Synthesis and evaluation of anticancer benzoxazoles and benzimidazoles related to UK-1

Article abstract of DOI:10.1016/S0968-0896(02)00327-9

UK-1 is a structurally unique bis(benzoxazole) natural product isolated from a strain of Streptomyces. UK-1 has been reported to possess anticancer activity but no activity against bacteria, yeast, or fungi. Previous work has also demonstrated the ability

Full text of DOI:10.1016/S0968-0896(02)00327-9

Bioorganic & Medicinal Chemistry 10 (2002) 3997–4004
Synthesis and Evaluation of Anticancer Benzoxazoles and
Benzimidazoles Related to UK-1
Devinder Kumar,a Melissa R. Jacob,b Michael B. Reynoldsa and Sean M. Kerwina,*
aDivision of Medicinal Chemistry andInstitute of Cellular andMolecular Biology, College of Pharmacy, The University of Texas atAustin,
Austin, TX 78712, USA
bNational Center for Natural Products Research, School of Pharmacy, Thad Cochran Research Center, The University of Mississippi,
University, MS 38677, USA
Received 28 May 2002; accepted 5 July 2002
Abstract—UK-1 is a structurally unique bis(benzoxazole) natural product isolated from a strain of Streptomyces. UK-1 has been
reported to possess anticancer activity but no activity against bacteria, yeast, or fungi. Previous work has also demonstrated the
ability of UK-1 to bind a variety of di- and tri-valent metal ions, particularly Mg2+ ions, and to form complexes with double-
stranded DNA in the presence of Mg2+ ions. Here we report the activity of UK-1 against a wide range of human cancer cell lines.
UK-1 displays a wide spectrum of potent anticancer activity against leukemia, lymphoma, and certain solid tumor-derived cell lines,
with IC50 values as low as 20 nM, but is inactive against Staphylococcus aureus, a methicillin-resistant strain of S. aureus, or Pseu-
domonas aeruginosa. A series of analogues of the bis(benzoxazole) natural product UK-1 in which the carbomethoxy-substituted
benzoxazole ring of the natural product was modified were prepared and evaluated for their anticancer and antibacterial properties.
An analogue of UK-1 in which the carbomethoxy-substituted benzoxazole ring was replaced with a carbomethoxy-substituted
benzimidazole ring was inactive against human cancer cell lines and the two strains of S. aureus. In contrast, a simplified analogue
in which the carbomethoxy-substituted benzoxazole ring was replaced with a carbomethoxy group was almost as active as UK-1
against the four cancer cell lines examined but lacked activity against S. aureus. Metal ion binding studies of these analogues
demonstrate that they both bind Zn2+ and Ca2+ ions about as well as UK-1. The non-cytotoxic benzimidazole UK-1 analogue
binds Mg2+ ions 50-fold weaker than UK-1, whereas the simple benzoxazole analogue binds Mg2+ ions nearly as well as UK-1.
These results support a role of Mg2+ ion binding in the selective cytotoxicity of UK-1 and provide a minimal pharmacophore for
the selective cytotoxic activity of the natural product.
# 2002 Elsevier Science Ltd. All rights reserved.
Introduction
positive or Gram-negative bacteria, yeast, or fungi at
concentrations as high as 250 mM.1 However, the semi-
synthetic derivatives methyl UK-1 (MUK-1) and deme-
thyl UK-1 (DMUK-1) both have activity against Gram-
positive and Gram-negative bacteria.3,4 MUK-1 is also
active against yeast and filamentous fungi.3
The bis(benzoxazole) natural products are a structurally
unique class of Streptomyces secondary metabolites that
have recently been reported in the literature.1ꢀ3 In the
course of a screening program for new bioactive com-
pounds, Taniguchi and co-workers isolated UK-1 from
the acetone extracts of Streptomyces sp. 517–02.1 Sub-
sequently, Tsuji and co-workers isolated AJI9561 from
Streptomyces sp. AJ956.2 Both UK-1 and AJI9561 were
reported to possess growth inhibitory activity against
the murine cancer cell line P388, with IC50 values in the
0.3–1.6 mM range.1,2 Despite its cancer cell cytotoxic
properties, UK-1 does not inhibit the growth of Gram-
The selective cytotoxicity of UK-1 towards cancer cells
versus bacteria and fungi indicates that UK-1 may have
*Corresponding author. Tel.: +1-512-471-5074; fax: +1-512-232-
2696; e-mail: skerwin@mail.utexas.edu
0968-0896/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00327-9

3998
D. Kumar et al. / Bioorg. Med. Chem. 10 (2002) 3997–4004
a unique mechanism of anticancer action. A recently
reported synthetic route for the preparation of UK-1
has aided studies related to the mechanism of action of
UK-1.5 The 2-(20-hydroxyphenyl)benzoxazole moiety
present in UK-1 is also present in a number of synthetic
metal ion chelators6,7 and is analogous to the 2-(20-
hydroxyphenyl)oxazole moiety present in a class of
microbial siderophores.8 The metal ion binding ability
of synthetically prepared UK-1 has been investigated.9
These studies indicate that UK-1 is capable of binding a
variety of biologically important metal ions, particularly
Mg2+ ions. Like the Mg2+-binding aureolic acid group
of antitumor antibiotics10,11 and synthetic antitumor
quinobenzoxazines,12,13 UK-1 binds DNA in a metal
ion-dependent fashion. One consequence of this inter-
action with DNA is the inhibition of topoisomerase II.9
A more recent investigation of the metal-mediated
DNA binding of UK-1 by ESI-MS demonstrated that
UK-1 forms complexes of the type [ds+UK-1+M2+
with a variety of metal ions including Ni2+, Co2+, and
Zn2+ 14 These results distinguish UK-10s metal medi-
]
.
Scheme 1. Synthesis of an aza-analogue of UK-1.
ated DNA binding from that observed for the aureolic
acids, which bind as [ds+2ꢁligand+M2+
] com-
plexes,14,15 and the quinobenzoxazines, which bind as
give the aza-analogue of UK-1 6 in 35% overall yield
from 2-benzyloxybenzoic acid.
