Array-based structure and gene expression relationship study of antitumor sulfonamides including N-[2-[(4-hydroxyphenyl) amino]-3-pyridinyl] -4-methoxybenzenesulfonamide and N-(3-chloro-7-indolyl)-1,4-benzenedisulfonamide

Article abstract of DOI:10.1021/jm0201060

Compounds from sulfonamide-focused libraries have been evaluated in cell-based antitumor screens using the COMPARE analysis with a panel of 39 human cancer cell lines and flow cytometric cell cycle analysis. Thus far, 2 (N-[2-[(4-hydroxyphenyl)amino]-3pyr

Full text of DOI:10.1021/jm0201060

J . Med. Chem. 2002, 45, 4913-4922
4913
Ar r a y-Ba sed Str u ctu r e a n d Gen e Exp r ession Rela tion sh ip Stu d y of An titu m or
Su lfon a m id es In clu d in g
N-[2-[(4-Hyd r oxyp h en yl)a m in o]-3-p yr id in yl]-4-m eth oxyben zen esu lfon a m id e a n d
N-(3-Ch lor o-7-in d olyl)-1,4-ben zen ed isu lfon a m id e
Takashi Owa,*,† Akira Yokoi,† Kanami Yamazaki,‡ Kentaro Yoshimatsu,§ Takao Yamori,‡ and Takeshi Nagasu†
Eisai Co., Ltd., 5-1-3 Tokodai, Tsukuba, Ibaraki 300-2635, J apan, and Cancer Chemotherapy Center,
J apanese Foundation for Cancer Research, 1-37-1 Kami-Ikebukuro, Toshima-ku, Tokyo 170-8455, J apan
Received March 6, 2002
Compounds from sulfonamide-focused libraries have been evaluated in cell-based antitumor
screens using the COMPARE analysis with a panel of 39 human cancer cell lines and flow
cytometric cell cycle analysis. Thus far, 2 (N-[2-[(4-hydroxyphenyl)amino]-3-pyridinyl]-4-
methoxybenzenesulfonamide (E7010)) and 3 (N-(3-chloro-7-indolyl)-1,4-benzenedisulfonamide
(E7070)) have been selected from the collections as potent cell cycle inhibitors, which have
progressed to clinical trials. Compound 2 is an orally active antimitotic agent disrupting tubulin
polymerization, whereas compound 3 belongs to a novel class of antiproliferative agents causing
a decrease in the S phase fraction along with G1 and/or G2 accumulation in various cancer
cell lines. Because both compounds exhibited preliminary clinical activities in the phase I
setting, we decided to examine further this series of oncolytic small molecules, particularly by
using high-density oligonucleotide microarray analysis. The array data have enabled us to
characterize these two classes of antitumor sulfonamides on the basis of gene expression
changes, illuminating the essential pharmacophore structure and drug-sensitive cellular
pathways for each class. Moreover, the dual character of 5 (N-(3-chloro-7-indolyl)-4-methoxy-
benzenesulfonamide (ER-67880)), resembling both 2 and 3, was revealed by array-based
transcription profiling, though the 3-type profile of this molecule had not been apparent in the
cell-based phenotypic screens. These results provide an example of the utility of structure and
gene expression relationship studies in medicinal genomics.
In tr od u ction
research efforts is considered to be personalized therapy,
i.e., specifying the most effective drugs against a
particular cancer in an individual patient.
The draft release of the human genome sequence has
tremendous potential benefits for human health care.1,2
Substantial endeavors of translational studies “from lab
to bedside” are currently in progress by means of
emerging new technologies such as bioinformatics,
transcriptomics (DNA microarray analysis), and pro-
teomics. Interdisciplinary research utilizing these tech-
nologies should provide a more rational basis for drug
discovery and development programs. A recent trial in
the Developmental Therapeutics Program (DTP) of the
National Cancer Institute (NCI) presents an advanced
example in which the correlation between gene expres-
sion and drug activity patterns in the NCI60 cancer cell
lines was evaluated to integrate large databases on gene
expression and anticancer drug pharmacology.3 The
COMPARE analysis in the DTP represents a bioinfor-
matic approach for categorizing anticancer agents ac-
cording to their antiproliferative “fingerprint” patterns
against a panel of diverse human cancer cell lines.4 DNA
microarray analysis in cancer research has allowed us
to identify molecular differences between normal and
cancer cells and to clarify the heterogeneity of cancer
on a genome-wide scale.5,6 The ultimate goal of these
The selection of the optimal treatment for each cancer
patient will require precise molecular diagnosis based
on gene expression profiles in cancerous tissues.5,6 At
the same time, the drug effect on an individual target
and/or its downstream markers may be used as molec-
ular biological endpoints for pharmacodynamic assess-
ment. Although tumor regression is a well-established
endpoint, it appears urgently necessary to provide a firm
molecular basis for the clinical benefit of “disease
stabilization” accompanied by improvement in progres-
sion-free survival (PFS) and quality of life (QOL). A
critical aspect of this problem is identifying tumor
changes in gene expression relevant to drug efficacy.
Array-based hybridization technology has proved to be
particularly useful for this purpose in preclinical studies
to date.7-9
Recent trends in cancer research strongly suggest the
increasing importance of accurate, systematic, and
molecular mechanism-based classification of anticancer
chemotherapeutics.10 In relation to this point, an inte-
grated database of chemosensitivity to 55 anticancer
drugs and gene expression profiles of 39 human cancer
cell lines has been reported very recently.11a There is a
growing possibility that comprehensive databases will
be developed further to connect antitumor activity
profiles and drug-induced transcriptional changes for
* To whom correspondence should be addressed. Phone: +81-298-
47-7196. Fax: +81-298-47-7614. E-mail: t-owa@hhc.eisai.co.jp.
† Laboratory of Seeds Finding Technology, Eisai Co., Ltd.
‡ J apanese Foundation for Cancer Research.
§ Discovery Research Laboratories II, Eisai Co., Ltd.
10.1021/jm0201060 CCC: $22.00 © 2002 American Chemical Society
Published on Web 09/24/2002

4914 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 22
Owa et al.
tubulin at the colchicine site.13,15 Because of its good
antitumor activity in vivo against both rodent tumors
and human tumor xenografts,16 2 entered clinical trials
as an orally active tubulin polymerization inhibitor.
Despite the drug-related adverse effect of peripheral
neuropathy, clinical activity was reported in 2 out of
16 patients in the phase I setting; spinal cord metastasis
was reduced by 74% in a patient with uterine sarcoma,
and a minor response was obtained in a patient with
pulmonary adenocarcinoma.17
Sch em e 1. Compounds from Sulfonamide-Focused
Libraries
A structure-activity relationship (SAR) study based
on template II resulted in the discovery of another novel
antitumor sulfonamide 3 (N-(3-chloro-7-indolyl)-1,4-
benzenedisulfonamide), which did not exhibit antimi-
totic action.18,19 Following compound 3 treatment, P388
murine leukemia cells accumulated in the G1 phase but
not in the M phase of the cell cycle.18,20 In an experiment
using HCT116 human colon cancer cells, 3 repressed
the expression of cyclin E and the phosphorylation of
CDK2, both of which are essential for G1/S transition.21
A more recent mechanistic study of 3 by Fukuoka et al.
