the data expressed as a percentage of inhibition of colchicine
binding. Symbols represent the following: O, thiocolchicine;
4, compound 41; 0, compound 29.
Exp er im en ta l Section
Melting points were determined with a Mel-Temp capillary
melting point apparatus and are uncorrected. Proton NMR
spectra were recorded on a Bruker VXR-300S or DRX-500
spectrometer. The chemical shift values are expressed in ppm
(parts per million) relative to tetramethylsilane as internal
standard: s ) singlet, d ) doublet, t ) triplet, m ) multiplet,
bs ) broad singlet. Microanalyses were performed at the
Purdue Microanalysis Laboratory, and all values were within
(0.4% of the calculated compositions. Column chromatography
was carried out using Merck silica gel (230-400 mesh).
Analytical thin-layer chromatography (TLC) was performed
on silica gel GF (Analtech) glass-coated plates (2.5 cm × 10
cm with 250 µm layer and prescored), and spots were visual-
ized with UV light at 254 nm. Most chemicals and solvents
were analytical grade and used without further purification.
Commercial reagents were purchased from Aldrich Chemical
Co. (Milwaukee, WI). The authors thank Drs. A. Brossi
(National Institute of Diabetes and Digestive and Kidney
Diseases) and M. G. Banwell (Australian National University)
for gifts of thiocolchicine and compound 41, respectively.
Tu bu lin Assa ys. Electrophoretically homogeneous tubulin
was purified from bovine brain as described previously.35 The
tubulin polymerization and colchicine binding assays were
performed as described previously27,36 except that Beckman
DU7400/7500 spectrophotometers equipped with “high-per-
formance” temperature controllers were used in the former
assay. Unlike the manual control possible with the previously
used Gilford spectrophotometers, the polymerization assays
required use of programs provided by MDB Analytical Associ-
ates, South Plainfield, NJ , since the Beckman instruments are
microprocessor-controlled. Temperature changes were more
rapid than in the Gilford instruments with the jump from 0
to 30 °C taking about 20 s and the reverse jump taking about
100 s.
precise pattern obtained also depends on the inhibitor/
colchicine/tubulin ratio used in the experiment.
In the experiment of Figure 2, the total inhibition of
colchicine binding by thiocolchicine persisted throughout
the experiment, reflecting the tighter binding of thio-
colchicine than colchicine (both faster kf and slower kr),29
as well as the 20:1 inhibitor/colchicine ratio used in this
experiment.
Compound 41, in contrast, was totally inhibitory only
at the earlier time points, and at the final time point, it
inhibited colchicine binding by only about 15%. The half-
life of this decay process was about 3 h under these
experimental conditions.
Compound 29 was not totally inhibitory even at the
earliest time point examined. At 2.5 min after colchicine
addition, the amount of [3H]colchicine bound was re-
duced by 62% in the presence of a 20-fold excess of 29.
The loss of inhibitory effect of 29 was even more rapid
than that of compound 41. The half-life of inhibition by
29 was only 50-60 min, about 3-fold faster than that
of 41. Compound 29 therefore must dissociate from
tubulin even more rapidly than compound 41. In
contrast, combretastatin A-4 dissociates from tubulin
more slowly than does compound 41.31
NIH Cytotoxicity Scr een .24,37 The human tumor cell lines
of the cancer screening panel were grown in RPMI 1640
medium containing 5% fetal bovine serum and 2 mM L-
glutamine. For a typical screening experiment, cells were
inoculated into 96-well microtiter plates in 100 µL at plating
densities ranging from 5000 to 40 000 cells/well depending on
the doubling time of individual cell lines. After cell inoculation,
the microtiter plates were incubated at 37 °C, 5% CO2, 95%
air, and 100% relative humidity for 24 h prior to addition of
experimental drugs.

1678 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 9
Mu et al.
After 24 h, two wells of each cell line were fixed in situ with
TCA to represent a measurement of the cell population for each
cell line at the time of drug addition (Tz). Experimental drugs
were solubilized in dimethyl sulfoxide at 400-fold the desired
final maximum test concentration and stored frozen prior to
use. At the time of drug addition, an aliquot of frozen concen-
trate was thawed and diluted to twice the desired final
maximum test concentration with complete medium contain-
ing 50 µg/mL gentamicin. Additional 4-fold, 10-fold, or 1/2 log
serial dilutions were made to provide a total of five drug
concentrations plus control. Aliquots of 100 µL of these
different drug dilutions were added to the appropriate micro-
titer wells already containing 100 µL of medium, resulting in
the required final drug concentrations.
Following drug addition, the plates were incubated for an
additional 48 h at 37 °C, 5% CO2, 95% air, and 100% relative
humidity. For adherent cells, the assay was terminated by the
addition of cold TCA. Cells were fixed in situ by the gentle
addition of 50 µL of cold 50% (w/v) TCA (final concentration,
10% TCA) and incubated for 60 min at 4 °C. The supernatant
was discarded, and the plates were washed five times with
tap water and air-dried. Sulforhodamine B (SRB) solution (100
µL) at 0.4% (w/v) in 1% acetic acid was added to each well,
and plates were incubated for 10 min at room temperature.
After the samples were stained, the unbound dye was removed
by washing five times with 1% acetic acid, and the plates were
air-dried. Bound stain was subsequently solubilized with 10
mM trizma base, and the absorbance was read on an auto-
mated plate reader at 515 nm. For suspension cells, the
methodology was the same except that the assay was termi-
nated by fixing settled cells at the bottom of the wells by gently
adding 50 µL of 80% TCA (final concentration, 16% TCA).
Using the seven absorbance measurements [time zero, (Tz),
control growth (C), and test growth in the presence of drug at
the five concentration levels (Ti)], the percentage growth was
calculated at each of the drug concentrations levels. Percentage
growth inhibition was calculated as
was isolated as a purple crystalline solid: mp 237-239 °C;
1H NMR (300 MHz, CDCl3) δ 10.39 (bs, 1 H, OH), 9.11 (s, 1 H,
OH), 9.02 (t, J ) 5.13 Hz, 1 H, NH), 8.80 (s, 1 H, OH), 7.93 (d,
J ) 1.50 Hz, 1 H), 7.56 (dd, J ) 8.64, 1.87 Hz, 1 H), 7.30-7.13
(m, 6 H), 7.00 (d, J ) 16.50 Hz, 1 H), 6.90 (d, J ) 2.41 Hz, 1
H), 6.85 (d, J ) 8.68 Hz, 1 H), 6.67 (d, J ) 8.61 Hz, 1 H), 6.51
(dd, J ) 8.61, 2.73 Hz, 1 H), 3.51 (q, J ) 7.07 Hz, 2 H), 2.84 (t,
J ) 7.31 Hz, 2 H); CIMS m/z 376 (MH+). Anal. (C23H21NO4)
C, H, N.
5-[2-(2,5-Dih ydr oxyph en yl)vin yl]-N-[2-(4-flu or oph en yl)-
eth yl]sa licyla m id e (11b). From compound 22b (0.3 g, 0.52
mmol) and BBr3 (3 mL, 1 M solution in CH2Cl2, 3 mmol), 11b
was prepared by a procedure similar to that described for 11a .
Compound 11b (0.145 g, 71%) was a dark-brown solid: mp
232-234 °C; 1H NMR (300 MHz, CDCl3) δ 12.59 (s, 1 H, OH),
9.03 (s, 1 H, OH), 9.01 (t, J ) 3.73 Hz, 1 H, NH), 8.71 (s, 1 H,
OH), 8.01 (d, J ) 1.33 Hz, 1 H), 7.57 (dd, J ) 8.50, 1.52 Hz, 1
H), 7.30 (m, 2 H), 7.24 (d, J ) 18.11 Hz, 1 H), 7.12 (t, J ) 8.93
Hz, 2 H), 6.98 (d, J ) 16.54 Hz, 1 H), 6.88 (m, 2 H), 6.68 (d,
J ) 8.49 Hz, 1 H), 6.52 (dd, J ) 8.54, 2.64 Hz, 1 H), 3.54 (q,
J ) 6.29 Hz, 2 H), 2.87 (t, J ) 7.34 Hz, 2 H); CIMS m/z 394
(MH+). Anal. (C23H20FNO) C, H, F, N.
