8-Aryl xanthines potent inhibitors of phosphodiesterase 5

Article abstract of DOI:10.1016/S0960-894X(02)00480-8

In clinical studies, several inhibitors of phosphodiesterase 5 (PDE5) have demonstrated utility in the treatment of erectile dysfunction. We describe herein a series of 8-aryl xanthine derivatives which function as potent PDE5 inhibitors with, in many cases, high levels of selectivity versus other PDE isoforms.

Full text of DOI:10.1016/S0960-894X(02)00480-8

Bioorganic & Medicinal Chemistry Letters 12 (2002) 2587–2590
8-Aryl Xanthines Potent Inhibitors of Phosphodiesterase 5
Ruth Arnold,a David Beer,a Gurdip Bhalay,a Urs Baettig,a Stephen P. Collingwood,a,*
Sarah Craig,a Nicholas Devereux,a Andrew Dunstan,a Angela Glen,a Sylvie Gomez,a
Sandra Haberthuer,a Trevor Howe,a Stephen Jelfs,a Heinz Moser,a Reto Naef,b
Paul Nicklin,a David Sandham,a Rowan Stringer,a Katharine Turner,a
Simon Watsona and Mauro Zurinib
aNovartis Horsham Research Centre, Wimblehurst Road, Horsham, West Sussex RH12 5AB, UK
bNovartis Pharma AG, Lichtstrasse 35, CH-4056 Basel, Switzerland
Received 22 April 2002; revised 23 May 2002; accepted 18 June 2002
Abstract—In clinical studies, several inhibitors of phosphodiesterase 5 (PDE5) have demonstrated utility in the treatment of erectile
dysfunction. We describe herein a series of 8-aryl xanthine derivatives which function as potent PDE5 inhibitors with, in many
cases, high levels of selectivity versus other PDE isoforms. # 2002 Elsevier Science Ltd. All rights reserved.
Inhibition of phosphodiesterase 5 (PDE5) is a clinically
proven concept for treatment of male erectile dysfunc-
tion (MED), with the PDE5 inhibitor sildenafil 1 laun-
ched in 1998, and great recent interest in the
development of new PDE5 inhibitors.1,2
cells within the eye, may be responsible for some of the
ocular side effects observed.4 In this communication, we
disclose the discovery of novel xanthine-based PDE5
inhibitors which exhibit good levels of in vitro potency
and differing selectivity over other PDE isoforms,
including PDE6.
3-Isobutyl-1-methylxanthine (IBMX) 2a has been long
known as a general, non-selective PDE inhibitor.5 More
recently, Corbin6 described a series of 8-substituted
derivatives of IBMX such as 2b and 2c, which exhibited
low nM potency against PDE5 and modest selectivity
against PDE1. In searching for a selective PDE5 inhi-
bitor, one strategy we employed involved a comparative
investigation of the similarity of xanthine templates
related to 2 and known PDE5 inhibitors such as silde-
nafil 1. The absolute binding conformations of these
molecules are unknown, and therefore it was sought to
determine those common features of known PDE5
inhibitors which were most likely to give rise to their
activity at this receptor. Using Catalyst#7 in hip-hop
mode, common features in 3-D space were identified
and aligned. This confirmed manual attempts at over-
laying our compounds where the 6 membered ring of
the pyrazolopyrimidine template was overlaid with the
five-membered ring of the xanthine template (Fig. 1).
The similarities between these two templates promp-
ted the design and subsequent synthesis of novel
hybrid 3a.
Despite the clear utility of 1, adverse effects such as
headaches, flushing and visual disturbance have been
noted in a small number of MED patients taking silde-
nafil,3 some of which could be associated with cross
reactivity on other PDE isoforms. Competitive inhibi-
tion of PDE6, which controls function of rod and cone
*Corresponding author. E-mail: steve.collingwood@pharma.novartis.
com
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00480-8

