Discovery of potent, selective, and orally bioavailable PDE5 inhibitor: Methyl-4-(3-chloro-4-methoxybenzylamino)-8-(2-hydroxyethyl)-7-methoxyqui nazolin-6-ylmethylcarbamate (CKD 533)

Article abstract of DOI:10.1016/j.bmcl.2009.10.071

In a continuing effort to discover novel PDE5 inhibitors, we have successfully found quinazolines with 4-benzylamino substitution as potent and selective PDE5 inhibitors. Initial lead compound (1) was found to be easily metabolized when incubated with hum

Full text of DOI:10.1016/j.bmcl.2009.10.071

Bioorganic & Medicinal Chemistry Letters 20 (2010) 383–386
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of potent, selective, and orally bioavailable PDE5 inhibitor:
Methyl-4-(3-chloro-4-methoxybenzylamino)-8-(2-hydroxyethyl)-
7-methoxyquinazolin-6-ylmethylcarbamate (CKD 533)
Hojin Choi a, Jaekwang Lee a, Young Hoon Kim a,b, Dai Sig Im a, In-Chang Hwang a, Soo Jin Kim a,
Seung Kee Moon a, Hong Woo Lee a, Sung Sook Lee a, Soon Kil Ahn a, Sang Woong Kim c,
Nam Song Choi a, Kyung Joo Lee a,
*
a Chong Kun Dang Research Institute, CKD Pharmaceuticals Inc., PO Box 74, Chonan, Republic of Korea
b Department of Chemistry, Soong Sil University, Seoul 156-743, Republic of Korea
c LeadGenex, Venture Town Dasan, 1687-2 Shinil-dong, Daedeok-gu, Daejeon, Republic of Korea
a r t i c l e i n f o
a b s t r a c t
Article history:
In a continuing effort to discover novel PDE5 inhibitors, we have successfully found quinazolines with 4-
benzylamino substitution as potent and selective PDE5 inhibitors. Initial lead compound (1) was found to
be easily metabolized when incubated with human liver microsomes mainly through C6 amide
hydrolysis. Blocking of this metabolic hot spot led to discovery of 10 (CKD533) which is highly potent,
selective and orally efficacious in conscious rabbit model for erectile dysfunction and now is undergoing
preclinical toxicology study.
Received 22 October 2008
Revised 15 October 2009
Accepted 16 October 2009
Available online 21 October 2009
Keywords:
PDE5
Ó 2009 Elsevier Ltd. All rights reserved.
PDE6
PDE11
Enzyme inhibitor
Metabolic stability
Quinazoline
Male erectile dysfunction
Male erectile dysfunction (MED), now assumed to affect >500
million in the world, was a largely unmet medical need before the
introduction of sildenafil (ViagraÒ) in 1998. Although it was discov-
ered serendipitously,1 the launch of Viagra marked the new age in
drug discovery by affecting one’s quality-of-life. Increased levels of
cGMP leads to decreased intracellular calcium in the cells of corpus
cavernosum, resulting in vasorelaxation, inflow of arterial blood,
and ultimately an erection.2 PDE5 inhibition blocks cGMP degrada-
tion, thus increases the cGMP concentration, enhancing the erec-
tion.3 To date, 11 families (PDE1–PDE11) of phosphodiesterase
were identified, and PDE5 is abundant in smooth muscle, lung and
platelets.4
Although successful commercially, sildenafil use is associated
with many adverse effects such as headache, nausea, flushing, and
visual disturbances, which are the result of low PDE1 and PDE6
(PDE6 is the sole cGMP PDE in rod and cone cells within the eye)
selectivity,5 and similar side effects (indigestion, back pain, etc.)
were seen with vardenafil (LevitraÒ)6 and Tadalafil (CialisÒ).7
The physiological significance of PDE11 is not clearly under-
stood, but several lines of evidence suggested the PDE11 inhibition
might have negative effect in male reproduction and muscle tissue,
since PDE11 is predominantly present in muscle, prostate and tes-
tes.8 Since tadalafil cross reacts with PDE11 at sub
lM range
(merely 10-fold selectivity ratio relative to PDE5), there is a grow-
ing concern over PDE11 selectivity. So, it is highly desirable to
maintain high PDE11 selectivity when developing PDE5 inhibitors.