5-[2-(2,5-Dih yd r oxyp h en yl)vin yl]-N-(n -h exyl)sa licyla -
m id e (11c). From compound 22c (0.35 g, 0.65 mmol) and BBr3
(3 mL, 1 M solution in CH2Cl2, 3 mmol), 11c was prepared by
a procedure similar to that described for 11a . Compound 11c
(0.11 g, 51%) was a purple crystalline solid: mp 218-220 °C;
1H NMR (300 MHz, DMSO-d6) δ 10.56 (s 1 H, OH), 9.04 (s, 1
H, OH), 8.95 (bs, 1 H, NH), 8.72 (s, 1 H, OH), 8.05 (s, 1 H),
7.59 (d, J ) 8.61 Hz, 1 H), 7.28 (d, J ) 16.52 Hz, 1 H), 7.03 (d,
J ) 16.46 Hz, 1 H), 6.89 (m, 2 H), 6.70 (d, J ) 8.52 Hz, 1 H),
6.52 (dd, J ) 8.95, 2.73 Hz, 1 H), 3.32 (m, 2 H), 1.56 (m, 2 H),
1.30 (m, 6 H), 0.87 (t, J ) 6.98 Hz, 3 H); CIMS m/z 356 (MH+).
Anal. (C21H25NO4) C, H, N.
5-[2-(2,5-Dih yd r oxyp h en yl)eth yl]-N-(2-p h en eth yl)sa li-
cyla m id e (12a ). Compound 22a (0.26 g, 0.46 mmol) was
hydrogenated over 10% palladium on activated carbon (wet,
contains 50% water, 0.05 g) in ethyl acetate (15 mL) at room
temperature and atmospheric pressure for 12 h. The mixture
was filtered through Celite and washed with ethyl acetate (10
[(Ti - Tz)/(C - Tz)] × 100
for concentrations for which Ti g Tz and as
[(Ti - Tz)/Tz] × 100
mL). The combined filtrate was evaporated on
evaporator to give the crude product, which was recrystallized
a rotary
from ethyl acetate/hexane. Compound 12a was obtained as an
1
off-white solid (0.15 g, 86%): mp 160-161 °C; H NMR (300
MHz, CDCl3) δ 12.33 (s, 1 H, OH), 8.88 (t, J ) 5.42 Hz, 1 H,
NH), 8.59 (s, 1 H, OH), 8.53 (bs, 1 H, OH), 7.69 (d, J ) 1.74
Hz, 1 H), 7.32-7.17 (m, 6 H), 6.80 (d, J ) 8.38 Hz, 1 H), 6.58
(d, J ) 8.45 Hz, 1 H), 6.46 (d, J ) 2.84 Hz, 1 H), 6.40 (dd, J )
8.42, 2.88 Hz, 1 H), 3.51 (q, bs, 2 H), 2.85 (t, J ) 7.68 Hz, 2
H), 2.69 (s, 4 H); CIMS m/z 378 (MH+). Anal. (C23H23NO4) C,
H, N.
for concentrations for which Ti < Tz.
Three dose-response parameters were calculated for each
experimental agent. Growth inhibition of 50% (GI50) was
calculated from [(Ti - Tz)/(C - Tz)] × 100 ) 50, which is the
drug concentration resulting in a 50% reduction in the net
protein increase (as measured by SRB staining) in control cells
during the drug incubation. The drug concentration resulting
in total growth inhibition (TGI) was calculated from Ti ) Tz.
The LC50 (concentration of drug resulting in a 50% reduction
in the measured protein at the end of the drug treatment
compared to that at the beginning), indicating a net loss of
cells following treatment, was calculated from [(Ti - Tz)/Tz] ×
100 ) -50. Values were calculated for each of these three
parameters if the level of activity was reached; however, if the
effect was not reached or was exceeded, the value for that
parameter was expressed as greater or less than the maximum
or minimum concentration tested.
5-[2-(2,5-Dih ydr oxyph en yl)eth yl]-N-[2-(4-flu or oph en yl)-
eth yl]sa licyla m id e (12b). From compound 22b (0.2 g, 0.35
mmol), H2, and palladium on activated carbon (wet, contains
50% water, 0.05 g), 12b was prepared by a procedure similar
to that described for 12a . Compound 12b was a white solid
1
(99.5 mg, 72%): mp 173-174 °C; H NMR (500 MHz, CDCl3)
δ 12.29 (s, 1 H, OH), 8.85 (t, J ) 5.38 Hz, 1 H, NH), 8.55 (s, 1
H, OH), 8.50 (s, 1 H, OH), 7.68 (s, 1 H), 7.27 (m, 2 H), 7.21 (d,
J ) 8.32 Hz, 1 H), 7.10 (t, J ) 8.68 Hz, 2 H), 6.78 (d, J ) 8.33
Hz, 1 H), 6.57 (d, J ) 8.46 Hz, 1 H), 6.40 (d, J ) 2.78 Hz, 1 H),
6.39 (dd, J ) 8.33, 2.87 Hz, 1 H), 3.49 (q, J ) 6.55 Hz, 2 H),
2.84 (t, J ) 7.24 Hz, 2 H), 2.69 (s, 4 H); CIMS m/z 396 (MH+).
Anal. (C23H22FNO4) C, H, F, N.
5-[2-(2,5-Dih yd r oxyp h en yl)vin yl]-N-(2-p h en eth yl)sa li-
cyla m id e (11a ). BBr3 (3 mL, 1 M solution in CH2Cl2, 3 mmol)
was added dropwise via syringe to compound 22a (0.30 g, 0.54
mmol) in anhydrous CH2Cl2 (10 mL) at -78 °C under argon.
The solution was warmed to room temperature and stirred for
4 h. After the mixture was recooled to -78 °C, H2O (10 mL)
was added, followed by ethyl acetate (20 mL), and the solution
was warmed to room temperature. The resulting solution was
extracted with ethyl acetate (3 × 50 mL), the combined organic
layer was washed with brine (10 mL) and dried over MgSO4,
and the solvent was removed to afford a crude oily residue,
which was further purified by flash chromatography (silica gel
85 g, ethyl acetate/hexane 1:2). The product 11a (99 mg, 49%)
5-[2-(2,5-Dih yd r oxyp h en yl)eth yl]-N-(n -h exyl)sa licyla -
m id e (12c). From compound 22c (0.20 g, 0.37 mmol), H2, and
palladium on activated carbon (wet, contains 50% water, 0.05
g), 12c was prepared by a similar procedure as that described
for 12a . Compound 12c was an off-white crystalline solid: mp
1
150-151 °C; H NMR (300 MHz, DMSO-d6) δ 12.50 (s, 1 H,
OH), 8.82 (t, J ) 4.50 Hz, 1 H, NH), 8.64 (s, 1 H, OH), 8.49 (s,
1 H, OH), 7.75 (s, 1 H), 7.19 (d, J ) 8.55 Hz, 1 H), 6.80 (d,
J ) 8.41 Hz, 1 H), 6.57 (d, J ) 8.82 Hz, 1 H), 6.37 (m, 2 H),
3.27 (m, 4 H), 1.53 (m, 2 H), 1.46 (d, J ) 7.18 Hz, 2 H), 1.28

Analogues of Lavendustin A
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 9 1679
(m, 6 H), 0.86 (t, J ) 6.77 Hz, 3 H); CIMS m/z 358 (MH+).
7.02 (t, J ) 8.68 Hz, 2 H), 6.74 (d, J ) 8.44 Hz, 1 H), 6.73 (d,
J ) 8.34 Hz, 1 H), 6.63 (d, J ) 2.14 Hz, 1 H), 6.60 (m, 1 H),
6.57 (d, J ) 11.72 Hz, 1 H), 6.44 (d, J ) 12.24 Hz, 1 H), 3.62
(q, J ) 6.65 Hz, 2 H), 2.91 (t, J ) 7.21 Hz, 2 H); positive ion
ESIMS m/z 394 (MH+); negative ion ESIMS m/z 392 (M - H)-.
Anal. (C23H20FNO4) C, H, F, N.
5-[(Z)-2-(2,5-Dih yd r oxyp h en yl)vin yl]-N-(n -h exyl)sa li-
cyla m id e (14c). From compound 13c (0.22 g, 0.55 mmol) and
BBr3 (2.5 mL, 1 M solution in CH2Cl2, 2.5 mmol), 14c was
prepared by a similar procedure as that described for 14a .
