Synthesis and biological evaluation of imidazol-2-one and 2-cyanoiminoimidazole derivatives: novel series of PDE4 inhibitors.

Article abstract of DOI:10.1016/S0960-894X(01)00817-4

This communication describes the synthesis and in vitro PDE4 inhibitory activity of a novel series of imidazol-2-one and 2-cyanoiminoimidazole derivatives. The compounds described were also tested in in vivo models to evaluate their anti-inflammatory activity after topical administration as well as their gastro-intestinal side effects. Several compounds proved to be potent PDE4 inhibitors and some 2-cyanoiminoimidazoles showed less pronounced gastro-intestinal side effects than reference compounds but maintained anti-inflammatory activity after topical administration.

Full text of DOI:10.1016/S0960-894X(01)00817-4

Bioorganic & Medicinal Chemistry Letters 12 (2002) 653–658
Synthesis and Biological Evaluation of Imidazol-2-one
and 2-Cyanoiminoimidazole Derivatives: Novel Series
of PDE4 Inhibitors
J. Ignacio Andres,a,* Jose M. Alonso,a Adolfo Dıaz,a Javier Fernandez,a Laura Iturrino,a
Pedro Martınez,a Encarna Matesanz,a Eddy J. Freyne,b Frederik Deroose,b
Gustaaf Boeckx,b Davy Petit,b Gaston Diels,b Anton Megens,b Marijke Somers,b
Jean Van Wauwe,b Paul Stoppie,b Marina Cools,b Fred De Clerck,b Danielle Peetersb
and Didier de Chaffoyb
aJanssen-Cilag, Basic Research Centre, Jarama s/n, 45007 Toledo, Spain
bJanssen Research Foundation, B2340, Belgium
Received 24 October 2001; accepted 4 December 2001
Abstract—This communication describes the synthesis and in vitro PDE4 inhibitory activity of a novel series of imidazol-2-one and
2-cyanoiminoimidazole derivatives. The compounds described were also tested in in vivo models to evaluate their anti-inflamma-
tory activity after topical administration as well as their gastro-intestinal side effects. Several compounds proved to be potent PDE4
inhibitors and some 2-cyanoiminoimidazoles showed less pronounced gastro-intestinal side effects than reference compounds but
maintained anti-inflammatory activity after topical administration. # 2002 Elsevier Science Ltd. All rights reserved.
efforts in this field.2 The therapeutic activity of the first
generation anti-inflammatory PDE4 inhibitors is,
Introduction
cAMP and cGMP are important second messengers with
a critical regulatory role in various biological processes.
The hydrolysis of these 30,50-cyclic nucleotides to the
corresponding nucleoside 50-monophosphates is cata-
lyzed by a superfamily of enzymes called phosphodi-
esterases (PDEs).1 PDE4 is a high-affinity cAMP
specific phosphodiesterase predominantly present in
virtually all immune and inflammatory cells. This PDE4
family consists of four genes (A–D) with a unique
chromosome localization. Elevation of cAMP through
inhibition of PDE4 results in a wide range of anti-
inflammatory effects in leukocytes. Given the increased
levels of cAMP in leukocytes from patients with atopic
dermatitis, asthma and allergic rhinitis, inhibition of
PDE4 is an attractive target for drug design, which has
generated great interest.2
however, limited by their side effects, especially gastro-
intestinal side effects such as emesis or acid secretion.
Although the anti-inflammatory activity is believed to
reside, for peripheral action, on the catalytic site of
PDE4, these inhibitors seem to interact with both
the catalytic site and the ‘high-affinity rolipram bind-
ing site’ (HARBS).3 It is thought that the interaction
and activity at this HARBS is responsible for the
undesirable CNS mediated side effects.4 Strategies to
overcome these side effects will depend upon the
identification of PDE4 inhibitors which exhibit low
affinity for the HARBS or PDE4 subtype-specific
inhibitors (Fig. 1).