Thus, there have been numerous efforts toward the discovery of
more isozyme selective PDE5 inhibitors,9 which is of great medic-
inal and commercial interest, although with limited success.
In this Letter, we detail our continuing efforts in PDE5 inhibitor,
which have yielded a very potent inhibitor with IC50 value of
picomolar range and highly selective against other PDE isozymes
as well as orally efficacious.
Previously, we reported the discovery of 6,7,8-substituted quin-
azolines as potent and selective PDE5 inhibitors with in vivo effi-
cacy.10 Optimization of PDE5 activity, isozyme selectivity and
physicochemical properties led us to identification of 1 (Fig. 1) as
a potent and selective PDE5 inhibitor (PDE5 IC50 = 1 nM, selectivity
ratio for PDE6 >470, and PDE11 >8600). Compound 1 demon-
strated excellent efficacy in conscious rabbit model when dosed
* Corresponding author. Tel.: +82 41 529 3320; fax: +82 41 558 3004.
E-mail address: kjlee@ckdpharm.com (K.J. Lee).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.10.071

384
H. Choi et al. / Bioorg. Med. Chem. Lett. 20 (2010) 383–386
amides prepared from 13 were further manipulated to provide ter-
tiary amides (NaH, MeI or EtI, THF), which afforded compounds
2–5 after desilylation. Finally, N-methylation of 14 followed by desi-
lylation afforded tertiary carbamate 10.13 In the case of 10, another
efficientprocedure was developed in whichup to 1 Kg of 10 was syn-
thesized for preliminary toxicology study, and its synthesis will be
reported elsewhere.
Initially, it was assumed that incorporation of tertiary amide
would render 1 more resistant toward hydrolysis by various ami-
dases in HLM. As shown in Table 1, N-methyl analog (2) had com-
Table 1
PDE5 activity and isozyme selectivity of 4-(3-chloro-4-methoxy)-benzylamino-7-
methoxy-8-hydroxyethyl quinazoline derivatives
Figure 1. Structures of 1 and metabolites (11, 12).
orally. However, disappointingly it was later found to be highly
unstable with just 3% of parent compound left when incubated in
human liver microsomes (HLM).11 This was in sharp contrast to
the results from rat and rabbit liver microsomes where 96% and
85% of 1 remained, respectively. A LC/MS analysis of incubation
mixture with human liver microsomes (1.25 mg/ml) revealed that
1 was rapidly converted to 11 (ca. 90% of all metabolites deter-
mined from peak areas), indicating amide hydrolysis was a major
culprit of instability (Fig. 1). Interestingly, negligible amount of
11 was detected from incubation mixtures with rat and rabbit liver
microsomes, and this may be attributed to different type and/or
content of proteases across species.12a It was clear that more stable
compound might be obtained if the rapid amide hydrolysis at C6 in
1 were blocked. To this end, several analogs that have a variation at
C6 position of 1 were prepared according to Scheme 1.
Compound
R6
PDE5a
0.001
PDE6a
0.46
PDE11a
10.5
1
2
3
4
0.001
0.001
0.003
0.32
ND
5.2
ND
ND
We utilized the already synthesized compound 110 to facilitate
synthesis, that is, compound 1 was hydrolyzed under acidic condi-
tion followed by TBS protection to provide common intermediate
13. Compounds 6 and 7 were prepared by reaction of appropriate
sulfonyl chloride or sufamoyl chloride followed by acidic desilyla-
tion. Alternatively, secondary carbamate was installed by reaction
of methyl- and ethyl chloroformate to provide 8 and 9. The suitable
1.1
5
0.001
0.011
0.012
0.003
0.017
0.0006
0.34
ND
ND
ND
ND
5.7
5.0
12.0
6
7
7.95
0.71
0.6
8
9
10
0.14
Tadalafil
Sildenafil
0.012
0.01
3.0
0.14
0.29
3.0
Scheme 1. Reagents and conditions: (a) (i) 2 M HCl, MeOH, reflux, 15 h; (ii)
imidazole, TBSCl, DCM, rt, 2 h; (b) MsCl (for 6) or dimethyl sulfamoyl chloride (for
7), pyridine, DCM, 0 °C to rt; (c) ClCO2Me or ClCO2Et, pyridine, DCM, rt, 2 h; (d) (i)
amide formation or ClCO2Me, rt, 2 h; (ii) MeI or EtI, NaH, THF, rt, 6 h; (e) 1 M HCl,
1,4-dioxane, rt, 1 h.