[ds+2ꢁligand+2ꢁM2+] complexes.12,13
In order to understand the structural basis for the
selective cytotoxicity of UK-1, a number of structural
analogues of this natural product have been prepared. A
comparison of the anticancer activity of these com-
pounds with their ability to inhibit the growth of
methicillin-sensitive and methicillin-resistant Staphylo-
coccus aureus demonstrates that a structurally simplified
analogue of UK-1 retains the natural product’s selective
activity against cancer cells. It was also found that
structurally conservative changes to UK-1 that diminish
Mg2+-binding ability result in a dramatic decrease in
cancer cell cytotoxicity. Taken together, these results
establish both a minimum structural pharmacophore as
well as a functional role for Mg2+-binding in the selec-
tive cytotoxicity of UK-1.
The simple benzimidazole analogue 9 was prepared as
shown in Scheme 2. Condensation of 2-benzyloxy-
benzoic acid with methyl 2,3-diaminobenzoate afforded
the amide 7, which was cyclized by heating in acetic acid
to give the benzimidazole ester 8. Hydrolysis of 8 affor-
ded benzimidazole carboxylate 9 in 49% overall yield.
The substituted benzoxazole 10 was prepared directly
from the amide 2 (Scheme 1) by p-TsOH-mediated
cyclization and deprotection, as shown in eq 1 (below).
When this same p-TsOH-mediated cyclization/depro-
tection reaction was performed on a large scale in the
synthesis of UK-1 from the precursor amide 11 (Scheme
3), a low yield of UK-1 was obtained, along with a side
product that was difficult to separate from the crude pro-
duct. Careful chromatography enabled the separations of
Results and Discussion
A number of analogues of UK-1 were prepared in
which the carbomethoxy-substituted benzoxazole ring
was either modified or deleted. The synthesis of these
UK-1 analogues was carried out in analogy with the
previously published total synthesis of the natural pro-
duct,5 with some improvements. An analogue of UK-1
in which the carbomethoxy-substituted benzoxazole
ring is replaced with a substituted benzimidazole ring
was prepared as shown in Scheme 1. Carbonyl diimida-
zole-coupling of 2-(benzyloxy)benzoic acid with methyl
anthranilic acid affords the amide 2 in 87% yield.
Pyrolysis of amide 2 at 230 ꢂC under vacuum affords
the benzoxazole 3 in 91% yield. Hydrolysis of the
ester functionality proceeded to give the benzoxazole
acid 4, which was coupled with methyl 2,3-diamino-
benzoate to afford the amide 5. Final cyclodehydra-
tion and deprotection were carried out by heating 5
in the presence of p-toluenesulfonic acid (p-TsOH) to
Scheme 2. Synthesis of a simplified benzimidazole UK-1 analogue.

D. Kumar et al. / Bioorg. Med. Chem. 10 (2002) 3997–4004
3999
The cancer cell cytotoxicity and antibacterial activity of
the UK-1 analogues were examined in order to deter-
mine the structural basis for the unique cancer cell
selective cytotoxicity of the natural product. The pre-
viously reported benzoxazole 135 and commercially
available 2-(20-hydroxyphenyl)benzoxazole (14) were
also included in this study. Table 2 summarizes the
results of cytotoxicity testing against MCF-7, HT-29,
HL60, and PC-3 cell lines along with the antibacterial
assays against S. aureus and a methicillin-resistant
strain of S. aureus. UK-1 shows a wide range of cyto-
toxic activity against the four cancer cells lines, due to
the relative resistance of the HT-29 cells to the action of
UK-1. The benzylated UK-1 analogue 12 demonstrates
a diminished cytotoxic effect against the cancer cell lines
relative to UK-1; however, the spectrum of activity of
this analogue is quite distinct. Compound 12 is more
active than UK-1 against HT-29 cells (IC50 5.2 mM vs 65
mM for UK-1), about 10-fold less potent than UK-1
against MCF-7 and PC-3 cells (IC50 13 and 4.5 mM vs
1.6 mM and 0.4 mM for UK-1, respectively), and 20-fold
less active than UK-1 against HL60 cells (IC50=27 mM
vs 0.32 mM for UK-1). Unfortunately, the limited
availability of compound 12 prevented the determina-
tion of its antibacterial properties. It remains to be seen
if substitution on the phenolic ring of UK-1 affects
cancer cell versus bacterial cell selectivity of UK-1.
Scheme 3. Formation of a benzylated analogue of UK-1.
a pure sample of the side product, which was identified
1
by H and 13C NMR, NOESY, COSY, and HRMS as
the previously unreported benzylated UK-1 analogue
12. Compound 12 apparently arises from interception of
a benzyl cation formed by acid-mediated cleavage of the
benzyl ester. In previous preparations of UK-1, this
cation was presumably intercepted by adventitious
water in the reaction mixture, but under more anhy-
drous conditions and larger scale, the cation is appar-
ently able to participate in Friedel–Crafts alkylation
chemistry, leading to the benzyl adduct 12. The forma-
tion of this side product can be avoided; carrying out
the transformation of 11 to UK-1 in the presence of 2–4
equivalents of water prevented the formation of 12, and
allowed for the isolation of UK-1 in high yields, as
reported earlier.5
The cytotoxicity of UK-1 against P388, B16, and HeLa
cell lines has been previously reported.1 In order to
determine the spectrum of cell lines whose growth is
inhibited by UK-1, a range of human cancer cell lines
were investigated. The results, shown in Table 1,
demonstrate that UK-1 displays a wide spectrum of
potent anticancer activity against leukemia, lymphoma,
and certain solid tumor-derived cell lines. In particular,
certain neuroblastoma cell lines (IMR-32 and NGP)
were very sensitive to UK-1, with IC50 values <100 nM.
Other cell lines, such as colon (HT-29) were much less
sensitive. Despite the potent anticancer activity of UK-
1, there was no indication of any antibacterial effect
against S. aureus, methicillin-resistant S. aureus, or
Pseudomonas aeruginosa at concentrations upto 50 mg/
mL of UK-1, in accord to previous reports of selective
cancer cell cytotoxicity of this natural product.1
Interestingly, the aza-analogue of UK-1, compound 6
lacks any anticancer activity against three of the four
cell lines examined, with IC50 values >100 mM against
MCF-7, HT-29, and PC-3 cells (Table 2). Compound 6
did demonstrate very modest activity against HL60 cells
(IC50=70 mM). This analogue lacks any antibacterial
activity against the two strains of S. aureus examined.