has further revealed that the compound disturbs the cell
cycle progression at multiple points, including both G1/S
and G2/M transitions, in human lung cancer cells.22 In
this report using A549 and a 3-resistant subline A549/
ER, 3 was shown to inhibit pRb phosphorylation, to
decrease the protein expression of cyclin A, cyclin B1,
CDK2, and CDC2, and to suppress CDK2 catalytic
activity with the induction of p53 and p21 proteins only
in the parental A549 cells. All these observations
indicate that this unique small molecule belongs to a
novel class of cell cycle inhibitors.23
Compound 3 displayed distinctive and promising
antitumor profiles both in vitro and in vivo.20 To assess
the toxicity, the maximum tolerated dose (MTD), and
the pharmacokinetics of this compound, phase I clinical
trials have been conducted by the European Organiza-
tion for Research and Treatment of Cancer (EORTC)s
Early Clinical Studies Group (ECSG).24-27 In these
studies, the major dose-limiting toxicity was determined
to be myelosuppression, such as thrombocytopenia and
neutropenia. Two partial responses with a more than
50% shrinkage of measurable tumors were observed in
patients with uterine adenocarcinoma and breast can-
cer. In addition, disease stabilizations and some minor
responses were also documented. Phase II trials are
currently ongoing in Europe and the U.S. Other sul-
fonamide compounds examined herein were all synthe-
sized by the procedures described earlier.14,18
each chemotherapeutic agent. Therefore, we decided to
scrutinize a series of sulfonamide antitumor agents
discovered in these laboratories based on the COM-
PARE algorithm of the Disease-Oriented Screening
(DOS) program11 and Affymetrix GeneChip analysis12
with HuGene FL (Hu6800) arrays. In this study, seven
compounds selected from our sulfonamide-focused li-
braries (Scheme 1) were characterized in terms of the
fingerprint patterns of cell growth inhibition and tran-
scriptional alteration.
Resu lts a n d Discu ssion
COMP ARE An a lysis. Compounds 2-8 were tested
for in vitro antiproliferative activity against a panel of
39 human cancer cell lines in the DOS.11 A 48 h assay
with sulforhodamine B was employed for the evalua-
tion.28 Mean graphs4 of the compounds were prepared
for the COMPARE algorithm based on the growth
inhibition parameter of GI50 (drug concentration re-
quired for 50% inhibition of cell proliferation). The
log GI50 (GI50 in units of M) data are available as
Supporting Information. In the NCI COMPARE analy-
sis,4 a comparable mean graph signature can be seen
among drugs with a common mode of action. Thus, 2
and 3 were used as seed compounds in COMPARE to
An titu m or Su lfon a m id es
Starting from the discovery of compound 1 as an
anticancer drug lead, a large number of sulfonamide
molecules have been synthesized so far using templates
I and II (Scheme 1).13,14 Compound 2 (N-[2-[(4-hydroxy-
phenyl)amino]-3-pyridinyl]-4-methoxybenzenesulfonamide),
an early drug candidate from the template I series,
potently inhibited microtubule assembly by binding to

Antitumor Sulfonamides
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 22 4915
Ta ble 1. COMPARE Analysis of a Series of Antitumor
ings indicated the N-(3-chloro-7-indolyl)benzenesulfona-
mide motif as a structural determinant for the activity
pattern of the compound 3 class. However, compound
5, possessing the same core structure, resembled 2 (r )
0.718) rather than 3 (r ) 0.473) in the mean graph
signature. The compound also contains the N-(2-ami-
nophenyl)-4-methoxybenzenesulfonamide motif associ-
ated with the 2 type of antimitotic activity, and thus,
this core structure would appear to be the dominant
pharmacophore for the activity profile of 5. In the
COMPARE analysis based on the mean graph finger-
print of 3, there were no standard agents with correla-
tion coefficients greater than 0.5. This supports the idea
that 3 is a novel type of antiproliferative agent with a
mechanism of action distinct from those of conventional
anticancer drugs in clinical use.18,20
Sulfonamidesa
b
ranking order
compound
r
(A) Based on the Mean Graph Fingerprint of 2
1
2
3
4
5
6
7
8
9
4
5
0.861
0.718
0.652
0.559
0.542
0.280
0.257
0.221
0.186
Navelbine
Taxol
Vincristine
6
7
3
8
(B) Based on the Mean Graph Fingerprint of 3
1
2
3
4
5
6
6
7
5
4
2
8
0.762
0.754
0.473
0.269
0.221
0.185
F low Cytom etr ic Cell Cycle An a lysis. To compare
the effects of compounds 2-8 on cell cycle progression,
fluorescence-activated cell sorter (FACS) analysis was
performed with HCT116-C9 cells (Figure 1).29 Cells were
treated with each compound at 8 µM for 12 h. Com-
pound 3 alone was tested further at extended time
points of 24, 36, and 48 h. In addition, the inhibitory
activity of each compound on tubulin polymerization
was measured (Table 2).
a
The mean graph fingerprints of 2 and 3 were individually
compared with those of other test compounds using the COMPARE
algorithm. The data on 200 standard drugs in the DOS were also
included in the calculations, and the three drugs with correlation
coefficients greater than 0.5 (P < 0.001) are presented here.
b
Pearson’s correlation coefficient.
categorize our sulfonamide agents on the basis of their
growth-inhibitory patterns. The calculation results are
presented in Table 1, where compounds are ordered in
accordance with Pearson’s correlation coefficient. The
mean graph signatures of 2 and 3 were also compared
with those of more than 200 standard drugs in the DOS
database. Among them, three drugs with correlation
coefficients greater than 0.5 (P < 0.001) are included
in Table 1.
Using 2 as an initial seed in COMPARE resulted in
the identification of 4 and 5 as highly related to 2 with
correlation coefficients greater than 0.7 (Table 1A). On
the basis of the mean of log GI50 values for 39 cell lines
(MG-MID), 2 (MG-MID ) -6.30) and 4 (MG-MID )
-6.49) were equally active and 5 (MG-MID ) -5.71)
was slightly less active. Furthermore, the COMPARE
run with the standard drugs assessed the top three
2-related agents (r > 0.5) to be navelbine, taxol, and
vincristine, all of which are well-known mitotic arrest
agents. This result is consistent with the fact that 2 is
an antimitotic agent, although, strictly speaking, only
taxol stabilizes microtubules. On the other hand, the
mean graph signatures of 3 and 6-8 showed poor
correlations (r < 0.3) compared with 2, suggesting that
all these agents are independent of anti-tubulin activity
as exhibited by 2. In light of the SAR study, the motif
of N-(2-aminophenyl or 2-amino-3-pyridinyl)-4-meth-
oxybenzenesulfonamide, which is common to 2, 4, and
5, was regarded as a pharmacophore for a 2 type of
microtubule-depolymerizing agent.
As expected from the COMPARE data, 4 and 5 caused
a clear G2/M arrest like 2. As shown in Table 2, all these
compounds were capable of inhibiting tubulin polym-
erization with low micromolar IC50 values. Furthermore,
a marked increase in the mitotic index was observed
microscopically in the cells treated with each of them
for 12 h (data not shown). On the other hand, 6 led to
a decrease of S and an increase of G2/M like 3. The cell
cycle effect of 7 was only marginal, corresponding to its
lower activity compared with 3 and 6. These three
compounds did not increase the mitotic index at all after
a 12 h treatment (data not shown), consistent with their
insignificant activities against tubulin polymerization
(IC50 > 100 µM). Hence, the drug-induced G2/M increase
is considered to be due to cell cycle perturbation in the
G2 phase. Compound 8, inferior and unrelated to both
2 and 3 in its activity profile, showed almost negligible
effects on the cell cycle progression of HCT116-C9 and
on tubulin polymerization.