Compound 14c (0.13 g, 67%) was a white solid: mp 180-181
°C; 1H NMR (500 MHz, acetone-d6) δ 10.68 (s, 1 H, OH), 7.96
(bs, 1 H, NH), 7.69 (d, J ) 1.42 Hz, 1 H), 7.68 (s, 1 H, OH),
7.60 (s, 1 H, OH), 7.32 (dd, J ) 8.55, 1.74 Hz, 1 H), 6.72 (d,
J ) 8.50 Hz, 1 H), 6.70 (d, J ) 8.62 Hz, 1 H), 6.63 (d, J ) 2.75
Hz, 1 H), 6.58 (dd, J ) 8.53, 3.00 Hz, 1 H), 6.56 (d, J ) 12.61
Hz, 1 H), 6.46 (d, J ) 12.23 Hz, 1 H), 3.39 (q, J ) 2.95 Hz, 2
H), 1.60 (quint, J ) 7.10 Hz, 2 H), 1.34 (m, 6 H), 0.88 (t, J )
6.52 Hz, 3 H); positive ion ESIMS m/z 356 (MH+), negative
ion ESIMS m/z 354 (M - H)-. Anal. (C21H25NO4) C, H, N.
Anal. (C21H27NO4) C, H, N.
5-[(Z)-2-(2,5-Dim eth oxyp h en yl)vin yl]-2-m eth oxy-N-(2-
p h en eth yl)ben za m id e (13a ). Compound 15a (0.41 g, 0.99
mmol) was hydrogenated over 5% palladium on calcium
carbonate, poisoned with lead (30 mg), in 3:1 ethanol/ethyl
acetate (40 mL) at room temperature and atmospheric pres-
sure for 2 h. The mixture was filtered through Celite and
washed with ethyl acetate (20 mL). The combined filtrate was
evaporated on a rotary evaporator to give the crude product.
Purification by flash chromatography (ethyl acetate/hexane
1:4) led to the recovery of the starting material (0.21 g) and
afforded the product 13a (0.20 g, 51%) as a white solid: mp
93-94 °C; 1H NMR (500 MHz, CDCl3) δ 8.12 (d, J ) 2.25 Hz,
1 H), 7.78 (bs, 1 H, NH), 7.33-7.31 (m, 1 H), 7.29-7.22 (m, 5
H), 6.80 (d, J ) 8.50 Hz, 1 H), 6.76 (d, J ) 2.91 Hz, 1 H), 6.72
(dd, J ) 7.41, 2.84 Hz, 1 H), 6.68 (d, J ) 8.66 Hz, 1 H), 6.63
(d, J ) 12.41 Hz, 1 H), 6.59 (d, J ) 12.34 Hz, 1 H), 3.80 (s, 3
H), 3.75 (t, J ) 6.70 Hz, 2 H), 3.70 (s, 3 H), 3.56 (s, 3 H), 2.92
(t, J ) 6.72 Hz, 2 H); ESIMS 418 (MH+), 440 (M + Na+). Anal.
(C26H27NO4) C, H. N.
5-[2-(2,5-Dim et h oxyp h en yl)et h yn yl]-2-m et h oxy-N-(2-
p h en eth yl)ben za m id e (15a ). To a solution of 34 (0.45 g, 1.44
mmol) in dry DMF (30 mL) was added EDCI (0.40 g, 2.1
mmol), HOBT (0.28 g, 2.1 mmol), and Et3N (0.89 mL, 6.3
mmol). After the mixture was stirred at room temperature for
1 h, 2-phenethylamine (0.87 mL, 6.4 mmol) was added drop-
wise, and the reaction continued for 48 h at room temperature
under argon. Water (200 mL) was then added, and the product
was extracted with ethyl acetate (3 × 100 mL). The combined
organic extracts were washed with brine (1 × 100 mL), dried
over sodium sulfate, and filtered, and the solvent was removed.
Purification was achieved by flash chromatography, eluting
with ethyl acetate/hexane (1:1 by volume) to yield pure 15a
(0.43 g, 73%) as a white crystalline solid: mp 112-113 °C; 1H
NMR (500 MHz, CDCl3) δ 8.35 (d, J ) 1.97 Hz, 1 H), 7.72 (bs,
1 H, NH), 7.53 (dd, J ) 8.50, 2.04 Hz, 1 H), 7.34-7.19 (m, 5
H), 6.97 (d, J ) 2.67 Hz, 1 H), 6.80 (d, J ) 8.54 Hz, 1 H), 6.78
(dd, J ) 8.60, 2.85 Hz, 1 H), 6.75 (d, J ) 8.64 Hz, 1 H), 3.78
(s, 3 H), 3.65-3.76 (m, 8 H), 2.78 (t, J ) 6.71 Hz, 2 H); ESIMS
m/ z 416 (MH+). Anal. (C26H25NO4) C, H, N.
5-[(Z)-2-(2,5-Dim eth oxyph en yl)vin yl]-N-[(4-flu or oph en -
yl)eth yl]-2-m eth oxyben za m id e (13b). From compound 15b
(0.4 g, 0.9 mmol), H2, and 5% palladium on calcium carbonate
poisoned with lead (10 mg), 13b was prepared by a procedure
similar to that described for 13a . Compound 13b (0.38 g, 97%)
1
was a yellowish liquid. H NMR (500 MHz, CDCl3) δ 8.13 (d,
J ) 1.54 Hz, 1 H), 7.80 (bs, 1 H, NH), 7.32 (dd, J ) 1.57, 8.59
Hz, 1 H), 7.23 (m, 2 H), 7.03 (t, J ) 8.51 Hz, 2 H), 6.84 (d,
J ) 8.72 Hz, 1 H), 6.80-6.75 (m, 2 H), 6.71 (d, J ) 8.60 Hz,
1 H), 6.66 (d, J ) 12.32 Hz, 1 H), 6.60 (d, J ) 12.34 Hz, 1 H),
3.80 (s, 3 H), 3.76 (s, 3 H), 3.72 (q, J ) 6.40 Hz, 2 H), 3.58 (s,
3 H), 2.91 (t, J ) 6.92 Hz, 2 H); ESIMS 436 (MH+), 458 (M +
Na+). Anal. (C26H26FNO4) C, H, F, N.
5-[(Z)-2-(2,5-Dim eth oxyp h en yl)vin yl]-2-m eth oxy-N-(n -
h exyl)ben za m id e (13c). From compound 15c (0.43 g, 1.09
mmol), H2, and 5% palladium on calcium carbonate poisoned
with lead (10 mg), 13c was prepared by procedure similar to
that described for 13a . Compound 13c (0.38 g, 97%) was a
1
yellow liquid. H NMR (300 MHz, CDCl3) δ 8.06 (d, J ) 2.37
Hz, 1 H), 7.72 (bs, 1 H, NH), 7.25 (dd, J ) 8.63, 2.42 Hz, 1 H),
6.77-6.68 (m, 4 H), 6.57 (s, 2 H), 3.84 (s, 3 H), 3.73 (s, 3 H),
3.54 (s, 3 H), 3.38 (q, J ) 6.96 Hz, 2 H), 1.55 (quint, J ) 7.21
Hz, 2 H), 1.28 (m, 6 H), 0.85 (t, J ) 7.02 Hz, 3 H); EIMS m/z
397 (M+); CIMS m/z 398 (MH)+. Anal. (C24H31NO4) C, H, N.
5-[2-(2,5-Dim eth oxyp h en yl)eth yn yl]-N-[(4-flu or op h en -
yl)eth yl]-2-m eth oxyben za m id e (15b). From compound 33
(0.45 g, 1.44 mmol), EDCI (0.40 g, 2.08 mmol), HOBT (0.28 g,
2.08 mmol), and Et3N (0.87 mL, 6.3 mmol), 15b was prepared
by a procedure similar to that described for 15a . Compound
15b was a white crystalline solid (0.45 g, 73%): mp 105-106
5-[(Z)-2-(2,5-Dih yd r oxyp h en yl)vin yl]-N-(2-p h en eth yl)-
sa licyla m id e (14a ). BBr3 (2.0 mL, 1 M solution in CH2Cl2,
2.0 mmol) was added dropwise via syringe to compound 13a
(0.18 g, 0.43 mmol) in anhydrous CH2Cl2 (30 mL) at -78 °C
under argon. The solution was warmed at -20 °C and stirred
for 2 h. After the mixture was recooled to -78 °C, H2O (10
mL) was added, followed by ethyl acetate (30 mL), and the
solution was warmed to room temperature. The resulting
solution was extracted with ethyl acetate (3 × 50 mL), and
the combined organic layer was washed with brine (10 mL)
and dried over MgSO4. Removal of the solvent gave a crude
oily residue, which was further purified by flash chromatog-
raphy (silica gel 80 g, ethyl acetate/hexane 1:5). The product
14a (0.079 g, 49%) was isolated as a solid: mp 179-180 °C;
1H NMR (500 MHz, CDCl3) δ 12.60 (s, 1 H, OH), 12.32 (s, 1 H,
OH), 9.03 (bs, 1 H, NH), 9.00 (s, 1 H, OH), 8.01 (d, J ) 1.60,
1 H), 7.56 (dd, J ) 8.55, 1.74 Hz, 1 H), 7.13-7.19 (m, 6 H),
6.97 (d, J ) 16.47 Hz, 1 H), 6.68 (m, 2 H), 6.66 (d, J ) 8.59
Hz, 1 H), 6.51 (dd, J ) 8.55, 2.78 Hz, 1 H), 3.53 (q, J ) 6.69
Hz, 2 H), 2.88 (t, J ) 7.62 Hz, 2 H); CIMS m/z 376 (MH+).