The majority of the other analogues of UK-1 examined
also lacked appreciable anticancer activity, although
two analogues, benzoxazole acid 4 and benzimidazole
acid 9, displayed some antibacterial activity against the
S. aureus strains. The activity of these acids against S.
aureus, although modest, is interesting in light of the
reported antibacterial activity of DMUK-1 but not the
methyl ester UK-1 (see above); compound 3, the methyl
ester of 4, also lacks antibacterial activity. The 2-(20-
hydroxyphenyl)benzoxazole (14) is neither cytotoxic to
cancer cells nor antibacterial (Table 2), in accord with
previous reports of the lack of antibacterial activity
(>100 mM) for this compound.16
Table 1. Cytotoxicity of UK-1 against cancer cell lines
a
Cell line
IC50 (mM)
MCF-7
HT-29
HL60
PC-3
MDA-231
BT-20
DU145
SKBR3
A549
SK-N-AS
SK-N-D7
SK-N-F1
SK-N-MC
SK-N-SH
CHP-212
IMR-32
NGP
1.6
65
0.32
0.4
0.5
0.17
0.2
0.1
1.9
24
0.17
7
1.1
7
Within the series of truncated UK-1 analogues exam-
ined, only one analogue, compound 10, demonstrated
cancer cell cytotoxicity (Table 2). Compound 10 also
displayed the selective activity against cancer cells ver-
sus bacterial cells that is characteristic of the natural
product. The anticancer activity of compound 10 is
somewhat less than that of UK-1 against HL60 and PC-
3 cells (IC50=5.7 mM vs 0.4 mM for UK-1, and 0.88 mM
vs 0.32 mM for UK-1, respectively), but equal to or
12
0.06
0.02
aDetermined using the alamarBlue assay. See Experimental for details.

4000
D. Kumar et al. / Bioorg. Med. Chem. 10 (2002) 3997–4004
Table 2. Antibacterial and Anticancer Activity of UK-1 and Analo-
gues
Mg2+ binding of ability of these compounds. The 4-
unsubstituted 2-(20-hydroxyphenyl)-benzoxazole (14)
does not complex Mg2+ ions (data not shown),
a
Compd
Antibacterial IC50
Cytotoxicity
IC50 rangeb (mM)
although 14 does form complexes with Zn2+ 17,18
.
Introduction of 4-carbomethoxy substituent, as in 10, or
a 4-(benzoxazo-2-yl) substituent (as in UK-1) on to the
2-(20-hydroxyphenyl)benzoxazole core imparts Mg2+
ion binding ability, but the 4-(benzimidazo-2-yl) sub-
stituent (as in 6) does not. Taken together, these results
support a role for Mg2+ ion binding in the cancer cell
cytotoxicity of these compounds, in that the cytotoxic
UK-1 and analogue 10 bind Mg2+ ions better than the
non-cytotoxic analogues 6 and 14. Moreover, the selec-
tive activity of compound 10 against cancer cells but not
bacteria reflects the selective activity of UK-1. Com-
pound 10 thus represents a minimum pharmacophore
for the selective anticancer activity of UK-1.
S. aureus
(mM)
Meth-resist
S. aureus
UK-1
12
6
3
4
9
10
13
14
NAc
ndd
NA
NA
43
102
NA
NA
NA
0.45
nd
NAc
ndd
NA
NA
29
102
NA
NA
NA
0.45
nd
0.32–65
4.5–27
7.0–>100
>100
>100
>100
0.88–9.1
>100
>100
nd
0.08–0.27
Ciprofloxacin
Mitomycin C
aSample concentration that affords 50% growth of the test organism.
bRange of IC50 values obtained by alamarBlue cytotoxicity assays
against MCF-7, HL-60, HT-29, and PC-3 cells after 72 h at 37 ꢂC.
cNo activity observed at concentrations upto 50 mg/mL.
dNot determined.
Conclusion
The relatively simple yet novel bis(benzoxazole) natural
products UK-11 and AJI95612 represent a new class of
cytotoxic secondary metabolites. Interest in this class of
bis(benzoxazoles) is fueled by the finding that UK-1 is
selectively cytotoxic to cancer cells but not bacterial
cells, yeast, or fungi.1 This selective cytotoxicity may be
mediated through the specific interaction of UK-1 with
an as yet undefined cancer cell-associated target. While
the putative target of UK-1 remains unknown, previous
studies have demonstrated that this natural product
forms complexes with a variety of physiologically rele-
vant metal ions, particularly magnesium ions.9 Further-
more, the metal ion binding of UK-1 increases the
affinity of this compound for DNA in a manner reminis-
cent of the aureolic acid class of antitumor anti-
biotics.9,14 Here it is shown that within a series of
analogues, the structural basis for Mg2+ ion binding and
selective cytotoxicity of UK-1 are the same; both Mg2+
ion binding and selective cytotoxicity require an appro-
priately 4-substituted 2-(20-hydrophenyl)benzoxazole
moiety. The elucidation of the structural basis for Mg2+
ion recognition by these 4-substituted 2-(20-hydroxy-
phenyl)benzoxazoles requires further investigation.
While it is unlikely that Mg2+ ion binding per se is the
origin of the selective cytotoxicity of UK-1 and analo-
gue 10, the ability of UK-1 to form metal ion complexes
that can bind to DNA9,14 and inhibit DNA-processing
enzymes9 indicates that Mg2+ ion binding by UK-1
may lead to biologically relevant complexes with a spe-
cific target in cancer cells. Further work with UK-1 and
analogues is required in order to determine if such tar-
geting is involved in the selective cytotoxicity of UK-1
and if the promising spectrum of in vitro anticancer
activity of UK-1 reported here is also reflected in vivo.
better than UK-1 against MCF-7 and HT-29 cells
(IC50=1.5 mM vs 1.6 mM for UK-1, and 9.1 mM vs 65
mM for UK-1, respectively). Compound 10 was not
active against S. aureus (Table 2) or against P. aerugi-
nosa, Cryptococcus neoformans, or Candida albicans
(IC50 >50 mg/mL, data not shown).