At the 12 h time point, no drugs produced the sub-
G1 fraction indicative of programmed cell death (apo-
ptosis),30 suggesting that the conditions of drug treat-
ment (at 8 µM for 12 h) are suitable for microarray
analysis to investigate early characteristic changes in
gene expression. In the time course experiment, 3
caused a decrease of S and an increase of G2/M in a
time-dependent manner, eventually resulting in the
appearance of the sub-G1 fraction after 48 h of treat-
ment. From the time point of 24 h, a drop in the early
S fraction was also evident, indicating perturbation at
the G1/S transition. This effect seems to be related to
the G1-targeting action of 3 as reported previously in
P388 murine leukemia18,20 and A549 human lung
carcinoma cells.22
Another round of COMPARE using 3 as a seed
revealed that 6 and 7 significantly correlated with 3 (r
> 0.7), although the growth-inhibitory potency of 7 (MG-
MID > -4.12) was rather low compared with those of
3 (MG-MID ) -4.81) and 6 (MG-MID ) -5.04) (Table
1B). A 3-chloro substituent on the indole moiety of this
series may interact with a putative cellular target(s),
leading to enhancement of the drug activity.18 Com-
pound 8, an amide analogue of 3, was not only inferior
(MG-MID ) -4.44) but also unrelated (r < 0.2) to 3 with
respect to the mean graph activity profile. These find-
High -Den sity Oligon u cleotid e Ar r a y Exp r ession
An a lysis. To profile compounds 2-8 on the basis of
their effects on gene expression, we used oligonucleo-
tide microarray analysis with Affymetrix HuGene FL
(Hu6800) arrays.12,29 HCT116-C9 cells were treated for

4916 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 22
Owa et al.
F igu r e 1. Flow cytometric cell cycle analysis. Drug effects on cell cycle progression of HCT116-C9 cells were examined according
to the procedure described in Experimental Section. The experiment was performed in triplicate with each test compound, and
representative DNA histograms are presented. The percentage of the cells in each cell cycle phase was calculated using the
ModFit LT software for the FACSCalibur flow cytometer (Becton-Dickinson) to afford the mean of three independent data. Errors
were within (10% of the reported values.
Ta ble 2. Inhibition of Tubulin Polymerizationa
antigen GA733-1 and caveolin 1) was common to
b
b
compounds 2-6 but not to the less active compounds 7
and 8, it might be explained by nonspecific chemosen-
sitivity of HCT116-C9 rather than by some specific
mechanism of drug action. There was no overlap be-
tween the profiles of 2 and 3 as to the down-regulation
of gene expression. Of the nine genes repressed by 2,
seven and five genes were also down-regulated by 4 and
5, respectively. It should be noted that all three com-
pounds significantly decreased the mRNA levels of two
R-tubulin subtypes. The accumulation of tubulin mono-
mers has been found to lead to the transcriptional
repression of tubulin genes by a feedback regulatory
mechanism.31,32 Therefore, we conclude that down-
regulation of the R-tubulin transcripts is a common
cellular response to the compound 2 class of microtu-
bule-depolymerizing agents. Other common features of
2, 4, and 5 were seen in the induction of Cyr61,
chondroitin sulfate proteoglycan 2, and clone 23933 and
in the repression of Id-2 and DEAD-box protein p72.
J udging from their synchronism in the three different
drug treatments, it would appear that all of these
expression changes, in addition to R-tubulin down-
regulation, have some relevance to the antimitotic and
compound
IC50 (µM)
compound
IC50 (µM)
2
4
5
3
2.2
2.1
9.5
6
7
8
>100
>100
>100
>100
a
Microtubule assembly was evaluated turbidimetrically at 350
nm 30 min after the reaction mixture was warmed from 4 to 37
°C. The experiment was carried out in triplicate with each test
compound. The IC50 values were calculated using the least-squares
method to afford the means of three independent experimental
data. Errors were within (15% of the reported values. Drug
concentration required for 50% inhibition of tubulin polymeriza-
tion.
b
12 h with each compound at 8 µM or with 0.015% DMSO
(control). Listed in Table 3 (in Supporting Information)
are genes up- or down-regulated at least 2-fold compared
with control cells. The criteria for the gene selection are
described in the table footnote.
Compound 2 treatment resulted in a 2-fold or more
induction and repression of 22 and 9 transcripts,
respectively. Of the 22 genes induced, 16 and 5 genes
were also up-regulated by 4 and 5, respectively, whereas
only 2 genes were up-regulated by 3. Because the
induction of these two genes (i.e., tumor-associated

Antitumor Sulfonamides
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 22 4917
Ta ble 4. Cell Growth Inhibition Assay Using 2- and 3-Resistant Cancer Cell Linesa
b
b
IC50 (µM)
IC50 (µM)
compoundd
P388
P388/4.0r-M
RRc
78
0.97
1.0
HCT116-C9
HCT116-C9-C1
RRc
2
3
5
0.19
0.61
1.2
15
0.59
1.2
0.90
0.21
0.37
0.96
29
2.2
1.1
1.4 × 102
5.9
a
Growth-inhibitory effects of 2, 3, and 5 were compared in the cell-based screen with P388, P388/4.0r-M (2-resistant), HCT116-C9,
and HCT116-C9-C1 (3-resistant) cancer cell lines. The MTT assays were performed twice in triplicate with each test compound. The IC50
values were calculated using the least-squares method to afford the means of six independent data. Errors were within (15% of the
b
reported values. Drug concentration to inhibit cell proliferation by 50% relative to untreated control cells after 72 h of continuous drug
exposure. c RR, relative resistance value, equals the IC50 value for the resistant cell line divided by the IC50 value for the parental cell
line.
antitumor properties of the 2-related agents. Interest-
ingly, trimethoxystilbenes (i.e., combretastatins) and
their analogues have been reported to be potent micro-
tubule-depolymerizing agents that bind to the colchicine
site as well.33-35 They have recently attracted a great
deal of attention because of their promising in vivo
antitumor efficacy and antivascular action.36,37 Microar-
ray analysis to compare the transcriptional effects of
these agents and 2 would also be of considerable
interest.
Compound 3 treatment resulted in a 2-fold or more
induction and repression of 18 and 163 genes, respec-
tively. This dominant transcriptional repression seems
to be characteristic of the expression profile of 3. Of the
total of 181 ()18 + 163) genes, 151 genes were also
found in the array data for 6 (12 induced and 139
repressed), corresponding to the activity data based on
the mean graph signature and the cell cycle effect. It is
of particular interest that 118 genes were also found in
the array data for 5 (12 induced and 106 repressed).
This unexpected result reveals that 5 exhibits the 3-type
profile as well as the 2-type profile on the gene expres-
sion basis, even though its activity fingerprint is similar
to that of 2. Compound 7, a des-chloro derivative of 3,
significantly affected the expression of 10 genes (2
induced and 8 repressed), all of which were included in
the genes altered by 3. In this experimental setting,
compound 8, an amide analogue of 3, did not change
any transcript level 2-fold or more. With respect to the
extent of the transcriptional effect, the array data for
3, 7, and 8 correlate well with the results of cell-based
antitumor screens as described above.
By visiting the DTP web site (http://dtp.nci.nih.gov),
one can see a variety of sulfonamide compounds pos-
sessing anticancer properties in the NCI60 activity
database. Some of the sulfonamide agents are structur-
ally related to 3 and may exhibit growth-inhibitory
effects through metal chelation (e.g., NSC compounds
641105, 643599, 647959, 676553, 676569, 676571,
687773, 687776, 687778, 687779, and 687780) or car-
bonic anhydrase (CA) inhibition (the data for represen-
tative compounds are available in recent papers38).
From an examination, there seemed to be no significant
correlation between any of these agents and 3 on the
basis of their mean graph signatures. However, it is still
possible that metal chelation and/or CA inhibition, at
least in part, might be involved in the mode of antitumor
action of 3. To address this possibility, array-based
transcription profiling of relevant compounds could be
informative, since it proved to be powerful for revealing
the 3-type profile latent in 5.
Cell-Ba sed Assa ys To Ver ify th e Bifu n ction a l
P r ofile of Com p ou n d 5. To provide further evidence
that 5 interacts with both cellular targets of 2 and 3 as
suggested by the gene expression data, we performed
additional cell-based assays described below. According
to our hypothesis, a mixture of 2 and 3 should be
indistinguishable from 2 and 5 in cell cycle analysis.
As expected, cotreatment of 2 (8 µM) and 3 (8 µM) for
12 h resulted in a clear G2/M arrest of HCT116-C9 cells,
corresponding to the histogram data for 2 and 5 (Figure
1). This result indicates that the G2/M arrest phenotype
is predominant over the 3-like cell cycle phenotype in
the FACS analysis of 5.