Anal. (C23H21NO4) C, H, N.
1
°C; H NMR (500 MHz, CDCl3) δ 8.41 (d, J ) 2.15 Hz, 1 H),
7.75 (bs, 1 H, NH), 7.62 (dd, J ) 8.50, 2.17 Hz, 1 H), 7.24 (m,
2 H), 7.03 (m, 3 H), 6.92 (d, J ) 8.59 Hz, 1 H), 6.85 (dd, J )
8.96, 2.88 Hz, 1 H), 6.85 (d, J ) 9.00 Hz, 1 H), 3.88 (s, 3 H),
3.82 (s, 3 H), 3.80 (s, 3 H), 3.74 (q, J ) 5.96 Hz, 2 H), 2.93 (t,
J ) 6.71 Hz, 2 H); EI m/z 433 (M+); CIMS m/z 434 (MH+).
Anal. (C26H24FNO4) C, H, F, N.
5-[2-(2,5-Dim et h oxyp h en yl)et h yn yl]-2-m et h oxy-N-(n -
h exyl)ben za m id e (15c). From compound 33 (0.48 g, 1.53
mmol), EDCI (0.60 g, 3.14 mmol), HOBT (0.42 g, 3.14 mmol),
and Et3N (0.87 mL, 6.3 mmol), 15c was prepared by a
procedure similar to that described for 15a . Compound 15c
was a white crystalline solid (0.57 g, 92%): mp 90-91 °C; 1H
NMR (300 MHz, CDCl3) δ 8.38 (d, J ) 2.26 Hz, 1 H), 7.71 (bs,
1 H, NH), 7.59 (dd, J ) 2.27, 8.55 Hz, 1 H), 7.01 (d, J ) 2.58
Hz, 1 H), 6.92 (d, J ) 8.62 Hz, 1 H), 6.84-6.77 (m, 2 H), 3.95
(s, 3 H), 3.84 (s, 3 H), 3.75 (s, 3 H), 3.42 (q, J ) 7.02 Hz, 2 H),
1.58 (quint, J ) 7.32 Hz, 2 H), 1.33 (m, 6 H), 0.89 (t, J ) 6.89
Hz, 3 H); CIMS m/z 396 (MH+). Anal. (C24H29NO4) C, H, N.
5-[2-(2,5-Dih yd r oxyp h en yl)et h yn yl]-2-h yd r oxy-N-(2-
p h en eth yl)ben za m id e (16a ). BBr3 (4.5 mL, 1 M solution in
CH2Cl2, 4.5 mmol) was added dropwise via syringe to com-
pound 15a (0.37 g, 0.89 mmol) in anhydrous CH2Cl2 (30 mL)
at -78 °C under argon. The solution was warmed at 0 °C and
stirred for 3 h. After the mixture was recooled to -78 °C, H2O
(10 mL) was added, followed by ethyl acetate (50 mL), and
the solution was warmed to room temperature. The resulting
solution was extracted with ethyl acetate (3 × 50 mL), and
5-[(Z)-2-(2,5-Dih yd r oxyp h en yl)vin yl]-N-[(4-flu or op h en -
yl)eth yl]sa licyla m id e (14b). From compound 13b (0.20 g,
0.46 mmol) and BBr3 (3.0 mL, 1 M solution in CH2Cl2, 3.0
mmol), 14b was prepared by a similar procedure as that
described for 14a . Compound 14b (0.094 g, 52%) was a solid:
1
mp 181-182 °C; H NMR (500 MHz, CDCl3) δ 12.73 (s, 1 H,
OH), 8.11 (bs, 1 H, NH), 7.69 (s, 1 H, OH), 7.64 (s, 1 H), 7.61
(s, 1 H, OH), 7.33 (d, J ) 8.58 Hz, 1 H), 7.29-7.26 (m, 2 H),

1680 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 9
Mu et al.
the combined organic layer was washed with brine (10 mL)
and dried over MgSO4. Removal of the solvent gave a crude
oily residue, which was further purified by flash chromatog-
raphy (silica gel 95 g, ethyl acetate/hexane 1:1). The product
16a (0.16 g, 48%) was isolated as a solid: mp 192-194 °C; 1H
NMR (500 MHz, CDCl3) δ 12.88 (s, 1 H, OH), 9.19 (s, 1 H,
OH), 9.03 (s, 1 H, NH), 8.90 (s, 1 H, OH), 8.02 (d, J ) 1.59, 1
H), 7.50 (dd, J ) 8.55, 1.64 Hz, 1 H), 7.28-7.23 (m, 5 H), 6.91
(d, J ) 8.51 Hz, 1 H), 6.70 (m, 2 H), 6.63 (dd, J ) 8.54, 2.58
Hz, 1 H), 3.51 (q, J ) 6.70 Hz, 2 H), 2.86 (t, J ) 7.62 Hz, 2 H);
CIMS m/z 374 (MH+). Anal. (C23H19NO4) C, H, N.
5-Iod o-N-(2-p h en eth yl)sa licyla m id e (18a ). To a solution
of 96% 5-iodosalicylic acid (17, 1.37 g, 5.0 mmol) in dry pyridine
(50 mL) was added DCC (1.24 g, 6.0 mmol) and â-phenethyl-
amine (0.63 mL, 5.0 mmol). The mixture was heated at 70-
80 °C for 5 h. On cooling, the precipitate was removed by
filtration. The pyridine was removed on a rotary evaporator.
The residue was treated with 10% HCl (50 mL), and the
mixture was stirred for 30 min. The yellow precipitate was
removed by filtration, washed with water (2 × 50 mL), and
dried in vacuo. Purification was achieved by flash chromatog-
raphy (ethyl acetate/hexane 1:4) to yield 18a as a white
crystalline solid (1.19 g, 63%): mp 143-145 °C; 1H NMR
(CDCl3) δ 12.20 (s, 1 H, OH), 7.64 (dd, J ) 8.71, 2.01 Hz, 1 H),
7.50 (d, J ) 2.09 Hz, 1 H), 7.40 (t, J ) 1.98, 1 H, NH), 7.31-
7.24 (m, 5 H), 6.79 (d, J ) 8.75 Hz, 1 H), 3.71 (q, J ) 6.95 Hz,
2 H), 2.95 (t, J ) 7.00 Hz, 2 H); CIMS m/z 368 (MH+). Anal.
(C15H14INO2) C, H, I, N.
filtration. The filtrate was concentrated in vacuo to afford an
oily residue, which was purified by flash chromatography
(ethyl acetate/hexane 1:5) to obtain a light-yellowish solid 21
(2.33 g, 94%): mp 34-35 °C; 1H NMR (CDCl3) δ 7.45-7.30
(m, 10 H), 7.16 (d, J ) 2.37 Hz, 1 H), 7.12 (d, J ) 6.64 Hz, 1
H), 6.87-6.80 (m, 2 H), 5.73 (dd, J ) 17.71, 1.33 Hz, 1 H),
5.27 (dd, J ) 11.10, 1.35 Hz, 1 H), 5.04 (s, 4 H); CIMS m/z 317
(MH+), 207, 117. Anal. (C22H20O2) C, H.
