The discovery of novel, potent and selective PDE5 inhibitors

Article abstract of DOI:10.1016/S0960-894X(01)00466-8

The design and synthesis of a novel scaffold for potent and selective PDE5 inhibitors are described. Compound 3a was more potent (PDE5 IC₅₀ = 0.31 nM) and selective (> 10,000-fold vs PDE1 and 160-fold selective vs PDE6) PDE5 inhibitor than sild

Full text of DOI:10.1016/S0960-894X(01)00466-8

Bioorganic & Medicinal Chemistry Letters 11 (2001) 2461–2464
The Discovery of Novel, Potent and Selective PDE5 Inhibitors
Yingzhi Bi,a,* Patrick Stoy,a Leonard Adam,b Bin He,b John Krupinski,b
Diane Normandin,b Ron Pongrac,b Laurie Seliger,b Andrew Watsonb
and John E. Macora
aDepartment of Discovery Chemistry, Bristol-Myers Squibb Pharmaceutical Research Institute, PO Box 5400,
Princeton, NJ 08543-5400, USA
bMetabolic and Cardiovascular Drug Discovery, Bristol-Myers Squibb Pharmaceutical Research Institute, PO Box 5400,
Princeton, NJ 08543-5400, USA
Received 9 May 2001; accepted 28 June 2001
Abstract—The design and synthesis of a novel scaffold for potent and selective PDE5 inhibitors are described. Compound 3a was
more potent (PDE5 IC50=0.31 nM) and selective (>10,000-fold vs PDE1 and 160-fold selective vs PDE6) PDE5 inhibitor than
sildenafil. # 2001 Elsevier Science Ltd. All rights reserved.
When a man is sexually stimulated, nitric oxide is
released from non-cholinergic, non-adrenergic neurons
in the penis. The nitric oxide released activates guanylyl
cylase, which produces cGMP. Increased levels of
cGMP lead to decreased intracellular calcium con-
centration in the cells of the corpus cavernosum, resulting in
relaxation of the smooth muscle of the penis. This relaxa-
tion results in enhanced arterial blood flow into the penis,
ultimately leading to an erection.1 Since phosphodiesterase
type 5 (PDE5) is the primary cGMP hydrolyzing enzy-
matic activity present in the corpus cavernosum, inhibition
of PDE5 elevates levels of cGMP, thereby potentiating the
signaling cascade that leads to an erection.2
have been noted, and their incidence increases with the
dose of the drug. Certain of these adverse events are
thought to be due to nonspecific inhibition of other
PDEs, specifically PDE1 and PDE6.5,6 Thus, the iden-
tification of more potent and more selective PDE5 inhi-
bitors is of substantial medicinal and commercial
interest. This report communicates the design and
Sildenafil (1, Viagra1, Fig. 1), the prototypical PDE5
inhibitor, has opened an entire field of drug discovery
focused on quality-of-life issues. The treatment of male
erectile dysfunction (ED) is one of those areas. It is not
surprising that significant interest in the discovery of
additional PDE5 inhibitors has emerged since the
introduction of Viagra1 as an efficacious orally active
agent for the treatment of ED.3,4 In spite of the efficacy
of 1 as a treatment for ED, there are notable drawbacks
associated with its use. Clinically significant adverse
effects such as nausea, headache, cutaneous flushing,
interactions with sources of NO and visual disturbances
*Corresponding author. Tel.: +1-609-818-3847; fax: +1-609-818-
3450; e-mail: yingzhi.bi@bms.com
Figure 1.
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00466-8

2462
Y. Bi et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2461–2464
synthesis of a novel template (3), which leads to very
potent and highly selective PDE5 inhibitors.
placed the key benzyl group in the middle ring, and
changed the imidazole to pyrazole.
Rotella and co-workers7 have described a series of N-3
substituted imidazoquinazolinones (2, Fig. 1) as potent
and selective PDE5 inhibitors. In this series, significant
improvement in PDE5 potency and PDE isozyme selec-
tivity compared to sildenafil was achieved by the addi-
tion of a benzyl moiety at the N-3 position of the
imidazole moiety of the imidazoquinazolinone template
(2). In an attempt to further develop the SARs for this
series, we designed a new template (3, Fig. 1) which
incorporated another nitrogen into the middle ring,
Commercially available 3-amino-4-carbethoxypyrazole
(4) was treated with p-methoxybenzyl (PMB) chloride
and sodium hydride in acetonitrile to give two regio-
isomers 5 and 6 (Scheme 1). The position of the p-
methoxybenzyl group in each regiomer was determined
based on NOE NMR studies on 5 and 6. Since both
isomers should ultimately give the same product after
deprotection, they were carried on together to their
respective pyrrazolopyridines (7 and 8) according to the
method of Sanghvi.8 However, only 7 appeared to be
Scheme 1. (a) PMBCl, NaH/CH3CN, 24 h, 71%; (b) CH2(CO2Et)2, NaOEt/HOEt, reflux, 20 h, 100%; (c) POCl3, reflux, 3 h, 55%; (d) NH3/HOEt,
rt, 7 days, 60%; (e) LiOH/H2O/THF/HOMe, 24 h, 78%; (f) (1) DCC/pentafluorophenol/DMF/EtOAc, 0 ꢀC, overnight; (2) NH3/THF, rt, overnight,
80%; (g) (1) 5[(3R)-(+)-(3-dimethylaminopyrrolidinyl)sulfonyl]-2-propoxybenzoyl chloride, KHMDS/THF, 3 h; (2) tBuOK/tBuOH, 80 ꢀC, 1 h,
22% for two steps; (h) (1) Pd/C, H2, 93%; (2) TFA, 60 ꢀC, 1 h, 60%; (i) benzylamine, DIEA/nPrOH, reflux, 4 h, ꢁ90%; (j) TFA, 60 ꢀC, 1 h, ꢁ80%.