In the compound 3 class, characteristic transcriptional
alteration (mainly repression) was observed in sets and
subsets of genes involved in cellular metabolism, cell
cycle control, signal transduction, transcription, mRNA
biogenesis, translation, immunomodulation, cell adhe-
sion, proteolysis, cytoskeleton, and membrane traffick-
ing. Among them, the down-regulation of subsets of
genes regulating mitochondrial functions and energy
metabolism implies cell cycle arrest at the G1/S check-
point (termed the restriction point), where nutritive
conditions are checked before committing to the DNA
replication.18 In yeast, cell size is monitored at this point
to determine whether sufficient nutrients are available
to complete a full cell cycle. The transcriptional repres-
sion of several DNA replication-associated and mitosis-
associated genes should also be noted because of its clear
correlation with the cell cycle effect of 3. Although the
precise mode of action of 3 has not yet been fully
understood, the microarray data shown here illuminate
drug-sensitive cellular pathways suggestive of potential
expression markers for the 3 type of antitumor activity.
We next compared growth-inhibitory effects of 2, 3,
and 5 in the MTT assay39 using P388 (parental), P388/
4.0r-M15 (subclone resistant to 2), HCT116-C9 (paren-
tal), and HCT116-C9-C129 (subclone resistant to 3). On
the IC50 basis, the relative resistances of P388/4.0r-M
and HCT116-C9-C1 to 2 and 3 were approximately 80-
and 140-fold, respectively (Table 4). These two subclones
showed no cross-resistance to other anticancer agents
such as adriamycin, etoposide, mitomycin C, cisplatin,
vincristine, and taxol (see ref 15 for P388/4.0r-M and
unpublished data taken by our group for HCT116-C9).
This suggests that their resistance mechanisms are
unrelated to the overexpression of P-glycoprotein caus-
ing a multidrug-resistance phenotype. Importantly, 5
displayed unchanged antiproliferative potency against
P388/4.0r-M (IC50 ) 1.2 µM) compared with the parental
P388 (IC50 ) 1.2 µM). Although HCT116-C9-C1 showed
a little cross-resistance to 5, the relative resistance value
of 5.9-fold was much less than the 140-fold to 3. In
addition, the IC50 value of 5 toward this subclone was
still at the low micromolar level (2.2 µM). These

4918 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 22
Owa et al.
F igu r e 2. Summary of the structure and gene expression relationships of a series of antitumor sulfonamides.
observations support the idea that 5 is effective against
the 2-resistant P388/4.0r-M by interacting with a cel-
lular target(s) of 3 and is also effective against 3-resis-
tant HCT116-C9-C1 by interacting with tubulin, a
cellular target of 2. Taken together, all of these results
are very consistent with the microarray data, leading
to a conclusion that 5 contains both antitumor profiles
of 2 and 3.
St r u ct u r e a n d Gen e E xp r ession R ela t ion sh ip
Stu d y. Exploiting the human genome database has
allowed the systematic construction of comprehensive
assay screens such as kinase and G-protein-coupled
receptor (GPCR) panels. High-throughput screening
using these assay panels facilitates detailed profiling
of biologically active small molecules on the basis of
their activity patterns (phenotypic fingerprints). It is
also possible to profile compounds in terms of their
genome-wide expression patterns (transcriptional fin-
gerprints) by using DNA microarray technology. Both
of these sophisticated characterization methods for
small-molecule collections are expected to be extremely
useful for rationalizing drug discovery and development
programs. In particular, array-based transcription pro-
filing of drug candidates will be increasingly in practical
and common use to provide validity for continuing
development. This is the basis for our work here to apply
microarray analysis to clarify the structure and gene
expression relationships of a series of antitumor sul-
fonamides.
The observed structure and gene expression relation-
ships are summarized in Figure 2. The major findings
are as follows: (i) The expression profile of the com-
pound 2 class of agents can be attributed to the N-(2-
aminophenyl or 2-amino-3-pyridinyl)-4-methoxybenze-
nesulfonamide motif based on the distinctive overlap
among the array data for 2, 4, and 5, including the
repression of two R-tubulin subtypes. (ii) The dual
character of 5, which has the antitumor characteristics
of both 2 and 3, is evident in the array-based profiling.
The compound 3-like activity profile of this molecule was
not seen clearly in the cell-based phenotypic screens
using COMPARE and FACS analyses because of the
predominant 2-like activity profile. (iii) Significant
overlap among the array data for 3, 5, and 6 indicates

Antitumor Sulfonamides
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 22 4919
that the 4 (para)-substituent on the phenyl moiety can
be varied without greatly influencing the expression
profile of the compound 3 class. (iv) Except for the
induction of two genes (tumor-associated antigen
GA733-1 and caveolin 1), there is no overlap between
the array data for 3 and 4 possessing a 2-chloro
substituent. This contrasts with the significant overlap
between the array data for 3 and 5 possessing a 3-chloro
substituent, suggesting that the 2-chloro substituent
obstructs an interaction with a cellular target(s) of the
compound 3 class. (v) Compared with 3, an overlapped
but very limited set of genes is affected by the des-chloro
derivative 7. From this observation and the SAR sum-
mary, it is conjectured that the 3-chloro substituent may
be critically involved in the interaction between 3 and
its molecular target(s). (vi) Because there are no genes
for which the mRNA levels are changed at least 2-fold
by 8, the sulfonamide linkage between two aromatic
rings is considered indispensable for the expression
profile of the compound 3 class. This is consistent with
the SAR summary. (vii) On the basis of points ii-vi, it
can finally be concluded that the N-(3-chloro-7-indolyl)-
benzenesulfonamide motif is an essential pharmacoph-
ore for the expression profile of the compound 3 class
of antitumor agents.
clinically achievable drug concentration, led to a 2-fold
or more repression of nine genes in common in three
different cancer cell lines, HCT116-C9 colon carcinoma,
MDA-MB-435 breast carcinoma, and MOLT-4 leukemia.
Importantly, these nine genes include DNA polymerase
R, cyclin H, and Rp8 homologue PDCD2 (unpublished
results of T. Owa, A. Yokoi, and T. Nagasu). The DNA
polymerase R protein is an essential enzyme for DNA
replication.41 The cyclin H protein forms part of the
CDK-activating kinase (CAK) together with CDK7 and
MAT1, playing a key role in coordinating cell cycle
progression, transcription, and DNA repair.42-44 The
significant down-regulation of these two genes correlates
with the cell cycle perturbation caused by 3. The precise
function of PDCD2 is still unclear, although its rat
homologue Rp8 is thought to be involved in apoptosis.45
Further microarray analyses and target-finding studies
of 3 are under way to establish the connection between
the functions of a target molecule(s) and the down-
stream effects on gene expression. Such efforts (medici-
nal genomic approaches) should allow us to utilize all
the above-mentioned expression changes by 3 as mo-
lecular biological endpoints for the pharmacodynamic
assessment of drug efficacy. Moreover, given the clinical
activity of 3, new important targets for future molecular-
based cancer therapeutics might be found among the
genes affected profoundly by this interesting compound.
Compound 3 has been evaluated in the phase I clinical
setting, displaying antitumor activities including dis-
ease stabilization.24-27 An appropriate assessment of
these clinical responses may require demonstration of
the drug effects on the primary target molecule(s) or
closely associated cellular pathways in cancer patients.
In relation to this point, preclinical measurements of
transcriptional changes in cancer cells (in vitro) or
tumors (in vivo) in response to compound 3 treatment
may provide potential expression markers indicative of
the drug efficacy. We therefore tried to narrow down
candidate genes implicated in compound 3 efficacy by
applying the following procedures to all our array
data: (i) eliminate genes also affected by structurally
related compounds with different mechanisms of action,
using the data for 2, 4, 8, and 3; (ii) select genes also
affected by different compounds with substantially the
same mechanism of action, using the data for 5-7 and
3. Although 7 is less active than 3, the fact that its mean
graph signature is similar to that of 3 suggests that the
compound still retains marginal ability to bind to a
cellular target(s) of 3. Thus, the array data for 7 were
used to reduce the number of putatively relevant genes.