2588
R. Arnold et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2587–2590
Figure 1. Schematic overview of the common features hypothesis methodology utilised to generate a hypothetical PDE5 inhibitor.
8-Substituted xanthine 3a was prepared by condensa-
tion of the known uracil 48 and benzoic acid 59 (Scheme
1). We were pleased to discover that the desired xan-
thine sildenafil hybrid 3a was a potent PDE5 inhibitor
(human platelet PDE5 IC50 20 nM)10 and displayed
similar selectivity to sildenafil versus PDEs1, 2, 3, 4 and
6. Its suitability as a lead structure was further assessed
(Fig. 2). Lipophilicity, permeability and solubility were
in acceptable ranges for an oral drug however clearance
was relatively high and the relatively high molecular
weight and polar surface area (PSA)11 were potentially
limiting for lead optimisation. Nevertheless given the
relative ease of synthetic access a modest lead optimisa-
tion programme was initiated.
The N1 methyl and N3 isobutyl groups of xanthine 3a
originate from IBMX 2a and have been found to be
relatively optimal for broad spectrum PDE inhibitory
potency.6 We were interested to determine whether this
relationship would be maintained in our new series.
Synthesis of the required N1, and N3 analogues of 3a
followed in direct analogy to Scheme 1, the required
diaminouracils analogous to 4 being either known
compounds or prepared using similar methodology.8
Removal of the N1 methyl lowered potency (Table 1).
Conservative modification at N3 largely retained activ-
ity with 4–5 carbon atom substituents being favoured
for potency at PDE5. Some potential for increased selec-
tivity versus PDE6 was noted for the larger R2 groups.
Table 1. PDE5 activity of 3a and derivatives
Scheme 1. Reagents and conditions: (i) acid, HATU or EDC; DIPEA,
DMF, rt; (ii) NaOH, MeOH, H2O, 60 ꢀC.
Compd R1
R2
hPDE5 bPDE6 bPDE1 hPDE3
0.020 0.10 3.38 >10
3a
3b
3c
3d
3e
3f
3g
3h
3i
Me
H
H
iPr
iPr
H
Et
0.315
>10
—
—
—
0.345
5.6
>10
—
—
—
>10
>10
—
—
—
>10
>10
H
1.36
0.10
0.10
Me
Me
Me
Me
Me
Me
Me
Et
Vinyl
cPr
tBu
CHMeEt
nPn
cHexyl
0.028 0.217 >10
0.091 2.38
0.009 0.083 >10
—
5.57
6.03
3j
3k
3l
0.13
0.16
—
—
—
—
—
—
Me 4-Methoxyphenyl
0.059 1.21
4.52 >10
IC50 values are reported in mM. Source of PDEs: human platelet
PDE5, bovine retina PDE6, bovine heart PDE1, human platelet
PDE3.
Figure 2. Physical chemistry and pharmacokinetic properties of 3a.