5-[2-(2,5-Dib en zyloxyp h en yl)vin yl]-N-(2-p h en et h yl)-
sa licyla m id e (22a ). A solution of 21 (0.63 g, 1.99 mmol), 18a
(0.62 g, 1.69 mmol), 98% palladium acetate (0.004 g, 0.018
mmol), 97% tri-o-tolylphosphine (0.023 g, 0.073 mmol), and
Et3N (1 mL, 7.2 mmol) in acetonitrile (10 mL) was heated at
100 °C under argon for 30 h. After the mixture was cooled to
room temperature, H2O (200 mL) was added and the mixture
was stirred for 10 min. It was then extracted with methylene
chloride (2 × 50 mL), and the combined organic layer was
washed with brine (10 mL), dried over MgSO4, and concen-
trated to furnish the crude product, which was further purified
by flash chromatography (silica gel 100 g, ethyl acetate/hexane
1:5). The product 22a (0.31 g, 33%) was isolated as a slightly
1
yellow crystalline solid: mp 126-127 °C; H NMR (CDCl3) δ
12.36 (s, 1 H, OH), 7.55 (dd, J ) 8.69, 1.98 Hz, 1 H), 7.45-
7.22 (m, 17 H), 7.18 (d, J ) 2.80 Hz, 1 H), 6.95 (d, J ) 5.20
Hz, 1 H), 6.94 (d, J ) 2.50 Hz, 1 H), 6.87 (d, J ) 8.86 Hz, 1 H),
6.81 (dd, J ) 8.96, 2.84 Hz, 1 H), 5.08 (s, 2 H), 5.05 (s, 2 H),
3.71 (q, J ) 6.15 Hz, 2 H), 2.93 (t, J ) 6.93 Hz, 2 H); CIMS
m/z 556 (MH+). Anal. (C37H33NO4) C, H, N.
5-Iod o-N-[(4-flu or op h en yl)et h yl]sa licyla m id e (18b ).
From compound 17 (1.38 g, 5.0 mmol), DCC (1.24 g, 6.0 mmol),
and 4-fluoro-2-phenethylamine (0.70 g, 5.0 mmol), 18b was
prepared by a procedure similar to that described for 18a .
Compound 18b was a light-yellow crystalline solid (1.18 g,
61%): mp 153-155 °C; 1H NMR (CDCl3) δ 12.24 (s, 1 H, OH),
7.63 (dd, J ) 8.77, 2.01 Hz, 1 H), 7.49 (d, J ) 2.11 Hz, 1 H),
7.19 (m, 2 H), 7.03 (m, 2 H), 6.78 (d, J ) 8.76 Hz, 1 H), 6.51
(bs, 1 H, NH), 3.67 (q, J ) 6.22 Hz, 2 H), 2.91 (t, J ) 7.08 Hz,
2 H); CIMS m/z 386 (MH+). Anal. (C15H13FINO2) C, H, F, I, N.
5-[2-(2,5-Diben zyloxyph en yl)vin yl]-N-[(4-flu or oph en yl)-
eth yl]sa licyla m id e (22b). From 21 (0.85 g, 2.7 mmol), 18b
(1.1 g, 2.8 mmol), 98% palladium acetate (0.010 g, 0.044 mmol),
97% tri-o-tolylphosphine (0.050 g, 0.18 mmol), and Et3N (1.5
mL, 10.8 mmol), 22b was prepared by a procedure similar to
that described for 22a . Compound 22b (0.64 g, 40%) was a
slightly yellow crystalline solid: mp 145-147 °C; 1H NMR
(CDCl3) δ 12.35 (s, 1 H, OH), 7.59 (dd, J ) 8.66, 1.94 Hz, 1 H),
7.49-7.34 (m, 12 H), 7.27 (d, J ) 1.96 Hz, 1 H), 7.24-7.19 (m,
3 H), 7.06 (d, J ) 8.64 Hz, 1 H), 7.01 (d, J ) 5.93 Hz, 1 H),
6.98 (d, J ) 2.06 Hz, 1 H), 6.92 (d, J ) 8.89 Hz, 1 H), 6.84 (dd,
J ) 8.87, 2.76 Hz, 1 H), 6.40 (t, J ) 3.05 Hz, 1 H, NH), 5.12 (s,
2 H), 5.08 (s, 2 H), 3.71 (q, J ) 6.26 Hz, 2 H), 2.93 (t, J ) 7.07
Hz, 2 H). Anal. (C37H32FNO4) C, H, F, N.
5-Iod o-N-(n -h exyl)sa licyla m id e (18c). From compound
17 (1.38 g, 5.0 mmol), DCC (1.24 g, 6.0 mmol), and n-
hexylamine (0.66 mL, 5.0 mmol), 18c was prepared by a
procedure similar to that described for 18a . Compound 18c
was a light-yellow crystalline solid (0.85 g, 49%): mp 58-60
1
5-[2-(2,5-Diben zyloxyp h en yl)vin yl]-N-(n -h exyl)sa licyl-
a m id e (22c). From 21 (1.45 g, 4.6 mmol), 18c (1.4 g 4.0 mmol),
98% palladium acetate (0.01 g, 0.04 mmol), 97% tri-o-
tolylphosphine (0.05 g, 0.18 mmol), and Et3N (1.5 mL, 10.8
mmol), 22c was prepared by a procedure similar to that
described for 22a . Compound 22c (0.6 g, 30%) was a white
crystalline solid: mp 98-100 °C; 1H NMR (CDCl3) δ 12.44 (s,
1 H, OH), 7.56 (dd, J ) 8.66, 1.79 Hz, 1 H), 7.45-7.33 (m, 10
H), 7.26 (d, J ) 8.43 Hz, 1 H), 7.20 (d, J ) 2.79 Hz, 1 H), 7.03
(d, J ) 16.48 Hz, 1 H), 6.97 (d, J ) 8.63 Hz, 1 H), 6.88 (d, J )
8.93 Hz, 1 H), 6.80 (dd, J ) 8.84, 2.88 Hz, 1 H), 6.30 (t, J )
5.29 Hz, 1 H, NH), 5.08 (s, 2 H), 5.04 (s, 2 H), 3.45 (q, J ) 6.32
Hz, 2 H), 1.62 (quint, J ) 7.10 Hz, 2 H), 1.37 (m, 6 H), 0.89 (t,
J ) 6.58 Hz, 3 H); CIMS m/z 536 (MH+). Anal. (C35H37NO4)
C, H, N.
°C; H NMR (CDCl3) δ 12.37 (s, 1 H, OH), 7.63 (d, J ) 2.03
Hz, 1 H), 7.60 (s, 1 H), 6.78 (d, J ) 8.52 Hz, 1 H), 6.20 (bs, 1
H, NH), 3.42 (q, J ) 6.31 Hz, 2 H), 1.65 (quint, J ) 6.47 Hz,
2 H), 1.37 (m, 6 H), 0.90 (t, J ) 6.61 Hz, 3 H); CIMS m/z 348
(MH+). Anal. (C13H18INO2) C, H, I, N.
2,5-Diben zyloxyben za ld eh yd e (20). K2CO3 (1.4 g, 10
mmol) and benzyl bromide (2.0 mL, 16.8 mmol) were added
to a solution of 2,5-dihydroxybenzaldehyde (19) (1.0 g, 7.1
mmol) in acetone (12 mL). The mixture was heated to reflux
under argon overnight, after which time TLC (silica gel, ethyl
acetate/hexane, 1:2) showed that the reaction was complete.
The potassium carbonate was removed by filtration, and the
acetone was evaporated in vacuo. The solid product 20 (2.05
g, 91%) was recrystallized from diethyl ether/hexane: mp 84-
86 °C; 1H NMR (CDCl3) δ 10.51 (s, 1 H), 7.44 (d, J ) 3.03 Hz,
1 H), 7.42-7.33 (m, 10 H), 7.18 (dd, J ) 9.05, 3.26 Hz, 1 H),
7.01 (d, J ) 9.07 Hz, 1 H), 5.15 (s, 2 H), 5.05 (s, 2 H); CIMS
m/z 319 (MH+), 291, 227, 213. Anal. (C21H18O3) C, H.