In consequence, characteristic alterations in gene ex-
pression in response to the compound 3 class were
observed in the induction of NCBP interacting protein
1 (NCBP2, CBP20) and defective mariner transposon
Hsmar2 and in the repression of cyclin H, DNA poly-
merase R, nuclear pore protein Nup160, Rp8 homologue
PDCD2, oxygen-regulated protein ORP150, and neuron-
specific (γ) enolase. The transcriptional regulation of all
these genes could be highly associated with the 3-sensi-
tive pathways, at least in HCT116-C9 cells.
Con clu sion
The microarray analysis presented herein has made
it possible to characterize the compound 2 and com-
pound 3 classes of antitumor sulfonamides on the basis
of drug-induced changes in gene expression. The array
data allowed us to identify the essential pharmacophore
structure and drug-sensitive cellular pathways for each
class. Furthermore, the dual character of 5, which
displays characteristics of both 2 and 3, was revealed
by array-based transcription profiling, even though the
3-type profile of this molecule was not clear in cell-based
phenotypic screens using the COMPARE analysis and
cell cycle monitoring. In conclusion, these results pro-
vide an example of the potential benefits of structure
and gene expression relationship studies in medicinal
genomics.
Exp er im en ta l Section
Gen er a l. All commercial solvents and reagents were used
without further purification. Reactions were generally con-
ducted under a nitrogen atmosphere using magnetic stirring.
Analytical thin-layer chromatography (TLC) was performed
on silica gel plates (Merck, No. 5715). Merck silica gel 60 (230-
400 mesh) was used for the flash chromatographic purification
of some compounds. Compounds 2, 3, 5, and 6 were synthe-
sized using procedures reported previously13,18 and character-
ized by obtaining spectroscopic and analytical data as de-
scribed earlier.13,18 The other new derivatives 4, 7, and 8 were
characterized according to the following general methods.
Melting points were determined on a Yanagimoto micromelt-
ing point apparatus and are reported uncorrected. Proton (1H)
NMR spectra were obtained at 400 MHz on a Varian UNITY
400 spectrometer in DMSO-d6. Chemical shifts are expressed
in δ (ppm) units relative to tetramethylsilane (TMS) reference.
Mass spectra (FABMS) were recorded on a J EOL J MS-
SX102AQQ using a direct introduction method in the FAB+
ion mode. Elemental analyses (C, H, N) were carried out at
the Eisai Analytical Chemistry Section, and the results were
within (0.4% of the theoretical values.
Besides the chemical genomic approach40 with a
variety of compounds as described above, conventional
approaches with different cancer species and varied
drug concentrations have also been employed for iden-
tifying the genes most relevant to compound 3 efficacy.
In such experiments, 0.8 µM of 3 (12 h treatment), a

4920 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 22
Owa et al.
N -(2-Ch lor o-7-in d olyl)-4-m e t h oxyb e n ze n e su lfon a -
m id e (4). To a solution of 2-chloro-7-nitroindole46 (460 mg, 2.34
mmol) in 2-propanol (8 mL) were added Fe powder (392 mg,
7.02 mmol) and an aqueous solution (1.6 mL) of NH4Cl (25
mg, 0.467 mmol). After the mixture was stirred at 60 °C for 2
h, activated charcoal (ca. 200 mg) was added to the reaction
mixture. The insoluble materials were filtered off and washed
with EtOAc (20 mL), followed by the immediate addition of
4-methoxybenzenesulfonyl chloride (484 mg, 2.34 mmol) and
pyridine (0.76 mL, 9.40 mmol). The reaction mixture was
stirred at room temperature for 3 h, diluted with EtOAc, and
washed successively with 0.5 N aqueous HCl, H2O, saturated
aqueous NaHCO3, and brine. The organic layer was dried over
anhydrous MgSO4, filtered, and concentrated under reduced
pressure. Purification by flash chromatography (eluted EtOAc-
hexane) gave 4 (652 mg, 83%) as a white solid: melting point
150-152 °C (after recrystallization from EtOAc-hexane); 1H
NMR (400 MHz, DMSO-d6) δ 3.78 (3H, s), 6.43 (1H, s), 6.74
(1H, dd, J ) 7.7, 0.92 Hz), 6.86 (1H, dd, J ) 7.9, 7.7 Hz), 7.00-
7.05 (2H, AA′BB′), 7.21 (1H, dd, J ) 7.9, 0.92 Hz), 7.65-7.70
(2H, AA′BB′), 9.52 (1H, br s), 11.48 (1H, br s); MS (FAB+)
m/z 336 (M+). Anal. (C15H13N2O3SCl) C, H, N.
MDA-MB-231); brain tumor (U251, SF-268, SF-295, SF-539,
SNB-75, and SNB-78); colon cancer (HCC2998, KM-12, HT-
29, HCT15, and HCT116); lung cancer (NCI-H23, NCI-H226,
NCI-H522, NCI-H460, A549, DMS273, and DMS114); mela-
noma (LOX-IMVI); ovarian cancer (OVCAR-3, OVCAR-4,
OVCAR-5, OVCAR-8, and SK-OV-3); renal cancer (RXF-631L
and ACHN); stomach cancer (St-4, MKN1, MKN7, MKN28,
MKN45, and MKN74); prostate cancer (DU-145 and PC-3). All
of these cell lines were grown in RPMI 1640 supplemented
with 5% fetal bovine serum, penicillin (100 units/mL), and
streptomycin (100 µg/mL) at 37 °C in humidified air containing
5% CO2. Measurements of cell growth inhibition and data
calculations were performed according to the reported meth-
ods.4,11 The sulforhodamine B assay28 with 48 h drug exposure
was employed for the chemosensitivity screen in the DOS.
Drug concentrations with 10-fold dilution ranging from 100
µM to 10 nM were tested in the assay. The main measure of
chemosensitivity used in the database and calculations of
Pearson’s correlation coefficients was GI50 (drug concentration
that inhibits cell proliferation by 50% relative to untreated
control cells).
Cell Clon in g a n d Cu ltu r e. HCT116 human colon carci-
noma was purchased from American Type Culture Collection
(Manassas, VA). HCT116-C9 was isolated as a 3-sensitive
subclone from the parental HCT116 cell line by using a
limiting dilution method.29 This cell line was used for flow
cytometry and oligonucleotide microarray analyses. HCT116-
C9-C1, a 3-resistant subclone, was obtained from a viable
colony of HCT116-C9 after continuous drug exposure with
serial dose escalation of 3.29 P388 murine leukemia was
supplied by the Cancer Chemotherapy Center, J apan Founda-
tion for Cancer Research (Tokyo, J apan). P388/4.0r-M, a
2-resistant subclone, was established at Tsukuba Research
Laboratories, Eisai Co., Ltd.15 HCT116-C9 and HCT116-C9-
C1 were maintained in RPMI 1640 medium containing 10%
fetal bovine serum, penicillin (100 units/mL), and streptomycin
(100 µg/mL) at 37 °C in humidified air containing 5% CO2.
For the culture of P388 and P388/4.0r-M, mercaptoethanol (5
× 10-2 mM) and sodium pyruvate (1 mM) were added to the
medium as supplements. All of these cell lines were used for
cell growth inhibition assays with 3-(4,5-dimethylthiazol-2-yl)-
2,5-diphenyltetrazolium bromide (MTT).39
F low Cytom etr ic Cell Cycle An a lysis.29 Exponentially
growing HCT116-C9 cells were seeded in six-well plates (4.5
× 105 cells/well) and precultured for 1 day. Test compounds
(2.0 mM stock solution in DMSO) were added to the plates at
the final concentration of 8 µM, and incubation was continued
for 12 h. Compound 3 was further tested at extended time
points of 24, 36, and 48 h to examine the time course of its
cell cycle effect. The combination effect of 2 and 3 on cell cycle
progression was also investigated at the 12 h time point. For
fluorescence-activated cell sorter (FACS) analysis, the cells
were fixed in 70% ethanol, treated with RNase (1 mg/mL), and
stained with propidium iodide (50 µg/mL). The DNA content
was quantified by detecting red fluorescence with a flow
cytometer (FACSCalibur, Becton-Dickinson). Data analysis
was done with CellQuest and ModFit LT software (Becton-
Dickinson). The experiment was carried out in triplicate for
each analysis.