We have discovered a novel series of imidazol-2-one
derivatives,5 potent PDE4 inhibitors, with activities
comparable to those displayed by some of the most
potent reference compounds, such as RP73401 (2) and
ariflo (3). Furthermore, we have prepared a series of 2-
cyanoiminoimidazole derivatives that show much
reduced gastro-intestinal side effects while keeping good
inhibitory activity.6
The archetypal PDE4 inhibitor rolipram (1) has been
the starting point for many of the medicinal chemistry
*Corresponding author. Fax: +34-925-23-2603; e-mail: iandres@
prdes.jnj.com
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00817-4

´
J. I. Andres et al. / Bioorg. Med. Chem. Lett. 12 (2002) 653–658
654
Figure 1.
Chemistry
procedures (Scheme 1). Reaction of amines 4 either with
phenylchloroformate or diphenyl N-cyanocarbonimidate
led to the corresponding intermediates 5 which, after reac-
tion with 2,2-dimethoxyethylamine, yielded the requisite
ureas or N-cyanoguanidines 6.9 The ureas 6 (X=O)
could also be obtained directly from amines 4 by reac-
tion with N-(2,2-dimethoxyethyl)-1-imidazol-1-carbox-
amide in THF at reflux temperature. The imidazol-2-
one derivatives 7 (X=O) were prepared by cyclization
of ureas 6, in the presence of dilute hydrochloric acid in a
mixture of methanol/water at room temperature.8 The
The general synthetic pathway for the preparation of
the target compounds is illustrated in Schemes 1 and 2.
There are very few literature references describing the
synthesis of 2-cyanoiminoimidazoles.7 We designed a
new synthetic pathway for 2-cyanoiminoimidazoles
based on the well known synthesis of imidazol-2-ones,8
so that we could use the same starting materials. The
starting 3,4-dialkyloxybenzaldehydes were converted
into the required phenylethylamines 4 using standard
Scheme 1. Reagents and conditions: (a) R1X, K2CO3, DMF, 80 ꢀC or R1OH, DIAD, PPh3, THF, rt; (b) NaBH4, CH3OH–THF, rt; (c) SOCl2,
DMF (cat.), toluene; (d) KCN, C2H5OH, H2O, DMF, 60 ꢀC, 2 h; (e) for R4=H: LDA, THF, R3X, ꢁ78 ꢀC to rt or 1) C2H5OCOOC2H5, NaOC2H5,
toluene-H2O, 110 ꢀC; 2) CH3X, rt; 3) NaOH, 60 ꢀC; for R3, R4=c-alkyl: BrCH2CH2Br, NaOH 50%, BTEAC, 60–70 ꢀC; (f) H2, RaNi, CH3OH–
NH3, rt; (g) PhOCOCl, (C2H5)3N, CH2Cl2, rt; (h) (PhO)2CNCN, C2H5OH, rt; (i) NH2CH2CH(OCH3)2, (C2H5)3N, DMAP, 1,4-dioxane, reflux;
(j) ImCONHCH2CH(OCH3)2, THF, reflux; (k) 0.5 N HCl (1.5 eq.), CH3OH/H2O (2:1, v/v), rt; (l) 0.5 N HCl (1.5 eq.), THF, reflux, 1 h; (m) NaH,
THF, reflux, 90%; (n) H2, Pd/C (10%), EtOH, rt, 92%; (o) 1 N aq. NaOH, CH3OH/THF (1:1, v/v), reflux, 52%; (p) 1) (C6H5O)2P(O)N3, Et3N,
toluene, 90 ꢀC; 2) NH2CH2CH(OCH3)2, rt, 82%.