Based on previous work that implicated metal ion
binding in the mechanism of action of UK-1, the ability
of selected UK-1 analogues to bind a variety of metal
ions was determined by the method of continuous var-
iation (Table 3). Methanolic solutions of UK-1 or ana-
logues and metal ion salts [Ca(NO3)2, Mg(NO3)2,
Zn(NO3)2, or Fe(NO3)3] at various molar ratios and 20
mM combined total concentration were prepared and
the absorbance at 418 nm was determined. Plots of
A418nm versus mole ratio of ligand demonstrated max-
ima at 0.5 mole ratio for all metal ions, indicating 1:1
stoichiometry for each of the metal ion complexes. The
aza-analogue 6, which lacks cancer cell cytotoxicity,
binds Mg2+ and Fe3+ ions much less well than UK-1,
while the Ca2+ and Zn2+ metal ion binding ability of
this analogue is slightly higher than that of UK-1. The
cancer cell cytotoxic analogue 10 retains much of the
Zn2+ metal ion binding ability of the natural product,
while the ability of 10 to bind Mg2+, Ca2+, and Fc2+
ion is slightly diminished relative to UK-1.
An appropriate 4-substituent on the 2-(20-hydroxy-
phenyl)benzoxazole core is important for the efficient
Table 3. Association constants for metal ion binding by UK-1 and
analogues
Compd
Metal ion
UK-1
6
10
Experimental
General
Mg2+
Ca2+
Zn2+
Fe3+
2.0ꢁ106 Mꢀ1
4.0ꢁ104 Mꢀ1
2.0ꢁ106 Mꢀ1
2.5ꢁ105 Mꢀ1
4.0ꢁ104 Mꢀ1
1.0ꢁ105 Mꢀ1
3.2ꢁ106 Mꢀ1
1.0ꢁ104 Mꢀ1
3.2ꢁ105 Mꢀ1
6.3ꢁ103 Mꢀ1
1.3ꢁ106 Mꢀ1
7.9ꢁ103 Mꢀ1
All reagents and solvents were purchased from Aldrich
Chemical Company and used without further purification,

D. Kumar et al. / Bioorg. Med. Chem. 10 (2002) 3997–4004
4001
unless noted. THF was distilled from sodium/benzo-
phenone immediately prior to use. CH2Cl2 and DMF
were distilled from CaH2 immediately prior to use. Tol-
uene was distilled from sodium metal. UV–Vis spectra
for the continuous variation plots were determined on a
Varian Cary 3E spectrophotometer. Melting points
(open capillary) are uncorrected. Unless otherwise
2-[2-(Benzyloxy)phenyl]-1,3-benzoxazole-4-carboxylic acid
(4). Compound 3 (6.0 g, 1.67 mol) was dissolved in 80
mL of THF and 40 mL of 5 M NaOH was added. The
reaction mixture was stirred at 60 ꢂC with stirring for 2
h. The reaction mixture was cooled to room tempera-
ture, diluted with EtOAc and acidified with con-
centrated HCl. The organic layer was washed with brine
two times and dried over anhydrous Na2S04. The sol-
vent was evaporated and the residue was dried under
vacuum to afford a white solid that was recrystallized
from EtOAc and hexanes (5.2 g, 90%): mp103–104 ꢂC;
1
noted, H and 13C NMR spectra were determined in
CDCl3 on a spectrometer operating at 300 and
75.5 MHz, respectively. All mass spectra were obtained
by chemical ionization using methane as the ionizing
gas. Chromatography refers to flash chromatography
on silica gel, and Rf values were determined using silica
gel-GF TLC plates (Merck) using the solvent system
indicated.
1
Rf 0.15 (40% EtOAc in hexanes); H NMR d 5.28 (s,
3H), 7.12–7.20 (m, 2H), 7.37–7.49 (m, 4H), 7.52–7.60
(m, 3H), 7.77 (dd, 1H, J=7.9, 0.9 Hz), 8.14 (dd, 1H,
J=7.6, 1.2 Hz), 8.21 (dd, 1H, J=7.8, 1.5 Hz); 13C
NMR d 70.98, 113.76, 114.61, 115.28, 120.07, 121.27,
125.48, 127.08, 127.31, 128.39, 128.90, 131.72, 134.44,
136.09, 141.13, 149.90, 158.27, 163.16, 165.00; CIMS m/
z 346 (MH+); HRMS m/z calcd for C21H16NO4:
346.1079, found 346.1078.
Methyl
2-{[2-(benzyloxy)benzoyl]amino}-3-hydroxy-
benzoate (2). Carbonyldiimidazole (CDI) (5.10 g, 31.5
mmol) was dissolved in 50 mL dry THF with stirring at
room temperature under an argon atmosphere, and 2-
(benzyloxy)benzoic acid (7.18 g, 31.5 mmol)19 was then
added carefully. After the evolution of CO2 ceased (ca. 5
min) the reaction mixture was stirred for an additional
10 min and then methyl 2-amino-3-hydroxybenzoate
(3.5 g, 21 mmol)20 was added. After stirring for an
addition 10 min at room temperature, the reaction mix-
ture was heated under reflux for 18 h. The reaction
mixture, brown in color, was concentrated and dis-
solved in a minimum volume of EtOAc. Silica gel (60–
100 mesh) was added to make the slurry and solvent
was evaporated to dryness. Column chromatography
was performed using 20% EtOAc in hexane to give a
Methyl 2-amino-3-({[2-(2-benzyloxyphenyl)-1,3-benzoxa-
zol-4-yl]carbonyl}aminobenzoate (5). To a solution of
compound 4 (1.035 g, 3 mmol) in 40 mL of dry CH2Cl2
was added 10 mL of freshly distilled oxalyl chloride.
