Critical structural motif for the catalytic inhibition of human topoisomerase II by UK-1 and analogs

Article abstract of DOI:10.1016/j.bmcl.2004.03.095

Three new analogs of UK-1 have been synthesized and their efficacies as topoisomerase II inhibitors have been determined. Results show that UK-1 and two of these analogs are catalytic inhibitors of topo II and identifies a critical structure motif necessa

Full text of DOI:10.1016/j.bmcl.2004.03.095

Bioorganic & Medicinal ChemistryLetters 14 (2004) 3221–3226
Critical structural motif for the catalytic inhibition of human
topoisomerase II by UK-1 and analogs
Ben B. Wang, Nima Maghami, Vanessa L. Goodlin and Paul J. Smith*
Department of Chemistry and Biochemistry, University of Maryland, Baltimore County, Baltimore, MD 21250, USA
Received 6 January2004; revised 29 March 2004; accepted 30 March 2004
Abstract—Three new analogs of UK-1 have been synthesized and their efficacies as topoisomerase II inhibitors have been deter-
mined. Results show that UK-1 and two of these analogs are catalytic inhibitors of topo II and identifies a critical structural motif
necessaryfor enzyme inhibition.
Ó 2004 Elsevier Ltd. All rights reserved.
1. Introduction
with a section of duplex DNA, termed the G-segment.
This signals the binding of two ATP molecules and
causes the N-terminal domains to dimerize. This con-
formational change traps the second DNA duplex (the
T-segment) and also induces cleavage of the G-segment.
Hydrolysis of the ATP molecules results in the passing
of the T-segment through the nicked G-segment and the
religation of the G-segment, after which the enzyme
dissociates from the DNA.6
DNA topoisomerases are nuclear enzymes responsible
for controlling the topological state of DNA, carrying
out a series of DNA cleavage and religation reactions.1
Two major forms of topoisomerases are found in the
eukaryotic cells, topoisomerase I (topo I) and topoiso-
merase II (topo II). Topo I creates single-stranded DNA
breaks, allowing the opposing intact strand to pass
through the nick, whereas topo II creates double
stranded breaks, allowing intact duplex DNA to pass
through the gap.2;3 These actions are essential during
DNA replication and transcription, where accumulated
topological stress must be alleviated. Topo II is partic-
ularlyimportant during cell replication, where the
intertwined strands of DNA in the chromosomes must
be separated before cell divisions can occur.
It is well known that compounds which interfere with
normal topoisomerase activityare highlycytotoxic; a
number of these such as topotecan7 (topo I) and eto-
poside8 (topo II) are currentlyin clinical use as anti-
cancer agents. Due to the complexityof the enzymatic
process, drug inhibition mechanisms can also be highly
complicated. Several compounds are known to interfere
with the enzyme in different ways (Fig. 1). For example,
etoposide inhibits topoisomerase byinterfering with the
religation of the G-segment, causing prolonged covalent
attachment of the enzyme to the DNA (the ‘cleavable
complex’).9 Compounds that act bythis mechanism are
known as topo II ‘poisons’, and cause the accumulation
of double-stranded DNA breaks, ultimatelyleading to
apoptotic cell death. (Interestingly, recent studies show
that etoposide can bind to the N-terminal ATPase
domain as well as the enzyme active site, adding com-
plexityto the mechanism of drug action. 10) DNA
binding compounds such as netropsin interfere with
topo II activitynot bystabilizing the cleavable complex,
but bypreventing the association of the enzyme with
DNA.11 This type of inhibition is a totally nonspecific
Topo II is a homodimer with each subunit containing
two domains, the ATPase domain and DNA binding
domain. The crystal structure of the DNA binding
domain of yeast topo II reveals a conserved tyrosine
residue located in the active site (tyrosine-783).4
Although the details concerning the mechanism of topo
II are not completelyclear, a working model has been
proposed based on the currentlyavailable biochemical
and structural information.5 The enzyme first associates
Keywords: Topoisomerase II; UK-1; Benzoxazole.