2,5-Diben zyloxystyr en e (21). Methyltriphenylphospho-
nium bromide (3.03 g, 8.1 mmol) was placed in a flask fitted
with a dropping funnel and a reflux condenser. The air in the
flask was replaced by argon, and dry, degassed THF (25 mL)
was added to the flask. n-Butyllithium (3.2 mL of a 2.5 M
solution in hexane, 8.0 mmol) was added dropwise to the
suspension, and the mixture was stirred for 1.5 h at room
temperature. Compound 20 (2.5 g, 7.8 mmol) in dry THF (10
mL) was then added, and the resulting mixture was stirred
for 1 h at room temperature. After the solution was stirred
for another 6 h at 70 °C, it was allowed to cool to room
temperature. Addition of diethyl ether (60 mL) and hexane
(60 mL) produced a precipitate, which was removed by
(2,5-Dim eth oxyph en yl)tr im eth ylsilyleth yn e (24). 1-Bro-
mo-2,5-dimethoxybenzene (23, 2.0 g, 9.2 mmol), (trimethylsi-
lyl)acetylene (2.6 mL, 18.4 mmol), PPh3 (0.048 g, 0.18 mmol),
and CuI (0.018 g, 0.009 mmol) were dissolved in piperidine
(35 mL). Palladium chloride (0.032 g, 0.18 mmol) was added,
and the resulting mixture was heated at reflux for 24 h. The
reaction mixture was concentrated in vacuo, and saturated
NaHCO3 (100 mL) was added. The product was extracted with
hexane (3 × 70 mL). The combined hexane extracts were
washed with brine (1 × 20 mL), dried over MgSO4, and
concentrated. The crude product was subjected to column
chromatography (silica gel 50 g, ethyl acetate/hexane 1:6) to
give 24 (1.92 g, 89%) as a brownish liquid, which was used
without further purification. 1H NMR (300 MHz, CDCl3) δ 7.12
(d, J ) 2.26 Hz, 1 H), 6.83 (dd, J ) 8.98, 2.76 Hz, 1 H), 6.77
(d, J ) 9.00 Hz, 1 H), 3.91 (s, 3 H), 3.76 (s, 3 H), 0.32 (s, 9 H).

Analogues of Lavendustin A
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 9 1681
(2,5-Dim eth oxy)p h en yla cetylen e (25). Compound 24
(1.92 g, 8.2 mmol) was dissolved in a mixture of THF (30 mL)
and H2O (2.5 mL). A 1.0 M solution of tetra-n-butylammonium
fluoride in THF (10 mL, 10 mmol) was added dropwise at 0
°C, and the reaction mixture was stirred for 1 h at 0 °C and 3
h at room temperature. The reaction mixture was concen-
trated, and H2O (100 mL) was added. The products were
extracted with ethyl acetate (2 × 50 mL), and the combined
extracts were dried over MgSO4 and concentrated. The product
was purified via column chromatography on silica gel (120 g).
The compound was eluted with ethyl acetate/hexane (1:8). The
solvent was evaporated to provide 25 (1.2 g, 90%) as a
yellowish liquid. 1H NMR (300 MHz, CDCl3) δ 7.04 (d, J )
2.97 Hz, 1 H), 6.92 (dd, J ) 9.04, 2.97 Hz, 1 H), 6.84 (d, J )
9.10 Hz, 1 H), 3.89 (s, 3 H), 3.79 (s, 3 H), 3.34 (s, 1 H). Anal.
(C10H10O2) C, H.
130.46, 122.96, 122.55, 120.27, 120.00, 117.15, 116.68, 60.58,
60.00, 46.15, 46.03, 44.39, 39.42; CIMS m/z 438 (MH+). Anal.
(C25H24FNO5) C, H, F, N.
5-[2-(2,5-Dim eth oxyp h en yl)eth yl]-N-[(4-flu or op h en yl)-
eth yl]sa licyla m id e (30). Compound 29 (0.50 g, 1.2 mmol)
was hydrogenated for 4 h over 10% palladium on activated
carbon (wet, contains 50% water, 0.1 g) in ethyl acetate (10
mL) at room temperature and atmospheric pressure. The
mixture was then filtered through Celite and washed with
ethyl acetate (10 mL). The combined filtrate was evaporated
on a rotary evaporator to give the crude product, which was
then recrystallized from ethyl acetate/hexane. Compound 30
(0.39 g, 77%) was obtained as a white crystalline solid: mp
110-112 °C; 1H NMR (300 MHz, CDCl3) δ 11.99 (s, 1 H, OH),
10.34 (m, 2 H), 7.22-7.15 (m, 3 H), 6.88 (d, J ) 8.45 Hz, 1 H),
6.83 (d, J ) 1.90 Hz, 1 H), 6.74 (d, J ) 8.81 Hz, 1 H), 6.68 (dd,
J ) 8.82, 2.90 Hz, 1 H), 6.59 (d, J ) 2.91 Hz, 1 H), 6.20 (bs, 1
H, NH), 3.72 (s, 3 H), 3.70 (s, 3 H), 3.62 (q, J ) 6.26 Hz, 2 H),
2.88 (t, J ) 7.01 Hz, 2 H), 2.78 (m, 4 H); CIMS m/z 424 (MH+).
Anal. (C25H26FNO4) C, H, F, N.
Meth yl 5-Iod o-2-m eth oxyben zoa te (26). To a solution of
5-iodosalicylic acid (17, 2.0 g, 7.27 mmol) in acetone (50 mL)
was added anhydrous K2CO3 (4.5 g, 32.5 mmol). The mixture
was cooled in an ice bath, and (CH3)2SO4 (1.4 mL, 15.27 mmol)
was added dropwise via syringe. The mixture was heated to
reflux and stirred overnight. After the mixture was cooled to
room temperature, water (200 mL) was added and the mixture
5-[2-(2,5-Dim eth oxyp h en yl)eth yn yl]-2-m eth oxyben zo-
ic Acid (34). Compound 27 (0.50 g, 1.53 mmol) was added to
a solution of 10% aqueous NaOH (20 mL) and methanol (10
mL). The mixture was heated at 60-70 °C for 3 h. After the
mixture was cooled to room temperature, the organic solvent
was removed on a rotary evaporator. The reaction mixture was
acidified with 1 N HCl, and the resulting precipitate was
extracted with CH2Cl2 (25 mL × 2), washed with water, and
dried over anhydrous MgSO4. Evaporation of the solvent
yielded 34 (0.48 g, 100%) as a yellowish solid: mp 139-140
was stirred for 1 h. Acetone was removed on
a rotary
evaporator, and the aqueous solution was extracted with ethyl
acetate (3 × 50 mL) and dried over anhydrous MgSO4. The
solvent was evaporated, yielding a yellow liquid, which was
purified by flash chromatography (silica gel 120 g, ethyl
acetate/hexane 1:5) to afford 26 (1.68 g, 79%) as white
crystalline solid: mp 48-50 °C; 1H NMR (300 MHz, CDCl3) δ
8.04 (d, J ) 2.32 Hz, 1 H), 7.72 (dd, J ) 8.78, 2.33 Hz, 1 H),
6.75 (d, J ) 8.81 Hz, 1 H), 3.86 (s, 6 H); CIMS m/z 293 (MH+),
261 (M - MeOH). Anal. (C9H9IO3) C, H, I.
Met h yl 5-[2-(2,5-Dim et h oxyp h en yl)et h yn yl]-2-m et h -
oxyben zoa te (27). Compound 25 (0.70 g, 4.3 mmol), com-
pound 26 (0.63 g, 2.15 mmol), bis(triphenylphosphine)palla-
dium(II) chloride (0.005 g, 0.007 mmol), and CuI (0.0041 g,
0.02 mmol) were dissolved in diethylamine (50 mL). The
mixture was heated at 50-60 °C for 8 h. The reaction mixture
was then concentrated in vacuo, and saturated NaHCO3 (100
mL) was added. The product was extracted with ethyl acetate
(3 × 80 mL). The combined organic extracts were washed with
brine (1 × 50 mL), dried over MgSO4, and concentrated. The
crude product was purified by flash chromatography (silica gel
150 g, ethyl acetate/hexane 1:2) to give 27 (0.66 g, 95%) as a
yellowish crystalline solid: mp 99-101 °C; 1H NMR (300 MHz,
acetone-d6) δ 7.66 (dd, J ) 8.75, 2.23 Hz, 1 H), 7.28 (d, J )
2.22 Hz, 1 H), 7.18 (d, J ) 8.68 Hz, 1 H), 7.03 (d, J ) 2.90 Hz,
1 H), 6.96 (d, J ) 8.92 Hz, 1 H), 6.91 (dd, J ) 9.05, 2.90 Hz,
1 H), 3.91 (s, 3 H), 3.84 (s, 6 H), 3.76 (s, 3 H); 13C NMR (300
MHz, acetone-d6) δ 166.00, 159.52, 155.44, 154.28, 136.87,
134.78, 122.01, 118.66, 116.44, 116.12, 113.60, 113.26, 92.51,
86.12, 56.61, 56.44, 55.99, 52.22; CIMS m/z 327 (MH+). Anal.
(C19H18O5) C, H.