ND = not determined.
Enzyme sources: see Ref. 18. IC50 values are reported in
>2 determinations).
a
lM (values are mean of

H. Choi et al. / Bioorg. Med. Chem. Lett. 20 (2010) 383–386
385
parable PDE5 inhibitory activity14 and PDE6, PDE11 selectivity
with compound 1. As anticipated, metabolic stability of 2 in HLM
was enhanced (16% of 2 left after 1 h incubation) relative to 1
(3%) while metabolic stability in rat and rabbit liver microsomes
remains excellent (>90%) as in the case of 1.15 It is interesting to
note that while hydrolyzed metabolite (11) was not detected, a
large amount of C7 de-methylated metabolite (12) was formed in
liver microsomes from all three species (Fig. 1). Although PDE5
inhibitory activities and isozyme selectivity of N-ethyl analog (3)
and N-methyl analog with different acyl groups (4, 5) were re-
tained, metabolic stability in HLM (5.5%, 4.8%, and 25% remained,
respectively) had not increased dramatically compared with 2.
The effect of other functional groups was explored in 6 and 7
and there was a 10-fold loss in PDE5 inhibition potency. Moreover,
both compounds were found to be unstable in HLM presumably
due to hydrolysis at C6 and/or de-methylation at C7, although met-
abolic study was not undertaken in this case. Low PDE5 inhibition
potency combined with unfavorable physicochemical property
such as high protein binding that is close to 100% precluded these
analogs from further profiling.
Figure 2. Docking model of 10 and tadalafil in PDE5 and PDE11 active site.
Based on high PDE5 activity and isozyme selectivity together
with preferable physicochemical properties such as good Caco-2
permeability12b (17 Â 10À6 cm/sec) and moderate protein binding
(95% vs 97% for sildenafil), excellent in vivo efficacy of 2 was ex-
pected. Indeed, 2 demonstrated equal efficacy relative to tadalafil
at 3 mg/kg when dosed orally in conscious rabbit model16 (data
not shown). Although 2 is effective in rabbit model, we felt that
metabolic stability in HLM was not satisfactory and might be a po-
tential liability in further development in human clinical trial.
Thus, we sought further modification to enhance metabolic stabil-
ity, by incorporating functional group that may be highly resistant
to hydrolysis from various proteases in human plasma. In this re-
gard, several carbamate analogs were prepared (8–10) since vari-
ous carbamate containing drugs are being used successfully.17
Gratifyingly, the simplest analog (8) was found to have potent
PDE5 inhibitory activity while isozyme selectivity remained excel-
lent. There was a slight loss of PDE5 inhibitory activity with ethyl
carbamate (9) with isozyme selectivity remained at the same level
as 8. A tertiary carbamate (10) which was prepared to further in-
crease metabolic stability was found to have highly potent PDE5
inhibitory activity in pM range. Compound 10 is not only 20-fold
more potent than tadalafil and sildenafil, it is significantly more
selective against other PDE isozymes as indicated in Table 1 (selec-
tivity ratio for PDE6 >230, and PDE11 >20,000). Compound 10
exhibited even greater selectivity against PDE1, PDE2, PDE3, and
moderate protein binding (same as sildenafil with 97%) and high
aqueous solubility (>2 mg/ml as methanesulfonate salt). Moreover,
10 was found to have negligible cell cytotoxicity (CC50 = 85
L-5178Y (murine leukemia) cells and did not block hERG channel
(<20% inhibition at 10 M). Oral exposure study in rat and rabbit
lM) in
l
was carried out to estimate intestinal absorption and in vivo effi-
cacy (Table 2). In rat, 10 (as HCl salt) was absorbed very rapidly
with Tmax of 0.2 h, and its moderate half-life (2.2 h) would be suit-
able for once-a day dosing in human. In contrast, a more sustained
exposure was achieved in rabbit with longer half-life >4 h.