* Corresponding author. Tel.: +1-410-455-2519; fax: +1-410-455-2608;
e-mail: pjsmith@umbc.edu
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.03.095

3222
B. B. Wang et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3221–3226
H3C
O
O
H
N
O
O
O
O
O
N
OH
HO
O
H
H
N
NH
N
O
O
O
HN
O
H3C
O
NH
NH
O
N
NH
HN
H2N
H3CO
OCH3
N
H
O
NH2
OH
B
C
A
Figure 1. Examples of some topo II inhibitors: (A) etopside, (B) netropsin, and (C) ICRF-187.
effect and not limited to topo II inhibition. Additionally,
drugs such as bisdioxopiperazine ICRF-187 interact
with the ATPase domains, preventing ATP hydrolysis
and enzyme turn-over.12 Clearly, it is important to fully
understand the mechanism of action for each class of
compounds in order to design more effective topoiso-
merase II inhibitors.
has been offered. Recently, Kumar et al. showed that
some truncated analogs of UK-1 share similar anti-
cancer activities.17 However, the exact mechanism of
UK-1 is still unclear. To further the understanding of
the anti-topo II activityof UK-1, three new analogs
have been synthesized with more modest structural
changes. These lack the methyl ester (1, Fig. 2), the or-
tho-hydroxyl (2), or both (3), allowing us to systemati-
callyevaluate the importance of these groups for
activity. The syntheses of these compounds and results
from topo II inhibition studies are presented below.
A more recentlydiscovered topoisomerase II inhibitor is
the natural product UK-1 (Fig. 2).13 First identified in
1993, UK-1 is a structurallyunique bis(benzoxazole)
metabolite of Streptomyces sp. 517-02.14 This compound
inhibits the growth of murine cancer cell line P388 with
an IC50 in the 0.3–1.6 lM range.15 UK-1 binds divalent
metal ions13 and binds DNA in a divalent metal ion-
Analogs of UK-1 were prepared in which either the
methyl-ester or the ortho-hydroxyl functionality was
deleted as shown in Figure 1. The syntheses used parallel
that previouslyreported for UK-1 with some modifica-
tions (Scheme 1).18 2-(Benzyloxy)benzoic acid (4)
was coupled with methyl anthranilic acid (5) using
1-hydroxybenzotriazole (HOBT), 2-(1-H-benzotriazol-
dependent manner.16 Among Mg2þ, Ca2þ, Co2þ, Ni2þ
,
and Zn2þ, Mg2þ was found to be the most effective to
promote association with DNA. However, no hypoth-
esis or data regarding the DNA binding mode of UK-1
HO
HO
CO2Me
CO2Me
N
O
N
O
N
O
N
O
N
N
N
O
N
O
O
O
UK-1
1
2
3
Figure 2. UK-1 and minimal analogs.
CO2H
OBn
+
CO2Me
CO2Me
CO2H
BnO
BnO
NH2
N
N
c
a, b
O
O
OH
4
5
6
7
60%
99%
HO
R2
O
OH
BnO
NH
R2
N
O
N
N
d
e
O
O
8: R1 = CO2Me
9: R1 = H
75%
87%
UK-1: R1 = CO2Me
1: R1 = H
40%
41%
Scheme 1. Reagents and conditions: (a) HOBT, HBTU, DIPEA, DMF, rt; (b) 230 °C; (c) 5 M NaOH, THF, reflux; (d) 5 or 2-aminophenol, HOBT,
HBTU, DIPEA, DMF, rt; (e) TsOH, toluene, reflux.

B. B. Wang et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3221–3226
3223
lyl)-1,1,3,3-tetramethyluronium
hexafluorophosphate
DNA will not be religated, resulting in the appearance
of linear DNA. If a compound simplyinterferes with the
enzyme, either by preventing DNA binding11 or disso-
ciation after completing the strand passage12, no linear
DNA would be observed, but there would be an increase
in supercoiled DNA (relative to drug free control). In
the case of UK-1 (Fig. 3A), no linear DNA is observed,
but a concentration dependent increase in the amount of
supercoiled DNA is seen.25 Similar results were observed
for compounds 1 and 2, suggesting these compounds are
inhibitors but not poisons.