5-[2-(2,5-Dim et h oxyp h en yl)a cet yl]-N-[(4-flu or op h en -
yl)eth yl]sa licyla m id e (29). Compound 25 (0.98 g, 2.5 mmol),
compound 18b (0.83 g, 5.1 mmol), PPh3 (0.027 g, 0.1 mmol),
and CuI (0.01 g, 0.05 mmol) were dissolved in piperidine (30
mL). Palladium chloride (0.018 g, 0.02 mmol) was added, and
the resulting mixture was heated at reflux for 24 h. The
reaction mixture was then concentrated in vacuo, and satu-
rated NaHCO3 (100 mL) was added. The product was extracted
with ethyl acetate (3 × 50 mL). The combined organic extracts
were washed with brine (1 × 20 mL), dried over MgSO4, and
concentrated. The crude product was purified by column
chromatography (silica gel 80 g, ethyl acetate/hexane 2:3) to
give 29 (1.92 g, 89%) as a yellowish liquid. 1H NMR (300 MHz,
acetone-d6) δ 12.95 (s, 1 H, OH), 8.10 (d, J ) 1.96 Hz, 1 H),
7.17 (m, 2 H), 7.14 (dd, J ) 8.70, 1.92 Hz, 1 H), 6.99 (m, 3 H),
6.75 (d, J ) 1.59 Hz, 1 H), 6.73 (dd, J ) 6.29, 2.36 Hz, 1 H),
6.69 (t, J ) 5.32 Hz, 1 H, NH), 4.13 (s, 2 H), 3.71 (s, 3 H), 3.67
(s, 3 H), 3.63 (q, J ) 6.23 Hz, 2 H), 2.88 (t, J ) 7.14 Hz, 2 H);
13C NMR (300 MHz, acetone-d6) δ 199.90, 168.20, 165.12,
158.77, 156.82, 140.46, 139.11, 135.73, 135.61, 133.39, 132.82,
1
°C; H NMR (300 MHz, CDCl3) δ 8.33 (d, J ) 2.18 Hz, 1 H),
7.71 (dd, J ) 2.20, 8.69 Hz, 1 H), 7.01 (d, J ) 8.62 Hz, 1 H),
7.00 (d, J ) 2.78 Hz, 1 H), 6.86 (dd, J ) 2.69, 8.93 Hz, 1 H),
6.81 (d, J ) 9.00 Hz, 1 H), 4.05 (s, 3 H), 3.85 (s, 3 H), 3.76 (s,
3 H); CIMS m/z 313 (MH+). Anal. (C18H16O5) C, H.
Meth yl 5-[(Z)-2-(2,5-Dim eth oxyp h en yl)vin yl]-2-m eth -
oxyben zoa te (35). From compound 27 (0.5 g, 1.53 mmol), H2,
and 5% palladium on calcium carbonate poisoned with lead
(30 mg), 35 was prepared by a procedure similar to that
described for 13a , Compound 35 (0.46 g, 92%) was a yellow
liquid. 1H NMR (300 MHz, CDCl3) δ 7.63 (d, J ) 2.30 Hz, 1
H), 7.39 (dd, J ) 2.33, 8.65 Hz, 1 H), 7.00 (d, J ) 8.71 Hz, 1
H), 6.94 (d, J ) 8.87 Hz, 1 H), 6.81 (dd, J ) 3.08, 8.88 Hz, 1
H), 6.75 (d, J ) 3.05 Hz, 1 H), 6.61 (d, J ) 12.30 Hz, 1 H),
6.55 (d, J ) 12.29 Hz, 1 H), 3.84 (s, 3 H), 3.76 (s, 3 H), 3.75 (s,
3 H), 3.56 (s, 3 H); CIMS m/z 329 (MH+), 328 (M+); ESIMS
m/z 329 (MH+), 351 (MNa+). Anal. (C19H20O5) C, H.
Meth yl 5-[(Z)-2-(2,5-Dih ydr oxyph en yl)vin yl]-2-h ydr oxy-
ben zoa te (36). From compound 35 (0.06 g, 0.18 mmol) and
BBr3, 36 was prepared by a procedure similar to that described
for 14a . Compound 36 (0.025 g, 48%) was an off-white
crystalline solid: mp 173-175 °C; 1H NMR (500 MHz, CDCl3)
δ 10.68 (s, 1 H, OH), 7.80 (d, J ) 2.10, 1 H), 7.67 (s, 1 H, OH),
7.58 (s, 1 H, OH), 7.44 (dd, J ) 8.75, 2.45 Hz, 1 H), 6.81 (d,
J ) 8.65 Hz, 1 H), 6.74 (d, J ) 8.23 Hz, 1 H), 6.72 (m, 3 H),
6.53 (d, J ) 12.26 Hz, 1 H), 3.90 (s, 3 H); ESIMS m/z 287
(MH+), 309 (MNa+). Anal. (C16H14O5) C, H.
Ack n ow led gm en t. This research was made possible
by grants from the Showalter Trust and Purdue Re-
search Foundation.
Refer en ces
(1) Onoda, T.; Iinuma, H.; Sasaki, Y.; Hamada, M.; Isshiki, K.;
Naganawa, H.; Takeuchi, T. Isolation of a Novel Tyrosine Kinase
Inhibitor, Lavendustin A, from Streptomyces griseolavendus. J .
Nat. Prod. 1989, 52, 1252-1257.
(2) Hsu, C.-Y. J .; Persons, P. E.; Spada, A. P.; Bednar, R. A.;
Levitzki, A.; Zilberstein, A. Kinetic Analysis of the Inhibition of
the Epidermal Growth Factor Receptor Tyrosine Kinase by
Lavendustin A and Its Analogue. J . Biol. Chem. 1991, 266,
21105-21112.
(3) Onoda, T.; Isshiki, K.; Takeuchi, T.; Tatsuta, K.; Umezawa, K.
Inhibition of Tyrosine Kinase and Epidermal Growth Factor
Internalization by Lavendustin A Methyl Ester in Cultured A431
Cells. Drugs Exp. Clin. Res. 1990, 16, 249-253.

1682 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 9
Mu et al.
(4) Agbotounou, W. K.; Umezawa, K.; J acquemin-Sablon, A.; Pierre,
J . Inhibition by Two Lavendustins of the Tyrosine Kinase
Activity of pp60F527 in Vitro and in Intact Cells. Eur. J .
Pharmacol. 1994, 269, 1-8.
(5) Imoto, M.; Sujikai, I.; Ui, H.; Umezawa, K. Involvement of
Tyrosine Kinase in Growth Factor-Induced Phospholipase C
Activation in NIH3T3 Cells. Biochim. Biophys. Acta 1993, 1166,
188-192.
(6) Liu, T.; Shirai, R.; Matsui, T.; Umezawa, K.; Iwasaki, S.
Synthesis and Biological Activity of 5-[2,5-Dihydroxybenzyl)-
amino]salicylic Acid Analogs as Inhibitors of EGF Receptor-
Associated Protein Tyrosine Kinase. Bioorg. Med. Chem. Lett.
1997, 7, 365-368.
(7) Chen, H.; Boiziau, J .; Parker, F.; Maroun, R.; Tocque, B.; Roques,
B. P.; Garbay-J aureguiberry, C. Synthesis and Structure-
Activity Studies of a Series of [(Hydroxybenzyl)amino]salicylates
as Inhibitors of EGF Receptor-Associated Tyrosine Kinase
Activity. J . Med. Chem. 1993, 36, 4094-4098.
(8) Chen, H.; Boiziau, J .; Parker, F.; Mailliet, P.; Commerc¸on, A.;
Tocque, B.; Le Pecq, J .-B.; Roques, B.-P.; Garbay, C. Structure-
Activity Relationships in a Series of 5-[(2,5-Dihydroxybenzyl)-
amino]salicylate Inhibitors of EGF-Receptor-Associated Tyrosine
Kinase: Importance of Additional Hydrophobic Aromatic Inter-
actions. J . Med. Chem. 1994, 37, 845-859.
(9) Nussbaumer, P.; Winiski, A. P.; Cammisuli, S.; Hiestand, P.;
Weckbecker, G.; Stu¨tz, A. Novel Antiproliferative Agents Derived
from Lavendustin A. J . Med. Chem. 1994, 37, 4079-4084.
(10) Cammisuli, S.; Winiski, A.; Nussbaumer, P.; Hiestand, P.; Stutz,
A.; Weckbecker, G. SDZ 281-977: a Modified Partial Structure
of Lavendustin A That Exerts Potent and Selective Antiprolif-
erative Activities in Vitro and in Vivo. Int. J . Cancer 1996, 65,
351-359.
(21) Lambert, C.; Utimoto, K.; Nozaki, H. Palladium(II) Catalyzed
Cyclization of Alkynoic Acids. Tetrahedron Lett. 1984, 25, 5323-
5326.