N-(7-In d olyl)-1,4-ben zen ed isu lfon a m id e (7). A solution
of 7-nitroindole (2.74 g, 16.9 mmol) in MeOH (75 mL) was
hydrogenated over 10% palladium on carbon (270 mg) under
H2 at 1 atm overnight. The catalyst was filtered off, and the
filtrate was evaporated to give almost pure 7-aminoindole. The
amine product was immediately dissolved in EtOAc (50 mL)
and pyridine (2.8 mL, 34.6 mmol) and then reacted with
4-sulfamoylbenzenesulfonyl chloride47 (4.32 g, 16.9 mmol) at
room temperature for 5 h. The reaction mixture was diluted
with EtOAc and washed successively with 0.5 N aqueous HCl,
H2O, saturated aqueous NaHCO3, and brine. The organic layer
was dried over anhydrous MgSO4, filtered, and concentrated
under reduced pressure. Purification by flash chromatography
(eluted with EtOAc-hexane) gave 7 (5.08 g, 86%) as a white
solid: melting point 218-219 °C (after recrystallization from
EtOH-hexane); 1H NMR (400 MHz, DMSO-d6) δ 6.41 (1H, dd,
J ) 3.1, 2.0 Hz), 6.68 (1H, dd, J ) 7.7, 0.92 Hz), 6.83 (1H, dd,
J ) 7.7, 7.9 Hz), 7.31 (1H, d, J ) 3.1 Hz), 7.33 (1H, dd, J )
7.9, 0.92 Hz), 7.56 (2H, br s), 7.93 (4H, s), 10.16 (1H, br), 10.80
(1H, br s); MS (FAB+) m/z 351 (M+). Anal. (C14H13N3O4S2) C,
H, N.
N-(3-Ch lor o-7-in d olyl)-4-su lfa m oylben za m id e (8). To a
solution of 3-chloro-7-nitroindole18 (1.50 g, 7.63 mmol) in
2-propanol (20 mL) were added Fe powder (1.28 g, 22.9 mmol)
and an aqueous solution (4 mL) of NH4Cl (82 mg, 1.53 mmol).
After the mixture was stirred at 60 °C for 2 h, activated
charcoal (ca. 500 mg) was added to the reaction mixture. The
insoluble materials were filtered off and washed with EtOAc
(200 mL). The obtained organic layer was washed with
saturated aqueous NaHCO3 and brine, dried over anhydrous
MgSO4, filtered, and concentrated under reduced pressure to
give almost pure 7-amino-3-chloroindole as a grayish-brown
solid. The amine product was immediately dissolved in THF
(20 mL), and then 4-sulfamoylbenzoic acid (1.54 g, 7.65 mmol)
and 1,1′-carbonyldiimidazole (1.36 g, 8.39 mmol) were added
to the solution. After being stirred under reflux overnight, the
reaction mixture was diluted with EtOAc and washed succes-
sively with 0.5 N aqueous HCl, H2O, saturated aqueous
NaHCO3, and brine. The organic layer was dried over anhy-
drous MgSO4, filtered, and concentrated under reduced pres-
sure. Purification by flash chromatography (eluted with EtOAc-
hexane) gave 7 (1.89 g, 71%) as a white solid: melting point
280-281.5 °C (after recrystallization from EtOH-hexane); 1H
NMR (400 MHz, DMSO-d6) δ 7.14 (1H, dd, J ) 7.9, 7.7 Hz),
7.38 (1H, dd, J ) 7.9, 0.92 Hz), 7.41 (1H, dd, J ) 7.7, 0.92
Hz), 7.56 (2H, br s), 7.56 (1H, s), 7.95-8.01 (2H, AA′BB′), 8.16-
8.22 (2H, AA′BB′), 10.39 (1H, br), 11.24 (1H, br s); MS (FAB+)
m/z 349 (M+). Anal. (C15H12N3O3SCl) C, H, N.
Tu bu lin P olym er iza tion Assa y.15,18 Test compounds dis-
solved at graded concentrations in DMSO were added to the
wells of 96-well microtiter plates containing 1 mg/mL micro-
tubule protein from bovine brain. A final DMSO concentration
of 5% (v/v) and compound concentrations of 0.30, 1.0, 3.0, 10,
30, and 100 µM were used for assays in polymerization buffer
(0.1 M 2-(N-morpholino)ethanesulfonic acid, 1 mM EGTA, 0.5
mM MgCl2, pH 6.8). After 30 min of incubation at 4 °C, the
mixture was warmed to 37 °C with 1 mM GTP to start tubulin
polymerization. Microtubule assembly was monitored by mea-
suring the turbidity kinetically at 350 nm with a microplate
reader (THERMO max, Molecular Devices) for 40 min. At the
time point of 30 min, the polymerization of the control reached
a plateau level. Therefore, the turbidity at 30 min was plotted
on a linear scale versus the concentration of the test compound
Hu m a n Ca n cer Cell Lin e P a n el, Cell Gr ow th In h ibi-
tion Assa y, a n d Da ta An a lysis. Characterization of 39
human cancer cell lines of the DOS has been described
elsewhere:11 breast cancer (HBC-4, BSY-1, HBC-5, MCF-7, and

Antitumor Sulfonamides
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 22 4921
Cancer Inst. 1989, 81, 1088-1092. (b) Monks, A.; Scudiero, D.;
Skehan, P.; Shoemaker, R.; Paull, K.; et al. Feasibility of a high-
flux anticancer drug screen using a diverse panel of cultured
human tumor cell lines. J . Natl. Cancer Inst. 1991, 83, 757-
766.
on a log scale. The experiment was carried out in triplicate
for each compound, and the drug concentration causing 50%
inhibition of polymerization was determined by the least-
squares method.
High -Den sity Oligon u cleotide Ar r ay Expr ession An aly-
sis.29 HCT116-C9 cells were plated at 5.0 × 106 cells/dish in
10 cm diameter dishes with 10 mL of fresh medium. After 24
h of preincubation, the cells were treated for 12 h with each
test compound at 8 µM or with 0.015% DMSO (control).
Sample preparation was performed according to established
protocols.12 In brief, total RNA was extracted from the cells
using Trizol (GIBCO BRL) and further purified with RNeasy
columns (Qiagen). Double-stranded cDNA was prepared from
10 µg of total RNA using the SuperScript Choice System
(GIBCO BRL) and T7-d(T)24 primers. The cDNA product was
purified by phenol-chloroform extraction. In vitro transcrip-
tion was carried out with an RNA transcript labeling kit
containing biotinylated UTP and CTP (Enzo Diagnostics).
After purification with RNeasy columns, the cRNA was
fragmented and then hybridized to Affymetrix GeneChip
HuGene FL (Hu6800) arrays. According to the EukGE-WS2
protocol, the probe arrays were washed and stained with
streptavidin-phycoerythrin and biotinylated goat antistrepta-
vidin on an Affymetrix fluidics station. Fluorescence intensities
were captured with a Hewlett-Packard confocal laser scanner.
All quantitative data were processed using the Affymetrix
GeneChip software.12 Up- or down-regulated genes were
selected according to the criteria as described in the legend to
Table 3 (Supporting Information).
(5) Ramaswamy, S.; Tamayo, P.; Rifkin, R.; Mukherjee, S.; Yeang,
C. H.; et al. Multiclass cancer diagnosis using tumor gene
expression signatures. Proc. Natl. Acad. Sci. U.S.A. 2001, 98,
15149-15154.