´
J. I. Andres et al. / Bioorg. Med. Chem. Lett. 12 (2002) 653–658
655
Scheme 2. Reagents and conditions: (a) R1X, K2CO3, DMF, 80 ꢀC or R1OH, DIAD, PPh3, THF, rt; (b) TMSCN, ZnI2, CH2Cl2, rt; (c) H2, RaNi,
CH3OH/NH3; (d) Im(CO)NHCH2CH(OCH3)2, THF, reflux; (e) 1 N HCl, CH3OH/H2O, rt; (f) DAST, CH2Cl2, ꢁ78 ꢀC; (g) ClCOCOCl, (C2H5)3N,
DMSO, CH2Cl2, ꢁ70 ꢀC to rt; (h) CH3MgCl, THF, 10 ꢀC.
best conditions found for the cyclization of N-cyano-
guanidines 6 to the 2-cyanoiminoimidazole derivatives 7
(X=N–CN) were the treatment with 1.5 equivalents of
dilute HCl in THF and heating for not more than 1 h
(Scheme 1). Longer reaction times resulted in extensive
decomposition. Compound 7f was synthesized by a dif-
ferent pathway starting from the acetophenone deriva-
tive 8,10 which was transformed in three steps into the
acid derivative 11. Curtius rearrangement of this
compound to the corresponding isocyanate derivative,
followed by in situ reaction with 2,2-dimethoxy-
ethylamine, yielded the urea derivative 6f, which upon
cyclization by the standard procedure afforded imida-
zol-2-one derivative 7f. The enantiomers of 7p were
obtained by chiral HPLC separation of the racemic
compound.
be transformed into the fluoro analogues 16 upon
treatment with DAST.11 Oxidation of the hydroxyl
group of 15 to the corresponding ketone 17, and sub-
sequent reaction with methylmagnesium chloride, yielded
methyl-hydroxy derivatives 18, which provided the
target methyl-fluoro substituted compounds 19 upon
treatment with DAST.
In principle compounds containing the cyanoiminoimi-
dazole moiety could exist as two tautomeric (‘imino’
and ‘amino’) structures. We only observed one unique
form for each compound. The structure of this cyano-
1
iminoimidazole ring was elucidated by H NMR NOE
experiments. The NOE effects observed between the
protons of the ring and the protons of the carbon linked
to the heterocyclic nitrogen led us to conclude that only
the ‘imino’ tautomer was present. Further NMR
experiments aimed at elucidating the E:Z isomery of the
iminic double bond were unsuccessful.
Target compounds in which R3=F or OH were syn-
thesized as shown in Scheme 2. Reaction of the starting
substituted benzaldehydes with trimethylsilyl cyanide
produced the corresponding silylated cyanohydrines 12,
which after reduction of the nitrile group and further
reaction with N-(2,2-dimethoxyethyl)-1-imidazol-1-car-
boxamide yielded the required ureas 14. Cyclization as
described in Scheme 1 provided the imidazol-2-ones 15,
containing a hydroxyl group on the spacer, which could
Biological Results and Discussion
Compounds were tested for their inhibitory potencies
against recombinant PDE4 B from human mononuclear
lymphocytes, which is
a PDE4 subtype broadly

´
J. I. Andres et al. / Bioorg. Med. Chem. Lett. 12 (2002) 653–658
656
Table 1. IC50s (mM) for inhibition of PDE4 B and LADs (mg/kg, po; n=3 per dose level) for inhibition of gastric emptying12
Compd
R1
R2
R3
R4
R5
X
PDE4 B IC50 (mM)
Gastric emptying (mg/kg)
7a
7b
7c
7d
7e
17a
7f
Cyclopentyl
Cyclopentyl
Cyclopentyl
Cyclopentyl
Cyclopentyl
Cyclopentyl
Cyclopentyl
H3C
H3C
F2HC
F3C
H3C
H3C
H3C
H
H
H
H
H
H3C
H
H
H
H
H
O
O
O
O
O
O
O
0.1096
0.0241
<0.01
>1
0.367
0.486
0.08
0.01
0.02
1.25
0.63
1.25
n. t.
