Synthesis and phosphodiesterase 5 inhibitory activity of new sildenafil analogues containing a phosphonate group in the 5′-sulfonamide moiety of phenyl ring

Article abstract of DOI:10.1016/j.bmcl.2004.02.040

Synthesis of new sildenafil analogues containing a phosphonate group in the 5^′-sulfonamide moiety of the phenyl ring, 12a-e, 13a-d, and 14a-d, and evaluation of their in vitro PDE5 inhibitory activity are disclosed. Enzyme assays revealed that

Full text of DOI:10.1016/j.bmcl.2004.02.040

Bioorganic & Medicinal Chemistry Letters 14 (2004) 2099–2103
Synthesis and phosphodiesterase 5 inhibitory activity of new
sildenafil analogues containing a phosphonate group in the
50-sulfonamide moiety of phenyl ring
Dae-Kee Kim,a,* Ju Young Lee,b Hyun-Ju Parkc and Khac Minh Thaic
aCollege of Pharmacy, Ewha Womans University, 11-1 Daehyun-dong, Seodaemun-gu, Seoul 120-750, South Korea
bLife Science Research Center, SK Chemicals, 600 Jungja-dong, Changan-gu, Suwon, Kyungki-do 440-745, South Korea
cCollege of Pharmacy, Sungkyunkwan University, Suwon, Kyungki-do 440-746, South Korea
Received 29 December 2003; revised 7 February 2004; accepted 10 February 2004
Abstract—Synthesis of new sildenafil analogues containing a phosphonate group in the 50-sulfonamide moiety of the phenyl ring,
12a–e, 13a–d, and 14a–d, and evaluation of their in vitro PDE5 inhibitory activity are disclosed. Enzyme assays revealed that
maximum 10-fold increase in PDE5 inhibitory activity, compared with sildenafil, was achieved by introducing a phosphonate group
in the 50-sulfonamide moiety. Docking model of (PDE5: 12d) complex shows that the PDE5-bound conformation of 12d matches
completely with that of sildenafil, while 12d is partially overlapped with cGMP with ethyl phosphonate group of 12d superimposed
onto the cyclic phosphate group of cGMP.
Ó 2004 Elsevier Ltd. All rights reserved.
Male erectile dysfunction (MED) is a common and
important medical problem, and over half of men at
40–70 years of age are reported to suffer from erectile
dysfunction according to a random community-based
sample study.1;2 Sildenafil citrate3 (Viagra; Fig. 1) is the
first orally effective phosphodiesterase type 5 (PDE5)
inhibitor available for the treatment of MED, and its
advent has spurred significant interest in the develop-
ment of additional PDE5 inhibitors.4 Despite the effi-
cacy of sildenafil as a treatment for MED, clinically
significant adverse side effects such as nausea, headache,
facial flushing, and visual disturbances have been
reported, and their incidence is dose-dependent. Certain
of these side effects are thought to be associated
with nonspecific inhibitions of other PDEs, specifically
PDE1, PDE3, and PDE6.5;6 Therefore, the search for
more potent and selective PDE5 inhibitors with fewer
PDE-related side effects and greater efficacy is of pri-
mary interest. In the earlier work of sildenafil series, it
was suggested that the 50-sulfonamide moiety of the
phenyl ring might reproduce the role of the cyclic
phosphate of cGMP in binding the enzyme active site,
thus leading to high and selective inhibition of PDE5.3c
In addition, it has been proposed by us that introduction
of a phosphonate group in the 50-sulfonamide moiety of
the phenyl ring might enhance PDE5 inhibitory activity,
probably by mimicking the role of the phosphate of
cGMP. In this paper, synthesis of new sildenafil ana-
logues containing a phosphonate group on the phenyl
ring, 12a–e, 13a–d, and 14a–d, evaluation of their in
vitro PDE5 inhibitory activity, and molecular modeling
study of 12d are disclosed.
O
N
Me
N
N
O
EtO HN
N
N
HN
H2N
N
O
O2S
N
O
O
.
citrate
O
OH
P
N
-
Me
O
cGMP
Sildenafil (Viagra)
Figure 1. Structures of sildenafil and cGMP.
Since it has been demonstrated that a n-propoxy group
at the 20-position of the phenyl ring is essential for high
PDE5 inhibitory activity,7a we decided to focus only on
the n-propoxy derivatives. Requisite piperidine deriva-
tives, 4c–e and 5c–e, were readily prepared according to
Keywords: PDE5 inhibitor.
* Corresponding author. Tel.: +82-2-3277-3025; fax: +82-2-3277-2467;
e-mail: dkkim@ewha.ac.kr
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.02.040

2100
D.-K. Kim et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2099–2103
a
b
BnN
O
BnN
OSO2CF3
BnN
PO(OEt)2
1
2
3
c
d
HN
HN
PO(OEt)2
e
c
BnN
BnN
X
PO(OEt)2
X
PO(OH)2
( )n
( )n
5a−e
4a−e
6
c
a: X = N, n = 1
b: X = N, n = 2
c: X = CH, n = 0
d: X = CH, n = 1
e: X = CH, n = 2
BnN
d
CHO
PO(OEt)2
7
8
Scheme 1. (a) LDA, DME; then PhN(SO2CF3)2; (b) diethyl phosphite, cat. Pd(PPh3)4, DMF; (c) Pd/C, H2 (1 atm), MeOH, rt; (d) CH2[PO(OEt)2]2,
NaH, toluene; (e) 4 N aqueous HCl, reflux.
Scheme 1 from the intermediates, 3, 6, and 8. Com-
pound 3 was prepared from 1-benzyl 4-piperidone (1) in
two steps, formation of the enol triflate 2 and Pd-cata-
lyzed coupling with diethyl phosphite, by employing the
literature procedure.8 Olefinations of 1 and 79 using
tetraethyl methylenediphosphonate (CH2[PO(OEt)2]2)
and sodium hydride (NaH) in toluene worked efficiently
at room temperature to give the a;b-unsaturated prod-
ucts, 6 and 8.10 Simultaneous removal of benzyl group
and reduction of the double bond in 3, 6, and 8,
occurred under a typical hydrogenation condition to
afford the piperidine derivatives 4c–e, respectively.
Treatment of the diethyl phosphonates 4c–e with 4 N
aqueous HCl solution under reflux produced the corre-
sponding phosphonic acids 5c–e. Diethyl piperazinyl
phosphonate derivatives 4a–b were obtained by apply-
ing the reported synthetic methods,10;11 and converted to
the phosphonic acids 5a–b as described in the prepara-
tion of 5c–e. All the target compounds, 12a–e, 13a–d,
and 14a–d, were synthesized in two straightforward
manners from the known chlorosulfonyl derivatives, 9
and 10,7b as depicted in Scheme 2. Mono-ethyl phos-
phonate compounds 12a–e were prepared in two steps
from compound 9. First, coupling reaction of 9 with an
appropriate cyclic amine 4a–e in EtOH readily produced
the sulfonamides 11a–e in 75–97% yields. Subsequent
cyclization of compounds 11a–e and the concomitant
partial hydrolysis of the diethyl phosphonate group
were efficiently affected under basic conditions at reflux
using a mixture of aqueous 2 N NaOH solution and
EtOH (2:1, v/v), and the corresponding mono-ethyl
phosphonate 12a–e were produced in 71–85% yields.
Diethyl phosphonate and phosphonic acid derivatives,
13a–d and 14a–d, were prepared in 61–90% yields by
simply mixing compound 10 with a cyclic amine, 4a–d or
HCl salt of 5a–d, in the presence of Et3N in a protic
O
Me
H2N
O
O
N
Me
n-PrO
H2N
O
N
N
n-PrO
a: X = N, n = 1
b: X = N, n = 2
c: X = CH, n = 0
d: X = CH, n = 1
e: X = CH, n = 2
N
H
N
a
N
H
O2S
N
O
P
OEt
OEt
SO2Cl
9
X
( )n
HN
11a−e
X
PO(OR)2
( )n
b
4a−e: R = Et
5a−d: R = H
O
Me
O
N
12: R1 = Et, R2 = H
13: R1 = R2 = Et
Me
N
n-PrO HN
N
N
n-PrO HN
N
N
14: R1 = R2 = H
a: X = N, n = 1
b: X = N, n = 2
c: X = CH, n = 0
d: X = CH, n = 1
e: X = CH, n = 2
c
O2S
N
O
P
OR1
OR2
SO2Cl
10
X
( )n
12a−e,13a−d, 14a−d
Scheme 2. (a) Et3N, EtOH, rt (for 5a–e); (b) 2 N NaOH in H2O, EtOH, reflux; (c) Et3N, DMF/H2O (for 5a–d), or EtOH (for 4a–d), rt.

D.-K. Kim et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2099–2103
Table 2. In vitro PDE inhibitory activity of compound 12da
2101
solvent such as EtOH or DMF/water mixture. Due to
the solubility problem of 5a–dÆHCl salts in an alcoholic
solvent, it was essential to use very polar system such
as a mixture of DMF and water.
Compound
IC50 (nM)
PDE5
PDE1
PDE3
PDE6
12d
156
(918)b
739
2600
(>15,000)c
14,600
0.17
(1)
2.58
(15)d
19.77
(11)d
PDE5 inhibitory activity was assayed in vitro using [3H]-
cGMP SPA kit,12 and the IC50 values for the target
compounds, 12a–e, 13a–d, and 14a–d, were determined
from concentration–response curves.13 As shown in
Table 1, most of the compounds turned out to be about
2–10-fold more potent PDE5 inhibitors than sildenafil.
Concerning the structure–activity relationship, three as-
pects are worthy of comments. First, the mono-ethyl
phosphonates 12a–e were found to exhibit higher PDE5
inhibitory activity than the diethyl phosphonate and
phosphonic acid derivatives, 13a–d and 14a–d. All the
mono-ethyl phosphonate derivatives 12a–e were 6–10-
fold more potent than sildenafil, and 12a and 12d proved
to be most active among them. Second, there seemed to
be no activity difference between the piperazinylsulf-
onamide (X ¼ N) and the corresponding piperidinyl
(X ¼ CH) derivatives (e.g., 12a–b vs 12d–e). Thirdly, it
was obvious that the PDE5 inhibitory activity was
affected by the length (n) of the methylene chain linking
the phosphonate group. Compounds with one methyl-
ene group (n ¼ 1) showed higher activity than the cor-
responding derivatives with no (n ¼ 0) or longer (n ¼ 2)
chain, regardless of the variations in the phosphonate
(R) and cyclic amine part (X).
Sildenafil
1.78
(1)
(415)b
(>8000)c
a PDEs were obtained from bovine heart (PDE1), rabbit platelet
(PDE3 and PDE5), and bovine retina (PDE6), respectively, and
assayed using [3H]-cGMP or [3H]-cAMP SPA kit. IC50 values were
determined from the logarithmic concentration–inhibition curve. The
value is the mean from three independent experiments.
b IC50 ratio of PDE1/PDE5.
c IC50 ratio of PDE3/PDE5.
d IC50 ratio of PDE6/PDE5.
For the docking analysis of 12d, the homology model of
PDE5 was used. Protein modeling and docking experi-
ments have been performed with the Sybyl 6.9 software
package (Tripos, Inc., St. Louis, MO, USA) on SGI
Octane II workstation. Protein sequence of human
cGMP-specific 30,50-cyclic phosphodiesterase 5 (PIR
Unique id: JW0106) was downloaded from the Pro-
tein Information Resource (http://pir.georgetown.edu/
pirwww/pirhome.shtml). The amino acid sequences,
from Met578 through Gln859, were selected for
homology modeling. Homology modeling was based on
known crystal structures of PDE4 (PDB entry: 1F0J,
1OYN, and 1MKD),14 and was performed using an
automatic web server, Geno-3D15 (http://geno3d-pbi-
l.ibcp.fr). Additional modeling and minimization were
Compound 12d showing the highest PDE5 inhibitory
activity was chosen, and its inhibitory activity toward
PDE1, PDE3, and PDE6 was determined, and its se-
lectivity (PDE1/PDE5, PDE3/PD5, PDE6/PDE5) over
these isozymes was compared with that of sildenafil. As
shown in Table 2, there were small increases in its
selectivity over PDE1 (2.2-fold), PDE3 (1.9-fold), and
PDE6 (1.4-fold) compared with those of sildenafil.
performed with the Composer and Base module of Sybyl
6.9. Reference protein coordinate was prepared by
docking of sildenafil into the active site of PDE5
homology model, taking into account the binding mode
determined by X-ray crystallography.16 X-ray data re-
veals that sildenafil binds into the active site pocket
mainly by various hydrophobic interactions between the
drug and protein, and with two hydrogen bonds. The
Table 1. In vitro PDE5 inhibitory activity of compounds, 12a–e, 13a–
d, and 14a–da
Compound
IC50 (nM)
PDE5
12a
12b
0.18
0.28
0.30
0.17
0.29
0.58
1.05
0.41
0.34
0.28
0.94
0.70
0.32
1.78
12c
12d
12e
13a
13b
13c
13d
14a
14b
14c
14d
Sildenafil
a PDE5 was prepared from rabbit platelet and assayed using [3H]-
cGMP SPA kit. IC50 values were determined from the logarithmic
concentration–inhibition curve. The value is the mean from three
independent experiments.
Figure 2. Docked alignment of cGMP (gray carbon), 12d (magenta
carbon), and sildenafil (green carbon) within the active site of PDE5.
Colors of other atoms are red (oxygen), blue (nitrogen), yellow (sul-
fur), and orange (phosphorus).

2102
D.-K. Kim et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2099–2103
hydrogen bonds are: between carbonyl oxygen of purine
ring of sildenafil and Ne2 of Gln 817, and between
adjacent NH group of sildenafil and Oe1 of Gln 817.16b
The ligands, sildenafil, cGMP, and 12d were flexibly
docked into the reference protein structure, respectively,
by using the FlexX program.17 As shown in Figure 3A,
FlexX performed well in reproducing the X-ray pose of
sildenafil. To compare the binding features of ligands,
the docked conformer of each ligand with top rank was
retrieved from FlexX results, and overlaid in the active
site of PDE5 (Fig. 2).17 The PDE5-bound conformation
of 12d matches completely with that of sildenafil, while
12d is partially overlapped with cGMP with ethyl
phosphonate group of 12d superimposed onto the cyclic
phosphate group of cGMP. To examine the binding
mode of cGMP (Fig. 3B), the molecule is embedded
inside the active site pocket, and positioned in line with
the polar amino acid residues, Asn 661 through Ser 663,
with forming various hydrogen bonding interactions.
For 12d (Fig. 3C), the purine ring moiety forms two
hydrogen bonds with Gln 817, just like sildenafil, and
additional hydrogen bond between pyrazolyl nitrogen
and OH of Tyr 612. The oxygen atoms of ethyl phos-
phonate group in 12d also forms two hydrogen bonds
with Asn 662, which shows the agreement with our
assumption that the phosphonate group of 12d may
substitute for the phosphate of cGMP in the interaction
with PDE5. In the binding of 12d into PDE5, the ligand
appears to act like a hybrid of sildenafil and cGMP by
utilizing both purine ring and phosphonate moiety for
formation of various hydrogen bonds with PDE5. The
resulting tight interactions between 12d and enzyme
would be related to the experimental data showing that
12d has more potent PDE5 inhibitory activity than
sildenafil.
In conclusion, our study on new sildenafil analogues
demonstrated that maximum 10-fold increase in PDE5
inhibitory activity, compared with sildenafil, could be
achieved by introducing a phosphonate group in the 50-
sulfonamide moiety. The piperidinyl derivative 12d,
which showed the highest PDE5 inhibitory activity,
has been selected for further biological investigations.
References and notes
1. (a) Saenz de Tejada, I. Contemp. Urol. 1992, 7, 52; (b)
Andersson, K.-E.; Wagner, G. Physiol. Rev. 1995, 75, 191.
2. Feldman, H. A.; Goldstein, I.; Hatzichristou, D. G.;
Krane, R. J.; McKinlay, J. B. J. Urol. 1994, 151, 54.
3. (a) Moreland, R. B.; Goldstein, I.; Kim, N. N.; Traish, A.
Trends Endocrinol. Metab. 1999, 10, 97; (b) Martel, A. M.;
Graul, A.; Rabasseda, X.; Castaner, R. Drugs Fut. 1997,
22, 138; (c) Terrett, N. K.; Bell, A. S.; Brown, D.; Ellis, P.
Bioorg. Med. Chem. Lett. 1996, 6, 1819; (d) Boolell, M.;
Allen, M. J.; Ballard, S. A.; Gepi-Attee, S.; Muirhead,
G. J.; Naylor, A. M.; Osterloh, I. H.; Gingell, C. Int. J.
Urol. Res. 1996, 8, 47.
4. (a) Eardley, I. Exp. Opin. Invest. Drugs 1997, 6, 1803; (b)
Garcia-Reboll, L.; Mulhall, J. P.; Goldstein, I. Drugs
Aging 1997, 11, 140; (c) Truss, M. C.; Stief, C. G. Drugs
Today 1998, 34, 805; (d) Rotella, D. P. Drugs Fut. 2001,
26, 153.
5. Beavo, J. A. Physiol. Rev. 1995, 75, 725.
6. Scrip 1997, October 9.
^
7. (a) Dumaıtre, B.; Dodic, N. J. Med. Chem. 1996, 39, 1635;
(b) Bell, A. S.; Brown, D.; Terrett, N. K. U.S. Patent
5,346,901, 1994.
8. Holt, D. A.; Erb, J. M. Tetrahedron Lett. 1989, 30, 5393.
9. Alfaro-Lopez, J.; Okayama, T.; Hosohata, K.; Davis, P.;
Porreca, F.; Yamamura, H. I.; Hruby, V. J. J. Med. Chem.
1999, 42, 5359.
10. Hoyle, W.; Howarth, G. A.; Brundish, D. E.; Kane, P. D.;
Walker, C. V.; Hayler, J.; Fullerton, J. D.; Smith, G. P.;
Wathey, W. B.; Le Grand, D. M; Allen, M. C.; Herold, P.
WO 9629327, 1996.
11. Kametani, T.; Kigasawa, K.; Hiiragi, M.; Wakisaka, K.;
Haga, S.; Sugi, H.; Tanigawa, K.; Suzuki, Y.; Fukawa, K.;
Irino, O.; Saita, O.; Yamabe, S. Heterocycles 1981, 16,
1205.
Figure 3. Comparison of binding poses of sildenafil (A), cGMP (B),
and 12d (C) in the active site of PDE5 (homology model), generated by
FlexX. The amino acid residues, which form hydrogen bonds within
the binding site are represented in line form, and the ligands are ren-
dered in capped stick. Yellow dotted lines are hydrogen bonding
ꢀ
interactions (<2.5 A).

D.-K. Kim et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2099–2103
2103
12. Determination of PDE1, PDE3, PDE5, and PDE6 inhib-
itory activity. Bovine heart PDE1 was purchased from
Sigma (St. Louis, MO, USA). PDE3 and PDE5 were
prepared from the rabbit platelet using the method
described by Hidaka et al. with minor modifications
(Biochim. Biophys. Acta. 1976, 429, 485). Fresh rabbit
whole blood was centrifuged at 360g to obtain the platelet-
rich plasma (PRP). Platelets were isolated from PRP by
centrifugation at 1200g, sonicated (20s/mL) in 50mM tris-
HCl buffer (pH 7.4) containing 1 mM MgCl2, and then
centrifuged at 40,000g for 2 h at 4 °C. The supernatant was
loaded on the DEAE-cellulose column with a bed volume
of 35 mL (Sigma) pre-equilibrated with equilibration
buffer (50mM tris-acetate containing 3.75 mM 2-mercap-
toethanol, pH 6.0). After the column was washed with
60mL of equilibration buffer, PDE3, and PDE5 were
eluted using a continuous gradient of 0–600 mM sodium
acetate in equilibration buffer with a total volume of
60mL. The bovine retina PDE6 was prepared using the
method described by Ballard et al. (J. Urol. 1998, 159,
2164) with minor modifications. Bovine retinas were
minced and homogenized in the homogenization buffer
(20mM HEPES containing 0.25 M sucrose, 1 mM EDTA,
1 mM phenylmethyl sulfonylfluoride (PMSF), pH 7.2)
using a Polytron PT 10/35 homogenizer (Kinematica
AG, Switzerland) at 5000 rpm with two bursts for 10 s.
The homogenate was then centrifuged at 40;000g for
60min at 4 °C. The supernatant was recovered and filtered
through 0.2 lm filter. The filtered sample was loaded on
the Hitrap Q column with a bed volume of 5 mL
(Pharmacia, Uppsala, Sweden) pre-equilibrated with
20mM HEPES buffer (pH 7.2) containing 1 mM EDTA
and 0.5 mM PMSF. The column was then washed with
25 mL of equilibration buffer. PDE6 was eluted using a
continuous gradient of 0–600 mM NaCl in equilibration
buffer with a total volume of 60mL. Fractions (1.0mL
each) collected at a flow rate of 60mL/h were character-
ized for cGMP (PDE5 and PDE6) or cAMP (PDE3)
hydrolytic PDE activities as described below. Fractions
comprising the main peaks of cGMP hydrolytic PDE
activity were pooled and stored at )20 °C in 50% glycerol
until the enzyme assay. Enzymatic activity was determined
using a PDE scintillation proximity assay (SPA) kit
(Amersham Pharmacia Biotech, Buckinghamshire, UK)
according to the protocol supplied by the manufacturer.
The reaction buffer contained [3H]-cGMP (5 lCi/mL) or
[3H]-cAMP (5 lCi/mL), 1.7 mM EGTA, and 8.3 mM
MgCl2 in 50mM Tris–HCl buffer (pH 7.5). After PDE
was added to the reaction buffer, the mixtures were
incubated at 30 °C for 30min. The reaction was then
stopped by the addition of 50 lL of SPA beads, and the
radioactivity was counted on the liquid scintillation
counter (Tri-Carb 1500, Packard Inc., Meriden, CT,
USA) after each sample was settled for 20min. For the
inhibitor studies, sildenafil and test compounds were
dissolved in DMSO and diluted with distilled water. The
final concentration of DMSO was less than 0.2% (v/v). All
the inhibition experiments were conducted under the
conditions where the level of cGMP or cAMP hydrolysis
did not exceed 15%, and the product formation increased
linearly with time and amount of enzyme. IC50 was defined
as the concentration of compounds to produce a 50%
inhibition of enzyme activity and calculated by quantal
probit analysis in Pharmacological Calculation System.
13. Tallarida, R. J.; Murray, R. B. Manual of Pharmacological
Calculations with Computer Program; Springer: New
York, 1987; pp 31–35.
14. (a) Xu, R. X.; Hassell, A. M.; Vanderwall, D.; Lambert,
M. H.; Holmes, W. D.; Luther, M. A.; Rocque, W. J.;
Milburn, M. V.; Zhao, Y.; Ke, H.; Nolte, R. T. Science
2000, 288, 1822; (b) Lee, M. E.; Markowitz, J.; Lee, J.-O.;
Lee, H. FEBS Lett. 2002, 530, 53; (c) Huai, Q.; Wang, H.;
Sun, Y.; Kim, H. Y.; Liu, Y.; Ke, H. Structure 2003, 11,
865.
15. (a) Combet, C.; Jambon, M.; Deleage, G.; Geourjon, C.
Bioinformatics 2002, 18, 213; (b) Geourjon, C.; Combet,
C.; Blanchet, C.; Deleage, G. Protein Sci. 2001, 10, 788.
16. (a) Sung, B.-J.; Hwang, K. Y.; Jeon, Y. H.; Lee, J. I.; Heo,
Y.-S.; Kim, J. H.; Moon, J.; Yoon, J. M.; Hyun, Y.-L.;
Kim, E.; Eum, S. J.; Park, S.-Y.; Lee, J.-O.; Lee, T. G.;
Ro, S.; Cho, J. M. Nature 2003, 425, 98; (b) Brown, D. G.;
Groom, C. R.; Hopkins, A. L.; Jenkins, T. M.; Kamp, S.
H.; Oꢀgara, M. M.; Ringrose, H. J.; Robinson, C. M.;
Taylor, W. E. WO 03038080, 2003.
17. Preparation of molecular structures. The structures of
cGMP, sildenafil, and 12d were prepared in MOL2 format
using the sketcher module and Gasteiger–Huckel charges
were assigned to the ligand atoms. The structures of
molecules are optimized by energy minimization, and
molecular dynamics using simulated annealing method.
The lowest-energy conformer of each molecule was
selected and stored in database. Preparation of target
protein structure and flexible docking. Initially, the refer-
ence molecule, sildenafil, was manually docked into the
homology model of PDE5, based on the crystallograph-
ically-determined binding mode of sildenafil in the active
site of PDE5. The resulting sildenafil-PDE5 docked
complex was used as a receptor for flexible docking of
cGMP, 12d, and sildenafil. The active site was defined as
ꢀ
all the amino acid residues enclosed within 6.5 A radius
sphere centered by the bound ligand, sildenafil. The
docking and subsequent scoring were performed using
the default parameters of the FlexX programs implanted
in the Sybyl 6.9. For the docking of ligand into the target
active site, the main settings are 1000 solutions per
iteration during the incremental construction algorithm.
Final scores for all FlexX solutions were calculated by a
standard scoring function, and used for database ranking.