(6) Su, A. I.; Welsh, J . B.; Sapinoso, L. M.; Kern, S. G.; Dimitrov,
P.; et al. Molecular classification of human carcinomas by use
of gene expression signatures. Cancer Res. 2001, 61, 7388-7393.
(7) Zimmermann, J .; Erdmann, D.; Lalande, I.; Grossenbacher, R.;
Noorani, M.; et al. Proteasome inhibitor induced gene expression
profiles reveal overexpression of transcriptional regulators ATF3,
GADD153 and MAD1. Oncogene 2000, 19, 2913-2920.
(8) Kudoh, K.; Ramanna, M.; Ravatn, R.; Elkahloun, A. G.; Bittner,
M. L.; et al. Monitoring the expression profiles of doxorubicin-
induced and doxorubicin-resistant cancer cells by cDNA mi-
croarray. Cancer Res. 2000, 60, 4161-4166.
(9) Martinez, E. J .; Corey, E. J .; Owa, T. Antitumor activity- and
gene expression-based profiling of ecteinascidin Et 743 and
phthalascidin Pt 650. Chem. Biol. 2001, 8, 1151-1160.
(10) Rabow, A. A.; Shoemaker, R. H.; Sausville, E. A.; Covell, D. G.
Mining the National Cancer Institute’s tumor-screening data-
base: identification of compounds with similar cellular activities.
J . Med. Chem. 2002, 45, 818-840.
(11) (a) Dan, S.; Tsunoda, T.; Kitahara, O.; Yanagawa, R.; Zembutsu,
H.; et al. An integrated database of chemosensitivity to 55
anticancer drugs and gene expression profiles of 39 human
cancer cell lines. Cancer Res. 2002, 62, 1139-1147. (b) Yamori,
T.; Matsunaga, A.; Sato, S.; Yamazaki, K.; Komi, A.; et al. Potent
antitumor activity of MS-247, a novel DNA minor groove binder,
evaluated by an in vitro and in vivo human cancer cell line panel.
Cancer Res. 1999, 59, 4042-4049. (c) Naasani, I.; Seimiya, H.;
Yamori, T.; Tsuruo, T. FJ 5002: a potent telomerase inhibitor
identified by exploiting the disease-oriented screening program
with COMPARE analysis. Cancer Res. 1999, 59, 4004-4011.
(12) Lockhart, D. J .; Dong, H.; Byrne, M. C.; Follettie, M. T.; Gallo,
M. V.; et al. Expression monitoring by hybridization to high-
density oligonucleotide arrays. Nat. Biotechnol. 1996, 14, 1675-
1680.
(13) Yoshino, H.; Ueda, N.; Niijima, J .; Sugumi, H.; Kotake, Y.; et
al. Novel sulfonamides as potential, systemically active antitu-
mor agents. J . Med. Chem. 1992, 35, 2496-2497.
(14) Owa, T.; Okauchi, T.; Yoshimatsu, K.; Sugi, N. H.; Ozawa, Y.;
et al. A focused compound library of novel N-(7-indolyl)benze-
nesulfonamides for the discovery of potent cell cycle inhibitors.
Bioorg. Med. Chem. Lett. 2000, 10, 1223-1226.
(15) Yoshimatsu, K.; Yamaguchi, A.; Yoshino, H.; Koyanagi, N.;
Kitoh, K. Mechanism of action of E7010: inhibition of mitosis
by binding to the colchicine site of tubulin. Cancer Res. 1997,
57, 3208-3213.
(16) Koyanagi, N.; Nagasu, T.; Fujita, F.; Watanabe, T.; Tsukahara,
K.; et al. In vivo tumor growth inhibition produced by a novel
sulfonamide, E7010, against rodent and human tumors. Cancer
Res. 1994, 54, 1702-1706.
(17) Yamamoto, K.; Noda, K.; Yoshimura, A.; Fukuoka, M.; Furuse,
K.; et al. Phase I study of E7010. Cancer Chemother. Pharmacol.
1998, 42, 127-134.
(18) Owa, T.; Yoshino, H.; Okauchi, T.; Yoshimatsu, K.; Ozawa, Y.;
et al. Discovery of novel antitumor sulfonamides targeting G1
phase of the cell cycle. J . Med. Chem. 1999, 42, 3789-3799.
(19) Owa, T.; Nagasu, T. Novel sulphonamide derivatives for the
treatment of cancer. Expert Opin. Ther. Pat. 2000, 10, 1725-
1740.
(20) Ozawa, Y.; Sugi, N. H.; Nagasu, T.; Owa, T.; Watanabe, T.; et
al. E7070, a novel sulfonamide agent with potent antitumor
activity in vitro and in vivo. Eur. J . Cancer 2001, 37, 2275-
2282.
(21) Watanabe, T.; Sugi, N.; Ozawa, Y.; Owa, T.; Nagasu, T.; et al. A
novel antitumor agent ER-35744, targeting G1 phase. III.
Studies of mechanism of action. Proc. Am. Assoc. Cancer Res.
1996, 37, A2667, 391.
(22) Fukuoka, K.; Usuda, J .; Iwamoto, Y.; Fukumoto, H.; Nakamura,
T.; et al. Mechanism of action of the novel sulfonamide antican-
cer agent E7070 on cell cycle progression in human non-small
cell lung cancer cells. Invest. New Drugs 2001, 19, 219-227.
(23) Owa, T.; Yoshino, H.; Yoshimatsu, K.; Nagasu, T. Cell cycle
regulation in the G1 phase: a promising target for the develop-
ment of new chemotherapeutic anticancer agents. Curr. Med.
Chem. 2001, 8, 1487-1503.
(24) Dittrich, C.; Dumez, H.; Calvert, H.; Hanauske, A. R.; Ravic,
M.; et al. Phase I and pharmacokinetic study of E7070 in
patients with solid tumors as single iv infusion, weekly × 4, q 6
weeks. Proc. Am. Assoc. Cancer Res. 2000, 41, A3875, 609.
Cell Gr ow th In h ibition Assa y Usin g Com p ou n d
2
Resista n t a n d Com p ou n d 3 Resista n t Ca n cer Cell Lin es.
P388/4.0r-M15 and HCT116-C9-C129 were used as subclonal
cell lines resistant to 2 and 3, respectively. P388 and P388/
4.0r-M cells were seeded at 1.25 × 103 cells/well in 96-well
round-bottomed plates, and HCT116-C9 and HCT116-C9-C1
cells were seeded at 3.0 × 103 cells/well in 96-well flat-
bottomed plates. Compounds 2, 3, and 5 were dissolved at 20
mM in DMSO and further diluted with the culture medium
(see Cell Cloning and Culture above) to prepare 3-fold serial
dilutions with the maximum concentration being 100 µM after
the addition into each well. After 24 h of preculture, the
obtained dilutions were added to the plates and then incuba-
tion was continued for 3 days. The antiproliferative activity
(IC50 value) was determined by the MTT colorimetric method.39
The assays were performed twice in triplicate with each test
compound. Relative resistance (RR) values were calculated as
the IC50 values for the resistant cell line divided by the IC50
values for the parental cell line.
Ack n ow led gm en t. We thank Dr. J . Kuromitsu and
Dr. T. Kawai for their helpful discussions and assistance
with the microarray analysis. Mass spectra and elemen-
tal analysis data were kindly provided by the Eisai
Analytical Chemistry Section.
Su p p or tin g In for m a tion Ava ila ble: The log GI50 (GI50
in units of M) data for compounds 2-8 in the DOS panel of 39
human cancer cell lines and Table 3 presenting the full details
of gene expression data for the same test compounds. This
material is available free of charge via the Internet at http://
pubs.acs.org..
Refer en ces
(1) Lander, E. S.; Linton, L. M.; Birren, B.; Nusbaum, C.; Zody, M.
C.; et al. Initial sequencing and analysis of the human genome.