Y. Bi et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2461–2464
2463
converted to its corresponding dichloro derivative (9)
under reflux in phosphorus oxychloride, since no
dichloro adduct from 8 was isolated or detected.
intermediates containing the p-methoxybenzyl protect-
ing group (13 and 15a) were potent PDE5 inhibitors
(PDE5 IC50’s=13 and 1.7 nM, respectively), suggesting
that there was an unexplored region of space where the
PMB moiety may be interacting with the enzyme. While
13 offered no PDE isozyme selectivity advantages versus
sildenafil, the p-fluorobenzyl substitution in 15a clearly
improved PDE1 selectivity (>104-fold) when compared
with sildenafil (140-fold). However, the selectivity of 15a
versus PDE6 was only slightly better than that of sildenafil.
Regiospecific displacement of the chlorine para to the
pyridine nitrogen with ammonia afforded 10. The
regioselectivity of this reaction was confirmed via X-ray
crystallographic studies on another crystalline derivative
of 9.9 The ester in 10 was converted to its corresponding
primary amide (12) via hydrolysis to acid 11 followed by
amide formation. The third ring of the heterocyclic
scaffold was generated by acylation of the amino group
with 5-[(3R)-(+)-(3-dimethylaminopyrrolidinyl)sulfo-
nyl]-2-propoxybenzoyl chloride,7 followed by base-
mediated cyclization to give pyrazolopyridopyrimidine
(13). (3R)-Dimethylaminopyrrolidine was chosen
because it gave Rotella and co-workers some of their
most potent PDE5 inhibitors. Removal of the chlorine
atom ortho to the pyridine nitrogen in 13 via catalytic
hydrogenation, followed by deprotection, afforded tem-
plate 14. Alternatively, displacement of the chlorine
ortho to the pyridine nitrogen with benzylamines affor-
ded 15, and subsequent deprotection yielded pyr-
azolopyridopyrimidines (3)10 with the desired benzyl
substitution pattern.
Derivatives 3a and 3b were prepared to complement the
SAR seen with the imidazoquinazolinone series of
PDE5 inhibitors.7 Namely, incorporation of an appro-
priately substituted benzyl group at R2 (Table 1) was
expected to lead to more potent and selective analogues.
Following the SAR seen by Rotella and co-workers,7
the PDE5 activity and selectivity for 3a and 3b was
substantially improved over the parent derivative (14).
The p-fluorobenzyl derivative (3a) was approximately 6-
fold more potent (PDE5 IC50=0.31 nM) than sildenafil,
whereas the p-methoxy-m-chlorobenzyl analogue
(PDE5 IC50=0.90 nM) was only slightly more potent
than sildenafil. However, the PDE isozyme selectivity
profiles of 3a and 3b were clearly superior to that of
sildenafil. Whereas sildenafil is only 140-fold selective
versus PDE1, both 3a and 3b were approximately
10,000-fold selective versus PDE1. The selectivity of 3a
and 3b versus PDE6 is especially noteworthy since the
vision disturbances associated with sildenafil use are
believed to be a result of its modest selectivity (8-fold)
versus PDE6. The approximate 200-fold selectivity
The in vitro activity of 3, 14, and selected intermediates
is summarized in Table 1. The pyrazolopyridopyr-
imidine scaffold clearly is a useful template for the pre-
paration of potent PDE5 inhibitors. Even the least
substituted analogue in the series (14) was similar to sil-
denafil in terms of potency and selectivity. Interestingly,
Table 1. PDE5 inhibition and isozyme selectivitiesa
Compd
R1
R2
IC50
PDE5 (nM)
IC50
Ratio 1/5
IC50
Ratio 2/5
IC50
Ratio 3/5
IC50
Ratio 4/5
IC50
Ratio 6/5
1
Sildenafil
1.6Æ0.5
140
>104
>105
3500
2600
8
3a
–H
0.31Æ0.14
>105
>105
>104
160
3b
–H
0.90Æ0.29
104
>104
>104
>104
180
13
14
–PMB
–H
—Cl
—H
13Æ1.4
2.1Æ1.2
130
400
1200
2000
1100
>104
310
3200
17
22
15a
–PMB
1.7Æ1.0
>104
>104
>104
2500
15
aAssays performed as described in ref 7; enzyme sources: PDE1: bovine heart; PDE2: rat kidney; PDE3: human platelet; PDE4: rat kidney; PDE5:
human platelet; PDE6: bovine retina; all IC50 determinations are averages based on three or more determinations.

2464
Y. Bi et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2461–2464
versus PDE6 for 3a and 3b represents a major advance
in the discovery of selective PDE5 inhibitors. To our
knowledge, compound 3a represents one of the most
potent and selective PDE5 inhibitors disclosed to date.
Should 3a be used in the treatment of ED, one would
expect a significantly reduced side-effect profile when
compared to sildenafil.
Drugs Future 1997, 22, 138. (b) Terrett, N. K.; Bell, A. S.;
Brown, D.; Ellis, P. Bioorg. Med. Chem. Lett. 1996, 6, 1819.
(c) 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.
In summary, we have identified a novel pyrazolopyr-
idopyrimidine template that provided PDE5 inhibitors
which are more potent and selective than sildenafil. Use
of an appropriately substituted benzylamino moiety
placed at the R2 region of the scaffold (Table 1) con-
ferred substantial improvement in both PDE5 potency
and selectivity in this series. One could hypothesize that
a similar substitution on the sildenafil molecule could
improve the selectivity of sildenafil (especially vs PDE1
and PDE6). In fact, a recent patent application exem-
plifies that type of sildenafil derivative.11 The improved
selectivity found in 3a may translate into an improved
PDE-related side-effect profile in vivo, based on experi-
ence to date with sildenafil. We are continuing to opti-
mize the activity of this series of molecules and will
report the results in due course.
5. Beavo, J. A. Physiol. Rev. 1995, 75, 725.
6. Scrip, October 9, 1997.
7. Rotella, D. P.; Sun, Z.; Zhu, Y.; Krupinski, J.; Pongrac, R.;
Seliger, L.; Normandin, D.; Macor, J. E. J. Med. Chem. 2000,
43, 1257.
8. Sanghvi, Y. S.; Larson, S. B.; Robins, R. K.; Revankar,
G. R. J. Chem. Soc., Perkin Trans. 1 1990, 2943.
9. Bi, Y.; Stoy, P.; Macor, J. Unpublished results.
10. The structures of all compounds were confirmed by NMR
1
and LC–MS. Compound 3a: H NMR (400 MHz, CD3OD) d
8.54 (d, J=2.4 Hz, 1H), 8.30 (s, 1H), 8.05 (dd, J=8.8, 2.4 Hz,
1H), 7.46 (dd, J=8.7, 5.4 Hz, 2H), 7.40 (d, J=8.8 Hz, 1H),
7.08 (t, J=8.7 Hz, 2H), 4.78 (s, 2H), 4.27 (t, J=6.3 Hz, 2H),
3.88 (m, 1H), 3.62–3.52 (m, 3H), 3.24 (dd, J=18.0, 8.1 Hz,
1H), 2.89 (s, 6H), 2.38 (m, 1H), 2.17 (m, 1H), 1.95 (m, 2H),
1.10 (t, J=7.4 Hz, 3H). MS (MH+) 621.35. 3b: 1H N MR
(400 MHz, CD3OD) d 8.50 (d, J=2.5 Hz, 1H), 8.19 (s, 1H),
8.07 (dd, J=9.1, 2.6 Hz, 1H), 7.45 (d, J=2.0 Hz, 1H), 7.43 (d,
J=9.1 Hz, 1H), 7.34 (dd, J=8.5, 2.0 Hz, 1H), 7.00 (d, J=8.6
Hz, 1H), 4.72 (s, 2H), 4.26 (t, J=6.6 Hz, 2H), 3.87 (s, 3H),
3.66–3.57 (m, 3H), 3.46 (m, 1H), 3.22 (dd, J=18.2, 8.1 Hz,
1H), 2.89 (s, 6H), 2.39 (m, 1H), 2.16 (m, 1H), 1.94 (m, 2H),
1.09 (t, J=7.3 Hz 3H). MS (MH+) 667.32 (100%), 669.30
(42%).
References and Notes
1. Andersson, K.-E.; Wagner, G. Physiol. Rev. 1995, 75, 191.
2. Gibson, A. Eur. J. Pharmacol. 2001, 411, 1.
3. (a) Martel, A. M.; Graul, A.; Rabasseda, X.; Castaner, R.
11. Bunnage, M. E.; Mathias, J. P.; Street, S. D. A.; Wood, A.
WO 9849166, 1998; Chem. Abstr. 1998, 129, 330741.