After stirring for 10 min, five drops of DMF was added
to the reaction mixture. The reaction mixture was stir-
red for 2 h at room temperature and the solvent was
removed by evaporation. The resultant yellow solid was
dried under vacuum for 3 h. The yellow solid was then
dissolved in 15 mL of dry CH2Cl2 and transferred via
canula to a solution of methyl 2,3-diaminobenzoate
(498 mg, 3 mmol)21 dissolved in 20 mL of dry CH2Cl2
and 2 mL of pyridine. The reaction mixture was stirred
at room temperature overnight and then diluted with
CH2Cl2. The solution was washed with 1% HCl, satd aq
NaHCO3 and satd aq NaCl. The organic layer was
dried over anhydrous Na2SO4 and the solvent was eva-
porated to give the residue which was recrystalized from
EtOAc to afford light yellow solid (1.077 g, 73%): mp
ꢂ
light yellow solid (6.9 g, 87%): mp104–105 C; Rf 0.328
1
(20% EtOAc in hexanes); H NMR d 3.80 (s, 3H), 5.50
(s, 2H), 7.05 (d, 1H, J=8.1 Hz), 7.10 (t, 1H), 7.24 (m,
2H), 7.34 (m, 3H), 7.47 (m, 3H), 7.61 (d, 1H, J=7.8
Hz), 8.27 (d, 1H, J=9.0 Hz), 9.28 (brs, 1H, NH), 12.27
(s, 1H, OH); 13C NMR d 52.20, 70.80, 113.34, 120.94,
121.27, 121.58, 122.97, 125.70, 126.05, 126.93, 128.01,
128.21, 128.60, 132.69, 133.93, 136.22, 150.76, 157.01,
165.51, 167.57; CIMS m/z 378 (MH+); HRMS m/z
calcd for C22H20NO5: 378.1341, found 378.1343.
178 ꢂC; Rf 0.447 (40% EtOAc in hexanes); H NMR d
1
3.91 (s, 3H, CH3), 5.30 (s, 2H, CH2), 6.23 (brs, 2H,
NH2), 6.75 (t, 1H, J=7.9 Hz), 7.14–7.30 (m, 6H), 7.45–
7.58 (m, 4H), 7.77 (dd, 1H, J=8.1, 0.7 Hz), 7.80 (dd,
1H, J=8.1, 1.5 Hz), 7.93 (dd, 1H, J=7.6, 1.2 Hz), 8.24
(dd, 1H, J=8.0, 1.7, 1.5 Hz), 8.28 (dd, 1H, J=7.6, 0.7
Hz), 10.64 (s, 1H, NH); 13C NMR d 51.59, 70.89,
111.92, 113.94, 113.97, 115.46, 115.77, 121.28, 123.82,
124.74, 125.02, 125.87, 1126.83, 127.81, 128.23, 128.41,
129.37, 131.55, 133.75, 136.12, 139.45, 139.45, 144.36,
150.33, 157.87, 162.77, 162.82, 168.64; CIMS m/z 494
(MH+); HRMS m/z calcd for C29H24N3O5: 494.1715,
found 494.1711.
Methyl 2-[2-(benzyloxy)phenyl]-1,3-benzoxazole-4-car-
boxylate (3). Compound 2 (6.9 g, 1.83 mol) was heated
neat at 230 ꢂC for 2 h with stirring in a long neck 50-mL
round-bottomed flask under vacuum, which was applied
slowly at small intervals (20 min) to remove the water
vapor generated in the reaction. Upon cooling, a light
orange solid mass was obtained, which was dissolved in
a minimum volume of EtOAc. To this solution, hexane
was added with stirring to precipitate the product
benzoazole, which was filtered to give a white solid (6.0
g, 91%): mp100–102 ꢂC; Rf 0.468 (40% EtOAc in hex-
1
anes); H NMR d 3.99 (s, 3H), 5.28 (s, 2H), 7.07–7.13
Methyl 2-[2-(2-hydroxyphenyl)-1,3-benzoxazol-4-yl]-1H-
benz imidazole-4-carboxylate (6). Compound 5 (493 mg,
1 mmol) was dissolved in 20 mL of toluene and p-
toluenesulfonic acid monohydrate (475 mg, 2.5 mmol)
was added. The reaction mixture was refluxed for 2 h
and the solvent was distilled off. The residue was treated
with CH2Cl2 and the solution was washed with satd aq
NaHCO3 and satd aq NaCl. The organic layer was dried
over anhydrous Na2SO4 and the solvent was evaporated
(m, 2H), 7.27–7.33 (m, 1H), 7.35–7.42 (m, 3H), 7.48
(ddd, 1H, J=7.8, 1.7 Hz), 7.63 (br d, 1H), 7.73 (dd, 1H,
J=8.1, 1.0 Hz), 8.02 (dd, 1H, J=7.8, 1.2, 1.0 Hz), 8.25
(br d, 1H); 13C NMR d 52.40, 70.56, 113.61, 114.67,
116.31, 121.03, 121.98, 124.12, 126.79 (2C), 127.65,
128.45, 131.92, 133.19, 136.66, 141.46, 151.24, 157.74,
163.60, 165.97; CIMS m/z 360 (MH+); HRMS m/z
calcd for C22H18NO4: 360.1235, found 360.1228.

4002
D. Kumar et al. / Bioorg. Med. Chem. 10 (2002) 3997–4004
to give a solid which was stirred with EtOAc and fil-
tered. The residue was dried and purified by preparative
TLC using 40% EtOAc in hexanes to afford a white
solid (200 mg, 68%): mp243 ꢂC; Rf 0.170 (40% EtOAc
359.1395, found 358.1393.
2-[2-(benzyloxy)phenyl]-1H-benzimidazole-4-carboxylic
acid (9). Compound 8 (1.0 g, 2.78 mmol) was dissolved
in 20 mL of THF and 8 mL of 5M NaOH was added.
The reaction mixture was stirred at 60 ꢂC with stirring
for 3 h. The reaction mixture was allowed to cool to
room temperature and acidified with concentrated HCl.
The white solid that was obtained was filtered, washed
with water, and dried under vacuum to afford the acid 9
(780 mg, 81%): mp188 ꢂC; Rf 0.125 (40% EtOAc in
1
in hexanes); H NMR d 4.20 (s, 3H, CH3), 7.10 (t, 1H,
J=8.1, 7.2 Hz), 7.24 (d, 1H, J=8.4 Hz), 7.41 (t, 1H,
J=7.8, 7.5 Hz), 7.52–7.61 (m, 2H), 7.74 (d, 1H, J=7.5
Hz), 8.02 (d, 1H, J=7.8 Hz), 8.09–8.14 (m, 2H), 8.56 (d,
1H, J=7.2 Hz), 10.53 (s, 1H, NH), 11.95 (brs, 1H, OH);
13C NMR d 52.65, 109.97, 112.49, 113.88, 117.84,
119.20, 120.19, 122.61, 124.30, 125.71, 126.08, 127.87,
134.10, 134.45, 134.54, 137.65, 142.83, 149.23, 149.39,
158.54, 163.96, 166.64; CIMS m/z 386(MH+); HRMS
m/z calcd for C22H16N3O4: 386.1141, found 386.1150.