In vivo efficacy of 10 in inducing penile erection using conscious
rabbit was then evaluated (Fig. 3).16 No penile erection was ob-
served in vehicle-treated animals, while the erectogenic effect
was potentiated by SNP injection (0.1 mg/kg), a nitric oxide donor
Table 2
Selected PK Parameters for compound 10 after oral dosinga
Speciesa
Tmax (h)
Cmax (ng/ml)
t1/2 (h)
AUClast (ng h/ml)
Rat
Rabbit
0.2 0.08
2.4 0.55
420 85
870 490
2.25 1.8
4.2 3.2
540 146
2760 1540
a
Dose: 10 mg/kg (0.5% Tween 80), n = 4.
PDE4 (IC50 = 12 lM, 32 lM, 43 lM, and 84 l
M, respectively).18
Most importantly, when incubated in HLM (1.25 mg/ml) both 8
and 10 were highly stable with 72% and 85% of parent compound
remained after 1 h incubation, indicating robustness of carbamate
relative to amide series against various proteases present in liver
microsomes.
To understand the underlying mechanism of the high selectivity
toward PDE11, the docking modes of 10 and tadalafil were com-
pared in the active sites of PDE5 and PDE11 (Fig. 2). HOMOLOGY
and AFFINITY modules of INSIGHT2000 were used for the modeling
and docking studies.19 Overall, the compositions of amino acids in
the active sites are highly conserved except that Phe820 in PDE5
was replaced with Trp298 in PDE11. Tadalafil makes favorable
p–p stacking interaction with Trp298 in PDE11 and with Phe820
in PDE5, which explains high affinity toward both PDE5 and
PDE11. In contrast, 10 snugly fits into the active site of PDE5, with
favorable interaction with Phe820, however, Trp298 in PDE11 kept
10 from tightly binding due to the steric hindrance with C8
hydroxyethyl group, thus illustrating very low affinity.
Figure 3. Effect of 10 on the length of penis in conscious rabbit model. The length of
penis was measured after oral administration of 10, followed by sodium
nitroprusside (SNP, 0.1 mg/kg) injection after 2 h (n = 4). See Ref. 16 for details.
Tertiary carbamate 10 possessed preferable physicochemical
a
properties such as good Caco-2 permeability (35 Â 10À6 cm/sec),

386
H. Choi et al. / Bioorg. Med. Chem. Lett. 20 (2010) 383–386
Lunden, S.; Sui, Z. J. Med. Chem. 2005, 48, 2126; (e) Xia, G.; Li, J.; Peng, A.; Lai, S.;
Zhang, S.; Shen, J.; Liu, Z.; Chen, X.; Ji, R. Bioorg. Med. Chem. Lett. 2005, 15, 2790.
10. Kim, Y. H.; Choi, H.; Lee, J.; Hwang, I-C.; Moon, S. K.; Kim, S. J.; Lee, H. W.; Im, D.
S.; Lee, S. S.; Ahn, S. K.; Kim, S. W.; Han, C. K.; Yoon, J. H.; Lee, K. J.; Choi, N. S.
Bioorg. Med. Chem. Lett. 2008, 18, 6279.
11. Van de Waterbeemd, H.; Smith, D. A.; Beaumont, K.; Walker, D. K. J. Med. Chem.
2001, 44, 1313.
12. (a) Coleman, M. D. Human Drug Metabolism; Wiley: Bermingham, 2005; (b)
Kerns, E. D.; Di, Li.. In Drug-like Properties, Concepts, Structure, Design and
Methods; Academic Press, 2008.
that is used as a sexual stimulant. Compound 10 demonstrated
equal efficacy at a dose (0.3 mg/kg) of one-tenth of tadalafil
(3 mg/kg) when dosed orally, which is in agreement with potent
PDE5 inhibitory activity. Compound 10 was advanced into preclin-
ical safety studies from these encouraging results together with
favorable physicochemical and pharmacokinetic profiles across
species (rat and rabbit).