0.01
H3C
H3C
H3C
H3C
O
H
CH3
H3C
H3C
H
H
0.277
0.0858
7g
H3C
H
O
7 h
7i
H3C
H3C
H3C
H3C
H
H
H
H
O
O
0.241
0.021
0.02
0.01
7j
7k
Cyclopentyl
Cyclopentyl
H3C
F2HC
–CH2-CH2–
–CH2-CH2–
H
H
O
O
0.341
0.0392
0.399
0.63
0.08
0.31
15b
H3C
HO
H
H
O
0.0243
0.08
16b
H3C
F
H
H
O
18c
19c
7l
H3C
H3C
HO
F
H3C
H3C
H
H
H
H
H
O
>1
0.31
0.31
1.25
0.08
O
0.327
0.0565
0.0224
H3C
H3C
H3C
N-CN
N-CN
7m
F2HC
H
7n
7o
Cyclopentyl
H3C
H3C
H3C
H3C
H
H
H
H
N-CN
N-CN
0.03
0.196
1.25
>1.25
7p
7q
H3C
H3C
H3C
H3C
H
H
H
H
N-CN
N-CN
0.019
2.5
0.0676
>1.25
7r
7s
Ph–(CH2)5–
Ph–(CH2)5–
H3C
F2HC
H3C
H3C
H
H
H
H
N-CN
N-CN
0.0361
0.0284
0.019
>1.25
>1.25
2.5
R-(+)-7p
H3C
H3C
H3C
H
H
N-CN
S-(ꢁ)-7p
H3C
H
H
N-CN
0.057
>5
Rolipram
RP 73401
Ariflo
0.741
0.0002
0.071
0.04
0.5
5

´
J. I. Andres et al. / Bioorg. Med. Chem. Lett. 12 (2002) 653–658
657
Table 2. Results obtained in the TPA-induced ear inflammation model
Dose
7a
7g
7i
7j
7l
7m
7n
7o
7p
7r
7s
(+)-7p
(ꢁ)-7p
Rolipram
RP 73401
Ariflo
2 mpk
1 mpk
81
32
74
56
90
34
66
18
78
38
52
22
84
50
71
38
82
51
77
36
69
41
43
6
69
43
88
69
56
54
% mean values of inhibition of inflammation at 2 mg/kg (50 mg) and 1 mg/kg (25 mg) after treatment of 6–8 animals with each compound.13
Table 3. Inhibition of [3H]-rolipram binding to rat forebrain mem-
greatly improve the therapeutic margin of these pro-
ducts. Compared with reference compounds many of
the imidazol-2-ones and 2-cyanoiminoimidazoles
showed good PDE4 inhibitory activities, while these
latter derivatives displayed gastro-intestinal side effects
in the same range as the best reference compound, ari-
flo, and the enantiomer S-(ꢁ) of one of the best com-
pounds, 7p, showed even a better activity/gastro-
intestinal side effects ratio.
branes16
Compd
Ki (nM)
7g
3.5
11
26
63
180
220
Rolipram
RP73401
R-(+)-7p
7p
S-(ꢁ)-7p
The cyanoiminoimidazoles and some imidazol-2-one
derivatives were also tested in the TPA model to evaluate
their anti-inflammatory activity after topical treatment.
As can be deduced from the data shown in Table 2 the
compounds tested showed more than 50% inhibitory
effect in the TPA-induced ear inflammation model, after
topical administration of 2 mg/kg of compound. Even
more, some of them kept over 50% inhibitory activity
after 1 mg/kg application, being comparable in that
respect to the reference compound ariflo.
distributed in tissues. In order to evaluate their gastro-
intestinal side effects, compounds were tested in vivo for
their inhibitory effect on gastric emptying in rats.12
Table 1 summarizes the results of a selection of some
compounds in order to discuss the structure–activity
relationships. The most potent compounds were also
tested against the other PDE4 isozymes (A, C and D) as
well as against PDE1, 3 and 5. All the compounds tested
were selective PDE4 inhibitors, while none of them were
PDE4 subtype-selective inhibitors. The topical anti-
inflammatory activity of several compounds was also
evaluated in the TPA induced ear inflammation model
to determine their potential as anti-inflammatory agents
for topical treatment.13
In order to evaluate a possible correlation between the
inhibition of gastric emptying and the affinity for the
high-affinity rolipram binding site (HARBS), a few
compounds were also evaluated for their ability to dis-
place [3H]-rolipram from its high-affinity binding site in
rat brain membranes.16 As can be deduced from the
data shown in Table 3 there is a certain degree of cor-
relation between the results obtained in both tests.