(HBTU), and diisopropylethylamine (DIPEA).19 The
amide product was isolated in 68% yield and subjected
to thermal cyclodehydration to give the benzoxazole 6 in
89% yield.20 The methylester was then hydrolyzed to the
carboxylic acid in refluxing 5 M NaOH/THF to give 7 in
essentiallyquantitative yield.
Using the previous coupling conditions (HOBT,
HBTU), acid 7 was reacted with either 5 or 2-amino-
phenol to give 8 and 9, respectively. Subsequent ring
closure and benzyl deprotection were achieved in a
single step byrefluxing the amides with p-toluenesulfonic
acid in toluene. The products, UK-1 and 1, were isolated
in 40% and 41% yield, respectively.
The gel data were used to compare the effectiveness of
each compound, using the ratio of fluorescence intensi-
ties for supercoiled DNA and relaxed circular DNA in
each lane. Figure 3B shows the graphical presentations
of the amount of supercoiled DNA as a function of drug
concentration for all the compounds. From these
graphs, IC50 values for UK-1, 1, and 2 were determined
to be 32, 20, and 38 lM, respectively.26 Notably, com-
pound 3 showed no measurable activity. Based on these
results, the order of activityis 1 > UK-1 > 2 ꢀ 3.27
For compounds 2 and 3, benzoyl chloride was reacted
with 5 in refluxing m-xylene to provide benzoxazole 10
in 73% yield in a single step (Scheme 2).21 Hydrolysis of
10 in refluxing 5 M NaOH/THF gave corresponding
acid 11 in 96% yield. The acid was activated using oxalyl
chloride and coupled with either 5 or 2-aminophenol to
yield 12 and 13, respectively. Refluxing the resulting
amides with p-toluenesulfonic acid in toluene afforded
the ring-closed deprotected products 2 and 3 in 42% and
77% yield, respectively.
The results from the topo II assayas shown in Figure
3A suggest UK-1 is a catalytic inhibitor of the enzyme
but not a topo II poison.28 Unfortunately, the assay
does not distinguish the remaining possible mechanisms
of inhibition. As previouslydescribed, compounds can
interfere with either DNA binding or enzyme dissocia-
tion, either of which would provide the observed results.
The fact that UK-1 has been shown to bind DNA would
tend to suggest that the former mechanism of topo II
inhibition is operative. It is, however, impossible to rule
out the possibilitythat UK-1 is capable of deactivating
the enzyme during catalysis without stabilizing the
cleavable complex.
The activityof UK-1 and analogs 1–3 against human
topo IIa were evaluated bydetermining their abilities to
interfere with the relaxation of supercoiled plasmid
DNA.22 The assaymonitors the DNA products after
30 min incubation of supercoiled DNA with topo II in
the absence and the presence of each compound.23 The
DNA products are separated on an agarose gel and
visualized byfluorescence. The results from these studies
are shown in Figure 3. A sample agarose gel from the
assayis shown in Figure 3A. This assayis capable of
determining if the compounds are topo II poisons.24 As
described earlier, poisons stabilize the cleavable complex
and prolong the covalent association between the en-
zyme and DNA. When this stabilized complex is treated
with SDS followed bydigestion with proteinase K, the
UK-1 has shown cytotoxicity at submicromolar con-
centrations against two cancer cell lines (HL60 and PC-
3, with IC50 values of 0.32 and 0.4 lM, respectively).17
These IC50 values for live cell studies are much lower
than those observed for topo II inhibition, reported
COCl
CO2Me
NH2
CO2Me
CO2H
N
N
b
a
+
O
O
OH
5
10
11
73%
96%
R
R
O
OH
O
N
NH
N
N
c, d
e
O
O
2: R = CO2Me
3: R = H
42%
77%
12: R = CO2Me
13: R = H
37%
40%
Scheme 2. Reagents and conditions: (a) TEA, pyridinium p-toluene sulfonate, m-xylene, reflux; (b) 5 M NaOH, THF, reflux; (c) (ClCO)2, DMF,
CH2Cl2; (d) 5 or 2-aminophenol, pyridine, CH2Cl2; (e) TsOH, toluene, reflux.