(22) Yanagihara, N.; Lambert, C.; Iritani, K.; Utimoto, K.; Nozaki,
H. Palladium-Catalyzed Regio- and Stereoselective Cyclized
Coupling of Lithium Alkynoates with Allyl Halides. J . Am.
Chem. Soc. 1986, 108, 2753-2754.
(23) Imi, K.; Imai, K.; Utimoto, K. Regioselective Hydration of
Alkynones by Palladium Catalysis. Tetrahedron Lett. 1987, 28,
3127-3130.
(24) Boyd, M. R.; Paull, K. D. Some Practical Considerations and
Applications of the National Cancer Institute In Vitro Anticancer
Drug Discovery Screen. Drug Dev. Res. 1995, 34, 91-109.
(25) Pettit, G. R.; Grealish, M. P.; J ung, M. K.; Hamel, E.; Pettit, R.
K.; Chapuis, J .-C.; Schmidt, J . M. Antineoplastic Agents. 465.
Structural Modification of Resveratrol: Sodium Resverastatin
Phosphate. J . Med. Chem. 2002, 45, 2534-2542.
(26) Cushman, M.; Nagarathnam, D.; Gopal, D.; He, H.-M.; Lin, C.
M.; Hamel, E. Synthesis and Evaluation of Analogues of (Z)-1-
(4-Methoxyphenyl)-2-(3,4,5-trimethoxyphenyl)ethene as Poten-
tial Cytotoxic and Antimitotic Agents. J . Med. Chem. 1992, 35,
2293-2306.
(27) Verdier-Pinard, P.; Lai, J . Y.; Yoo, H.-D.; Yu, J .; Marquez, B.;
Nagle, D. G.; Nambu, M.; White, J . D.; Falck, J . R.; Gerwick,
W. H.; Day, B. W.; Hamel, E. Structure-Activity Analysis of
the Interaction of Curacin A, the Potent Colchicine Site Anti-
mitotic Agent, with Tubulin and Effects of Analogs of the Growth
of MCF-7 Breast Cancer Cells. Mol. Pharmacol. 1998, 53, 62-
76.
(28) Hastie, S. B. Interactions of Colchicine with Tubulin. Pharmacol.
Ther. 1991, 51, 377-401.
(29) Kang, G.-J .; Getahun, Z.; Muzaffar, A.; Brossi, A.; Hamel, E.
N-Acetylcolchinol O-Methyl Ether and Thiocolchicine, Potent
Analogs of Colchicine Modified in the C Ring: Evaluation of the
Mechanistic Basis for Their Enhanced Biological Properties. J .
Biol. Chem. 1990, 265, 10255-10259.
(30) Grover, S.; Boye´, O.; Getahun, Z.; Brossi, A.; Hamel, E. Chloro-
acetates of 2- and 3-Demethylthiocolchicine: Specific Covalent
Interaction with Tubulin with Preferential Labeling of the
â-Subunit. Biochem. Biophys. Res. Commun. 1992, 187, 1350-
1358.
(31) Lin, C. M.; Ho, H. H.; Pettit, G. R.; Hamel, E. Antimitotic Natural
Products Combretastatin A-4 and Combretastatin A-2: Studies
on the Mechanism of Their Inhibition of the Binding of Colchi-
cine to Tubulin. Biochemistry 1989, 28, 6984-6991.
(32) Bane, S.; Puett, D.; Macdonald, T.; Williams, R. C., J r. Binding
to Tubulin of the Colchicine Analog 2-Methoxy-5-(2′,3′,4′-tri-
methoxyphenyl)tropone: Thermodynamic and Kinetic Aspects.
J . Biol. Chem. 1984, 259, 7391-7398.
(33) Engelborghs, Y.; Fitzgerald, T. J . A Fluorescence Stopped Flow
Study of the Competition and Displacement Kinetics of Podo-
phyllotoxin and the Colchicine Analog 2-Methoxy-5-(2′,3′,4′-
trimethoxyphenyl)tropone on Tubulin. J . Biol. Chem. 1987, 262,
5204-5209.
(34) Banerjee, A.; Barnes, L. D.; Luduen˜a, R. F. The Role of the B
Ring of Colchicine in the Stability of the Colchicine-Tubulin
Complex. Biochim. Biophys. Acta 1987, 913, 138-144.
(35) Hamel, E.; Lin, C. M. Separation of Active Tubulin and Micro-
tubule-Associated Proteins by Ultracentrifugation and Isolation
of a Component Causing the Formation of Microtubule Bundles.
Biochemistry 1984, 23, 4173-4184.
(36) Lin, C. M.; Kang, G. J .; Roach, M. C.; J iang, J . B.; Hesson, D.
P.; Luduen˜a, R. F.; Hamel, E. Investigation of the Mechanism
of the Interaction of Tubulin with Derivatives of 2-Styrylquinazo-
lin-4(3H)-one. Mol. Pharmacol. 1991, 40, 827-832.
(11) Mu, F.; Coffing, S. L.; Riese, D. J . I.; Geahlen, R. L.; Verdier-
Pinard, P.; Hamel, E.; J ohnson, J .; Cushman, M. Design,
Synthesis, and Biological Evaluation of a Series of Lavendustin
A Analogues That Inhibit EGFR and Syk Tyrosine Kinases, as
Well as Tubulin Polymerization. J . Med. Chem. 2001, 44, 441-
452.
(12) Paull, K. D.; Hamel, E.; Malspeis, L. Prediction of Biochemical
Mechanism of Action from the in Vitro Antitumor Screen of the
National Cancer Institute. Cancer Chemotherapeutic Agents;
American Chemical Society: Washington, DC, 1995; pp 9-45.
(13) Riese, D. J . I.; van Raaij, T. M.; Plowman, G. D.; Andrews, G.
C.; Stern, D. F. The Cellular Response to Neuregulins Is
Governed by Complex Interactions of the ErbB Receptor Family.
Mol. Cell. Biol. 1995, 15, 5770-5776.
(14) Geahlen, R. L.; McLaughlin, J . L. Piceatannol (3,4,3′,5′-Tetrahy-
droxy-trans-stilbene) Is a Naturally Occurring Protein Tyrosine
Kinase Inhibitor. Biochem. Biophys. Res. Commun. 1989, 165,
241-245.
(15) Smyth, M. S.; Stefanova, I.; Hartman, F.; Horad, I. D.; Osherov,
N.; Levitzki, A.; Burke, T. R. J . Non-Amine Based Analogues of
Lavendustin A as Protein-Tyrosine Kinase Inhibitors. J . Med.
Chem. 1993, 36, 3010-3014.
(16) Thakkar, K.; Geahlen, R. L.; Cushman, M. Synthesis and
Protein-Tyrosine Kinase Inhibitory Activity of Polyhydroxylated
Stilbene Analogs of Piceatannol. J . Med. Chem. 1993, 36, 2950-
2955.
(17) Pettit, G. R.; Singh, S. B.; Hamel, E.; Lin, C. M.; Alberts, D. S.;
Garcia-Kendall, D. Isolation and Structure of the Strong Cell
Growth and Tubulin Inhibitor Combretastatin A-4. Experientia
1989, 45, 209-211.
(18) Cushman, M.; Nagarathnam, D.; Gopal, D.; Chakraborti, A. K.;
Lin, C. M.; Hamel, E. Synthesis and Evaluation of Stilbene and
Dihydrostilbene Derivatives as Potential Anticancer Agents That
Inhibit Tubulin Polymerization. J . Med. Chem. 1991, 34, 2579-
2588.
(37) Alley, M. C.; Scudiero, D. A.; Monks, A.; Hursey, M. L.;
Czerwinski, M. J .; Fine, D. L.; Abbott, B. J .; Mayo, J . G.;
Shoemaker, R. H.; Boyd, M. R. Feasibility of Drug Screening
with Panels of Human Tumor Cell Lines Using a Microculture
Tetrazolium Assay. Cancer Res. 1988, 48, 589-601.
(19) Cushman, M.; He, H.-M.; Lin, C. M.; Hamel, E. Synthesis and
Evaluation of a Series of Benzylaniline Hydrochlorides as
Potential Cytotoxic and Antimitotic Agents Acting by Inhibition
of Tubulin Polymerization. J . Med. Chem. 1993, 36, 2817-2821.
(20) Sinhababu, A. K.; Kawase, M.; Borchardt, R. T. tert-Butyldi-
methylsilyl Ethers of Phenols: Their One-Step Conversion to
Benzyl or Methyl Ethers and Utility in Regioselective ortho-
Lithiation. Tetrahderon Lett. 1987, 28, 4139-4142.
J M020292+