In conclusion, we have discovered a potent and selective PDE5
inhibitor possessing novel quinazoline scaffold. An analysis of pre-
viously identified potent PDE5 inhibitor (1) posed a metabolic
instability issue in HLM, mainly through C6 amide hydrolysis. This
was circumvented by switching to tertiary carbamate, resulting in
substantial increase in metabolic stability and ultimately led to
identification of 10. Compound 10 is not only more potent (20-
fold) PDE5 inhibitor than tadalafil, but highly selective against
other isozymes including PDE6 and PDE11. Based on favorable pro-
files of this molecule, 10 (CKD533) was selected for preclinical
safety evaluations as a potential candidate for the treatment of
male erectile dysfunction.
13. Selected data for 2: 1H NMR (400 MHz, DMSO-d6) d 8.63 (t, J = 5.9 Hz, 1H), 8.49
(s, 1H), 8.17 (s, 1H), 7.41 (s, 1H), 7.29 (d, J = 8.5 Hz, 1H), 7.08 (d, J = 8.5 Hz, 1H),
4.74 (br, 1H), 4.60–4.69 (m, 2H), 3.80 (s, 3H), 3.74 (s, 3H), 3.59 (br, 2H), 3.20–
3.30 (m, 2H), 3.19 (s, 3H), 1.97–2.12 (m, 2H), 0.91 (t, J = 7.3 Hz, 3H); MS (ESI) m/
z 459 (M++1). For 8: 1H NMR (400 MHz, DMSO-d6) d 9.09 (s, 1H), 8.65 (t,
J = 5.8 Hz, 1H), 8.41 (s, 1H), 8.30 (s, 1H), 7.39 (d, J = 2.0 Hz, 1H), 7.28 (dd, J = 6.4,
2.0 Hz, 1H), 7.06 (d, J = 8.5 Hz, 1H), 4.79 (br, 1H), 4.65 (d, J = 5.8 Hz, 2H), 3.79 (s,
3H), 3.76 (s, 3H), 3.70 (s, 3H), 3.59 (br, 2H), 3.20 (t, J = 7.7 Hz, 2H); MS (ESI) m/z
447 (M++1). For 10: 1H NMR (400 MHz, DMSO-d6) d 8.64 (t, J = 4.8 Hz, 1H), 8.47
(s, 1H), 8.11 (s, 1H), 7.41 (d, J = 1.9 Hz, 1H), 7.29 (dd, J = 8.5, 1.9 Hz, 1H), 7.06 (d,
J = 8.8 Hz, 1H), 4.80 (t, J = 7.8 Hz, 1H), 4.65 (d, J = 5.9 Hz, 2H), 3.80 (s, 3H), 3.75
(s, 3H), 3.56–3.59 (m, 5H), 3.22 (m, 2H), 3.19 (s, 3H); 13C NMR (100 MHz,
DMSO-d6) d 157.9, 156.1, 155.3, 153.9, 148.9, 133.0, 129.7, 129.4, 127.9, 122.1,
121.2, 113.0, 111.8, 61.6, 61.1, 56.4, 53.2, 43.1, 37.9, 29.2; MS (ESI) m/z 461
(M++1); Anal. Calcd for C22H25ClN4O5: C, 57.33; H, 5.47; N, 12.16. Found: C,
57.29; H, 5,43; N, 12.36.
14. PDE5 assays were carried out as described in: Turko, I. V.; Ballard, S. A.; Francis,
S. H.; Corbin, J. D. Mol. Pharmacol. 1999, 56, 124. assays for other isozymes were
conducted in a similar manner.
References and notes
1. (a) Martel, A. M.; Graul, A.; Rabasseda, X.; Castaner, R. Drugs Future 1997, 22,
138; (b) Terret, N. K.; Bell, A. S.; Brown, D.; Ellis, P. Bioorg. Med. Chem. Lett. 1996,
6, 1819.