1
hexanes); H NMR (DMSO-d6) d 5.32 (s, 2H, CH2),
7.15 (t, 1H, J=7.5 Hz), 7.29–7.41 (m, 5H), 7.49 (t, 1H,
J=8.1 Hz), 7.56 (d, 2H, J=7.4 Hz), 7.78 (d, 1H, J=7.7
Hz), 7.92 (d, 1H, J=7.5 Hz), 8.32 (dd, 1H, J=7.9, 1.8
Hz), 11.85 (brs, 1H); 13C NMR (DMSO-d6) d 70.27,
107.28, 114.01, 117.73, 121.49, 121.66, 124.21, 123,68,
127.85, 128.12,128.64,130.00, 131.80, 134.68, 136.13,
141.02, 150.23, 155.89, 167.11; CIMS m/z 345 (MH+);
HRMS m/z calcd for C22H17N2O3: 345.1239, found
345.1233,
Methyl 2-amino-3-{[2-(benzyloxy)benzoyl]amino}benzo-
ate (7). Carbonyldiimidazole (2.43 g, 15 mmol) was dis-
solved in 30 mL dry THF with stirring under argon at
room temperature and 2-(benzyloxy)benzoic acid (3.42
g, 15 mmol) was then added carefully. Heavy evolution
of CO2 was observed for 3–4 min. After stirring for 10
min, methyl 2,3-diaminobenzoate (1.66 g, 10 mmol) was
added and stirring continued further for 10 min. The
reaction mixture was then heated under reflux for 5
days. The reaction mixture, dark brown in color, was
concentrated and dissolved in a minimum volume of
EtOAc. Silica gel (60–100 mesh) was added to make a
slurry and the solvent was evaporated. The resulting
solid was added to the topof a dry-packed chromato-
graphy column that was eluted with 20% EtOAc in
hexane to give the product as a white solid (2.86 g,
76%): mp111–113 ꢂC; Rf 0.373 (40% EtOAc in hex-
Methyl 2-(2-hydroxyphenyl)-benzoxazole-4-carboxylate
(10). Compound 2 (103 mg, 0.3 mmol) was dissolved in
4 mL of toluene and p-toluenesulfonic acid mono-
hydrate (142 mg, 0.75 mmol) was added. The reaction
mixture was refluxed for 1.5 h and after cooling at room
temperature was diluted with EtOAc. The organic layer
was washed with satd aq NaHCO3 and satd aq NaCl
and dried over anhydrous Na2SO4 and the solvent was
evaporated to afford the benzoxazole as a solid (45 mg,
58%): mp134–135 ꢂC; Rf 0.578 (30% EtOAc in hex-
1
1
anes); H NMR d 3.86 (s, 3H, CH3), 5.25 (s, 2H, CH2),
anes); H NMR d 4.05 (s, 3H, CH3), 7.01 (t, 1H), 7.13
5.58 (s, 2H, NH2), 6.66 (t, 1H, J=8.0 Hz), 7.13–7.18 (m,
2H), 7.43–7.58 (m, 7H), 7.75 (d, 1H, J=7.8 Hz), 8.27 (d,
1H, J=7.8 Hz), 9.20 (brs, 1H, NH); 13C NMR d 51.68,
71.72, 112.57, 112.59, 116.10, 121.75, 121.83, 124.76,
128.49, 128.77, 129.12 (2C), 130.09, 132.72, 133.32,
135.01, 144.51, 156.54, 163.98, 168.40; CIMS m/z 377
(MH+); HRMS m/z calcd for C22H21N2O4: 377.1501,
found 377.1499.
(d, 1H, J=8.1 Hz), 7.4–7.5 (m, 2H), 7.78 (d, 1H, J=8.1
Hz), 8.01 (d, 1H, J=7.5 Hz), 8.06 (d, 1H, J=8.1 Hz);
13C NMR d 52.38, 109.91, 114.86, 117.63, 119.55,
121.24, 124.69, 127.15, 127.40, 134.19, 139.36, 149.70,
159.33, 164.26, 165.52; CIMS m/z 270 (MH+); HRMS
m/z calcd for C15H12NO4: 270.0766, found 270.0766.
Methyl 3-hydroxy-2-{[2-(benzyloxyphenyl)-1,3-benzoxa-
zol-4-yl] carbonyl}aminobenzoate (11). To a solution of
compound 4 (3.45 g, 10 mmol) in 60 mL of dry CH2Cl2
was added 10 mL of freshly distilled oxalyl chloride.
The reaction mixture was stirred for 10 min and 0.3 mL
of dimethyl formamide was added to initiate the reac-
tion. The reaction mixture was stirred for 2 h at room
temperature and the solvent was removed by evapora-
tion. The resultant yellow solid was dried under vacuum
for 3 h. The yellow solid was then dissolved in 20 mL of
dry CH2Cl2 and transferred via canula to a solution of
methyl 2-amino-3-hydroxybenzoate (1.67 g, 10 mmol)
dissolved in 40 mL of dry CH2Cl2 and 3 mL of pyridine.