3224
B. B. Wang et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3221–3226
Figure 3. (A) An example of an agarose gel, showing concentration dependent inhibition of topo II-mediated relaxation of supercoiled pBR322 by
UK-1. Abbreviations: Supercoiled (SC), nicked circular (NC), linear (L), and relaxed circular (RC) DNA. Controls were run in the first four lanes.
Lane 1 contained plasmid DNA only, Lane 2 contained plasmid DNA with 1% DMSO, 1% SDS, and proteinase K, lane 3 contained linearized
DNA, lane 4 contained plasmid DNA and topo II, and lanes 5–11 contained 1% DMSO with or without UK-1 (the final concentration of UK-1 is
shown on the top of each lane). (B) The quantified data for the % supercoiled DNA present in each lane. Data from three separate assays were
23
plotted and IC50 values were obtained byfitting the pooled data.
OMe
O
OMe
O
a
a
b
H
O
H
O
O
N
a
H
O
O
N
a
b
O
N
a
MeO
O
a
b
b
a
a
a
N
N
N
O
N
O
O
O
UK-1
14
1
2
ꢀ
Figure 4. A unique structural motif for UK-1 and analogs based on molecular modeling. Distances indicated by a varied between 2.875 and 3.045 A
30
ꢀ
and those indicated by b varied between 4.524 and 4.727 A for all the structures shown.
here. The reason for the apparent discrepancyis still
unclear. It is possible that topo II might not be the
primarycellular target of UK-1. Alternativel,y it is
possible that UK-1 interferes with multiple cellular
processes and cytotoxicity is enhanced by multiple
mechanisms. It has been observed with topoisomerase I
inhibitors such as Hoechst dyes that the mode of inhi-
bition changes between low drug concentration
(<1.0 lM) and high drug concentration (>1.0 lM).29 At
low concentration, the ligands exhibit topo I poisoning
activity, inducing DNA cleavage, while at high con-
centration, the ligands act as catalytic supressors by
preventing association of enzyme with DNA. Such a
change in mode of inhibition could significantlyalter the
apparent cytotoxicity of each ligand. Although no evi-
dence for such behavior has been presented for topo II
inhibitors, work is currentlyunderwayto explore such a
possibilityfor UK-1 and analogs.
Our data show that the removal of either the methyl
ester or ortho-hydroxyl group does not alter the anti-
topo II activitysignificantly. However, the activityis

B. B. Wang et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3221–3226
3225
6. Baird, C. L.; Harkins, T. T.; Morris, S. K.; Lindsley, J. E.
Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 13685–13690.
7. Garcia-Carbonero, R.; Supko, J. G. Clin. Cancer Res.
2002, 8, 641–661.
8. Hande, K. R. Euro. J. Cancer 1998, 34, 1514–1521.
9. Chen, A. Y.; Liu, L. F. Annu. Rev. Pharmacol. Toxicol.
1994, 34, 191–218.
10. Vilain, N.; Tsai-Pflugfelder, M.; Benoit, A.; Gasser, S. M.;
Leroy, D. Nucl. Acids Res. 2003, 31, 5722–5741.
11. Bell, A.; Kittler, L.; Lober, G.; Zimmer, C. J. Mol.
Recogn. 1997, 10, 245–255.
12. Classen, S.; Olland, S.; Berger, J. M. Proc. Natl. Acad. Sci.
U.S.A. 2003, 100, 10629–10634.
13. Reynolds, M. B.; DeLuca, M. R.; Kerwin, S. M. Bioorg.
Chem. 1999, 27, 326–337.
14. Ueki, M.; Ueno, K.; Miyadoh, S.; Abe, K.; Shibata, K.;
Taniguichi, M.; Oi, S. J. Antibiot. 1993, 46, 1089–1094.