R. Arnold et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2587–2590
2589
Scheme 2. Reagents and conditions: (i) WSCD, HOBt, CH2Cl2, H2O, 40 ꢀC; (ii) NaOH, MeOH, H2O, 60 ꢀC; (iii) neat ClSO3H, 40–60 ꢀC, 30 min;
(iv) amine, DMF, RT, 18 h.
thines 7 was made (Table 2) before preferred compounds
(7d, e, f, and g) were converted to the reactive sulphonyl
chlorides 8 and thence to the sulphonamides 9. O-De-
alkylation was problematic during chlorosulphonic acid
treatment of 7f, presumably due to the greater stability
of the iPr cation, hence an alternative sequence analo-
gous to Scheme 1 was used for the preparation of 9f.
For derivatives 7, optimal PDE5 potency is achieved
We next probed the effect of variation of the 8-aryl
group substituents. Planning to later synthesise a range
of sulphonamides suggested a more attractive route to
the derivatives wherein the sulphonamides could be
introduced in the final step of the synthesis (Scheme 2).
Ortho substituted benzoic acids 6 were either commer-
cially available or synthesised by alkylation of salicylic acid
methyl ester with the corresponding alkyl bromide and
saponification. Condensation of 6 with diamino uracil 4
gave the simple C8 aryl xanthines 7. An initial appraisal of
the PDE5 inhibitory activity of the various 8-aryl xan-
when R3 is a small alkyl ether which is a similar trend to
that disclosed by Terrett et al. in the optimisation of
sildenafil.2a Interestingly, ethers appear preferred over
thioethers (7c vs 7o) both of which have higher potency
that the nitrogen substituents (7m and 7n). Introduction
of an additional hetero atom into the alkyl group results
in a substantial loss of activity (7g vs 7k). These com-
pounds show little selectivity over PDE6. Again, simi-
larly to sildenafil, introduction of the piperazinyl
sulphonamide increased potency at PDE5, however, lit-
tle obvious trend in selectivity for PDE6 was apparent
(Table 3). Interestingly these SAR trends also mimic
that of sildenafil2a providing additional convincing evi-
dence to support our overlay hypothesis.
Table 2. PDE5 activity and isoform selectivity: variation of sub-
stituents at C20 of the 8 phenyl IMBX skeleton
The ethyl and propyl ethers were selected for further
optimisation at the sulphonamide group. Key inter-
mediates 8 (R=OEt and OnPr) were synthesised on
gram scale and a library of sulphonamides prepared in
96-deep-well plates, with purification by preparative
HPLC (Scheme 3). Representative examples are shown
Table 4. In general the most potent PDE5 inhibitors con-
tained a propyl ether at C20 (e.g., 3a vs 11d, 11e vs 10e, 11f
vs 10f and 11h vs 10h). In the case of piper-
azinylsulphonamides selectivity over PDE6 was generally
poor, the simpler derivatives showing best selectivity (up to
7.5-fold) when R3=OEt. High potency and improved
PDE6 selectivity could nevertheless be obtained with
N-dimethyl-ethylaminosulphonamides, for example 11c.
Compd
R3
hPDE5
bPDE6
Ratio
PDE5/PDE6
7a
7b
7c
7d
7e
7f
7g
7h
7i
7j
7k
7l
7m
7n
7o
H
OH
OMe
OEt
O-nPr
O-iPr
>10
>10
1.81
—
—
—
0.227
0.087
0.214
0.436
0.073
—
—
—
—
3.1
1.4
2.8
3.2
0.7
—
0.073
0.060
0.077
0.135
0.105
0.551
>10
1.50
>10
>10
>10
5.42
O-nBu
O-iBu
O-nPn
OCH2Ph
OCH2CH2OMe
F
—
—
—
—
NHMe
NMe2
SMe
Biopharmaceutical charaterisation of some representa-
tive derivatives is presented in Table 5. Potent com-
IC50 values are reported in mM.
Table 3. PDE5 activity and isoform selectivity: variation of sub-
stituents at C20 of the sildenafil–IBMX hybrid skeleton
Compd
R3
hPDE5
bPDE6
Ratio
PDE5/PDE6
3a
9e
9f
OEt
0.020
0.010
0.016
0.093
0.100
0.006
0.077
0.102
4.9
0.6
5.0
1.1
O-nPr
O-iPr
O-iBu
9h
IC50 values are reported in mM.
Scheme 3. Reagents and conditions: (i) amine, DMF, rt, 18 h.

2590
R. Arnold et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2587–2590
Table 4. PDE5 activity and isoform selectivity: variation of sub-
stituents at C50 of the 8 phenyl IMBX skeleton
pounds were associated with high PSA values; however,
this did not always result in poor permeability.
Xanthines containing a basic amino residue gave sig-
nificantly better solubility. A more recurrent problem
was high intrinsic and in vivo clearance resulting in
short half-lives and poor bioavailability. We felt that
this together with the high molecular weight of the
template precluded likely progression of this series.
Shortly after the termination of our studies a closely
related series of compounds was published by workers
from Almirall Prodespharma.12
Compd
R4
hPDE5 bPDE6
Ratio
PDE6/PDE5
11a
11b
11c
NH2
NEt(CH2CH2NMe2)
NMe (CH2CH2NMe2)
0.005
0.011
0.016
0.016
0.242
1.15
3
22
72
11d
11e
10e
11f
10f
11g
11h
10h
11i
0.010
0.0055
0.026
0.0035
0.019
0.012
0.024
0.063
0.003
0.006
0.003
0.203
0.015
0.068
0.091
0.037
0.146
0.017
0.6
0.5
7.5
4
In conclusion, we have shown that xanthines containing
an 8-(2-alkyloxy-5-aminosulphonyl) phenyl substituent
can serve as potent inhibitors of PDE5. Small alkyl
groups are favoured substituents at the N1 and N3
positions of the xanthine whereas a large variety of sul-
phonamides are tolerated at C50 of the 8 aryl group.
Variation at this position provides potent PDE5 inhibi-
tors and allows selectivity up to 70-fold to be achieved
versus the closely related PDE6 isoform.
3.5
7.5
1.5
2.5
5.5
References and Notes
1. (a) Maw, G. N. In Annual Reports in Medicinal Chemistry;
Bristol, J. A., Ed.; Academic: New York, 1999; p 71. (b)
Bivalaqua, T. J.; Champion, H. C.; Hellstrom, W. J. G.;
Kadowitz, P. J. Trends Pharmacol. Sci. 2000, 21, 484.
2. (a) Terrett, N. K.; Bell, A. S.; Brown, D.; Ellis, P. Bioorg.
Med. Chem. Lett. 1996, 6, 1819. (b) Eardley, I.; Br, J. Urol.
1998, 81, 122. (c) Rotella, D. P. Drugs Future 2001, 26, 153.
3. Osterloh, I. H.; Collins, M.; Wicker, P.; Wagner, G.; Int, J.
Clin. Pract. 1999, 102 (Suppl.), 3.
11j
0.0035
0.014
4.0
IC50 values are reported in mM.
4. Marmor, M. F.; Kessler, R. Surv. Ophthalmol. 1999, 44,
153.
5. Beavo, J. A.; Rogers, N. L.; Crofford, O. B.; Hardman,
J. G.; Sutherland, E. W.; Newman, E. V. Mol. Pharmacol.
1970, 6, 597.
6. Sekhar, K. R.; Grondin, P.; Francos, S. H.; Corbin, J. D. In
Phosphodiesterase Inhibitors; Schudt, C., Dent, S., Rabe, K. F.
Eds. Academic: New York, 1996; p 137.
Table 5. Biopharmaceutical properties of selected PDE5 inhibitors
3i
11j
11i
11c
Physical properties
Mr
cLog P
518.6
3.81
575.7
4.04
547.7
3.67
534.7
3.90
PSA
92.9
107.3
115.1
96.6
7. Catalyst, version 4.1; Accelrys, Inc. (previously known as
MSI): 9685 Scranton Road, San Diego, CA 92121, USA, 2000.
8. Traube, W. Chem. Ber. 1900, 33, 3035. Merlos, M.; Gomez,
L.; Vericat, M. L.; Bartroli, J.; Garcia-Rafanell, J.; Forn, J.
Eur. J. Med. Chem. 1990, 25, 653.
9. Dunn, P. J.; Wood, A. S. Eur. Pat. Appl., 1997; EP 812845
A1. Dale, D. J.; Dunn, P. J.; Golightly, C.; Hughes, M. L.;
Levett, P. C.; Pearce, A. K.; Searle, P. M.; Ward, G.; Wood,
A. S. Org. Process Res. Dev. 2000, 4, 17.
Kinetic solubility
pH 7.4 (mg/L)
pH 5 (mg/L)
ca. 100
>200
ca. 25
ca. 25
ca. 5
ca. 5
—
—
Caco2 permeability
Papp ꢁ 10ꢂ5 cm sꢂ1
In vitro microsome
metabolism Clint
(mL/min/mg) rat
human
3.45
2.29
1.32
0.03
:
178
114
185
71
226
High
—
—
10. Inhibition of PDE5 enzyme activity was determined with a
[3H]cGMP scintillation proximity assay using ytttrium silicate
beads (Amersham Pharmacia Biotech). Assays for other PDE
isoforms were conducted similarly.
11. Ertl, P.; Rohde, B.; Selzer, P. J. Med. Chem. 2000, 43, 3714.
12. Vega Noverola, A.; Gracia Ferrer, J.; Feixas Gras, J.;
Prieto Soto, J. M. PCT Int. Appl. WO 9962905 A1, 1999.
Rat in vivo PK
t1/2 (min)
Vss (L/kg)
CL (mL/kg)
BAV id
a
a
a
92ꢃ8
22ꢃ3
—
—
—
—
9.7ꢃ0.5
94ꢃ6
11%
3.4ꢃ0.5
140ꢃ10
0%
0%
aClearance too fast to determine reliably.