The reaction mixture was stirred at room temperature
overnight and then diluted with dichlomethane. The
solution was washed with 1% HCl, satd aq NaHCO3
and satd aq NaCl. The organic layer was dried over
anhydrous Na2SO4 and the solvent was evaporated to
give a residue which was subjected to column chroma-
tography using CH2Cl2 as eluant to afford a light yellow
solid (3.8 g, 77%): mp157–158 ꢂC; Rf 0.329 (CH2Cl2);
1H NMR d 3.79 (s, 3H), 5.29 (s, 2H), 7.13–7.39 (m, 8H),
7.50–7.55 (m, 4H), 7.65 (dd, 1H, J=7.5, 1.5 Hz), 7.80
Methyl 2-[2-(benzyloxy)phenyl]-1H-benzimidazole-4-car-
boxylate (8). Compound 7 (2.86 g, 76 mmol) was
heated in glacial acetic acid (20 mL) at 120 ꢂC for 30
min with stirring in a 50 mL round bottom flask. The
reaction mixture was allowed to cool and then poured
into ice cold water and extracted with CH2Cl2. The
combined organic layers were washed with satd aq
NaHCO3, brine and dried (Na2SO4). The solvent was
removed by evaporation and the residue was purified by
chromatography on silica gel (60–100 mesh) using 20 to
40% EtOAc in hexane to afford benzimidazole ester 8
as a white solid (2.3 g, 80%): mp100–101 ꢂC; Rf 0.412
(40% EtOAc in hexanes); 1H NMR d 3.67 (s, 3H, CH3),
5.36 (s, 2H, CH2), 7.17–7.23 (m, 2H), 7.33 (t, 1H, J=7.7
Hz), 7.40–7.49 (m, 4H), 7.56 (dd, 2H, J=7.6, 1.7 Hz),
7.90 (dd, 1H, J=7.6, 1.0 Hz), 8.03 (d, 1H, J=8.4 Hz),
8.62 (dd, 1H, J=7.8, 1.8 Hz), 11.56 (brs, 1H); 13C NMR
d 51.69, 71.29, 112.95, 113.40, 117.93, 121.68, 121.90,
124.30, 124.52, 127.95, 128.51, 128.92, 130.51, 131.58,
134.16, 135.64, 143.74, 150.91, 156.33, 166.33; CIMS m/
z 359 (MH+); HRMS m/z calcd for C22H19N2O3:

D. Kumar et al. / Bioorg. Med. Chem. 10 (2002) 3997–4004
4003
(brd, 1H, J=8.1 Hz), 8.31 (brd, 1H, J=7.8 Hz), 8.50
(dd, 1H, J=7.5, 1.8 Hz); 13C NMR d 52.23, 70.71,
113.76, 114.71, 115.76, 121.18, 122.51, 122.98, 123.62,
124.73, 125.62, 126.35, 126.50, 126.69, 127.68, 127.87,
132.31, 133.66, 136.28, 139.91, 150.87, 151.20, 157.89,
163.46, 164.40, 167.11 ; CIMS m/z 495 (MH+); HRMS
m/z calcd for C29H23N2O6: 495.1556, found 495.1553.
IC50 values, the average concentration required to
produce a decrease of fluorescent intensity of 50%
relative to vehicle-treated controls in two separate
determinations.
Antimicrobial bioassay
S. aureus ATCC 29213 and methicillin-resistant S. aureus
ATCC 43300 (MRS) were obtained from the American
Type Culture Collection (Rockville, MD, USA) and
stored on agar slants at 4 ꢂC until needed. Susceptibility
testing was performed using a modified version of the
NCCLS methods.23,24 Microorganisms were subcultured
prior to the assay by suspending cells from the slant in
Eugon broth (BBL, MD, USA) and incubating for 24 h
at 37ꢂC. Inocula were prepared by diluting the sub-
cultured organism in its incubation broth after compar-
ison to the 0.5 McFarland Standard (a BaSO4 suspension)
to afford final inocula ranges of S. aureus: 1.0–5.0ꢁ105
and MRS: 0.2–6.0ꢁ105 CFU/mL. Test compounds were
dissolved in DMSO, serially-diluted using normal saline,
and transferred in duplicate to 96-well flat-bottomed
microtiter plates. The microbial inocula were added to the
samples to achieve a final volume of 200 mL and final
compound concentrations starting with 50 mg/mL. Drug
[Ciprofloxacin (ICN Biomedicals, OH, USA)] as well as
growth and blank (media only) controls were added to
each test plate. Plates were read at 630 nm using the
EL-340 Biokinetics Reader (Bio-Tek Instruments, VT,
USA) prior to and after incubation at 37 ꢂC for 24 h.
Percent growth was calculated and p lotted versus con-
centration to afford the IC50, or sample concentration that
affords 50% growth of the organism relative to controls.
Methyl
20-(3-benzyl-2-hydroxyphenyl)-2,40-bi-1,3-
benzoxazole-4-carboxylate (12). A mixture of dry tolu-
ene (20 mL) and p-toluenesulfonic acid monohydrate
(950 mg, 5 mmol) was refluxed under a reflux con-
denser. Compound 11 (988 mg, 2 mmol) was added and
the reaction mixture was refluxed for an additional 2 h.
After the solvent was removed by distillation, the resi-
due was dissolved in CH2Cl2 and the solution was
washed with satd aq NaHCO3 and satd aq NaCl. The
organic layer was dried over anhydrous Na2SO4 and the
solvent was evaporated to give 432 mg of solid that was
1
a mixture of UK-1 and compound 12 (TLC and H
NMR) in the ratio of 1.6:1.0. Careful column chroma-
tography using 5 to 20% EtOAc in hexane afforded 10
mg pure compound 12 (1%) along with 300 mg UK-1
(13%) and a large mixed fraction (122mg). Compound
12: mp204–205 ꢂC; Rf 0.357 (30% EtOAc in hexanes);
1H NMR d 4.09 (s, 3H, CH3), 4.15 (s, 2H, CH2), 6.96 (t,
1H, J=7.6 Hz, C5-H), 7.20–7.14 [m, 1H, C3-benzyl-
H(para)], 7.26 [s, 2H, C3-benzyl-H(meta)], 7.27 [s, 2H,
C3-benzyl-H(ortho)], 7.30 (dd, 1H, J=7.5, 1.2 Hz, C4-
H), 7.46 (t, 1H, J=7.9 Hz, C6-H), 7.52 (t, 1H, J=8.0
Hz, C6 -H), 7.77 (dd, 1H, J=8.1, 1.0 Hz, C7 -H), 7.86
(dd, 1H, J=8.1, 1.0 Hz, C7-H), 7.96 (dd, 1H, J=7.9, 1.6
Hz, C6-H), 8.10 (dd, 1H, J=7.7, 1.0 Hz, C5-H), 8.31
(dd, 1H, J=7.8, 0.9 Hz, C5 -H); 13C NMR d 35.83,
0
0
0
52.63, 109.72, 113.82, 115.09, 117.61, 119.33, 122.57,
124.93, 125.11, 125.24, 125.51, 126.07, 127.57, 128.42,
128.81, 129.77, 135.26, 138.76, 140.38, 141.56, 150.01,
151.22, 157.65, 161.71, 165.05, 166.32 : CIMS m/z 477
(MH+); HRMS m/z calcd for C29H21N2O5: 477.1450,
found 477.1442.