15. Shibata, K.; Kashiwada, M.; Ueki, M.; Taniguichi, M.
J. Antibiot. 1993, 46, 1095–1100.
dramaticallydecreased when both functional groups are
deleted (even at 80 lM, no inhibition is observed for 3).
It is notable that UK-1 and analogs 1 and 2 all possess a
triangular arrangement of Lewis-basic sites comprised
of heterocyclic nitrogen atoms, an ester oxygen atom,
and/or a phenolic hydroxyl group (Fig. 4). Truncated
UK-1 analog 14, reported byKumar et al., possesses a
similar arrangement of basic sites and shows cytotoxic-
ityat similar concentrations to that of the parent com-
pound (although anti-topo II activitywas apparently
not evaluated).17 Molecular mechanics-minimized
structures for all these compounds point to a verywell-
defined arrangement of the basic sites, which effectively
ꢀ
defines an isosceles triangle that is 2.875–3.045 A on two
ꢀ
sides and 4.524–4.727 A on the remaining side. Our data
argue that this unique structural motif is necessaryfor
enzyme inhibition.
16. Reyzer, M. L.; Brodbelt, J. S.; Kerwin, S. M.; Kumar, D.
Nucl. Acids Res. 2001, 29, e103.
17. Kumar, D.; Jacob, M. R.; Reynold, M. B.; Kerwin, S. M.
Bioorg. Med. Chem. 2002, 10, 3994–4004.
18. DeLuca, M. R.; Kerwin, S. M. Tetrahedron Lett. 1997, 38,
199–202.
19. Hudson, D. J. Org. Chem. 1988, 53, 617–624.
20. Dunwell, D. W.; Evans, D.; Hicks, T. A. J. Med. Chem.
1975, 18, 53–58.
21. Frakas, L.; Vermes, B.; Nogradi, M. Tetrahedron 1967, 23,
741–744.
22. Bjergbaek, L.; Kingma, P.; Nielsen, I. S.; Wang, Y.;
Westergaard, O.; Osheroff, N.; Andersen, A. H. J. Biol.
Chem. 2000, 275, 13041–13048.
23. Human topoisomerase II alpha was purchased from USB
Corporation. Supercoiled pBR322 plasmid DNA was
purchased from New England Biolabs. Reaction buffers
were freshlyprepared according to the manufacturer
recommended concentrations. All drugs were diluted in
dimethyl sulfoxide to give 10 mM stock solutions. Each
solution was further diluted with DMSO to give the
desired final concentration, and each reaction contained
1% DMSO. The assaywas carried out byincubating
2 units (0.045 lg) of human topo IIa, 0.2 lL of appropriate
ligand, and 0.111 lg of supercoiled pBR322 DNA in a
total volume of 20 lL assaybuffer (10 mM Tris–HCl,
pH 7.9, 50 mM NaCl, 50 mM KCl, 5 mM MgCl2, 1 mM
EDTA, 15 lg/mL BSA, 1 mM ATP). The sequence of
addition was water, reaction buffer, DNA, ligand, and
then enzyme. Reactions were incubated at 37 °C for 30 min
and stopped bythe addition of 2.2 lL 10% SDS and 1 lL
of 1.4 mg/mL Proteinase K solution. The samples were
incubated at 37 °C for an additional hour before subject-
ing to electrophoresis in 1.5% agarose gel in TAE (40 mM
Tris–acetate, pH 8.3, 2 mM EDTA) containing 0.5 lg/mL
ethidium bromide. Gels were run at 40 V for 6 h and
destained for at least 1 h before visualizing on a Typhoon
9200 (Amersham Biosciences). The bands were quantified
using ImageQuant Version 5.2 (Molecular Dynamics).
The data were plotted and analyzed with Origin version
6.0 (Mircrocal) imbedded DoseResp Pharmacologycurve
fitting.
We would like to suggest that the role of these groups
involves metal ion chelation, which has been shown to
be important for DNA binding and correlated to cyto-
toxicity.17 For magnesium, this type of chelation is ob-
served in chlorophylls, where Mg2þ is coordinated to the
four nitrogens in the porphine core structure.31 It is
possible that UK-1 and the active analogs described
here form similar square planar complexes, with a donor
site from DNA occupying the fourth coordination site.