Nature 2001, 409, 860-921.
(2) Venter, J . C.; Adams, M. D.; Myers, E. W.; Li, P. W.; Mural, R.
J .; et al. The sequence of the human genome. Science 2001, 291,
1304-1351.
(3) Scherf, U.; Ross, D. T.; Waltham, M.; Smith, L. H.; Lee, J . K.; et
al. A gene expression database for the molecular pharmacology
of cancer. Nat. Genet. 2000, 24, 236-244.
(4) (a) Paull, K. D.; Shoemaker, R. H.; Hodes, L.; Monks, A.;
Scudiero, D. A.; et al. Display and analysis of patterns of
differential activity of drugs against human tumor cell lines:
development of mean graph and COMPARE algorithm. J . Natl.

4922 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 22
Owa et al.
(25) Droz, J . P.; Roche, H.; Zanetta, S.; Terret, C.; Blay, J . Y.; et al.
Phase I trial of five-days continuous infusion E7070 [N-(3-chloro-
7-indolyl)-1,4-benzenedisulfonamide] in patients (pts) with solid
tumors. Proc. Am. Assoc. Cancer Res. 2000, 41, A3876, 609.
(26) (a) Punt, C. J . A.; Fumoleau, P.; Walle, B. V. D.; Deporte-Fety,
R.; Bourcier, C.; et al. Phase I and pharmacokinetic (PK) study
of E7070, a novel sulphonamide, given daily × 5 every 3 weeks
as a 1-hour intravenous infusion in patients (pts) with advanced
solid tumors. Proc. Am. Assoc. Cancer Res. 2000, 41, A3877, 609.
(b) Punt, C. J .; Fumoleau, P.; van de Walle, B.; Faber, M. N.;
Ravic, M.; et al. Phase I and pharmacokinetic study of E7070, a
(37) Beauregard, D. A.; Hill, S. A.; Chaplin, D. J .; Brindle, K. M. The
susceptibility of tumors to the antivascular drug combretastatin
A4 phosphate correlates with vascular permeability. Cancer Res.
2001, 61, 6811-6815.
(38) (a) Supuran, C. T.; Scozzafava, A. Carbonic anhydrase inhibitors
and their therapeutic potential. Expert Opin. Ther. Pat. 2000,
10, 575-600. (b) Scozzafava, A.; Supuran, C. T. Carbonic
anhydrase inhibitors: Synthesis of N-morpholyl-thiocarbonyl-
sulfenylamino aromatic/heterocyclic sulfonamides and their
interaction with isozymes I, II and IV. Bioorg. Med. Chem. Lett.
2000, 10, 1117-1120. (c) Supuran, C. T.; Briganti, F.; Tilli, S.;
Chegwidden, W. R.; Scozzafava, A. Carbonic anhydrase inhibi-
tors: sulfonamides as antitumor agents? Bioorg. Med. Chem.
2001, 9, 703-714.
(39) Mosmann, T. Rapid colorimetric assay for cellular growth and
survival: application to proliferation and cytotoxicity assays. J .
Immunol. Methods 1983, 65, 55-63.
(40) Stockwell, B. R.; Hardwick, J . S.; Tong, J . K.; Schreiber, S. L.
Chemical genetic and genomic approaches reveal a role for
copper in specific gene activation. J . Am. Chem. Soc. 1999, 121,
10662-10663.
(41) Foiani, M.; Lucchini, G.; Plevani, P. The DNA polymerase alpha-
primase complex couples DNA replication, cell-cycle progression
and DNA damage response. Trends Biochem. Sci. 1997, 22, 424-
427.
(42) Nigg, E. A. Cyclin-dependent kinase 7: at the cross-roads of
transcription, DNA repair and cell cycle control? Curr. Opin. Cell
Biol. 1996, 8, 312-317.
(43) Yankulov, K. Y.; Bentley, D. L. Regulation of CDK7 substrate
specificity by MAT1 and TFIIH. EMBO J . 1997, 16, 1638-1646.
(44) Araujo, S. J .; Tirode, F.; Coin, F.; Pospiech, H.; Syvaoja, J . E.;
et al. Nucleotide excision repair of DNA with recombinant
human proteins: definition of the minimal set of factors, active
forms of TFIIH, and modulation by CAK. Genes Dev. 2000, 14,
349-359.
(45) Kawakami, T.; Furukawa, Y.; Sudo, K.; Saito, H.; Takami, S.;
et al. Isolation and mapping of a human gene (PDCD2) that is
highly homologous to Rp8, a rat gene associated with pro-
grammed cell death. Cytogenet. Cell Genet. 1995, 71, 41-43.
(46) Nakashima, T.; Suzuki, I. Ring contraction of 3-hydroxyquino-
lines to oxindoles with hydrogen peroxide in acetic acid. Chem.
Pharm. Bull. 1969, 17, 2293-2298.
novel sulfonamide, given at
a daily times five schedule in
patients with solid tumors. A study by the EORTC-early clinical
studies group (ECSG). Ann. Oncol. 2001, 12, 1289-1293.
(27) Raymond, E.; ten Bokkel-Huinink, W. W.; Taieb, J .; Beijnen, J .
H.; Mekhaldi, S.; et al. Phase I and pharmacokinetic (PK) study
of E7070, a new chloroindolylsulfonamide, given as a 1-hour
infusion every 3 weeks in patients (pts) with advanced solid
tumors. Proc. Am. Assoc. Cancer Res. 2000, 41, A3889, 611.
(28) Skehan, P.; Storeng, R.; Scudiero, D.; Monks, A.; McMahon, J .;
et al. New colorimetric cytotoxicity assay for anticancer-drug
screening. J . Natl. Cancer Inst. 1990, 82, 1107-1112.
(29) Yokoi, A.; Kuromitsu, J .; Kawai, T.; Nagasu, T.; Sugi, N. H.; et
al. Profiling novel sulfonamide antitumor agents with cell-based
phenotypic screens and array-based gene expression analysis.
Mol. Cancer Ther. 2002, 1, 275-286.
(30) Ro¨dicker, F.; Stiewe, T.; Zimmermann, S.; Pu¨tzer, B. M. Thera-
peutic efficacy of E2F1 in pancreatic cancer correlates with TP73
induction. Cancer Res. 2001, 61, 7052-7055.
(31) Ben-Ze’ev, A.; Farmer, S. R.; Penman, S. Mechanisms of regulat-
ing tubulin synthesis in cultured mammalian cells. Cell 1979,
17, 319-325.
(32) Rosania, G. R.; Chang, Y.-T.; Perez, O.; Sutherlin, D.; Dong, H.;
et al. Myoseverin, a microtubule-binding molecule with novel
cellular effects. Nat. Biotechnol. 2000, 18, 304-308.
(33) Pettit, G. R.; Rhodes, M. R. Antineoplastic agents 389. New
syntheses of the combretastatin A-4 prodrug. Anticancer Drug
Des. 1998, 13, 183-191.
(34) Ohsumi, K.; Nakagawa, R.; Fukuda, Y.; Hatanaka, T.; Morinaga,
Y.; et al. Novel combretastatin analogues effective against
murine solid tumors: design and structure-activity relation-
ships. J . Med. Chem. 1998, 41, 3022-3032.
(35) Wang, L.; Woods, K. W.; Li, Q.; Barr, K. J .; McCroskey, R. W.;
et al. Potent, orally active heterocycle-based combretastatin A-4
analogues: synthesis, structure-activity relationship, pharma-
cokinetics, and in vivo antitumor activity evaluation. J . Med.
Chem. 2002, 45, 1697-1711.
(36) Adams, J .; Elliott, P. J . New agents in cancer clinical trials.
Oncogene 2000, 19, 6687-6692.
(47) Holland, G. F.; Funderburk, W. H.; Finger, K. F. Preparation
and anticonvulsant activity of N-substituted benzenedisulfona-
mides. J . Med. Chem. 1963, 6, 307-312.
J M0201060