Continuous variation plots
A 100 mM solution of UK-1 or analogue and 100 mM
solutions of Ca(NO3)2, Mg(NO3)2, Zn(NO3)2, or
Fe(NO3)3 were prepared in methanol. From these stock
solutions, 15 samples (1 mL) of varying mole fraction of
the metal ion were prepared at a constant, combined
concentration of 20 mM for the ligand and the metal
ion. The change in absorbance was monitored from 500
to 200 nm for each sample using as a reference a 1 mL
solution containing the same concentration of ligand as
the sample being measured but without the metal ion.
The maximum absorbance change was typically around
418 nm but varied somewhat for each metal ion. The
absorbance at 418 nm was plotted as a function of the
mole fraction of the metal ion and the maximum
absorbance change was determined. The absorbance of
a sample containing the metal ion concentration corre-
sponding to this maximum change and a 15-fold excess of
metal ion at this ligand concentration, subtracted from a
reference containing the same concentration of ligand but
no metal ion was used to normalize the curve and obtain
the conditional formation constant as described.25
Cytotoxicity assays
Cell culture cytotoxicity assays were performed using
the alamarBlue method, as previously described.22
Briefly, aliquots of 100 mL of cell suspension (1.0–
2.5ꢁ103 cells) were placed in 96-well microtiter plates in
an atmosphere of 5% CO2 at 37 ꢂC. After 24 h, 100 mL
of culture medium and 2 mL of compound in vehicle
(culture media with 40% pyridine) or vehicle alone were
added, and the plates incubated an additional 72 h. The
final pyridine concentration in all cases was 0.4%; at
this pyridine concentration, there was no affect on the
growth of cells compared to cells in culture media with-
out added pyridine. Compounds, along with mitomycin
C as positive control, were evaluated in duplicate at
final concentrations ranging from 0.001 to 100 mM.
After the culture media had been removed from each
well, 200 mL of fresh media and 20 mL of 90% alamar-
Blue reagent were added, followed by an additional 6 h
incubation. The fluorescent intensity was measured
using a SpectrafluorPlus plate reader with excitation at
530 nm and emission at 590 nm. Results are reported as
Acknowledgements
This work was supported in part by the Robert Welch
Foundation (F-1298 to S.M.K.) and the United States

4004
D. Kumar et al. / Bioorg. Med. Chem. 10 (2002) 3997–4004
Department of Agriculture, Agricultural Research Service
Specific Cooperative Agreement No. 58-6408-2-0009.
14. Reyzer, M. L.; Brodbelt, J. S.; Kerwin, S. M.; Kumar, D.
Nucleic Acid Res. 2001, 29, e103.
15. Silva, D.; Goodnow, R.; Kahne, D. Biochemistry 1993, 32,
463.
References and Notes
16. Cetiinkaya, E.; Alici, B.; Gok, Y.; Durmaz, R.; Gunal, S.
J. Chemother. 1999, 11, 83.
1. Ueki, M.; Ueno, K.; Miyadoh, S.; Abe, K.; Shibata, K.;
Taniguchi, M.; Oi, S. J. Antibiot. 1993, 46, 1089.
2. Sato, S.; Kajiura, T.; Noguchi, M.; Takehana, K.;
Kobayasho, T.; Tsuji, T. J. Antibiot. 2001, 54, 102.
3. Ueki, M.; Shibata, K.; Taniguchi, M. J. Antibiot. 1998, 51, 883.
4. Ueki, M.; Taniguchi, M. J. Antibiot. 1997, 50, 788.
5. DeLuca, M. R.; Kerwin, S. M. Tetrahedron Lett. 1997, 38,
199.
17. Qin, W.; Obare, S. O.; Murphy, C. J.; Angel, S. M. Ana-
lyst 2001, 126, 1499.
18. Nakamura, N.; Wakabayashi, S.; Miyairi, K.; Fujii, T.
Chem. Lett. 1994, 1741.
19. Cavallito, C. J.; Buck, J. S. J. Am. Chem. Soc. 1943, 65,
2140.
20. Kelly, T. R.; Martinez, C.; Mears, R. J. Heterocycles 1997,
45, 87.
6. Hoveyda, H. R.; Rettig, S. J.; Orvig, C. Inorg. Chem. 1993,
32, 4909.
7. Tanaka, K.; Kumagai, T.; Aoki, H.; Deguchi, M.; Iwata, S.
J. Org. Chem. 2001, 66, 7328.
8. Ratledge, C.; Snow, G. A. Biochem. J. 1974, 139, 407.
9. Reynolds, M. B.; DeLuca, M. R.; Kerwin, S. M. Bioorg.
Chem. 1999, 27, 326.
10. Aich, P.; Dasgupta, D. Biochemistry 1995, 34, 1376.
11. Banville, D. L.; Keniry, M. A.; Shafer, R. H. Biochemistry
1990, 29, 9294.
21. Kubo, K.; Kohara, Y.; Imamiya, E.; Sugiura, Y.; Inada,
Y.; Furukawa, Y.; Nishikawa, K.; Naka, T. J. Med. Chem.
1993, 36, 2182.
22. Ahmed, S. A.; Gogal, R. M., Jr.; Walsh, J. E. J. Immun.
Methods 1994, 170, 211.
23. Ferraro, M. J. Methods for Dilution Antimicrobial Sus-
ceptibility Tests for Bacteria that Grow Aerobically; Natl.
Comm. Clin. Lab. Stds.: Wayne, 1997. M7-A4, Vol. 17, No.
2.
24. Galgiani, J. N. Reference Method for Broth Dilution Anti-
fungal Susceptibility Testing of Yeasts, Approved Standard;
Natl. Comm. Clin. Lab. Stds.: Wayne, 1997; M27-A, Vol. 17,
No. 9.
12. Fan, J.-Y.; Sun, D.; Yu, H.; Kerwin, S. M.; Hurley, L. H.
J. Med. Chem. 1995, 38, 408.
13. Yu, H.-T.; Hurley, L. H.; Kerwin, S. M. J. Am. Chem.
Soc. 1996, 118, 7040.
25. Likussar, W.; Boltz, D. F. Anal. Chem. 1971, 43, 1265.