It is important to point out that the anti-topoisomerase
activityof UK-1 could involve direct interaction of the
drug with the enzyme. By a similar scenario, then, it is
possible that binding with enzyme-associated magne-
sium could also lead to inhibition (with functionality
from the enzyme occupying a coordination site).
This work demonstrates that UK-1 does not inhibit
human topoisomerase II bystabilizing the cleavable
complex and identifies a minimal structural motif nec-
essaryfor activity. Magnesium binding experiments and
cytotoxicity studies, currently in progress, should
establish a more clear relationship between magnesium
binding, topo II inhibition, and cytotoxicity.
Acknowledgements
This investigation was supported, in part, byUMBC
and the National Science Foundation’s Research
Experiences for Undergraduates (REU) Sites Program
Award DBI-0139619.
References and notes
1. Champoux, J. J. Annu. Rev. Biochem. 2001, 70, 369–413;
Wang, J. C. Annu. Rev. Biochem. 1996, 65, 635–692.
2. Wang, J. C. J. Mol. Biol. 1971, 55, 523–533.
3. Champoux, J. J.; Dulbecco, R. Proc. Natl. Acad. Sci.
U.S.A. 1972, 69, 143–146.
4. Berger, J. M.; Gamblin, S. J.; Harrison, S. C.; Wang, J. C.
Nature 1996, 379, 225–232.
5. Roca, J.; Berger, J. M.; Harrison, S. C.; Wang, J. C. Proc.
Natl. Acad. Sci. U.S.A. 1996, 93, 4057–4062.
24. The assaywas first carried out with known topo II poison
etopside (VP-16) to evaluate the efficiencyof the assay.
Under the same assayconditions, a 15% increase in linear
DNA was observed at 100 lM drug concentration.
25. The linear DNA standard was obtained bytreating
supercoiled pBR322 DNA with EcoRI (Sigma).
26. IC50 values were obtained the byfitting the data with the
Dose Response function in Microcal Origin 6.0.

3226
B. B. Wang et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3221–3226
Fluorescence intensityof supercoiled DNA
supercoiled DNA range from 5% to 40%. The reason
behind this phenomenon is still unclear. It is also unclear
whycomplete inhibition is not observed at high ligand
concentration.
Y ¼
Fluorescence intensityof supercoiled DNA þ relaxed DNA
A2 þ A1
Y ¼ A1 þ
1 þ 10ðlog X0ÀXÞp
29. Chen, A. Y.; Yu, C.; Gatto, B.; Liu, L. F. Proc. Natl.
Acad. Sci. U.S.A. 1993, 90, 8131–8135.
Y ¼ fraction of supercoiled DNA, X ¼ ligand concentration,
A1 ¼ bottom asymptote, A2 ¼ top asymptote.
30. These data are based on energy-minimized structures
using MacroModel 7.1 and the MMFF94S forcefield
parameters. Mohamadi, F.; Richard, N. G. J.; Guida, W.
C.; Liskamp, R.; Lipton, M.; Caufield, C.; Chang, G.;
Hendrickson, T.; Still, W. C. MacroModel––an Integrated
Software System for Modeling Organic and Bioorganic
Molecules Using Molecular Mechanics. J. Comput. Chem.
1990, 11, 440.
27. The slopes of these plots at low ligand concentrations
suggest a similar order of efficacy; the values are 0.043,
0.023, 0.010 for 1, UK-1, and 2, respectively. We do not
believe that the lack of activityof three results from poor
solubilityof this compound under the assayconditions
based on the fact that all reaction mixtures were visually
homogeneous.
28. It should be noted that at concentration higher than
80 lM, a decrease in the intensityof supercoiled DNA was
often observed. The drops in the relative concentration of
31. Cotton, F. A.; Wilkinson, G. Advanced Inorganic Chem-
istry, 5th ed.; John Wiley& Sons: New York, NY, 1988;
pp 156–157.