Coumarin-based inhibitors of HIV integrase

Article abstract of DOI:10.1021/jm960450v

The structures of a large number of HIV-1 integrase inhibitors have in common two aryl units separated by a central linker. Frequently at least one of these aryl moieties must contain 1,2-dihydroxy substituents in order to exhibit high inhibitory potency. The ability of o-dihydroxy-containing species to undergo in situ oxidation to reactive quinones presents a potential limitation to the utility of such compounds. The recent report of tetrameric 4-hydroxycoumarin-derived inhibitor 5 provided a lead example of an inhibitor which does not contain the catechol moiety. Compound 5 represents a large, highly complex yet symmetrical molecule. It was the purpose of the present study to determine the critical components of 5 and if possible to simplify its structure while maintaining potency. In the present study, dissection of tetrameric 5 (IC₅₀ = 1.5 μM) into its constituent parts showed that the minimum active pharmacophore consisted of a coumarin dimer containing an aryl substituent on the central linker methylene. However, in the simplest case in which the central linker aryl unit consisted of a phenyl ring (compound 8, IC₅₀ = 43 μM), a significant reduction in potency resulted by removing two of the original four coumarin units. Replacement of this central phenyl ring by more extended aromatic systems having higher lipophilicity improved potency, as did the addition of 7- hydroxy substituents to the coumarin rings. Combining these latter two modifications resulted in compounds such as 3,3'-(2- naphthalenomethylene)bis[4,7-dihydroxycoumarin] (34, IC₅₀ = 4.2 μM) which exhibited nearly the full potency of the parent tetramer 5 yet were structurally much simpler.

Full text of DOI:10.1021/jm960450v

2
42
J . Med. Chem. 1997, 40, 242-249
Cou m a r in -Ba sed In h ibitor s of HIV In tegr a se1
†
‡
†
‡
†
‡
He Zhao, Nouri Neamati, Huixiao Hong, Abhijit Mazumder, Shaomeng Wang, Sanjay Sunder,
†
‡
,†
George W. A. Milne, Yves Pommier, and Terrence R. Burke, J r.*
Laboratories of Medicinal Chemistry and Molecular Pharmacology, Division of Basic Sciences, National Cancer Institute,
National Institutes of Health, Bethesda, Maryland 20892
X
Received J une 21, 1996
The structures of a large number of HIV-1 integrase inhibitors have in common two aryl units
separated by a central linker. Frequently at least one of these aryl moieties must contain
1
,2-dihydroxy substituents in order to exhibit high inhibitory potency. The ability of
o-dihydroxy-containing species to undergo in situ oxidation to reactive quinones presents a
potential limitation to the utility of such compounds. The recent report of tetrameric
4
-hydroxycoumarin-derived inhibitor 5 provided a lead example of an inhibitor which does not
contain the catechol moiety. Compound 5 represents a large, highly complex yet symmetrical
molecule. It was the purpose of the present study to determine the critical components of 5
and if possible to simplify its structure while maintaining potency. In the present study,
dissection of tetrameric 5 (IC50 ) 1.5 µM) into its constituent parts showed that the minimum
active pharmacophore consisted of a coumarin dimer containing an aryl substituent on the
central linker methylene. However, in the simplest case in which the central linker aryl unit
consisted of a phenyl ring (compound 8, IC50 ) 43 µM), a significant reduction in potency resulted
by removing two of the original four coumarin units. Replacement of this central phenyl ring
by more extended aromatic systems having higher lipophilicity improved potency, as did the
addition of 7-hydroxy substituents to the coumarin rings. Combining these latter two
modifications resulted in compounds such as 3,3′-(2-naphthalenomethylene)bis[4,7-dihydroxy-
coumarin] (34, IC50 ) 4.2 µM) which exhibited nearly the full potency of the parent tetramer
5
yet were structurally much simpler.
Development of resistance is a major problem associ-
ated with traditional therapeutic approaches toward
AIDS. One promising means of overcoming this is to
direct agents at multiple points which are required for
with the dimeric 6,7-dihydroxynaphtho-2-yl analogue 4
exhibiting the best potency of the series.
8
2
HIV replication. The thought here is that the simul-
taneous mutation of more than one critical enzyme,
which would be necessary for this type of multidrug
resistance, either would be quite rare or would result
in decreased virulence. Among HIV enzymes, three
have been identified as major targets for therapeutic
development. The first two, reverse transcriptase and
protease, have inhibitors either on the market or in
clinical trial. The third enzyme, integrase, is equally
appealing as a site for antagonist action; however, this
enzyme has not yet been as extensively investigated as
the first two. Recent reports have appeared on HIV
integrase inhibitory potency of a variety of compounds.
In many cases, multiple aromatic rings and aryl ortho-
hydroxylation are required for good inhibitory potency.
Examples of these compounds include flavones, such as
3
3
quercetin (1), caffeic acid phenethyl ester (CAPE, 2),
4
5
and analogues as well as certain “tyrphostins” (3),
arctigenin-based compounds, and bis-catechols such as
â-conidendrol. These inhibitors may be described in
6
7
general as consisting of two aryl units, at least one of
which contains the 1,2-dihydroxy pattern, separated by
an appropriate linker segment. Based on this model, a
series of bis-arylamides have recently been prepared,
The practical utility of catechol-containing inhibitors
is significantly diminished by cytotoxicity, presumably
attributable in part to the in situ oxidation of the
catechol moiety to reactive quinone species. This is
exemplified by the finding that CAPE-like compounds
can cross-link cellular protein at micromolar concentra-
tions, most likely via Michael-type addition of protein
nucleophiles to such quinone intermediates.9 In an
effort to derive new inhibitors of the general type 1-4,
*
To whom correspondence should be addressed at: Building 37,
Room 5C06, NIH, Bethesda, MD 20892. Fax: (301) 402-2275. E-mail:
tburke@helix.nih.gov.
†
Laboratory of Medicinal Chemistry.
Laboratory of Molecular Pharmacology.
Abstract published in Advance ACS Abstracts, J anuary 1, 1997.
‡
X
S0022-2623(96)00450-5 This article not subject to U.S. Copyright. Published 1997 by the American Chemical Society

Coumarin-Based Inhibitors of HIV Integrase
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 2 243
Ta ble 1. Examination of Constituent Components of Tetramer
5
Ta ble 2. Rotation of the Site of Coumarin Attachment about a
a
Central Phenyl Linkera
a
IC50 values were obtained against HIV-1 integrase as outlined
in the Experimental Section. Results are expressed as the average
of three independent experiments.
which do not contain catechol functionality, note was
taken of the high HIV integrase inhibitory potency of
certain coumarin-based analogues such as coumermycin
1
0
11
A1 and, most notably, the tetrameric compound 5.
Inhibitor 5 bears several planes of symmetry which
allow it to be viewed in multiple orientations. More
importantly, though, is the absence of catechol func-
tionality. Based on these considerations, a study was
undertaken to delineate structural features of 5 neces-
sary for high inhibitory potency and to explore modifica-
tions which might simplify the structure without losing
potency.
a
IC50 values were obtained against HIV-1 integrase as outlined
in the Experimental Section. Where standard errors are given,
results are expressed as the average of three independent experi-
ments. In other cases, results of two independent experiments
are explicitly listed.
Syn th eses
Synthesis of final products was accomplished by one
of two general methods. Geminal dicoumarins were
prepared by reaction of an appropriate aldehyde linker
with either 4-hydroxycoumarin or 4,7-dihydroxycou-
marin, while nongeminally substituted analogues were
synthesized by heating coumarin monomers neat with
an R,R′-dibromoxylene.
represent a dissection of the parent tetramer 5 into its
possible dimeric components (7-9) as well as the
monomeric building block, 4-hydroxycoumarin (6). None
of the analogues retained the full potency of the parent
tetramer 5 (IC50 ) 1.5 µM), with the monomeric 6 and
partial “horizontal” dimer 7 being inactive within the
test range. Inclusion of the phenyl ring in the horizontal
dimer (compound 8, IC50 ) 43 µM) exhibited approxi-
mately 2-fold better inhibitory potency than the corre-
sponding “vertical” dimer 9 (IC50 ) 81 µM).
The above results suggested that both the nature of
the phenyl portion of the central “linker” and the
position of attachment of the coumarin units affected
affinity. In order to more closely examine the effect of
attachment position, additional analogues were there-
fore prepared, 10 and 11, which placed the coumarins
in 1,2- and 1,3-orientations relative to the phenyl spacer.
The resulting set of isomeric dimers (8-11) sequentially
rotated the coumarins about the phenyl spacer from the
Resu lts a n d Discu ssion
One purpose of this study was to examine features of
tetracoumarin analogue 5 which contribute to its high
integrase inhibitory potency. A general arrangement
of two aromatic units separated by a central linker has
emerged as a common motif for a large number of HIV
integrase inhibitors. As a prelude, it was necessary to
first determine the optimum arrangement of structural
components of 5. An initial series of compounds were
therefore assembled (6-9) to address this question
(Table 1). These four coumarin-containing analogues

2
44 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 2
Zhao et al.
Ta ble 3. Examination of the Effect of Central Linkers on Inhibitory Potencya
a
IC50 values were obtained against HIV-1 integrase as outlined in the Experimental Section. Results of two independent experiments
are explicitly expressed.
1
,1- to the 1,4-positions (Table 2). While inhibition of
of 2 (compound 9). Because rotation about the phenyl
spacer has the double effect of both increasing inter-
coumarin spacing and torquing relative coumarin ori-
entations, the modest effect on inhibitory potency
indicates surprising latitude in both of these param-
eters. Finally, since the 1,3-orientation was more potent
integration is slightly increased in going sequentially
from the 1,1- to 1,2- to 1,3-orientations (compounds
8
ing effect on 3′-processing until the 1,4-orientation is
achieved, at which point potency decreases by a factor
-11, respectively), there is no significant correspond-

Coumarin-Based Inhibitors of HIV Integrase
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 2 245
Ta ble 4. Examination of the Effect of 7-Hydroxylation on Inhibitory Potencya
aIC50 values were obtained against HIV-1 integrase as outlined in the Experimental Section. Where standard errors are given, results
are expressed as the average of three independent experiments. In other cases, results of two independent experiments are explicitly
listed.
than the 1,4-orientation in this series of dimers, it was
also interesting to note that the corresponding 1,3-
arranged tetramer (12, IC50 ) 2.9 µM) did not exhibit
greater inhibitory potency than the 1,4-substituted
parent 5 (Table 2).
Having examined positional isomerism, it was next
of interest to study effects of variation in the chemical
structure of the central linker. Since the “horizontal”
dimer is more potent than the “vertical” dimer (com-
pounds 8 and 9, respectively, Table 1), the former
pattern was utilized, and a series of aryl-containing
geminal coumarins (13-27) was prepared (Table 3).
Using as a reference compound 8, which has an unsub-
stituted benzyl central linker, addition of hydroxyl
The number of aryl rings on the central linker was
increased next. Utilizing a 3-quinolyl ring (compound
20, IC50 ) 35 µM) resulted in a slight increase in potency
relative to 8, while removal of the nitrogen, yielding the
2-naphthyl ring system (compound 21, IC50 ) 20 µM),
gave a further enhancement. In the next series, the
length of the aryl system was extended first as the 4′-
biphenyl 22, then as the closely related fluorenyl 23,
and finally as the stilbene analogue 24. Each modifica-
tion sequentially enhanced inhibitory potency, with
analogue 24 (IC50 ) 7.3 µM) being approximately 6-fold
more potent than starting phenyl compound 8. Re-
placement of the styryl component of 24 with a benzyl-
oxy group (compound 25) caused a 2.5-fold loss of
potency. In this case, both 24 and 25 have phenyl rings
separated by 2-unit spacers; however, the vinyl bridge
of 24 is significantly more rigid than the methyleneoxy
bridge of 25. Addition of a second benzyloxy group
(compound 26) restored most of the lost potency, with
the compound now appearing reminiscent of a simplified
version of the tetrameric structure of parent 5. Finally,
ferrocene analogue 27 displayed potency equivalent to
(compounds 13 and 14) or amino (compound 15) func-
tionality to the 4-position reduced inhibitory activity
approximately 3-4-fold, while either nitro or carboxyl
substituents at this position had little effect (compounds
1
6 and 17, respectively). Replacement of the phenyl
ring with a heteroaryl ring decreased inhibitory potency,
with 2-thiopheno (18) being less potent than the 2-fura-
no analogue 19.

2
46 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 2
Zhao et al.
The above results firmly established the value of
-hydroxylation, and it was of further interest to
7
examine the relative effect of this modification when
combined with alterations in the central linker. Re-
placement of the phenyl ring of 30 with 2- and 3-pyridyl
rings (32 and 33, IC50 ) 81 and 89 µM, respectively)
decreased inhibitory potency from 4- to 5-fold. Exten-
sion of the phenyl ring as the 2-naphthyl and 4-stilbene
analogues (34, IC50 ) 4.2 µM; 35, IC50 ) 7 µM,
respectively) increased inhibitory potency, similar to the
effect observed in compounds 21 and 24, which lacked
the 7-hydroxyl substituents.
As a final study, compounds 36 and 37 were prepared
as conformationally constrained analogues in which the
aryl ring of the central linker is locked via an ether
bridge to one of the coumarin units. The cumulative
effect of this ether bridge is to lock both center linker
and the coumarin ring into fixed conformations. This
type of restraint proved deleterious to inhibitory potency
(Table 4).
In summarizing the findings of Tables 1-4, it can be
concluded that adding 7-hydroxy groups to the coumarin
rings increased inhibitory potency across a wide range
of analogues. Additionally, modification of the central
linker by extending its aryl component also enhanced
affinity. This latter effect may be dependent on both
the shape and the hydrophobicity of the central linker
unit. In order to ascertain the possible role hydropho-
bicity may have in differential potency, the logarithm
of the partition coefficient in an n-octanol/water system
F igu r e 1. Inhibition of the related retroviral integrases from
HIV-1, HIV-2, and SIV by five representative coumarin deriva-
tives: phosphorimager pictures of the concentration-dependent
inhibition of HIV-1 (A), HIV-2 (B), and SIV (C) integrases. The
DNA strand transfer products (STP) are indicated by brackets,
while the 3′-processing product (19-mer) and the DNA sub-
strate (21-mer) are shown by arrows. Coumarin concentrations
(log P) was approximated using the PROLOGP program,
which calculates a theoretical log P based on chemical
structure. In several systems quantitative consider-
ations of lipophilicity have proven to be useful param-
eters in the bioactivity. As seen in Table 5, a general
trend was observed in which increased hydrophobicity
of the central linker substituent was associated with
increased potency. Compounds fell within two groups:
Those having IC50 > 30 µM exhibited ∆log P < 2, while
those with IC50 e 20 had ∆log P > 2.5. The lack of strict
correlation between hydrophobicity and potency may
indicate that additional parameters must be considered.
These may include steric factors and specific bonding
interactions which could occur between the enzyme and
the central linker moiety. Additionally, the possibility
of multiple binding orientations within a given binding
site, as well as multiple binding sites, may all contribute
to the lack of a strict correlation.
It was hoped that this study would clarify some of
the structural requirements for the high inhibitory
potency of parent tetramer 5 and, in so doing, allow the
generation of smaller, simpler coumarin-based struc-
tures which retained high inhibitory potency. Several
of the analogues reported in this study have reached
this latter objective. Of note are compounds 34 and 35,
which have effectively replaced the “bottom” half of
parent tetramer 5 with simple aryl rings without
significantly sacrificing potency.
(µM) are indicated above each lane: lane 1, DNA alone; lanes
2
7
, 15, and 24, plus integrase without inhibitor; lanes 3-6,
-10, 16-19, and 20-23, plus integrase in the presence of
coumarin inhibitors 8, 13, 24, 30, and 31, respectively.
stilbene analogue 24, which was the best compound of
the series to this point.
Having examined modifications of the central linker,
the next group of analogues (Table 4) was designed to
assess the effect of altering the coumarin ring itself by
addition of a 7-hydroxyl moiety. Precedence for hy-
droxylation at the 7-position was established by the
recent finding that a 7-hydroxycoumarin-containing
1
1
analogue showed good integrase inhibitory potency.
In the present study, monomeric 4,7-dihydroxycoumarin
8 showed extremely weak potency (IC50 ) 300 µM);
2
however, conversion to the simple dimer 29 significantly
enhanced potency (IC50 ) 46 µM). Addition of a phenyl
ring (compound 30, IC50 ) 17 µM) more than doubled
the potency, while elaboration as the full tetramer 31
resulted in a 4-fold enhancement of inhibitory potency
(
IC50 ) 0.37) relative to the parent compound 5. For
comparative purposes five representative compounds (8,
3, 24, 30, and 31) were tested for inhibition of related
1
Exp er im en ta l Section
retroviral integrases from HIV-2 and Simian immuno-
deficiency virus (SIV). Gels for these assays as well as
for an HIV-1 assay are shown in Figure 1. All five
compounds inhibited both 3′-processing and strand
transfer activities against HIV-1, HIV-2, and SIV inte-
grases, with no remarkable selectivity noted.
P r epar ation of Oligon u cleotide Su bstr ates. The HPLC-
purified oligonucleotides AE117 (5′-ACTGCTAGAGATTTTC-
CACAC-3′) and AE118 (5′-GTGTGGAAAATCTCTAGCAGT-
3
′) were purchased from Midland Certified Reagent Co.
(Midland, TX). Purified recombinant wild-type HIV-1 was
prepared as described.12 The extent of 3′-processing and

Coumarin-Based Inhibitors of HIV Integrase
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 2 247
Ta ble 5. Comparison of Inhibitory Potency Determined against HIV-1 versus Hydrophobicity of the Central Linkera
a
∆log P values represent the hydrophobic contribution of the R group, taking the value of R ) H as zero.
strand transfer was determined using 5′-end-labeled substrate,
with 5′-labeled AE118 having been prepared using T4 poly-
rimager (Sunnyvale, CA). Percent inhibition was calculated
using the equation: 100 × [1 - (D - C)/(N - C)], where C, N,
and D are the fractions of 21-mer substrate converted to 19-
mer (3′-processed product) or strand transfer product for DNA
alone, DNA plus integrase, and integrase plus drug, respec-
tively. Determination of IC50 values was achieved by plotting
drug concentration versus percent inhibition and measuring
the concentration at which 50% inhibition occurred.
3
2
nucleotide kinase (Gibco BRL) and [γ- P]ATP (DuPont NEN).
The kinase was heat inactivated prior to adding AE117 to the
same final concentration. The resulting mixture was heated
to 95 °C, allowed to cool slowly to room temperature, and run
on a G-25 Sephadex quick spin column (Boehringer Mannheim,
Indianapolis, IN) to separate double-stranded oligonucleotide
from unincorporated label.
HIV-1 In tegr a se Assa y. Integrase was preincubated with
inhibitor at 30 °C for 30 min at a final concentration of 200
nM in reaction buffer [50 mM NaCl, 1 mM HEPES, pH 7.5,
HIV-1 a n d SIV In tegr a se Assa ys. A plasmid encoding
the HIV-2 integrase was generously provided by Dr. R. H. A.
Plasterk (Netherlands Cancer Institute). This integrase was
1
2
purified essentially as described for the HIV-1 integrase.
5
0 µM EDTA, 50 µM dithiothreitol, 10% glycerol (w/v), 7.5 mM
Purified recombinant wild-type SIV integrase was a generous
gift of R. Craigie (NIDDK). Assays were run as indicated
above for HIV-1, except that a final enzyme concentration of
600 nM was employed for SIV assays.
MnCl
2
, 0.1 mg/mL bovine serum albumin, 10 mM 2-mercap-
toethanol, 10% dimethyl sulfoxide, and 25 mM MOPS, pH 7.2].
Following preincubation, 5′-end 32P-labeled linear oligonucle-
otide substrate was added, and incubation was continued for
an additional 1 h at 30 °C. Reactions were quenched by
addition of an equal volume (16 µL) of loading dye (98%
deionized formamide, 10 mM EDTA, 0.025% xylene cyanol,
Molecu la r Mod elin g. log P values were calculated using
1
3
the program PROLOGP, contained in the PALLAS package.
PALLAS is a package of powerful tools for the prediction of
certain physicochemical parameters of organic compounds
based solely on structural information. PROLOGP is an
optional component of the PALLAS system, which predicts the
logarithm of the partition coefficient (log P) of organic com-
pounds in an n-octanol/water system based on chemical
structure.
0
.025% bromophenol blue), and an aliquot (5 µL) was electro-
phoresed on a denaturing 20% polyacrylamide gel (0.09 M Tris-
borate, pH 8.3, 2 mM EDTA, 20% acrylamide, 8 M urea). Gels
were dried, exposed in a Molecular Dynamics phosphorimager
cassette, and analyzed using a Molecular Dynamics phospho-

2
48 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 2
Syn th esis. Melting points were determined on a Mel Temp
Zhao et al.
+
6.42 (s, 1H), 3.81 (s, 2H); FABMS m/ z 501 (MH ). Anal.
1
II melting point apparatus and are uncorrected. Elemental
analyses were obtained from Atlantic Microlab Inc., Norcross,
GA, and are within 0.4% of theoretical values unless otherwise
indicated. Fast atom bombardment mass spectra (FABMS)
were acquired with a VG Analytical 7070E mass spectrometer
(C32
H
20
O
6
‚ /
4
H
2
O) C, H.
3,3′-(4-St ilb e n om e t h yle n e )b is[4-h yd r oxycou m a r in ]
(24): yield 90%; mp 238-240 °C; IR (KBr) 3448 (br), 3026 (br),
-
1 1
1656, 1618, 1568, 1348, 766 cm ; H NMR (DMSO-d ) δ 7.88
6
(dd, J ) 7.8, 1.2 Hz, 2H), 7.16 (m, 4H), 7.44 (d, J ) 8.3 Hz,
1
under the control of a VG 2035 data system. H NMR data
2H), 7.32 (m, 6H), 7.16 (m, 5H), 6.33 (s, 1H); FABMS m/ z 515
+
were obtained on a Bruker AC250 (250 MHz) spectrometer
and are reported in ppm relative to TMS and referenced to
the solvent in which they were run. Anhydrous solvents were
obtained commercially and used without further drying.
Meth od A: Syn th esis of Cou m a r in Dim er s a n d Tet-
r a m er s by Con d en sa tion of Ald eh yd es w ith 4-Hyd r oxy-
cou m a r in or 4,7-Dih yd r oxycou m a r in . 4-Hydroxycoumarin
or 4,7-dihydroxycoumarin (2 or 4, for tetramers, mmol) was
dissolved in 6 mL of hot ethanol, 1 mmol of the corresponding
aldehyde was added, and the mixture was refluxed for 24 h.
After cooling to room temperature, the solid was filtered off
and crystallized to give the product.
(MH ). Anal. (C33
H
22
O
6
) C, H.
3,3′-[4-(Ben zyloxy)b en zylid en e]bis[4-h yd r oxycou m a -
r in ] (25): yield 85%; mp 180-182 °C; IR (KBr) 3448 (br), 3070
(br), 1671, 1619, 1607, 1569, 1351, 763 cm ; H NMR (DMSO-
-
1 1
d
6
) δ 7.89 (d, J ) 7.8Hz, 2H), 7.59 (m, 2H), 7.37 (m, 9H), 7.05
(d, J ) 8.4 Hz, 2H), 6.68 (d, J ) 8.4 Hz, 2H), 6.27 (s, 1H), 5.01
+
(s, 2H); FABMS m/ z 519 (MH ). Anal. (C32
22 7
H O ) C, H.
1
4
3,3′-[3,5-Bis(ben zyloxy)ben zylid en e]bis[4-h yd r oxycou -
m a r in ] (26): yield 85%; mp 207-209 °C; IR (KBr) 3028 (br),
-
1 1
1663, 1618, 1561, 1349, 764 cm ; H NMR (DMSO-d ) δ 7.85
6
(dd, J ) 7.8, 1.2 Hz, 2H), 7.54 (m, 2H), 7.27 (m, 14H), 6.46 (s,
1H), 6.32 (s, 2H), 6.20 (s, 1H), 4.93 (s, 4H); FABMS m/ z 625.6
+
3
,3′-Ben zylid en ebis[4-h yd r oxycou m a r in ] (8): prepared
(MH ). Anal. (C39
H
28
O
8
) C, H.
1
5
as previously described.
3,3′-(F e r r oce n om e t h yle n e )b is[4-h yd r oxycou m a r in ]
27): yield 35%; mp >360 °C; IR (KBr) 3448 (br), 2365, 2344,
(
1
3
,3′,3′′,3′′′-Isop h th a la lid en etetr a k is[4-h yd r oxycou m a -
-
1 1
6
655, 1608, 1560, 1398, 759 cm ; H NMR (DMSO-d ) δ 7.74
r in ] (12): yield 77%; mp 230-234 °C; IR (KBr) 3448 (br), 1655,
1
-
1
1
(s, 2H), 7.42 (s, 2H), 7.15 (s, 4H), 6.21 (s, 1H), 4.00 (m, 9H);
618, 1561, 1342, 760 cm ; H NMR (DMSO-d
6
) δ 7.68 (d, J
+
FABMS m/ z 520.4 (MH ).
)
7.8 Hz, 4H), 7.54 (m, 4H), 7.25 (m, 4H), 7.22 (m, 4H), 6.95
+
3,3′-Ben zyliden ebis[4,7-dih ydr oxycou m ar in ] (30): yield
(
m, 4H), 6.29 (s, 2H); FABMS m/ z 747.6 (MH ). Anal.
‚H O) C, H.
,3′-(4-H yd r oxyb en zylid en e)b is[4-h yd r oxycou m a r in ]
7
1
2%; mp 287-289 °C; IR (KBr) 3409 (br), 3087 (br), 1655, 1619,
(
C
26
H
18
O
8
2
-
1 1
572, 1354, 793 cm ; H NMR (DMSO-d ) δ 7.71 (d, J ) 8.7
6
3
1
5
Hz, 2H), 7.18 (m, 3H), 7.11 (m, 2H), 6.76 (dd, J ) 8.7, 2.2 Hz,
(
13): prepared as previously described.
,3′-(4-H yd r oxy-3-m et h oxyb en zylid en e)b is[4,7-d ih y-
d r oxycou m a r in ] (14): yield 77%; mp 217-219 °C; IR (KBr)
2
(
H), 6.67 (d, J ) 2.2 Hz, 2H), 6.26 (s, 1H); FABMS m/ z 445
3
+
1
MH ). Anal. (C25
3,3′,3′′,3′′′-(1,4-Dim et h ylen op h en yl)t et r a k is[4,7-d ih y-
d r oxycou m a r in ] (31): yield 72%; mp >380 °C; IR (KBr) 3180
H
16
O
8
‚ /
2 2
H O) C; H: calcd, 3.78; found, 4.19.
-1 1
3
(
1
3
495 (br), 3070 (br), 1671, 1617, 1561, 1271, 768 cm ; H NMR
DMSO-d ) δ 7.90 (dd, J ) 7.8, 1.1 Hz, 2H), 7.59 (dd, J ) 7.8,
.3 Hz, 2H), 7.38 (m, 4H), 6.67 (m, 3H), 6.24 (s, 1H), 3.57 (s,
6
-
1 1
(
6
br), 1652, 1619, 1569, 1256, 780 cm ; H NMR (DMSO-d ) δ
+
7.69 (d, J ) 8.7 Hz, 4H), 6.95 (s, 4H), 6.74 (dd, J ) 8.8, 2.1
H); FABMS m/ z 459 (MH ). Anal. (C26
18 8
H O ) C, H.
Hz, 4H), 6.65 (d, J ) 2.0 Hz, 4H), 6.20 (s, 2H); FABMS m/ z
3
,3′-[4-(Dim eth ylam in o)ben zyliden e]bis[4-h ydr oxycou -
+
1
1
6
811.1 (MH ). Anal. (C26
H
18
O
8
2 2
‚3 / H O) C, H.
m a r in ] (15): prepared as previously described.
3
,3′-(2-P yr id in om et h ylen e)b is[4,7-d ih yd r oxycou m a -
r in ] (32): yield 71%; mp >320 °C; IR (KBr) 3448 (br), 3179
3
,3′-(4-N it r o b e n zy lid e n e )b is [4-h y d r o x y c o u m a r in ]
1
5
(
16): prepared as previously described.
-
1 1
(
1
)
br), 1686, 1611, 1560, 1404, 760 cm ; H NMR (DMSO-d
6
) δ
3
,3′-(4-Ca r b oxyb en zylid en e)b is[4-h yd r oxycou m a r in ]
0.21 (s, 2H), 8.66 (d, J ) 5.6 Hz, 1H), 8.52 (s, 1H), 8.25 (d, J
(
1
17): yield 72%; mp 280-282 °C; IR (KBr) 3398 (br), 1684,
8.0 Hz, 1H), 7.88 (dd, J ) 8.1, 5.7 Hz, 1H), 7.59 (d, J ) 8.7
-
1 1
6
654, 1618, 1560, 1340, 758 cm ; H NMR (DMSO-d ) δ 7.82
Hz, 2H), 6.67 (dd, J ) 8.6, 2.2 Hz, 2H), 6.60 (d, J ) 2.1 Hz,
(
2
dd, J ) 7.8, 1.4 Hz, 2H), 7.78 (d, J ) 8.3 Hz, 2H), 6.76 (m,
+
2
15
H), 6.30 (s, 1H); FABMS m/ z 446 (MH ). Anal. (C24H -
+
H), 7.30 (m, 6H), 6.34 (s, 1H); FABMS m/ z 456 (MH ). Anal.
1
NO
8
‚ /
2 2
H O) C, H, N.
(C
26
H
16
O
8
) C, H.
3
,3′-(3-P yr id in om et h ylen e)b is[4,7-d ih yd r oxycou m a -
3
,3′-(2-Th iop h e n om e t h yle n e )b is[4-h yd r oxycou m a -
r in ] (33): yield 54%; mp 306-308 °C; IR (KBr) 3448 (br), 2364,
1
7
r in ] (18): prepared as previously described.
-1 1
2
344, 1654, 1617, 1561, 1410, 777 cm ; H NMR (DMSO-d
δ 10.26 (s, 2H), 8.59 (d, J ) 5.5 Hz, 1H), 8.41 (m, 1H), 7.80
m, 2H), 7.60 (d, J ) 8.6 Hz, 2H), 6.68 (dd, J ) 8.6, 2.1 Hz,
H), 6.62 (d, J ) 2.1 Hz, 2H), 6.39 (s, 1H); FABMS m/ z 446
6
)
3
,3′-(2-Fu r a n om eth ylen e)bis[4-h yd r oxycou m a r in ] (19):
1
7
prepared as previously described.
(
2
(
3
,3′-(4-Qu in olin om et h ylen e)b is[4-h yd r oxycou m a r in ]
20): yield 78%; mp 279-281 °C; IR (KBr) 3448 (br), 3069 (br),
686, 1609, 1559, 1389, 756 cm-1; 1H NMR (DMSO-d
) δ 9.11
(
1
(
(
+
3
MH ). Anal. (C24
,3′-(2-Na p h th a len om eth ylen e)bis[4,7-d ih yd r oxycou -
m a r in ] (34): yield 51%; mp 284-285 °C; IR (KBr) 3356 (br),
H
15NO
8
‚1 /
4 2
H O) C, H, N.
6
3
d, J ) 1.6 Hz, 1H), 8.87 (s, 1H), 8.31 (d, J ) 8.2 Hz, 1H), 8.20
d, J ) 5.0 Hz, 1H), 8.03 (m, 1H), 7.87 (d, J ) 7.6 Hz, 1H),
7
-1
1
1
(
654, 1610, 1560, 1364, 786 cm ; H NMR (DMSO-d
m, 5H), 7.58 (s, 1H), 7.42 (d, J ) 3.3 Hz, 1H), 7.40 (d, J ) 3.3
Hz, 1H), 7.28 (d, J ) 8.6 Hz, 1H), 6.76 (dd, J ) 8.6, 2.1 Hz,
6
) δ 7.73
.79 (dd, J ) 7.8, 1.5 Hz, 2H), 7.55 (m, 2H), 7.24 (m, 4H), 6.53
1
+
(
s, 1H); FABMS m/ z 464 (MH ). Anal. (C28
H, N.
,3′-(2-Na p h t h a len om et h ylen e)b is[4-h yd r oxycou m a -
r in ] (21): yield 97%; mp 292-294 °C; IR (KBr) 3058 (br), 1663,
H
17
O
6
‚ /
2 2
H O) C,
+
2
H), 6.69 (d, J ) 2.1 Hz, 2H), 6.40 (s, 1H); FABMS m/ z (MH )
3
495. Anal. (C H O ) C, H.
2
9
18
8
3,3′-(4-St ilb en om et h ylen e)b is[4,7-d ih yd r oxycou m a -
r in ] (35): yield 46%; mp 215-217 °C; IR (KBr) 3083 (br), 1655,
-
1 1
1
615, 1561, 1347, 765 cm ; H NMR (DMSO-d
7.6 Hz, 2H), 7.77 (m, 3H), 7.62 (d, J ) 7.6 Hz, 2H), 7.58 (m,
H), 7.40 (m, 3H), 7.36 (m, 2H), 7.32 (m, 2H), 6.49 (s, 1H);
6
) δ 7.90 (d, J
)
1
1619, 1561, 1355, 813 cm-1
; H NMR (DMSO-d ) δ 7.75 (d, J
1
6
) 8.6 Hz, 2H), 7.57 (d, J ) 7.4 Hz, 2H), 7.47 (d, J ) 8.3 Hz,
+
FABMS m/ z 463 (MH ). Anal. (C29
,3′-(4-P h e n ylb e n zylid e n e )b is[4-h yd r oxycou m a r in ]
22): yield 88%; mp 227-229 °C; IR (KBr) 3448 (br), 3028 (br),
671, 1606, 1567, 1347, 761 cm-1; 1H NMR (DMSO-d
) δ 7.89
dd, J ) 7.9, 1.2 Hz, 2H), 7.60 (m, 3H), 7.49 (m, 4H), 7.39 (d,
J ) 8 Hz, 2H), 7.33 (m, 4H), 7.21 (d, J ) 8.1 Hz, 2H), 6.36 (s,
18 6
H O ) C, H.
2H), 7.35 (m, 2H), 7.17 (m, 5H), 6.79 (dd, J ) 8.7, 2.2 Hz, 2H),
+
3
6.71 (d, J ) 2.1 Hz, 2H), 6.29 (s, 1H); FABMS m/ z 546 (MH ).
1
(
1
(
Anal. (C33
H
22
O
8
‚ /
4
H
2
O) C, H.
6
Meth od B: Syn th esis of Cou m a r in Dim er s by Alk yla -
tion of 4-Hyd r oxycou m a r in w ith r,r′-Dibr om oxylen e.
4-Hydroxycoumarin (324 mg, 2 mmol) and R,R′-dibromoxylene
(1 mmol) were heated in the molten state at a temperature of
130-180 °C for 2 h under argon, and the solid was formed.
Then the solid was dissolved in hot ethanol and gave crystals
after cooling.
+
1
1
H); FABMS m/ z 489 (MH ). Anal. (C32
,3′-(2-F lu or en om et h ylen e)b is[4-h yd r oxycou m a r in ]
23): yield 97%; mp 279-280 °C; IR (KBr) 3448 (br), 3071 (br),
H
20
O
6
‚ /
4 2
H O) C, H.
3
(
656, 1618, 1567, 1351, 760 cm-1; H NMR (DMSO) δ 7.90
dd, J ) 7.9, 1.2 Hz, 2H), 7.80 (d, J ) 7.2 Hz, 1H), 7.72 (d, J
7.9 Hz, 1H), 7.60 (m, 2H), 7.52 (d, J ) 7.2 Hz, 1H), 7.38 (m,
H), 7.35 (m, 2H), 7.32 (m, 2H), 7.24 (m, 1H), 7.20 (m, 1H),
1
1
(
)
2
3,3′-p-Xylen ebis[4,7-dih ydr oxycou m ar in ] (9): yield 73%;
mp 320-322 °C; IR (KBr) 3219 (br), 1668, 1631, 1570, 1171,
-1 1
6
759 cm ; H NMR (DMSO-d ) δ 7.95 (dd, J ) 8.4, 1.6 Hz, 2H),

Coumarin-Based Inhibitors of HIV Integrase
.60 (m, 2H), 7.37 (m, 4H), 7.11 (s, 4H), 3.81 (s, 4H); FABMS
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 2 249
7
acid phenethlyl ester (CAPE) and related compounds. Biochem.
Pharmacol. 1994, 48, 595-608.
4) Burke, T. R.; Fesen, M. R.; Mazumder, A.; Wang, J .; Carothers,
A. M.; Grunberger, D.; Driscoll, J .; Kohn, K.; Pommier, Y.
Hydroxylated aromatic inhibitors of HIV-1 integrase. J . Med.
Chem. 1995, 38, 4171-4178.
+
1
m/ z 427 (MH ). Anal. (C26
,3′-o-Xylen eb is[4,7-d ih yd r oxycou m a r in ] (23): yield
7%; mp 326-328 °C; IR (KBr) 3230 (br), 1678, 1635, 1570,
H
18
O
6
‚ /
4 2
H O) C, H.
(
3
2
1
2
3
-
1
1
395, 751 cm ; H NMR (DMSO-d
6
) δ 7.99 (d, J ) 7.8 Hz,
H), 7.64 (m, 2H), 7.38 (m, 4H), 7.01 (m, 2H), 6.94 (m, 2H),
.99 (s, 4H); FABMS m/ z 427 (MH ). Anal. (C26
(5) Mazumder, A.; Gazit, A.; Levitzki, A.; Nicklaus, M.; Yung, J .;
Kohlhagen, G.; Pommier, Y. Effects of tyrphostins, protein
kinase inhibitors, on human immunodeficiency virus type 1
integrase. Biochemistry 1995, 34, 15111-15122.
+
1
H
18
O
6
4 2
‚ / H O)
C, H.
,3′-m -Xylen ebis[4,7-d ih yd r oxycou m a r in ] (11): yield
0%; mp 316-317 °C; IR (KBr) 3240 (br), 1670, 1635, 1570,
3
(
6) Eckart, E.; Pertz, H.; Kaloga, M.; Schulz, J .; Fesen, M. R.;
Mazumder, A.; Pommier, Y. (-)-Arctigenin as a lead structure
for inhibitors of human immunodeficiency virus type-1 integrase.
J . Med. Chem. 1996, 39, 86-95.
2
1
-
1 1
396, 754 cm ; H NMR (DMSO-d ) δ 11.60 (br, 2H), 7.92 (dd,
6
J ) 8.6, 1.6 Hz, 2H), 7.61 (m, 2H), 7.35 (m, 4H), 7.14 (m, 2H),
7
+
(7) LaFemina, R. L.; Graham, P. L.; LeGrow, K.; Hastings, J . C.;
Wolfe, A.; Young, S. D.; Emini, E. A.; Hazuda, D. Inhibition of
human immunodeficiency virus integrase by bis-catechols. An-
timicrob. Agents Chemother. 1995, 39, 320-324.
(8) Zhao, H.; Neamati, N.; Mazumder, A.; Pommier, Y.; Burke, T.
R., J r. Arylamide inhibitors of HIV-1 integrase. J . Med. Chem.,
in review.
.02 (m, 2H), 3.83 (s, 4H); FABMS m/ z 427 (MH ). Anal.
4
(C
26
H
18
O
6
‚
1
/ H
2
O) C, H.
Meth od C: Syn th esis of 7,7′-Dicou m a r ol (7). 4,7-
Dihydroxycoumarin (356 mg, 2 mmol) was dissolved in 25 mL
of hot water to which an excess of 36% formalin (0.1 mL) was
added. The precipitate was produced immediately and con-
tinued refluxing for 0.5 h. After cooling, the solid was filtrated
and crystallized from ethanol to give the title compound in
(
9) Stanwell, C.; Ye, B.; Yuspa, S. H.; Burke, T. R., J r. Cell protein
cross-linking by erbstatin and related compounds. Biochem.
Pharmacol. in press.
8
1
1% yield: mp 326-328 °C; IR (KBr) 3327, 3105 (br), 1627,
(10) Mazumder, A.; Neamati, N.; Sommadossi, J . P.; Gosselin, G.;
Schinazi, R. F.; Imbach, J . L.; Pommier, Y. Effects of nucleotide
analogues on human immunodeficiency virus type 1 integrase.
Mol. Pharmacol. 1996, 49, 621-628.
597, 1572, 1355, 1312, 1246, 1158, 795 cm-1
) δ 10.51 (br, 2H), 7.74 (d, J ) 8.6 Hz, 2H), 6.80 (dd,
J ) 8.7, 2.2 Hz, 2H), 6.69 (d, J ) 2.2 Hz, 2H), 3.68 (s, 2H);
;
1
H NMR
(DMSO-d
6
(
11) Mazumder, A.; Wang, S.; Neamati, N.; Nicklaus, M.; Sunder,
S.; Chen, J .; Milne, G. W. A.; Rice, W. G.; Burke, T. R., J r.;
Pommier, Y. Antiretroviral agents as inhibitors of both human
immunodeficiency virus type 1 and protease. J . Med. Chem.
1996, 39, 2472-2481.
+
FABMS m/ z 369 (MH ). Anal. (C26
-[6-Oxo(1)ben zop yr a n o[4,3-b]-(1)10-h yd r oxyben zop y-
r a n -7-yl]-4-h yd r oxycou m a r in (36): prepared as previously
18 6
H O ) C, H.
3
1
5
described.
Cou m a r in tr im er 37: prepared as previously described.15
(12) J enkins, T. M.; Engleman, A.; Ghirlando, R.; Craigie, R. A soluble
active mutant of HIV-1 integrase. J . Biol. Chem. 1996, 271,
7
712-7718.
Ack n ow led gm en t. We thank Dr. R. H. A. Plasterk
for providing the plasmid encoding HIV-2 integrase and
Ms. Pamela Russ and Dr. J ames Kelley of the LMC for
mass spectral analysis.
(
13) PALLAS FRAME MODULE, a prediction tool of physiochemical
parameters, is supplied by CompuDrug Chemistry, Ltd., P.O.
Box 23196, Rochester, NY 14692.
(
14) Buckle, D. R.; Cantello, B. C. C.; Smith, H.; Spicer, B. A.
Antiallergic activity of 4-hydroxy-3-nitrocoumarins. J . Med.
Chem. 1975, 18, 391-394.
Refer en ces
(15) Arora, R. B.; Krishnaswamy, N. R.; Seshadri, T. R.; Seth, S. D.
S.; Sharma, B. R. Structure and anticoagulant activity of bridge-
substituted dicoumarols. J . Med. Chem. 1967, 10, 121-124.
(
1) A preliminary account of this work was presented. Zhao, H.;
Neamati, N.; Mazumder, A.; Pommier, Y.; Burke, T. R., J r.
Design, synthesis and structure-activity relationship of HIV-1
integrase inhibitors. 212th National ACS Meeting, Orlando, FL,
Aug. 25-29, 1996, MEDI 21.
(
16) Sullivan, W. R.; Huebner, C. F.; Stahmann, M. A.; Link, K. P.
Studies on 4-hydroxycoumarins. II. The condensation of alde-
hydes with 4-hydroxycoumarins. J . Am. Chem. Soc. 1943, 65,
2
288-2291.
(
2) De Clercq, E. Toward improved anti-HIV chemotherapy: Thera-
peutic strategies for intervention with HIV infections. J . Med.
Chem. 1995, 38, 2491-2517.
(
17) Litvan, F.; Stoll, W. G. On ring closure reactions of 3,3′-methylen-
bis-(4-hydroxycoumarin). Helv. Chim. Acta 1959, 42, 878-
8
86.
(3) Fesen, M. R.; Pommier, Y.; Leteurtre, F.; Hiroguchi, S.; Yung,
J .; Kohn, K. Inhibition of HIV-1 integrase by flavones, caffeic
J M960450V