2. (a) Gibson, A. Eur. J. Pharmacol. 2001, 411, 1; (b) Andersson, K.-E.; Wagner, G.
Physiol. Rev. 1995, 75, 191.
3. Manallack, D. T.; Hughes, R. A.; Thompson, P. E. J. Med. Chem. 2005, 48, 3449.
4. Francis, S. H.; Turko, I. V.; Corbin, J. D. Prog. Nucl. Acid Res. Mol. Biol. 2001, 65, 1.
5. (a) Osterloh, I. H.; Riley, A. Br. J. Clin. Pharmacol. 2002, 53, 219; (b) Beavo, J. A.
Physiol. Rev. 1995, 75, 725.
15. Human, rat, and rabbit liver microsomes were purchased from BD Sciences, Inc.
To a mixture of compound (5 lM) and liver microsomes in phosphate buffer
(pH 7.4) was added a solution of NADPH-regenerating system (2.5 mM MgCl2,
2.5 mM glucose-6-phosphate, 1 mM NADP, 1 U/ml glucose-6-phosphate
dehydrogenase) and incubated at 37 °C for 1 h. Samples were withdrawn,
reaction terminated by adding acetonitrile, then extracted and analyzed by LC/
MS.
16. (a) Bischoff, E.; Niewoehner, U.; Haning, H.; Es Sayed, M.; Schenke, T.;
Schlemmer, K. H. J. Urol. 2001, 165, 1315; (b) Kang, K. K.; Ahn, G. J.; Ahn, B.
O.; Yoo, M.; Kim, W. B. Eur. Urol. 2003, 43, 689. in brief, New Zealand white
rabbits weighing 3–4 kg (n = 4) were orally dosed (dissolved in distilled water)
with compounds, followed by injection of SNP (0.1 mg/0.1 ml/kg, dissolved in
saline) after 2 h, and the length of penis (covered + mucosa area) was measured
with sliding calipers, recorded in cm for total 2 h after administration of test
compounds.
17. Ray, S.; Pathak, S. R.; Chaturvedi, D. Drugs Future 2005, 30, 161.
18. Enzyme sources: PDE1: bovine heart, PDE2: bovine heart, PDE3: bovine
platelet, PDE4: human recombinant, PDE5: bovine platelet, PDE6: bovine
retina, and PDE11: human recombinant.
6. Bischoff, E. Int. J. Impot. Res. 2004, 16, S34.
7. Porst, H. EAU Update Series 2004, 2, 56.
8. (a) Makhlouf, A.; Kshirsagar, A.; Niederberger, C. Int. J. Impot. Res. 2006, 18, 501;
(b) Francis, S. H. Int. J. Impot. Res. 2005, 17, 467; (c) Seftel, A. D. J. Urol. 2005, 174,
1043; (d) Baxendale, R.; Burslem, F.; Phillips, S. J. Urol. 2001, 165, 340.
9. (a) Arnold, N. J.; Beer, D.; Bhalay, G.; Craig, S.; Farr, D.; Glen, A.; Hatto, J. D.;
Jones, D.; Moser, H. E.; Naef, R.; Turner, K. L.; Tweed, M. F.; Watson, S. J.; Zurini,
M. Bioorg. Med. Chem. Lett. 2007, 17, 2376; (b) Giovannoni, M. P.; Vergelli, C.;
Biancalani, C.; Cesari, N.; Graziano, A.; Biagini, P.; Gracia, J.; Gavalda, A.; Dal
Piaz, V. J. Med. Chem. 2006, 49, 5363; (c) Haning, H.; Niewoehner, U.; Schenke,
T.; Lampe, T.; Hillish, A.; Bischoff, E. Bioorg. Med. Chem. Lett. 2005, 15, 3900; (d)
Jiang, W.; Guan, J.; Macielag, M. J.; Zhang, S.; Kraft, Y.; John, T. M.; H-J, D.;
19. INSIGHT 2000, Accelrys Inc., San Diego, CA.