Imidazo[1,2-a]quinoxalines: Synthesis and cyclic nucleotide phosphodiesterase inhibitory activity

Article abstract of DOI:10.1016/S0223-5234(01)01213-2

A group of imidazo[1,2-a]quinoxalines have been synthesised from quinoxaline by condensation of an appropriate haloester or intramolecular cyclisation of a keto moiety on an intracyclic nitrogen atom. The reactivity of the heterocycle was explored through diverse reactions such as electrophilic substitution, lithiation and halogen-metal exchange to give access to a new series of derivatives. Confirmation of their structure was mainly performed by NMR, after careful assignment of the signals in comparison to previous attributions made on the parent imidazo[1,2-a]quinoxaline and discussion of available data in the literature. The cyclic nucleotide phosphodiesterase inhibitor activity of some of these derivatives has been assessed on isoenzymes type III and type lV. Compound 15, 4-(methylamino)imidazo[1,2-a]quinoxaline-2-carbonitrile, exhibited potent relaxant activity on smooth muscle, with a potency similar to the one measured with SCA 40, its structural analogue in the imidazo[1,2-a]pyrazine series.

Full text of DOI:10.1016/S0223-5234(01)01213-2

Eur. J. Med. Chem. 36 (2001) 255–264
´
© 2001 Editions scientifiques et me´dicales Elsevier SAS. All rights reserved
PII: S0223-5234(01)01213-2/SCO
Short communication
Imidazo[1,2-a]quinoxalines: synthesis and cyclic nucleotide phosphodiesterase
inhibitory activity
Ste´phanie Parraa, Florence Laurentb, Guy Subraa, Carine Deleuze-Masquefaa,
Veronique Benezecha, Jean-Roch Fabreguettesa, Jean-Pierre Vidalc, Tristan Pocockd,
Keith Elliottd, Roger Smalld, Roger Escalea, Alain Michelb, Jean-Pierre Chapata,
Pierre-Antoine Bonneta*
aPharmacochimie & Biomole´cules, EA 2414, Laboratoire de Chimie Organique, Faculte´ de Pharmacie,
15 Avenue Charles Flahault, 34060 Montpellier Cedex 2, France
bLaboratoire de Pharmacodynamie, Faculte´ de Pharmacie, 15 Avenue Charles Flahault, 34060 Montpellier Cedex 2, France
cUMR 5074 Chimie Biomole´culaire et Interactions Biologiques, Faculte´ de Pharmacie, 15 Avenue Charles Flahault,
34060 Montpellier Cedex 2, France
dPharmacology Department, University of Manchester, Oxford Road, Manchester, UK
Received 2 August 2000; revised 28 December 2000; accepted 28 December 2000
Abstract – A group of imidazo[1,2-a]quinoxalines have been synthesised from quinoxaline by condensation of an appropriate
haloester or intramolecular cyclisation of a keto moiety on an intracyclic nitrogen atom. The reactivity of the heterocycle was
explored through diverse reactions such as electrophilic substitution, lithiation and halogen–metal exchange to give access to a new
series of derivatives. Confirmation of their structure was mainly performed by NMR, after careful assignment of the signals in
comparison to previous attributions made on the parent imidazo[1,2-a]quinoxaline and discussion of available data in the literature.
The cyclic nucleotide phosphodiesterase inhibitor activity of some of these derivatives has been assessed on isoenzymes type III and
type lV. Compound 15, 4-(methylamino)imidazo[1,2-a]quinoxaline-2-carbonitrile, exhibited potent relaxant activity on smooth
muscle, with a potency similar to the one measured with SCA 40, its structural analogue in the imidazo[1,2-a]pyrazine series. © 2001
´
Editions scientifiques et me´dicales Elsevier SAS
imidazo[1,2-a]quinoxaline / lithiation / halogen–metal exchange / phosphodiesterase inhibition / smooth muscle relaxant activity
1. Introduction
It was then decided to study a new series of com-
pounds, with an imidazo[1,2-a]pyrazine structural ba-
SCA40, 6-bromo-8-(methylamino)imidazo[1,2-a]-
sic frame condensed with a phenyl on the 6–7 face,
pyrazine-2-carbonitrile (figure 1), has been shown to
i.e. imidazo[1,2-a]quinoxaline derivatives.
be a potent smooth muscle relaxant agent in vitro as
Therefore, substitutions on positions 1, 2, 4 (equiv-
well as in vivo [1–9].
alent positions 2, 3,
pyrazinic series) were modified in this imidazo[1,2-
8 in the imidazo[1,2-a]-
Structure–activity relationships studies in the imi-
dazo[1,2-a]pyrazine series have demonstrated that
substitution on position 2, 3 and 8 have an influence
on PDE III and/or IV inhibitory activity and, then,
on smooth muscle relaxant activity [10].
Abbreviations: PDE III, phosphodiesterase type III; PDE IV, phos-
phodiesterase type IV; cAMP, cyclic adenosine monophosphate;
DME, ethylene glycol dimethyl ether; DMF, dimethylformamide;
n-Buli, n-butyllithium; DMSO, dimethylsulfoxide.
* Correspondence and reprints
E-mail address: pabonnet@pharma.univ-montp1.fr (P.-A. Bonnet).
Figure 1. SCA 40.

256
S. Parra et al. / European Journal of Medicinal Chemistry 36 (2001) 255–264
a]quinoxalinic series in order to study their influence
on PDE III and/or IV inhibitory activity as well as on
smooth muscle relaxation.
(figure 4) was followed by oxidation of the hydroxyl
function of the intermediate alcohol 17 using the
sulfur trioxide trimethylamine complex as the oxi-
dant. The carbonyl compound 18 was then cyclised in
a
trifluoroacetic anhydride/trifluoroacetic acid
2. Chemistry
medium to give the 4-chloro-1-methylimidazo[1,2-
a]quinoxaline 19.
The 4-substituted imidazo[1,2-a]quinoxalines are
synthesised by an intramolecular cyclisation reaction
on an aromatic position. This cyclisation may occur
in either of two reactions leading to the formation of
either a quinoxaline [11, 12] or an imidazole ring.
In this study, we chose to build the imidazolic ring
starting from the quinoxaline. The classical Tschits-
chibabin condensation of an a-halogeno carbonyl
compound with 2-amino-3-chloroquinoxaline 2 led to
the formation of the 4-chloroimidazo[1,2-a]-
quinoxaline 3. Such strategy allows the access to
derivatives with some variety of substitutions on the
two-position of the heterocycle (figures 2 and 3).
Hydrogenation of 3 using Pd/C as catalyst led to the
imidazo[1,2-a]quinoxaline 8 for a basic NMR study.
In another strategy, the condensation of an
aminoalcohol with 2,3-dichloroquinoxaline 1 [11]
Diversity on position 1 was also obtained through a
reactivity study of the tricyclic heterocycle, either
under electrophilic substitution, lithiation or halogen-
metal exchange conditions [13]. These studies were
preferably performed on 4-methoxy- and not on 4-
chloroimidazo[1,2-a]quinoxaline in order first to in-
crease the reactivity of the aromatic position towards
electrophilic reagents and second, to avoid secondary
possible reactions on position 4 such as nucleophilic
substitution or halogen–metal exchange.
Hydroxymethylation was performed using the clas-
sical electrophilic substitution already described for
azaindolizines [14] and which have given interesting
results in some of these series known to be quite
reluctant to electrophilic substitution (32 and 41%
yield on imidazo[1,2-a]pyrimidine and imidazo[1,2-
a]pyrazine, respectively). Following the same proce-
dure,
1-(hydroxymethyl)-4-methoxyimidazo[1,2-a]-
quinoxaline 22 was obtained by direct reaction of
formaldehyde on 7 (figure 5) but with a very poor
yield of 5%. The formylation reaction was studied
under Vilsmeier–Haack conditions [15] but, unfortu-
nately, no substitution occurred, with only degrada-
tion products being obtained.
Since, as it is quite usual in these polyazaheterocy-
cles, electrophilic substitutions were difficult to
achieve, we were then interested in lithiation and
halogen–metal exchange reactions. A detailed study
of the reactivity of the heterocycle towards such reac-
tions is available elsewhere [13]. We then applied
already determined optimal conditions for the lithia-
tion of 4-methoxyimidazo[1,2-a]quinoxaline (2 equiv.
of n-BuLi at −75°C, followed by the addition of 10
equiv. of the electrophile at −75°C) [13] to obtain
various derivatives with substitutions on position 1 as
indicated in figure 5. The formylation, which was
shown to be impossible through Vilsmeier–Haack
conditions, was realised after lithiation and subse-
quent quenching with DMF [15]. The formylated
compound 23 was so obtained with a 60% yield. Alkyl
halides were also tested as alternative electrophilic
reagents in the quenching of intermediate lithiated
derivatives [16]. However, no reaction occurred with
Figure 2. Synthesis of 2–8.
Figure 3. Synthesis of 9–16.

S. Parra et al. / European Journal of Medicinal Chemistry 36 (2001) 255–264
257
The halogen–metal exchange reaction was also ap-
plied to 1-bromo-4-methoxyimidazo[1,2-a]quinoxa-
line 26, obtained upon N-bromosuccinimide bromina-
tion in dichloromethane of 4-methoxyimidazo[1,2-
a]quinoxaline [17]. The exchange was performed
through previously defined optimal lithiation condi-
tions [13], and with methyl iodide used as the elec-
trophile, 4-methoxy-1-methylimidazo[1,2-a]quinoxa-
line 27 was obtained with a 91% yield.
2.1. NMR study
Figure 4. Synthesis of 17–21.
1H- and 13C-NMR were used to unambiguously as-
sign the position of substitution on the heterocycle
following above described reactivity studies and, in a
more general concern, the exact chemical structure of
the newly synthesised imidazo[1,2-a]quinoxaline deriva-
tives. Our first attempt was to base our assignments on
data generated by Cobb and Cheeseman in 1985 [18] for
diverse two-ring and three-ring heterocycles, and in
particular for imidazo[1,2-a]quinoxaline (table I). In
1
their H-NMR study, which was performed in CDCl3,
H-6 (8.05 ppm) was more deshielded than H-9 (7.83
ppm). From our carbon–proton correlation study, per-
formed in DMSO-d6, we observed, for the same imi-
dazo[1,2-a]quinoxaline and in contradiction with the
above data, a correlation between C-6 at 129.8 ppm and
the shielded proton at 8.06 ppm and, by the way,
another correlation between C-9 at 116 ppm and the
deshielded proton at 8.38 ppm.
In order to confirm all carbon and proton NMR
assignments, we carried out a n.O.e. study. In this study,
irradiation of H-1 at 8.81 ppm showed correlations with
two protons at 7.85 and 8.38 ppm. The proton at 8.38
ppm also exhibited correlations with two protons at 8.81
and 7.75 ppm. On the other hand, the proton at 8.06
Figure 5. Synthesis of 22–27.
bromoalkanes and, even when the quenching reaction
was performed at increasing temperature (−75°C to
room temperature), only starting material was
recovered.
Table I. Proton and carbon-13 chemical shifts (ppm) for the imidazo[1,2-a]quinoxaline.
a
b
b
b
l
(CDCl3)
l
(CDCl3)
l
(DMSO-d6)
Proton assignment
l
(DMSO-d6)
Carbon-13 assignment
9.057
8.063
8.051
7.828
7.759
7.592
7.526
9.11
8.84
8.06
8.40
7.86
7.76
7.64
9.095
8.81
8.06
8.38
7.85
7.75
7.64
H-4
H-1
H-6
H-9
H-2
H-8
H-7
143.9
134.2
129.8
129.0
126.4
116.0
114.1
C-4
C-2
C-6
C-8
C-7
C-9
C-1
a Data and assignments from Cobb and Cheeseman [18].
b Our study.

258
S. Parra et al. / European Journal of Medicinal Chemistry 36 (2001) 255–264
Table II. Imidazo[1,2-a]quinoxalines.
Compound
R1
R2
R4
M.p. (°C)
Formula
3
4
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CO2Et
CO2Et
Cl
NH2
192
200
178
106
153
122
203
194
\250
\250
220
180
182
186–187
139
184
210
167
C10H6N3Cl
C10H8N4
C11H10N4
C12H12N4
C11H9N3O
C10H7N3
C14H14N4O2
C15H16N4O2
C12H11N5O
C13H13N5O
C12H9N5
C13H11N5
C11H8N3Cl
C12H12N4
5
NHCH3
NHC2H5
OCH3
H
NHCH3
NHC2H5
NHCH3
NHC2H5
NHCH3
NHC2H5
Cl
NHCH3
NHC2H5
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
6
7
8
11
12
13
14
15
16
19
20
21
22
23
24
25
26
27
CONH2
CONH2
CN
CN
H
H
H
H
H
H
H
CH3
CH3
CH3
CH2OH
CHO
CH(OH)–C2H5
CH(OH)–CH2Ph
Br
C13H14N4
C12H11N3O2
C12H9N3O2
C14H15N3O2
C19H17N3O2
C11H8N3OBr
C12H11N3O
202
146
165
H
H
CH3
zymes of cyclic nucleotide phosphodiesterase was
measured in vitro. The results are summarised in table
III. All tested compounds exhibited relative potency
as inhibitors of isoenzyme type IV, their activities
being similar or higher to that of the selective in-
hibitor Ro 20-1724 (table III). Most of them pre-
sented a very low PDE III inhibitory activity as
compared to the reference compound SKF 94120
(table III). Only compound 15, with the cyano group
on position 2, showed moderate inhibitory activity
against isoenzyme type III. 15 exhibited an unique
profile since it appeared to be PDE III and IV in-
hibitor, without selectivity. These results could be
compared to those previously obtained in a imi-
dazo[1,2-a]pyrazine series [10]. It has been published
that these compounds exhibited PDE IV activity
higher than the PDE III activity with the exception of
two compounds, with a cyano group on position 2,
ppm exhibited a correlation with only one proton at
7.64 ppm.
These results led us to assign with absolute certainty
the deshielded peak at 8.38 ppm to H-9, which corre-
lates with H-1 (8.81 ppm) and H-8 (7.75 ppm), and the
shielded peak at 8.06 ppm to H-6 which correlates only
with H-7 (7.64 ppm). Such study reveals that Cobb and
Cheeseman had confused the NMR assignment of the
protons H-6 and H-9, the proton H-9 being actually
more deshielded than the proton H-6.
All synthesised imidazo[1,2-a]quinoxaline derivatives
1
are described in table II. Their H-NMR assignments
are listed in the experimental part.
3. Biological results and discussion
The inhibitory activity of type III and/or IV isoen-

S. Parra et al. / European Journal of Medicinal Chemistry 36 (2001) 255–264
259
which showed potent inhibitory activity against isoen-
zyme type III. It is interesting to note that the only
compound exhibiting potent PDE III inhibitory activ-
ity exhibits a large structural analogy (cyano and
aminomethyl groups on equivalent positions) with
SCA40 which was shown, in the imidazo[1,2-
a]pyrazine series, to be the more potent PDE III
inhibitor [10]. We can also note that the presence, in
the imidazoquinoxaline series, of a supplementary
phenyl on face e of the parent imidazo[1,2-a]pyrazinic
heterocycle induced a dramatic loss of inhibitory ac-
tivity against PDE III and a one to half a log decrease
against PDE IV in comparison with the imidazopy-
razine series.
bromine substituent, appears to be the most potent
PDE IV selective inhibitor.
The relaxant activity of guinea-pig isolated trachea
was measured for two characteristic compounds: 15,
unselective PDE III and IV inhibitor and structural
analogue to SCA40 in the imidazo[1,2-a]pyrazine se-
ries, and 26, the more potent PDE IV inhibitor of the
series (table IV). These two compounds exhibited a
higher or quite similar activity compared to milri-
none, a selective PDE III inhibitor. The most potent
derivative was shown to be compound 15 with a pD2
of 6.02+0.10, similar to the value measured in the
same conditions for SCA40. So, like in the imi-
dazo[1,2-a]pyrazine series [10], the potent relaxant
activity of these compounds is related in a complex
way to its simultaneous inhibition of PDE isoenzymes
III and IV.
Inhibitory activity against isoenzyme type IV de-
pends strongly on the nature of the substitution on
position 1. The methyl, hydroxypropyl or bromine
groups on position 1 favoured potent inhibitory activ-
ity against this isoenzyme. Compound 26, with a
4. Experimental
4.1. Chemistry
Table III. Phosphodiesterase inhibitory activity of the imi-
dazo[1,2-a]quinoxaline derivatives a.
All melting points were determined using a Ko¨fler hot
plate melting point apparatus and are uncorrected.
Purification was checked by thin-layer chromatography
on silica gel (230–240 mesh from Merck). 1H-NMR
spectra were recorded using a Brucker AC 100 or 200
spectrometer. Elemental analysis were performed by the
Microanalytical Center (Montpellier, France). Analyses
indicated by the symbols of the elements or functions
were within 0.4% of the theorical values.
Compound −log IC50 PDE III
−log IC50 PDE IV
SKF 94120 5.1490.06
Ro 20-1724 3.8390.13
B3.0
4.6390.20
4.8990.12
4.7990.21
4.8390.21
5.0890.23
5.5090.05
4.7090.15
4.8890.18
5.4590.15
4.6590.05
6.1990.09
5
7
B3.0
B3.0
11
15
20
22
23
24
25
26
B3.0
5.2290.07
B3.0
B3.0
B3.0
4.1.1. 2-Amino-3-chloroquinoxaline (2)
B3.0
A mixture of 2,3-dichloroquinoxaline 1 (9.2 g, 46.2
mmol) in 150 mL of ammonia solution in DMF was
stirred at 0°C for 1 h. The mixture was filtered and
concentrated in vacuo. The residue was chro-
matographed on a silica gel column (eluent: methanol–
dichloromethane: 5/95) to give 2 (7.5 g, 90%); m.p.
B3.0
4.2990.09
a Data indicate mean (9S.E. mean) values of −log IC50 for
each compound.
1
Table IV. Activity of the imidazo[1,2-a]quinoxaline deriva-
139°C; H-NMR (CDC13) l (100 MHz): 7.61 (m, 4H,
tives as relaxants of guinea-pig isolated trachea a.
H5+H6+H7+H8), 5.51 (br s, 2H, NH2). Anal.
(C8N3H6Cl) C, H, N.
Compound
n
−log EC50
4.1.2. 4-Chloroimidazo[1,2-a]quinoxaline (3)
Milrinone
Ro 20-1724
SCA 40
15
6
6
10
10
10
5.2990.15
3.5090.23
6.0690.14
6.0290.10
4.6090.03
A mixture of bromoacetaldehyde diethyl acetal (18.7
g, 95 mmol), concentrated hydrobromic acid (2.8 mL),
and water (11.3 mL) was heated under reflux for 1 h.
The mixture was then cooled, diethylether was added,
and the layers were separated. The organic phase was
dried and poured into a solution of 2 (7 g, 39 mmol) in
26
a Data are mean (9S.E. mean) values (n=6 or 10) of pD2
(−log EC50).

260
S. Parra et al. / European Journal of Medicinal Chemistry 36 (2001) 255–264
DME (7.5 mL). The reaction mixture was stirred at
room temperature for 2 days and the precipitate was
separated by filtration. A mixture of this precipitate in
dry ethanol (75 mL) was heated under reflux for 2 h.
The solvent was removed in vacuo and the residue was
purified by chromatography on silica (eluent:
methanol–dichloromethane: 5/95) to give 3 (3.89 g,
H7), 7.78 (br t, 1H, NH), 3.69 (m, 2H, CH2), 1.34 (t, 3H,
CH3). Anal. (C12N4H12) C, H, N.
4.1.6. 4-Methoxyimidazo[1,2-a]quinoxaline (7)
A methanolic solution of sodium methylate was pre-
pared from sodium (0.23 g, 10 mmol) in dry methanol
(70 mL). 3 (1 g, 4.9 mmol) was added and the resulting
solution was heated under reflux for 2 h and stirred
overnight at room temperature. The reaction mixture
was evaporated to dryness in vacuo and the residue was
redissolved in dichloromethane (150 mL), washed with
sodium chlorure (100 mL), water (100 mL), dried and
concentrated to give the crude product. Purification by
column chromatography (eluent: dichloromethane–
methanol: 98/2) gave the compound 7 (0.93 g, 95%);
1H-NMR (CDC13) l (200 MHz): 8.77 (d, 1H, H1), 8.27
(dd, 1H, H9), 7.81 (dd, 1H, H6), 7.74 (d, 1H, H2), 7.56
(m, 2H, H7+H8), 4.15 (s, 3H, OCH3). Anal. (C11N3H9O)
C, H, N.
1
49%); H-NMR (DMSO-d6) l (200 MHz): 9.05 (d, 1H,
H1), 8.51 (dd, 1H, H1), 8.1 (dd, 1H, H6), 7.99 (d, 1H,
H2), 7.89 (td, 1H, H8), 7.76 (td, 1H, H7). Anal.
(C10N3H6Cl) C, H, N.
4.1.3. 4-Aminoimidazo[1,2-a]quinoxaline (4)
A mixture of 3 (2.03 g, 100 mmol) in 30% ammonia
solution (20 mL) was heated at 100°C for 4 h in a 125
mL autoclave. The mixture was then cooled at room
temperature and extracted with dichloromethane. The
organic phase was dried over anhydrous sodium sulfate.
After evaporation in vacuo, the residue was chro-
matographed on silica gel (eluent: dichloromethane–
methanol: 90/10) to give 4 (0.47 g, 80%); 1H-NMR
(DMSO-d6) l (200 MHz): 8.71 (d, 1H, H1,), 8.20 (dd,
1H, H9), 7.73 (d, 1H, H2), 7.65 (dd, 1H, H6), 7.5 (td, 1H,
H8), 7.39 (td, 1H, H7), 7.3 (br s, 2H, NH2). Anal.
(C10N4H8) C, H, N.
4.1.7. Imidazo[1,2-a]quinoxaline (8)
A mixture of 3 (1 g, 4.9 mmol) solubilised in anhy-
drous methanol was hydrogenated in presence of 10%
Pd/C at air-pressure. The mixture was filtrated on celite.
The filtrate was evaporated, then the residue was solu-
bilised in minimum methanol and precipited by ether
addition. The precipitate was filtrated, solubilised in
15% sodium hydroxide solution and extracted by
dichloromethane. The organic phase was dried over
calcium chloride and evaporated to give 8 (0.81 g, 98%);
1H-NMR (CDCl3) l (200 MHz): 9.11 (s, 1H, H4), 8.84
(d, 1H, H1), 8.40 (dd, 1H, H9), 8.06 (dd, 1H, H6), 7.86
(d, 1H, H2), 7.76 (ddd, 1H, H8), 7.64 (ddd, 1H, H7).
Anal. (C10N3H7) C, H, N.
4.1.4. 4-(Methylamino)imidazo[1,2-a]quinoxaline (5)
A 40% aqueous methylamine solution (2.55 mL, 29.7
mmol) was added to a solution of 3 (1 g, 4.9 mmol) in
absolute ethanol (25 mL). The resulting solution was
stirred at room temperature for 24 h. Ethanol was
removed in vacuo. The residue was redissolved in
dichloromethane (80 mL), washed with 5% sodium hy-
drogen carbonate (50 mL) and water (50 mL), dried and
concentrated to give the crude product. Purification by
column chromatography (eluent: ethyl acetate–hexane:
4.1.8. 2-Amino-3-(methylamino)quinoxaline (9)
A mixture of 2 (10.4 g, 58 mmol) in a 40% aqueous
methylamine solution (36 mL, 419 mmol) was stirred at
1
30/70) gave the compound 5 (0.81 g, 83%); H-NMR
(DMSO-d6) l (200 MHz): 8.68 (d, 1H, H1), 8.18 (dd,
1H, H9), 7.7 (dd, 1H, H6), 7.7 (d, 1H, H2), 7.5 (td, 1H,
H8), 7.41 (td, 1H, H7), 7.63 (br q, 1H, NH), 3.13 (d, 3H,
CH3). Anal. (C11N4H10) C, H, N.
room
temperature
overnight.
Extraction
with
dichloromethane and chromatography on a silica gel
column (eluent: methanol–methylene chloride: 5/95)
provided the desired compound 9 (5.16 g, 51%); 1H-
NMR (DMSO) l (100MHz): 7.45 (m, 4H, H5+H6+
H7+H8), 6.95 (m, 1H, NH), 6.64 (s, 2H, NH2), 3.00 (d,
3H, CH3). Anal. (C9N4H10) C, H, N.
4.1.5. 4-(Ethylamino)imidazo[1,2-a]quinoxaline (6)
It was prepared in an analogous manner of 5 from
4-chloroimidazo[1,2-a]quinoxaline 3 and a 70% aqueous
ethylamine
solution.
Chromatography
(eluent:
methanol–dichloromethane: 10/90) yielded the com-
pound 6 (0.93 g, 89%); 1H-NMR (DMSO-d6) l (200
MHz): 8.67 (d, 1H, H1), 8.17 (dd, 1H, H9), 7.7 (d, 1H,
H2), 7.67 (dd, 1H, H6), 7.48 (td, 1H, H8), 7.36 (td, 1H,
4.1.9. 2-Amino-3-(ethylamino)quinoxaline (10)
It was prepared in an analogous manner of 9 from 2
and a 70% aqueous ethylamine solution. Chromatogra-
phy (eluent: methanol–methylene chloride: 10/90) pro-

S. Parra et al. / European Journal of Medicinal Chemistry 36 (2001) 255–264
261
1
vided the desired compound 10 (6.86 g, 63%); H-NMR
(CDCl3) l (100 MHz): 7.47 (m, 4H, H5+H6+H7+H8),
4.59 (br m, 3H, NH+NH2), 3.57 (m, 2H, CH2), 1.31 (t,
3H, CH3). Anal. (C10N4H12) C, H, N.
boxylate (12). Recrystallisation from methanol afforded
14 (0.52 g, 56%). Anal. (C13N5H13O) C, H, N.
4.1.14. 4-(Methylamino)imidazo[1,2-a]quinoxaline-2-
carbonitrile (15)
4.1.10. Ethyl-4-(methylamino)imidazo[1,2-a]-
quinoxaline-2-carboxylate (11)
A mixture of 13 (0.7 g, 2.9 mmol) in phosphorus
oxychloride (25 mL) was heated under reflux for 1 h 30.
After distillation of phosphorus oxychloride (23 mL),
the residue was poured carefully onto ice. After the
destruction of excess phosphorus oxychloride, the mix-
ture was alkalinised with sodium carbonate, and the
nitrile was extracted with dichloromethane. The solvent
was removed in vacuo and the residue purified by
2-Amino-3-(methylamino)quinoxaline (9) (5 g, 28.7
mmol) was dissolved in DME (30 mL) and DMF (20
mL). A solution of ethyl-3-bromopyruvate (3.64 mL,
28.7 mmol) in DME (6 mL) was added dropwise and
the resulting solution was heated at 70°C for 2 h. The
precipitate obtained was separated by filtration, dis-
solved in dry ethanol and heated under reflux for 1 h.
The filtrate was evaporated to dryness in vacuo, the
residue was dissolved in dry ethanol and heated under
reflux for 1 h. Upon cooling, the two ethanolic solutions
gave a precipitate. After filtration and chromatography
on a silica gel column (eluent: methanol–dichloro-
methane: 2/98) yielded 11 (2.72 g, 35%); 1H-NMR
(DMSO-d6) l (200 MHz): 9.29 (s, 1H, H1), 8.27 (dd, 1H,
H9), 8.07 (br q, 1H, NH), 7.61 (dd, 1H, H6), 7.44 (td,
1H, H8), 7.28 (td, 1H, H7), 4.37 (q, 2H, CH2), 3.01 (d,
3H, CH3), 1.35 (t, 3H, CH3). Anal. (C14N4H14O2) C, H,
N.
chromatography on
a silica gel column (eluent:
dichloromethane–methanol: 95/5) to give 15 (0.56 g,
1
87%); H-NMR (CDCl3) l (200 MHz): 9.51 (s, 1H, H1),
8.13 (br q, 1H, NH), 8.10 (dd, 1H, H9), 7.63 (dd, 1H,
H6), 7.48 (td, 1H, H8), 7.34 (td, 1H, H7), 3.02 (d, 3H,
CH3). Anal. (C12N5H9) C, H, N.
4.1.15. 4-(Ethylamino)imidazo[1,2-a]quinoxaline-2-
carbonitrile (16)
It was prepared in an analogous manner of 15 from
4-(ethylamino)imidazo[1,2-a]quinoxaline-2-carboxamide
(14). Chromatography (eluent: dichloromethane–
1
methanol: 95/5) yielded to 16 (0.58 g, 85%); H-NMR
4.1.11. Ethyl-4-(ethylamino)imidazo[1,2-a]-
quinoxaline-2-carboxyIate (12)
(CDCl3) l (200 MHz): 9.51 (s, 1H, H1), 8.13 (br t, 1H,
NH), 8.11 (dd, 1H, H9), 7.60 (dd, 1H, H6), 7.47 (td, 1H,
H8), 7.33 (td, 1H, H7), 3.55 (q, 2H, CH2), 1.24 (t, 3H,
CH3). Anal. (C13N5H11) C, H, N.
It was prepared in an analogous manner of 11 from
2-amino-3-(ethylamino)quinoxaline (10). Chromatogra-
phy (eluent: methanol–dichloromethane: 2/98) yielded
1
12 (2.77 g, 34%); H-NMR (CDCl3) l (100 MHz): 8.52
4.1.16. 2-(2-hydroxypropylamino)-3-chloroquinoxaline
(17)
(s, 1H, H1), 7.66 (m, 2H, H9+H6), 7.42 (m, 2H, H8+H7),
6.29 (br t, 1H, NH), 4.47 (q, 2H, CH2), 3.73 (q, 2H,
CH2), 1.43 (t, 3H, CH3), 1.33 (t, 3H, CH3). Anal.
(C15N4H16O2) C, H, N.
1-Aminopropan-2-ol (11.6 mL, 150.75 mmol) was
added to a solution of 2,3-dichloroquinoxaline (1) (12 g,
60.3 mmol) in dioxane (60 mL). The resulting solution
was heated under reflux for 7 h. Dioxane was removed
in vacuo. The residue was redissolved in dichloro-
methane (100 mL), washed with 10% sodium carbonate
(60 mL), water (60 mL), dried and concentrated to give
the product 17 (9 g, 63%); m.p. 104°C; 1H-NMR
(CDCl3) l (100 MHz): 7.53 (m, 4H, H5+H6+H7+H8),
5.99 (br t, 1H, NH), 4.17 (br d, 1H, OH), 3.73 (m, 1H,
CH), 3.48 (dd, 2H, CH2), 1.28 (d, 3H, CH3). Anal.
(C11N3H12OCl) C, H, N.
4.1.12. 4-(Methylamino)imidazo[1,2-a]-
quinoxaline-2-carboxamide (13)
A mixture of 11 (1 g, 3.7 mmol) in ammonia solution
in methanol (500 mL) and sodium cyanide (0.1 g, 2
mmol) was stirred at 0°C for 16 h. The resulting solu-
tion was filtered and concentrated in vacuo. The residue
was purified by recrystallisation from methanol to give
13 (0.47 g, 53%). Anal. (C12N5H11O) C, H, N.
4.1.13. 4-(Ethylamino)imidazo[1,2-a]quinoxaline-2-
carboxamide (14)
It was prepared in an analogous manner of 13 from
ethyl - 4 - (ethylamino)imidazo[1,2 - a]quinoxaline - 2 - car-
4.1.17. 2-(2-Oxopropylamino)-3-chloroquinoxaline (18)
Sulfur trioxyde trimethylamine complex (25 g, 180
mmol) was added slowly to a solution of 17 (9 g, 37.9
mmol) in triethylamine (60 mL) and DMSO (60 mL).

262
S. Parra et al. / European Journal of Medicinal Chemistry 36 (2001) 255–264
The resulting solution was stirred for 16 h at room
temperature. The mixture was then poured carefully
onto ice and extracted with dichloromethane. The sol-
vent was removed in vacuo and the resulting brown oil
purified by chromatography on a silica gel column
(eluent: dichloromethane–ether: 50/50) to give 18 (6.4 g,
4.1.21. 1-(Hydroxymethyl)-4-methoxyimidazo[1,2-a]-
quinoxaline (22)
A mixture of 7 (2 g, 10 mmol), sodium acetate (32.6 g,
397 mmol), acetic acid (25.2 mL, 441 mmol), and 37%
solution of formaldehyde (225.8 mL, 8130 mmol) in
water was heated in an autoclave at 120°C for 2.5 h. On
cooling, water (120 mL) was added to the reaction
mixture. After alkalinisation with sodium carbonate and
extraction with dichloromethane, the organic phase was
dried and evaporated. The residue was chro-
matographed on a silica gel column (eluent: methanol–
dichloromethane: 4/96) to give the unreacted product 7
(1.05 g). Further chromatography (eluent: methanol–
1
72%); m.p. 154°C; H-NMR (CDCl3) l (100 MHz): 7.58
(m, 4H, H5+H6+H7+H8), 6.42 (br t, 1H, NH), 4.46 (d,
2H, CH2), 2.33 (s, 3H, CH3). Anal. (C11N3H10OC1) C,
H, N.
4.1.18. 4-Chloro-1-methylimidazo[1,2-a]quinoxaline (19)
18 (6.35 g, 27 mmol) were dissolved in trifluoroacetic
acid (19 mL) under nitrogen. Then, trifluoroacetic anhy-
dride (10 mL) were added slowly. The resulting solution
was stirred for 5 h at room temperature under nitrogen.
After evaporation in vacuo, the residue was redissolved
in dichloromethane (50 mL), washed with 10% sodium
hydrogen carbonate (30 mL), water (30 mL), dried and
concentrated. Chromatography on a silica gel column
(eluent: dichloromethane–ether: 80/20) provided the de-
1
dichloromethane: 10/90) afforded 22 (0.115 g, 5%); H-
NMR (DMSO-d6) l (100 MHz): 8.42 (m, 1H, H9), 7.80
(m, 1H, H6), 7.61 (s, 1H, H2), 7.54 (m, 2H, H7+H8),
5.68 (t, 1H, OH), 5.01 (d, 2H, CH2), 4.14 (s, 3H,
O-CH3). Anal. (C12N3H11O2) C, H, N.
4.1.22. 1-Formyl-4-methoxyimidazo[1,2-a]quinoxaline
(23)
1
sired compound 19 (2.35 g, 40%); H-NMR (CDCl3) l
n-BuLi (2.5 mL, 4 mmol) was added dropwise to a
solution of 7 (0.4 g, 2 mmol) in dry THF (15 mL) under
nitrogen at −75°C. The resulting solution was stirred for
1 h at −75°C. Freshly distilled DMF (1.55 mL, 20
mmol) was then added and the resulting solution was
stirred at −75°C for further 15 min before hydrolysis
(−75°C) by water. The precipitate obtained was sepa-
rated by filtration, washed with water and ether. The
crude product was chromatographed on a silica gel
column (eluent: methanol–methylene chloride: 2/98) to
(100 MHz): 8.14 (dd, 1H, H9), 7.97 (dd, 1H, H6), 7.73
(td, 1H, H8), 7.65 (td, 1H, H7), 7.64 (s, 1H, H2), 2.94 (s,
3H, CH3). Anal. (C11N3H8Cl) C, H, N.
4.1.19. 1-Methyl-4-(methylamino)imidazo[1,2-a]-
quinoxaline (20)
A mixture of 19 (1 g, 4.6 mmol) in 40% aqueous
methylamine solution (2.37 mL, 27.6 mmol) was stirred
at room temperature overnight. Extraction with
dichloromethane and chromatography on a silica gel
column (eluent dichloromethane–ether: 50/50) provided
the desired compound 20 (0.8 g, 82%); 1H-NMR
(CDCl3) l (100 MHz): 8.16 (dd, 1H, H9), 7.70 (br q, 1H,
NH), 7.62 (dd, 1H, H6), 7.39 (td, 1H, H8), 7.33 (s, 1H,
H2), 7.25 (td, 1H, H7), 3.02 (d, 3H, CH3), 2.84 (s, 3H,
CH3). Anal. (C12N4H12) C, H, N.
1
give 23 (0.27 g, 60%); H-NMR (CDCl3) l (100 MHz):
9.98 (s, 1H, CHO), 9.39 (m, 1H, H9), 8.38 (s, 1H, H2),
7.87 (m, 1H, H6), 7.58 (m, 2H, H7+H8), 4.26 (s, 3H,
OCH3). Anal. (C12N3H9O2) C, H, N.
4.1.23. 1-(1-Hydroxypropyl)-4-methoxyimidazo[1,2-a]-
quinoxaline (24)
n-BuLi (5 mL, 8 mmol) was added dropwise to a
solution of 7 (0.8 g, 4 mmol) or 26 (1.11 g, 4 mmol) in
dry tetrahydrofuran (16 mL) under nitrogen at −75°C.
The resulting solution was stirred for 1 h at −75°C.
Propanal (2.88 mL, 40 mmol) was then added and the
resulting solution was stirred at −75°C for further 1 h
before hydrolysis (−75°C) by tetrahydrofuran–ethanol–
35% HCl (1/1/1). The reaction mixture was then washed
with saturated aqueous sodium hydrogen carbonate so-
lution, extracted with dichloromethane, dried and the
solvent was removed in vacuo. The residue was purified
by chromatography on a silica gel column (eluent:
4.1.20. 4-(Ethylamino)-1-methylimidazo[1,2-a]-
quinoxaIine (21)
It was prepared in an analogous manner of 20 from
4-chloro-1-methylimidazo[1,2-a]quinoxaline 19 and a
70% aqueous ethylamine solution. Chromatography
(eluent: ether–dichloromethane: 20/80) provided the de-
1
sired compound 21 (0.71 g, 68%); H-NMR (CDCl3) l
(100 MHz): 8.15 (dd, 1H, H9), 7.60 (dd, 1H, H6), 7.54
(br t, 1H, NH), 7.39 (td, 1H, H8), 7.33 (s, 1H, H2), 7.25
(td, 1H, H7), 3.57 (m, 2H, CH2), 2.85 (s, 3H, CH3), 1.20
(d, 3H, CH3). Anal. (C13N4H14) C, H, N.

S. Parra et al. / European Journal of Medicinal Chemistry 36 (2001) 255–264
263
methanol–methylene chloride: 1/99) to give 24 (0.95 g,
92%); H-NMR (DMSO-d6) l (100 MHz): 8.63 (m, 1H,
4.2. Biochemistry and pharmacology
1
H9), 7.88 (m, 1H, H6), 7.73 (s, 1H, H2), 7.61 (m, 2H,
H7+H8), 5.78 (d, 1H, OH), 5.16 (m, 1H, CH), 4.20 (s,
3H, OCH3), 2.09 (m, 2H, CH2), 1.10 (t, 3H, CH3). Anal.
(C14N3H15O2) C, H, N.
4.2.1. Measurement of inhibitory activity against cyclic
nucleotide phosphodiesterase (PDE) isoenzymes
Isoenzymes of cyclic nucleotide PDE were isolated
from guinea-pig cardiac ventricles (type III) and bovine
trachealis (type IV) as previously described [19]. The
identities of the isoenzymes were confirmed by their
sensitivities to SKF 94120 and Ro 20-1724, respective
PDE III and PDE IV selective inhibitors (table III), and
cyclic GMP (inhibitor of PDE III but not PDE IV). The
isoenzymes used for kinetic analysis were at least 90%
pure based on their sensitivities to the above agents.
The activity of isoenzymes III and IV was measured
essentially by the method of Thompson and Appleman
[20] as modified by Rutten et al. [21]. Assays of enzyme
activity were performed in a final volume of 100 mL
comprising 25 mL of the isoenzyme solution, 50 mL of
assay buffer and 25 mL of twice-distilled water or PDE
inhibitor (imidazo[1,2-a]pyrazine derivative) solution.
The assay buffer (pH 8.0) contained 0.2 mCi [3H]-cAMP
and yielded final concentrations of 1 mM cAMP, 40 mM
Tris-HCl, 2.5 mM MgCl2 and 3.75 mM b-mercapto–
ethanol in the reaction mixture. 10 mM stock solutions
of imidazo[1,2-a]pyrazine derivatives were prepared in
ethanol. Dilutions from these stock solutions were pre-
pared using twice-distilled water. In tests of enzyme
inhibition, the reaction mixture contained concentra-
tions of imidazo[1,2-a]pyrazine derivatives in the range
of 10 nM–100 mM.
The reagents were mixed on ice and the reaction was
initiated by transferring the mixture to a water bath at
37°C. Following 30 min incubation, the reaction was
stopped by transferring the reaction tubes to a bath of
boiling water for 3 min. After cooling on ice, 20 mL of
a 1 mg mL−1 solution of Ophiophagus hannah venom
was added to each tube and the mixture was incubated
at 37°C for 10 min. Unreacted [3H]-cAMP was removed
by the addition of 400 mL of a 1 in 3 suspension of
Dowex resin (1×8-400) and incubation on ice for 30
min. Each tube was then centrifuged (10.000 g) for 2
min and 200 mL of the supernatant was removed for
liquid scintillation counting. Less than 10% of the triti-
ated cAMP was hydrolysed in any assay.
4.1.24. 1-(1-hydroxy-2-phenylethyl)-4-methoxyimidazo-
[1,2-a]quinoxaline (25)
It was prepared in an analogous manner of 24 from 7
and phenylacetaldehyde. Chromatography (eluent:
dichloromethane) yielded 25 (0.64 g, 50%); 1H-NMR
(DMSO-d6) l (100 MHz): 8.52 (m, 1H, H9), 7.75 (m,
1H, H6), 7.74 (s, 1H, H2), 7.50 (m, 2H, H7+H8), 7.23
(m, 5H, HAr), 5.88 (d, 1H, OH), 5.42 (m, 2H, CH+CH),
4.09 (s, 3H, OCH3). Anal. (C19N3H17O2) C, H, N.
4.1.25. 1-Bromo-4-methoxyimidazo[1,2-a]quinoxaline
(26)
A solution of 7 (0.5 g, 2.5 mmol) and N-bromosuccin-
imide (0.44 g, 2.5 mmol) in CHCl3 (25 mL) was heated
under reflux for 2h. The resulting green solution was
cooled, washed with 5% sodium hydrogen carbonate (15
mL), dried and concentrated. The residue was purified
by recrystallisation from chloroform to give 26 (0.64 g,
91%); 1H-NMR (CDCl3) l (100 MHz): 9.04 (m, 1H,
H9), 7.81 (m, 1H, H6), 7.75 (s, 1H, H2), 7.50 (m, 2H,
H7+H8), 4.20 (s, 3H, OCH3). Anal. (C11N3H8OBr) C,
H, N.
4.1.26. 4-Methoxy-1-methylimidazo[1,2-a]quinoxaline
(27)
n-BuLi (5 mL, 8 mmol) was added dropwise to a
solution of 26 (1.11 g, 4 mmol) in dry tetrahydrofuran
(16 mL) under nitrogen at -75°C. The resulting solution
was stirred for 1 h at −75°C. Methyl iodide (2.49 mL,
40 mmol) was then added and the resulting solution was
stirred at −75°C for further 1 h before hydrolysis
(−75°C) by tetrahydrofuran–ethanol–35% HCl (1/1/1).
The reaction mixture was then washed with saturated
aqueous sodium hydrogenocarbonate solution, extracted
with dichloromethane, dried and the solvent was re-
moved in vacuo. The residue was purified by chro-
matography on a silica gel column (eluent: methanol–
methylene chloride: 1/99) to give 27 (0.77 g, 91%);
1H-NMR (CDCl3) l (100 MHz): 7.92 (m, 1H, H9),
7.65 (m, 1H, H6), 7.27 (m, 3H, H2+H7+H8), 4.14 (s,
3H, OCH3), 2.77 (s, 3H, CH3). Anal. (C12N3H11O) C, H,
N.
The test compound concentration producing 50% in-
hibition of PDE activity (IC50 value) was determined for
each compound by use of a non-linear regression curve
fitting program.

264
S. Parra et al. / European Journal of Medicinal Chemistry 36 (2001) 255–264
[4] Naline E., Cui Y.Y., Michel A., Bonnet P.A., Bakdach H.,
The results presented in table III are means S.E.M.
of at least four experiments.
Advenier C., Fundam. Clin. Pharmacol. 10 (1996) 368–378.
[5] Cortijo J., Villargrasa V., Navarette C., Sanz C., Berto L.,
Michel A., Bonnet P.A., Morcillo E.J., Br. J Pharmacol. 119
(1996) 99–106.
4.2.2. Guinea-pig isolated trachea preparation
Adult male Dunkin–Hartley guinea-pigs (Iffa Credo,
France), weighing 400–500 g, were killed by a blow to
the head. Tracheae were excised and cleaned of adhering
adipose and connective tissue. The contractility of tra-
cheal segments (four tracheal rings in all cases) was
measured by adapting the method previously described
[22]. Tissue segments were suspended in Krebs’ solution
maintained at 37°C and gassed with a mixture of 95%
O2, 5% CO2. All tissues were connected to an isometric
force-displacement transducer under a basal tension of
0.5 g. The force-displacement transducer provided input
to a Physiograph Narco Bio-system. The bronchocon-
strictor agent (1 mM carbachol) induced a contraction
which reached a plateau within 5 min. Cumulative log
concentration response curves were constructed for each
test compound, relaxation being measured as the per-
centage reduction in carbachol-induced contraction. Re-
laxant potency of each compound was expressed as the
negative log EC50 (pD2), where EC50 is the concentra-
tion producing 50% inhibition of the contraction. The
EC50 values were calculated by linear regression analysis
applied to the linear portion of each concentration–re-
sponse curve. The results presented in table IV are
means S.E.M. of six or ten determinations.
[6] Buchheit K.H., Hofmann A., Pfannkuche H., J. Naunyn-
Schmiedeberg’s Arch. Pharmacol. 355 (1997) 217–223.
[7] Mu¨ller-Schweinitzer E., Fozard J.R., Br. J. Pharmacol. 120
(1997) 1241–1248.
[8] Pocock T., Laurent F., Isaac L.M., Chiu P., Elliott K.R.F.,
Poster R.W., Michel A., Bonnet P.A., Small R.C., Eur. J.
Pharmacol. 334 (1997) 75–81.
[9] Bonnet P.A., Michel A., Laurent F., Sablayrolles C., Rechencq
E., Mani J.C., Boucard M., Chapat J.P., J. Med. Chem. 35
(1992) 3353–3358.
[10] Vitse O., Laurent F., Pocock T.M., Be´ne´zech V., Zanik L.,
Keith R.F.E., Subra G., Portet K., Bompart J., Chapat J.P.,
Small R.C., Michel A., Bonnet P.A., Bioorg. Med. Chem. 7
(1999) 1059–1065.
[11] Lumma Jr W.C., Randall W.C., Cresson E.L., Huff J.R.,
Hartam R.D., Lyon T.F., J. Med. Chem. 26 (1983) 357–363.
[12] Meurer L.C., Tolman R.L., Chapin E.W., Saperstein R., Vi-
cario P.P., Zrada M.M., MacCoss M., J. Med. Chem. 35 (1992)
3845–3857.
[13] S. Parra, O. Vitse, V. Be´ne´zech, C. Deleuze-Masque´fa, G.
Subra, J. Bompart, R. Escale, J.P. Chapat, P.A. Bonnet, J.
Heterocyclic Chem. 38 (2001) in press.
[14] J.C. Teulade, P.A. Bonnet, J.N. Rieu, H. Viols, J.P. Chapat, G.
Grassy, A. Carpy, J. Chem. Res. (S) 6 (1986) 202–203.
[15] Ollis W.D., Stanforth S.P., Ramsden C.A., J. Chem. Soc.
Perkin Trans. I (1989) 961–963.
[16] Paudler W.W., Shin H.G., J. Org. Chem. 33 (1968) 1638–1639.
[17] Bradac J., Furek Z., Janezic D., Molan S., Smerkolj I.,
Stanovnik B., Tisler M., Vercek B., J. Org. Chem. 42 (1977)
4197–4201.
[18] Cobb J., Cheeseman G.W.H., Magn. Reson. Chem. 24 (1986)
References
231–237.
[19] Elliott K.R.F., Berry J.L., Bate A.J., Foster R.W., Small R.C.,
[1] Laurent F., Michel A., Bonnet P.A., Chapat J.P., Boucard M.,
Enzyme Inhib. 4 (1991) 245–251.
Br. J. