NMR-based discovery of lead inhibitors that block DNA binding of the human papillomavirus E2 protein

Article abstract of DOI:10.1021/jm9703404

The E2 protein is required for the replication of human papillomaviruses (HPVs), which are responsible for anogenital warts and cervical carcinomas. Using an NMR-based screen we tested compounds for binding to the DNA-binding domain of the HPV-E2 protein. Three classes of compounds were identified which bound to two distinct sites on the protein. Biphenyl and biphenyl ether compounds containing a carboxylic acid bind to a site near the DNA recognition helix and inhibit the binding of E2 to DNA. Benzophenone- containing compounds which lack a carboxylic acid group bind to the β- barrel formed by the dimer interface and exhibit negligible effects on the binding of E2 to DNA. Structure-activity relationships from the biphenyl and biphenyl ether compounds were combined to produce a compound [5-(3'- (3'',5''-dichlorophenoxy)-phenyl)-2,4-pentadienoic acid] with an IC₅₀ value of approximately 10 μM. This compound represents a useful lead for the development of antiviral agents that interfere with HPV replication and further illustrates the usefulness of the SAR by NMR method in the drug discovery process.

Full text of DOI:10.1021/jm9703404

3
144
J . Med. Chem. 1997, 40, 3144-3150
Expedited Articles
NMR-Ba sed Discover y of Lea d In h ibitor s Th a t Block DNA Bin d in g of th e
Hu m a n P a p illom a vir u s E2 P r otein
Philip J . Hajduk, J u¨ rgen Dinges, Gregory F. Miknis, Megan Merlock, Tim Middleton, Dale J . Kempf,
†
David A. Egan, Karl A. Walter, Terry S. Robins, Suzy B. Shuker, Thomas F. Holzman, and Stephen W. Fesik*
Pharmaceutical Products Division, Abbott Laboratories, Abbott Park, Illinois 60064-3500
X
Received May 23, 1997
The E2 protein is required for the replication of human papillomaviruses (HPVs), which are
responsible for anogenital warts and cervical carcinomas. Using an NMR-based screen, we
tested compounds for binding to the DNA-binding domain of the HPV-E2 protein. Three classes
of compounds were identified which bound to two distinct sites on the protein. Biphenyl and
biphenyl ether compounds containing a carboxylic acid bind to a site near the DNA recognition
helix and inhibit the binding of E2 to DNA. Benzophenone-containing compounds which lack
a carboxylic acid group bind to the â-barrel formed by the dimer interface and exhibit negligible
effects on the binding of E2 to DNA. Structure-activity relationships from the biphenyl and
biphenyl ether compounds were combined to produce a compound [5-(3′-(3′′,5′′-dichlorophenoxy)-
phenyl)-2,4-pentadienoic acid] with an IC50 value of approximately 10 µM. This compound
represents a useful lead for the development of antiviral agents that interfere with HPV
replication and further illustrates the usefulness of the SAR by NMR method in the drug
discovery process.
In tr od u ction
Ta ble 1. Initial Ligands for the E2 Protein Discovered Using
SAR by NMR
Papilloma viruses are a group of small DNA tumor
viruses that infect a wide variety of mammalian cells.
They are the causative agents of benign proliferative
1
lesions of the skin or mucosa, and certain strains have
been implicated in the development of cervical carcino-
mas and other malignancies.2 All of the papillomavi-
ruses encode a DNA-binding protein, E2, that regulates
the transcription of several viral genes and, in conjunc-
tion with the E1 protein, is required for viral repli-
cation.3
,3
-7
Many approaches have targeted the E2 protein for
developing antiviral agents against the human papil-
lomaviruses (HPVs). One strategy is to block the
expression of the HPV-E2 protein, which has been
demonstrated using antisense oligonucleotides targeted
a
Dissociation constants were derived from an analysis of the
changes in amide chemical shifts as a function of added compound.
Binding sites were determined by mapping the observed chemical
b
8
,9
at E2 mRNA.
Another approach is to target the E2
shift changes onto the structure of the protein as shown in Figure
c
2
.
Inhibition of DNA binding was determined at compound
protein itself. Truncated forms of the HPV-E2 protein
containing the DNA-binding domain (DBD) can act as
trans-activating repressors by blocking the homo-dimer-
concentrations of 1 mM using a filter-binding assay.
1
0,11
the discovery of small, novel therapeutic agents against
the human papillomaviruses, we used an NMR-based
approach called SAR by NMR (for structure-activity
relationships by nuclear magnetic resonance)15 to iden-
tify ligands for the DNA-binding domain of the E2
protein. Using this approach, small molecules which
bind to proteins can be identified in an NMR-based
screen by monitoring changes in the N/ H amide
chemical shifts of the protein upon the addition of
compounds. We have previously reported on the ap-
plication of the SAR by NMR method for the discovery
of high-affinity ligands for the FK506 binding protein15
and the catalytic domain of the matrix metalloprotein-
ization of E2.
In addition, oligonucleotides which
are recognized by the E2 protein can compete with DNA
1
2
binding and inhibit viral cell growth.
These data
suggest that the E2 protein can serve as a viable target
for the development of therapeutics against HPVs.
The X-ray structure of the DNA-binding domain of
bovine E2 when complexed to its cognate DNA has been
1
3
15
1
reported, and we have recently solved the NMR
1
4
structure of the HPV-31 E2 DBD. In order to aid in
†
Current address: Specialty Laboratories, Inc., 2211 Michigan
Avenue, Santa Monica, California 90404-3900.
*
Address for correspondence: Dr. Stephen W. Fesik, Abbott
Laboratories, D-47G, AP-10, 100 Abbott Park Rd, Abbott Park, IL
0064-3500. Phone: (847) 937-1201. Fax: (847) 938-2478. E-mail:
6
16
ase stromelysin. Here we describe the application of
the SAR by NMR technique to the discovery of com-
fesik@steves.abbott.com.
X
Abstract published in Advance ACS Abstracts, September 1, 1997.
S0022-2623(97)00340-3 CCC: $14.00 © 1997 American Chemical Society

Lead Inhibitors that Block DNA Binding
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 20 3145
1
5
F igu r e 1. [ N]HSQC spectra of E2 DBD in the absence of added ligand (black) and in the presence of (A) 3 mM 1 (red) and (B)
mM 3 (red). Some of the residues which exhibit significant chemical shift changes upon addition of ligand are labeled. Chemical
2
1
15
1
15
1
15
shifts were considered significant if the average weighted H/ N chemical shift difference (∆δ( H/ N) ) ∆δ( H) + ∆δ( N)/5) was
greater than 0.04 ppm.
F igu r e 2. Ribbons28 representation of the NMR structure of the DNA-binding domain of the E2 dimer from HPV-31.14 DNA
(
top) is also depicted from the X-ray structure of bovine HPV-E2 complexed with DNA.13 For clarity, the DNA was translated
from its original position. Colored in yellow are those residues whose amide chemical shifts were significantly perturbed upon
binding compound 1, while those colored in magenta are those significantly perturbed upon binding compound 3. Chemical shifts
1
15
1
15
1
15
were considered significant if the average weighted H/ N chemical shift difference (∆δ( H/ N) ) ∆δ( H) + ∆δ( N)/5) was greater
than 0.04 ppm at compound concentrations of 3 and 2 mM for 1 and 3, respectively.
pounds which bind to the DNA-binding domain of the
E2 protein and inhibit its interaction with DNA.
presence and absence of potential ligands. Three classes
of compounds were identified that bind weakly to two
distinct sites on the E2 protein (Table 1). Biphenylcar-
boxylic acids (e.g., 1) and biphenyl ether carboxylic acids
Id en tifica tion of In itia l Lea d Liga n d s
(
e.g., 2) caused significant chemical shift changes in the
To identify compounds that bind to the E2 DBD from
HPV strain 31b, we screened a library of small organic
backbone amides of residues 305-310, 312, 313, 358-
360, 364, and 371 of the protein. These residues are
located in the DNA recognition helix or the adjacent loop
(Figure 1A and yellow in Figure 2). The third class of
1
7
15
molecules by recording two-dimensional N-hetero-
1
5
1
nuclear single-quantum correlation ( N/ H HSQC)
spectra18 of uniformly N-labeled E2 DBD in the
15

3
146 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 20
Hajduk et al.
Sch em e 1. General Synthesis of Biphenylcarboxylates
Ta ble 2. Binding Data for Biphenylcarboxylate Compounds
no.
R
KD (mM)a
F igu r e 3. Ribbons28 representation of the NOE-based model
of the complex between E2 DBD and 1 (colored by atom type).
Shown in blue are side chains of E2 DBD which make NOE
contacts with the ligand. The secondary structural elements
are labeled as in Figure 2 for aid in visualization.
5
6
7
8
9
2′-CH3
3′-CF3
3′-NH2
3′-F
2.7
0.32
>10
1.6
3′-Cl
1.2
1
1
1
1
14
15
1
1
1
0
1
2
3
3′-NO2
4′-CH3
4′-CF3
4′-OCH3
4′-t-Bu
4′-OPh
4′-F
8.9
1.5
1.5
8.8
compounds, which contain a benzophenone moiety and
lack a carboxy functionality (e.g., 3), caused significant
changes in the amide chemical shifts of residues 296-
>10
3
01, 322, 328, 339-342, 344, 366, 367, 369, and 373
2.1
1.5
1.6
3.5
6
7
8
(Figure 1B and magenta in Figure 2). These compounds
4′-Cl
induced little or no changes in the chemical shifts of
residues comprising the DNA recognition helix. Thus,
the NMR data indicate that this third class of com-
pounds bind to residues located near the â-barrel formed
by the interface of the two monomer subunits (Figure
2′,4′-Cl2
3′-Cl,4′-F
3′,5′-CF3
3′,5′-Cl2
2′-OCH3
3′-CH3
4′-CONH2
4′-CN
19
3.4
2
2
2
2
24
2
2
0
1
2
3
>10
0.06
1.7
1.9
2
).
>10
>10
>10
4.5
5
6
The initial lead ligands from the three classes were
4′-OH
further tested in a filter-binding assay to evaluate their
ability to inhibit the binding of E2 to DNA. It was found
that 1 and 2 inhibit the binding of E2 to DNA at
concentrations of 1 mM (Table 1). These results were
consistent with the observed chemical shift perturba-
tions, since compounds which bind near the recognition
helix may sterically occlude complex formation. In
contrast, the compound (3) which binds near the â-bar-
rel region did not affect E2/DNA complex formation.
Thus, although this compound binds to the E2 DBD, it
does not inhibit the binding of DNA either by directly
blocking the DNA binding site or by disrupting the
homodimerization of E2.
27
4′-NH2
4′-N(CH3)2
4′-NO2
3′-Cl,4′-NO2
3′,4′-Cl2
3′-N(CH3)2
2
2
3
3
8
9
0
1
6.8
>10
>10
6.6
32
>10
a
Dissociation constants were derived from an analysis of the
changes in amide chemical shifts as a function of added compound.
Op tim iza tion of th e Bip h en ylca r boxyla tes
To explore the SAR of the biphenyl leads, a series of
biphenylcarboxylate analogs were synthesized via Su-
zuki coupling using solid phase parallel synthesis19 as
shown in Scheme I. Since the number of commercially
available boronic acids was very limited, the traditional
strategy of attaching p-bromobenzoic acid to resin
method A) was switched to Guiles’ method of reacting
aryl halides with an immobilized aryl boronic acid
method B), which had to be modified in the case of
the 3′-NMe2-substituted biphenylcarboxylate (method
To confirm the ligand binding site and to provide a
structural basis for the inhibition of DNA binding by
the biphenyl and biphenyl ether carboxylates, structural
studies were performed on the complex of E2 and 1
using two-dimensional isotope-edited NMR experi-
ments. A model of the complex based on 23 intermo-
lecular NOEs is shown in Figure 3. The compound
binds to a hydrophobic groove between the DNA-
recognition helix and the adjacent loop, and the biaryl
moiety makes hydrophobic contacts with residues
Leu306, Leu309, Leu313, Val358, Ile360, and Pro361
of the E2 protein. This location is consistent with the
observed amide chemical shift perturbations (Figure 2)
and indicates that the inhibition exhibited by this class
of compounds is due an occlusion of the DNA-binding
site on the E2 protein.
(
2
0
(
C). To obtain the maximum amount of SAR with a
minimum number of reagents, the members of the
Topliss scheme
21,22
for aromatic substitution were syn-
thesized. This set of compounds is based on an optimum
discrimination between hydrophobic (π), electronic (σ),
and steric (ES) effects of the substituents on the phenyl
ring. The dissociation constants for a variety of biphe-
nylcarboxylates were determined by NMR (Table 2).
Biphenyls containing polar substituents (e.g., 7, 24-
26) exhibited a significant decrease in binding affinity
relative to the starting unsubstituted biphenylcarboxylic

Lead Inhibitors that Block DNA Binding
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 20 3147
Ta ble 3. Binding Data for Biphenyl Ether Compounds
a
Dissociation constants were derived from an analysis of the
changes in amide chemical shifts as a function of added compound.
acid 1 (KD ) 2.5 mM). The most active compound in
the series was the 3′,5′-dichloro-substituted compound
F igu r e 4. Summary of the discovery of a 10 µM inhibitor of
HPV-E2 DBD starting from millimolar leads identified using
SAR by NMR.
2
1 (KD ) 0.06 mM, IC50 ) 0.15 mM), which exhibited
more than a 40-fold enhancement in activity. The 3′-
CF3-substituted analogue 6 was the second most active
compound found in the series, but the activity decreased
by a factor of 15 when an additional CF3 group in the
the rigid trans,trans-butadiene spacer (39), which ex-
hibited a KD of 0.35 mM and an IC50 of 75 µM in the
filter binding assay.
5
′-position was added (20).
Op tim iza tion of Bip h en yl Eth er Lea d s
Com bin in g th e SAR of Both Lea d Liga n d s
Analogs of the biphenyl ether compound 2 were also
pursued and tested for binding to the DBD of the E2
protein. Compounds from commercial sources with
shorter linkers between the biphenyl ether moiety and
the carboxylic acid group (e.g. 33-36) exhibited a
reduced binding affinity for E2 DBD (Table 3). On the
basis of these results, several compounds were synthe-
sized (Scheme 2) which increased the distance between
the biphenyl ether and carboxyl functionality. Increas-
ing the length of the spacer to four methylene units (38)
resulted in a 2-fold increase in affinity for the E2 DBD
When added to the E2 protein, the biphenyl and
biphenyl ether leads caused the same set of amide
crosspeaks to shift, indicating that both classes of
compounds bind to the same location on E2 DBD. Thus,
we reasoned that the SAR from both of these classes of
compounds could be combined into a single molecule.
Indeed, a compound (40) which incorporated both the
butadiene spacer and the dichloro substituents, as
synthesized according to Scheme 2, was found to be the
most potent compound of the series with an IC50 value
of 10 µM in a filter binding assay (Figure 4).23
(Table 3). Further improvements were obtained with
Sch em e 2. Synthesis of Biphenyl Ether Carboxylates 38, 39, and 40

3
148 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 20
Hajduk et al.
Con clu d in g Rem a r k s
TFA/H
2
O (5.0 mL) and agitation at room temperature for 2 h.
2 2
The resin was filtered off and washed with CH Cl (2.0 mL).
An important aspect of drug discovery is the identi-
fication of lead compounds that can be used as starting
points in the design of high-affinity ligands for proteins
or other therapeutic targets. Using conventional high-
throughput screening of large libraries of molecules, it
is not always possible to find leads of suitable potency.
In fact, using an E2-DNA binding assay, high-through-
put screening of more than 100 000 compounds from our
corporate library produced no inhibitors with activities
better than 10 µM. As an alternative to finding high-
affinity ligands from existing libraries, suitable lead
compounds may be constructed from structure-activity
relationships obtained on small, weakly binding mol-
ecules. A critical aspect of such an approach, however,
is the ability to obtain reliable SAR for compounds
which bind to the target with affinities in the millimolar
range. Most conventional screening methods are se-
verely limited for identifying compounds with such weak
affinities due to background signals from the high
concentrations of ligand required. However, the use of
The combined filtrates were concentrated in vacuo, and the
crude material was recrystallized from EtOAc/n-hexane to give
the products in >90% overall yield. MS and analytical reverse
phase HPLC analyses showed that the products were obtained
in purities > 90%.
Gen er a l Meth od for th e Cou p lin g of Ar yl Ha lid es w ith
Resin -Bou n d Ar ylbor on ic Acid 4b (Meth od B: 22-31).
Benzoyldioxaborinane was prepared via the procedure of
23
Takahashi, loaded onto Wang resin, and coupled with a series
2
0
of aryl bromides (method B). The products were obtained
as described above.
3
′-(N,N-Dim eth yla m in o)-4-bip h en ylca r boxylic Acid 32
Met h od C). 1,1′-Bis(diphenylphosphino)ferrocene (67 mg,
.12 mmol) and palladium(II)acetate (14 mg, 0.06 mmol) were
(
0
stirred in degassed toluene at room temperature for 10 min.
The solution was then added to a slurry of the phenyldiox-
aborinane on Wang resin 4b (1.33 g, 0.75 mmol/g theoretical
load), prepared via the procedure of Guiles,20 and 3-(N,N-
2
5
dimethylamino)phenyl trifluoromethanesulfonate (0.29 g,
.08 mmol) in toluene/EtOH (10:1, 11.0 mL). The mixture was
stirred for further 10 min, then 2 M aqueous Na CO (2.8 mL)
was added, and the reaction mixture was heated to reflux for
d. After filtration, washing of the resin, cleavage of the
1
2
3
2
1
1
5
5
N spectral editing allows the facile observation of the
N-labeled protein signals even in the presence of
product, and recrystallization as described above, 100 mg
(44%) of the product was obtained as pale yellow crystals.
5-(4′-P h en oxyp h en yl)-2,4-p en ta d ien oa te (37). LiHMDS
(15 mL, 15.13 mmol, 2.0 equiv) was added dropwise to a 30
mL solution of triethyl 4-phosphonocrotonate (3.35 mL, 15.13
millimolar (or higher) compound concentrations. The
1
5
use of [ N]HSQC spectra allows not only the detection
of weakly bound ligands, but also the ability to dif-
ferentiate between multiple binding sites on the protein
surface. In the case of E2 DBD, two distinct sites were
observed, and independent structure-activity relation-
ships could be developed at each location. Since ligand
binding at only one of the sites (near the DNA-recogni-
tion helix) inhibited E2/DNA binding, we were able to
focus our design efforts only on those compounds. One
compound, which combined the SAR obtained on two
lead series, inhibits the binding of the E2 protein to
DNA at low micromolar concentrations and serves as a
useful lead for an antiviral agent against the human
papillomavirus.
2
mmol, 2.0 equiv) under N at 0 °C. The resulting solution was
stirred at 0 °C for 30 min before adding 4-phenoxybenzalde-
hyde (1.5 g, 7.56 mmol. 1.0 equiv) in 30 mL of THF. After 3
h of stirring at 0 °C, the reaction was quenched by addition of
1 M HCl and extraction with ethyl acetate. The organic
4
solution was washed with water and dried over MgSO . The
solution was filtered, concentrated to a white solid, and carried
on without further purification. Purification of the esters could
be performed by flash column chromatography using 10-20%
diethyl ether/hexanes. Recrystallization using ethyl acetate/
petroleum ether afforded 1.8 g of colorless crystals: 85% yield;
1
H NMR (CDCl ) δ 7.47-7.38 (5H, m), 7.15-7.09 (1H, m),
3
7.05-6.94 (4H, m), 6.90-6.73 (2H, m), 5.95 (1H, d, J ) 15.3
1
3
Hz), 4.22 (2H, q, J ) 7.1 Hz), 1.33 (3H, t, J ) 7.1 Hz);
NMR (75.47 MHz, CDCl ) δ 167.0 (C), 158.21 (C), 156.45 (C),
44.57 (CH), 139.51 (CH), 131.00 (C), 129.78 (CH), 128.63
CH), 125.18 (CH), 123.74 (CH), 120.73 (CH), 119.29 (CH),
118.63 (CH), 60.12 (CH ), 14.28 (CH ); MS (M + H) 295, (M +
C
3
1
(
Exp er im en ta l Section
Solvents and other chemicals were reagent grade and used
2
3
1
+
4
NH ) 312.
as received. H-NMR spectra for analysis of synthesized
compounds were recorded at 300 MHz and expressed as ppm
downfield from tetramethylsilane (TMS) as an internal stan-
dard. Column chromatography was performed on Kiesel gel
5-(4′-P h en oxyp h en yl)-2,4-p en ta d ien oic Acid (38). The
ester 37 (1.0 g, 3.4 mmol, 1.0 equiv) was dissolved in 15 mL of
95% ethanol, and LiOH (81 mg, 3.4 mmol, 1.0 equiv) was
added. The solution was stirred at room temperature for 12
h. The reaction was concentrated under vacuum and the
residue taken up in water, acidified with 1 M HCl, and
extracted with ethyl acetate. The organic solution was washed
6
0 silica (EM Science). Wang resin was obtained from
Advanced ChemTech. Elemental analyses were performed by
Robertson Microlit Laboratories, Inc. (Madison, NJ ). Melting
points were obtained using a Thomas Hoover UniMelt. Com-
pounds 1 (Aldrich), 2, 3 (Aldrich), 33 (Lancaster), 34 (Aldrich),
4
with water, dried over MgSO , filtered, and concentrated to a
3
5 (Lancaster), and 36 (Maybridge) were obtained from our
thick gum. Purification was carried out by flash column
chromatography (2-10% methanol/chloroform + 0.5% acetic
acid). Evaporation of the solvent afforded the acid as a white
corporate library and used without further purification.
Gen er a l Meth od for th e Su zu k i Cou p lin g of Resin -
Bou n d p-Br om oben zoic Acid 4a (Meth od A: 5-21).
Wang resin-bound p-bromobenzoic acid 4a (0.3 g, 0.83 mmol/g
theoretical load), prepared by standard procedure (DCC,
solid: recrystallized from ethyl acetate/petroleum ether; mp
1
176-178 °C; H NMR (DMSO-d
6
) δ 7.58 (2H, d, J ) 8.5 Hz),
7.46-7.32 (3H, m), 7.28-7.19 (1H, m), 7.08-6.95 (6H, m), 6.00
(1H, d, J ) 15.3 Hz); 13C NMR (75.47 MHz, DMSO-d
) δ 167.49
DMAP, CH
preswollen in degassed DME (5.0 mL) for 10 min. Pd(PPh
29 mg, 0.025 mmol) was then added, and the mixture was
2
Cl
2
/THF (4:1), room temperature, 16 h), was
6
3
)
4
(C), 157.96 (C), 144.32 (CH), 138.99 (CH), 131.14 (C), 130.03
(
(CH), 128.88 (CH), 125.57 (CH), 123.83 (CH), 121.62 (CH),
+
agitated for 15 min under nitrogen. A solution of the ap-
propriate boronic acid (0.5 mmol) in EtOH (1.0 mL) was added,
and the mixture was agitated for further 10 min followed by
119.02, (CH), 118.41 (CH); MS (M + H) 267, (M + NH
4
) 284.
Anal. Calcd for C17
H, 5.28.
14 3
H O : C, 76.67; H, 5.29. Found: C, 76.95;
treatment with 2 M aqueous Na
slurry was refluxed for 20 h, cooled, filtered, washed succes-
sively with DME/H O (1:1), H O, 2 N HOAc, H O, DME,
2
CO
3
(0.5 mL). The resulting
5-(4′-P h en oxyp h en yl)p en ta n oic Acid (39). The dieneoic
ester 37 (500 mg, 1.7 mmol, 1.0 equiv) was dissolved in 20
mL of ethanol, and 10% Pd/C (50 mg) was added. The ester
was hydrogenated under atmospheric pressure (i.e. balloon)
for 3 h. After the mixture was purged with N , Celite was
2
added, and the solution was filtered and concentrated to a
thick oil/glass. The crude ester was taken up in 15 mL of
2
2
2
EtOAc, EtOAc/MeOH (1:1), MeOH, 2 × [0.5% diethyldithio-
carbamic acid, sodium salt trihydrate and 0.5% DIEA in
2
4
DMF, THF], and 3 × THF (10.0 mL each) and dried in vacuo.
The product was cleaved from the resin by addition of 95%

Lead Inhibitors that Block DNA Binding
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 20 3149
dioxane, and 3 M NaOH (680 mL, 2.04 mmol, 1.2 equiv) was
added. The solution was stirred at room temperature over-
night. The solution was concentrated, and the residue was
taken up in water, acidified with 1 M HCl, extracted with ethyl
E2 DBD were known,14 and assignments for the protein in the
complexed form were obtained by titrating 1 from 0.0 to 2.0
1
5
13
mM into a sample of N/ C-labeled E2 DBD and following
1
3
the changes in [ C]HSQC spectra as a function of added
1
3
acetate, washed with water, dried over MgSO
4
, filtered, and
ligand. NOEs between 1 and E2 were obtained from 2D C-
filtered26 NOESY spectra with mixing times ranging from 80
to 350 ms. A total of 23 intermolecular NOEs between the
ligand and the protein were unambiguously assigned.
concentrated to a white solid. The solid was further purified
by flash column (2-10% methanol/chloroform + 0.5% acetic
acid). Recrystallization from ethyl acetate/petroleum ether
1
afforded colorless needles: mp 58-59 °C; H NMR (DMSO-
A model of the complex was obtained by manually docking
d
6
) δ 7.38 (2H, t, J ) 7.4 Hz), 7.19 (2H, d, J ) 8.5 Hz), 7.11
14
1
onto the NMR-derived structure of the E2 DBD dimer
(1H, t, J ) 7.4 Hz), 6.97 (2H, t, J ) 7.46 Hz), 6.92 (2H, d, J )
followed by energy minimizations with the XPLOR 3.1 pro-
8
1
.5 Hz); 2.57 (2H, t, J ) 7.1 Hz), 2.24 (2H, t, J ) 7.1 Hz), 1.64-
27
gram on a Silicon Graphics computer. The XPLOR Frepel
1
3
.5 (4H, m); C NMR (75 MHz, DMSO-d ) δ 174.38 (C), 157.06
6
function was used to simulate van der Waals interactions with
(
(
(
C), 154.41 (C), 137.20 (C), 129.89 (CH), 129.66 (CH), 123.02
CH), 118.68 (CH), 118.17 (CH), 34.05 (CH ), 33.50 (CH ), 30.45
CH ), 24.08 (CH ); MS (M + H) 271, (M + NH ) 288. Anal.
: C, 75.53; H, 6.71. Found: C, 75.56; H,
-1
a force constant of 4.0 kcal mol and with atomic radii set to
2
2
0
.8 times their CHARMM values. A total of 23 intermolecular
+
4
2
2
distance restraints were were employed with a square well
Calcd for C17
.90.
-(3′-(3′′,5′′-Dich lor op h en oxy)p h en yl)-2,4-p en t a d ien -
H
18
O
3
-1
-2
potential (FNOE ) 50 kcal mol
Å ) and given lower and upper
6
bounds of 1.8 and 5.0 Å, respectively. Experimental hydrogen
bonds and dihedral angle restraints from the NMR structure
of E2 DBD were included in the calculations. After mini-
5
oic Acid (40). LHMDS (7.5 mL, 7.48 mmol, 2.0 equiv) was
added dropwise to a 20 mL solution of triethyl 4-phosphono-
crotonate (1.7 mL, 7.48 mmol, 2.0 equiv) under N at 0 °C.
2
The resulting solution was stirred at 0 °C for 30 min before
14
mization, the energetic penalty from the intermolecular dis-
tance violations was less than 0.2 kcal/mol, and only minor
changes were observed in the structure of E2 DBD (rmsd of
adding 3-(3′,5′-dichlorophenoxy)benzaldehyde (1.0 g, 3.74 mmol,
0
.60 Å for the C
R
trace).
1
.0 equiv) in 20 mL of THF. After 3 h of stirring at 0 °C, the
In h ibition Assa ys. Inhibition of binding of the E2 DBD
reaction was quenched by addition of 1 M HCl and extraction
with ethyl acetate. The organic solution was washed with
water and dried over MgSO . The reaction was filtered,
4
concentrated to a white solid, and carried on without further
purification. Purification of the esters could be performed by
flash column chromatography using 10-20% diethyl ether/
hexanes. Recrystallization using diethyl ether/hexane afforded
to DNA was assayed using a nitrocellulose filter binding assay.
Plasmid p1, containing the HPV11 genome (four E2 binding
sites) inserted into the BamH1 site of pSP65, was linearized
with EcoR1 and end-labeled with polynucleotide kinase to a
5
specific activity of approximately 2 × 10 dpm/fmol. Binding
reactions were performed at room temperature for 10 min in
2
5 mL of binding buffer (20 mM Tris-HCl, pH 7.5, 100 mM
1
.1 g, 80% yield. The resulting ester (700 mg, 1.9 mmol, 1.0
equiv) was dissolved in 15 mL of dioxane, and LiOH (51 mg,
.3 mmol, 1.2 equiv) was added. The solution was stirred at
NaCl, 5 mM DTT, 1 mM EDTA, 50 mg/mL BSA) containing 4
nM DNA, 400 nM E2 DBD, and various concentrations of
compounds added from DMSO stocks. DNA fragments bound
by E2 were captured by filtration through a MultiSCreen-HA
2
room temperature for 12 h. The reaction was concentrated
under vacuum, and the residue taken up in water, acidified
with 1 M HCl, and extracted with ethyl acetate. The organic
solution was washed with water, dried over MgSO
4
, filtered,
and concentrated to a thick gum. Purification was carried out
9
1
6-well nitrocellulose plate (Millipore) and washed once with
00 µL of binding buffer lacking BSA. Next, 30 µL of
Microscint O (Packard) was added to each well, and plates
were counted in a TopCount scintillation counter (Packard).
IC50 values were determined by plotting the compound con-
centration versus percent inhibition to determine the concen-
tration required for 50% inhibition.
by flash column chromatography (2-10% methanol/chloroform
+
0.5% acetic acid). Evaporation of the solvent afforded the
acid as a white solid. Recrystallization from ethyl acetate/
petroleum ether afforded 581 mg of a white solid: 89% yield;
1
mp 133-139 °C; H NMR (DMSO-d
6
) δ 7.38-7.32 (3H, m),
7
.21-7.02 (5H, m), 7.49-7.44 (2H, m), 6.04 (1H, d, J ) 15.3
Refer en ces
13
Hz); C NMR (75.47 MHz, DMSO-d
6
) δ 167.38 (C), 158.53 (C),
55.38 (C), 143.83 (CH), 138.53 (CH), 138.44 (C), 134.92 (C),
30.71 (CH), 127.83 (CH), 123.79 (CH), 122.97 (CH), 122.91
1
1
(
(
1) Beckter, T. M.; Stone, K. M.; Alexander, E. R. Genital Human
Papillomavirus Infection:A Growing Concern. Obstet. Gynecol.
Clin. North Am. 1987, 14, 389-396.
+
4
CH), 119.99 (CH), 117.98 (CH), 116.88 (CH); MS (M + NH )
(2) Pfister, H. Human Papillomaviruses and Skin Cancer. Cancer
Biol. 1992, 3, 263-271.
3
12 3 2
52. Anal. Calcd for C17H O Cl : C, 60.91; H, 3.60. Found:
C, 61.19; H, 3.73.
Detection of Liga n d Bin d in g Usin g NMR. The DNA-
(
3) Garcia-Carranca, A.; Gariglio, P. V. Molecular Aspects of Human
Papillomaviruses and their relation to uterine cervix cancer. Rev.
Invest. Clin. (Mexico) 1993, 45, 85-92.
binding domain of the HPV-31b E2 protein was expressed and
1
4
purified as previously described. NMR samples were com-
(4) Prakash, S. S.; Grossman, S. R.; Pepinsky, R. B.; Laimins, L.
A.; Androphy, E. J . Amino Acids Necessary for DNA Contact
and Dimerization Imply Novel Motifs in the Papillomavirus E2
Trans-activator. Genes Dev. 1992, 6, 105-116.
1
5
posed of uniformly N-labeled E2 DBD at 0.3 mM in a H
O (9:1) solution containing 20 mM phosphate, 10 mM DTT,
pH 6.5. Ligand binding was detected at 25 °C by acquiring
2
O/
D
2
(
5) Haugen, T. H.; Cripe, T. P.; Ginder, G. D.; Karin, M.; Turek, L.
P. Trans-activation of an Upstream Early Gene Promoter of
Bovine Papilloma Virus-1 by a Product of the Viral E2 Gene.
EMBO J . 1987, 6, 145-152.
1
5
18
sensitivity-enhanced [ N]HSQC spectra on 400 µL of 0.3 mM
HPV-E2 DBD in the presence and absence of added compound.
Compounds were added as solutions in perdeuterated DMSO.
A Bruker sample changer was used on a Bruker AMX500
spectrometer. Compounds were initially tested at 1.0 mM
each, and binding was determined by monitoring changes in
the [ N]HSQC spectra. Dissociation constants were obtained
for selected compounds by monitoring the chemical shift
changes as a function of ligand concentration. Data were fit
using a single binding site model. A least-squares grid search
(
6) Ustav, M.; Stenlund, A. Transient Replication of BPV-1 Requires
Two Viral Polypeptides Encoded by the E1 and E2 Open Reading
Frames. EMBO J . 1991, 10, 449-457.
(
7) Mohr, I. J .; Clark, R.; Sun, S.; Androphy, E. J .; MacPherson, P.;
Botchan, M. R. Targeting the E1 Replication Protein to the
Papillomavirus Origin of Replication by Complex Formation with
the E2 Transactivator. Science 1990, 250, 1694-1699.
8) Cowsert, L. M.; Fox, M. C.; Zon, G.; Mirabelli, C. K. In vitro
evaluation of phosphorothioate oligonucleotides targeted to the
E2-messanger-RNA of Papillomavirus-potential treatment for
genital warts. Antimicrob. Agents Chemother. 1993, 37, 171-
177.
1
5
(
D
was performed by varying the values of K and the chemical
shift of the fully saturated protein.
NOE-Ba sed Mod el of th e E2/1 Com p lex. NMR experi-
(
9) Crooke, S. T.; Mirabelli, C. K.; Ecker, D. J .; Cowsert, L. M.
Antisense Oligonucleotide Inhibition of Papillomavirus. PCT
International, WO 9320095, Oct 14, 1993, ISIS Pharmaceuticals.
ments were performed at 30 °C on a Bruker AMX500 spec-
trometer. NMR samples were composed of uniformly 1 N/ C-
5
13
labeled E2 DBD at 1.0 mM in a 100% D
2
6
2
O solution containing
.0 mM 1, 20 mM phosphate, 1 mM perdeuterated DTT, pH
.5 (uncorrected). Backbone and side chain assignments for
(
10) Androphy, E. J .; Barsoum, J . G. Inhibitors of the trans-activating
function of Papillomavirus E2 proteins, PCT International. WO
9212728, Aug 6, 1992, Biogen Inc.

3
150 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 20
Hajduk et al.
(
11) Pepinsky, R. B.; Androphy, E. J .; Corina, K.; Brown, R.;
Barsoum, J . Specific Inhibition of a Human Papillomavirus E2
Trans-Activator by Intracelluar Delivery of Its Receptor. DNA
Cell Biol. 1994, 13, 1011-1019.
(19) Frenette, R.; Friesen, R. W. Biaryl Synthesis via Suzuki Coupling
on Solid Support. Tetrahedron Lett. 1994, 35, 9177-9180.
(20) Guiles, J . W.; J ohnson, S. G.; Murray, W. V. Solid-Phase Suzuki
Coupling for C-C Bond Formation. J . Org. Chem. 1996, 61,
5169-5171.
(21) Topliss, J . G.; Martin, Y. C. Drug Design; Ariens, E. J ., Ed.;
Academic Press: New York, 1975; Vol. 5, pp 1-21.
(22) Hansch, C.; Unger, S. H.; Forsythe, A. B. Strategy in Drug
Design. Cluster Analysis as an Aid in the Selection of Substit-
uents. J . Med. Chem. 1973, 16, 1212-1222.
(23) This compound was not in fast exchange on the NMR time scale.
Therefore, chemical shifts could not be monitored as a function
of ligand concentration and dissociation constants could not be
obtained.
(24) Matsubara, H.; Steo, K.; Tahara, T.; Takahashi, S. 2,5-Diaryl-
1,3,2-dioxaborinanes - A New Series of Liquid Crystals. Bull.
Chem. Soc. J pn. 1989, 62, 3896.
(
(
(
12) Androphy, E. J .; Lowy, D. R.; Schiller, J . T. Inhibition of
Papillomavirus Expression with Oligonucleotides or Blocking
Proteine. Eur. Pat. Appl., EP 302758, Feb 8, 1989, New England
Medical Center Hospitals, Inc.
13) Hegde, R. S.; Grossman, S. R.; Laimins, L. A.; Sigler, P. B.
Crystal Structure at 1.7A of the Bovine Papillomavirus-1 E2
DNA-Binding Domain Bound to its DNA Target. Nature 1992,
3
59, 505-512.
14) Liang, H.; Petros, A. M.; Meadows, R. P.; Yoon, H. S.; Egan, D.
A.; Walter, K.; Holzman, T. F.; Robins, T.; Fesik, S. W. Solution
Structure of the DNA-Binding Domain of a Human Papilloma-
virus E2 Protein: Evidence for Flexible DNA-Binding Regions.
Biochemistry 1996, 35, 2095-2103.
(
15) Shuker, S. B.; Hajduk, P. J .; Meadows, R. P.; Fesik, S. W.
Discovering High-Affinity Ligands for Proteins: SAR by NMR.
Science 1996, 274, 1531-1534.
(25) Prepared according to Kates, S. A.; Daniels, S. B.; Albericio, F.
Automated Allyl Cleavage for Continuous-Flow Synthesis of
Cyclic and Branched Peptides. Anal. Biochem. 1993, 212,
303.
(
16) Hajduk, P. J .; Sheppard, G.; Nettesheim, D. G.; Olejniczak, E.
T.; Shuker, S. B.; Meadows, R. P.; Steinman, D. H.; Carrera, G.
M.; Marcotte, P. A.; Severin, J .; Walter, K.; Sith, H.; Gubbins,
E.; Simmer, R.; Holzman, T. F.; Morgan, D. W.; Davidsen, S. K.
Discovery of Potent Nonpeptide Inhibitors of Stromelysin Using
SAR by NMR. J . Am. Chem. Soc. 1997, 119, 5818-5827.
17) Approximately 2000 compounds were screened against the DNA-
binding domain of E2.
(26) Gemmecker, G.; Olejniczak, E. T.; Fesik, S. W. An Improved
1
2
Method for Selectively Observing Protons Attached to C in the
1
13
presence of H- C Spin Pairs. J . Magn. Reson. 1992, 96, 199-
204.
(27) Br u¨ nger, A. T. X-PLOR Version 3.1: A System for X-ray Crystal-
lography and NMR; Yale University Press: New Haven, CT,
1992.
(28) Carson, M. Ribbon Models of Macromolecules. J . Mol. Graph.
1987, 5, 103.
(
(
18) Kay, L. E.; Keifer, P.; Saarinen, T. Pure Absorption Gradient
Enhanced Heteronuclear Single Quantum Correlation Spectros-
copy with Improved Sensitivity. J . Am. Chem. Soc. 1992, 114,
1
0663-10665.
J M9703404