Compound S-(ꢁ)-7p showed the lowest affinity for
HARBS while showing the least gastro-intestinal side
effects in our test.
In general, the alkoxy substitution confirmed the known
structure–activity relationship of rolipram-derived
PDE4 inhibitors.14,15 Substitution of the methyl group
at R2 by a difluoromethyl group resulted in a significant
increase in PDE4 inhibitory activity. By contrast, intro-
duction of a trifluoromethyl group led to a complete
loss of activity. Among substituents at R1 position a
variety of bulky substituents are allowed. Thus cyclo-
propylmethyl, bicyclo[2.2.1]hept-2-yl, phenylpentyl and
indan-2-yl proved to give compounds of the same order
of activity as cyclopentyl moiety, while tetrahydrofuran-
3-yl showed a tenfold decrease in potency. Regarding
the modifications in the spacer, the presence of a methyl
group at R3 benzylic position gave more potent com-
pounds than those in which R3=R4=R5=H but the
introduction of further methyl groups at R4 and/or R5
positions proved to be detrimental for inhibitory activ-
ity. Other functional groups such as ketone or hydroxyl
instead of methyl at R3 proved to give less active com-
pounds as well, while the introduction of a fluorine
atom at that position led to compounds showing com-
parable activity to those with R3=CH3.
In conclusion, we have reported the synthesis of a series
of imidazol-2-one derivatives and of 2-cyanoiminoimi-
dazole derivatives which showed PDE4 inhibitory
activities comparable to reference compounds. Those
compounds were topically active in an in vivo model of
anti-inflammatory activity and in the case of the 2-cy-
anoiminoimidazole series with much reduced gastro-
intestinal side effects. Further pharmacological evalu-
ation of compound S-(ꢁ)-7p will be reported elsewhere.
References and Notes
1. Beavo, J. A.; Conti, M.; Heaslip, R. J. Mol. Pharmacol.
1994, 46, 399.
2. Dyke, H. J.; Montana, J. G. Exp. Op. Invest. Drugs 1999, 8,
1301.
Regarding gastro-intestinal side effects it can be con-
cluded that the substitution of the imidazol-2-one ring
by a 2-cyanoiminoimidazole reduced to a larger degree
the gastro-intestinal side effects in vivo without affect-
ing the PDE4 inhibitory potency in vitro. This more
pronounced decrease in side-effect potential might
3. (a) Muller, T.; Engels, P.; Fozard, J. R. Trends Pharmacol.
Sci. 1996, 17, 294. (b) Kelly, J. J.; Barnes, P. J.; Giembycz,
M. A. Biochemistry 1996, 318, 425. (c) Jacobitz, S.;
McLaughlin, M. M.; Livi, G. P.; Burman, M.; Torphy, T. J.
Mol. Pharmacol. 1996, 50, 891.

´
J. I. Andres et al. / Bioorg. Med. Chem. Lett. 12 (2002) 653–658
658
4. (a) Barnette, M. S.; Torphy, T. J.; Christensen, S. B. Com-
pounds, WO 9500139, 1995. (b) Duplantier, A. J.; Biggers,
M. S.; Chambers, R. J.; Cheng, J. B.; Cooper, K.; Damon,
D. B.; Eggler, J. F.; Kraus, K. G.; Marfat, A.; Masamune, H.;
Pillar, J. S.; Shirley, J. T.; Umland, J. P.; Watson, J. W. J.
Med. Chem. 1996, 39, 120.
5. Freyne, E. J.; Diels, G. S.; Andres-Gil, J. I.; Fernandez-
Gadea, F. J. WO 9631485, WO 9631486, WO 9631487, 1996.
6. (a) Freyne, E. J.; Fernandez-Gadea, F. J.; Andres-Gil, J. I.
WO 9814432, 1998. (b) Freyne, E. J.; Diels, G. S.; Matesanz-
Ballesteros, M. E.; Dıaz-Martınez, A. WO 9950262, 1999.
7. (a) SmithKline Beckman Corp. JP 62153274, 1987. (b)
Ikesu, S.; Kita, H.; Kaneko, Y. JP 07128824, 1995. (c) Kris-
tinsson, H.; Rempfler, H.; Nussbaumer, H. WO 9800408,
1999. (d) Wigerinck, P. T.; Verschueren, W. G.; Schroven, M.
F.; De Bruyn, M. F. WO 9929687, 1999.
est active dose) was determined, that is, the lowest dose that
induced inhibition of gastric emptying in at least two out of
three tested rats.
13. Female CD1 mice (20–22 g) were divided in groups of 6–8
animals. Solutions of tetradecanoyl phorbol acetate (TPA, 100
mg/mL) with or without experimental compound (5 mg/mL)
were prepared in 200 mL of acetone. Ten mL quantities (con-
taining 1 mg TPA and 50 mg compound) were applied to the
inner surface of the left ear. The right ear received 10 mL of
acetone and served as the control. Three hours later, the ears
were removed and an 8-mm punch was taken. The degree of
ear edema was calculated by substracting the weight of the
right (control) ear punch from that of the left (treated) ear in
each treatment group. Differential ear weights were then
compared between the TPA and TPA+compound groups to
determine the percentage of inhibition.
8. Wong, O.; Tsuzuki, N.; Richardson, M.; Rytting, H.;
Konishi, R.; Higughi, T. Heterocycles 1987, 26, 3153.
9. Webb, R. L.; Labaw, C. S. J. Heterocyclic Chem. 1982, 19,
1205.
10. Bender, P. E.; Christensen, S. B.; Esser, K. M.; Forster, C.
J.; Ryan, M. D.; Simon, P. L. WO 9200968, 1991.
11. Middleton, W. J. US 3914265, 1976.
12. Inhibition of gastric emptying in rats: a 2-mL caloric test
meal (commercial Shak ISO from Roussel UCLA/nutrition
France; 1 kcal /mL) was given to male rats (200–250 g) pre-
treated orally for 30 min with test compound or solvent. The
weight of the stomach content was measured as the difference
in weight between the full and empty stomach 1 h after the
administration of the meal. Stomachs containing food residues
were discarded. Criterion for drug-induced inhibition of gas-
tric emptying: stomach contents >1.15 g (0% false positive
control rats). Compounds were prepared as solutions in poly-
ethyleneglycol 200 (PEG200) and administered in a volume of
1.25 mL/kg (higher volumes of PEG200 given orally inhibit
gastric emptying per se). For each compound, the LAD (low-
14. Marivet, M. C.; Bourguignon, J. J.; Lugnier, C.; Mann, A.;
Stoclet, J. C.; Wermuth, C. G. J. Med. Chem. 1989, 32, 1450.
15. Cheng, J. B.; Cooper, K.; Duplantier, A. J.; Eggler, J. F.;
Kraus, K. G.; Marshall, S. C.; Marfat, A.; Masamune, H.;
Shirley, T. J.; Tickner, J. E.; Umland, J. P. Bioorg. Med.
Chem. Lett. 1995, 5, 1969.
16. Inhibition of [3H]-rolipram binding to rat forebrain mem-
branes by selected compounds was tested as follows: to a mix
of 50 mM adenosine monophosphate and 0.05% BSA in 50
mM Tris buffer, pH 7.5, containing 5 mM MgCl2, [3H]-roli-
pram racemate (Amersham TRK1055, 85Ci/mmol, 2.35 mM)
and compound were added. The reaction was started by add-
ing membranes (1.45 mg/mL protein content) and incubating
the mixture for 30 min at 25 ꢀC. The reaction was stopped by
filtering the mixtures over GF/B filters (Whatman), presoaked
in 1 mg/mL BSA. The filtered protein was washed, dried and
the radioactivity on the filters quantified by liquid scintilliation
counting. Membranes were prepared from 12-g forebrains
from 10 rats of 250 g this is sufficient to do all the compounds
different times.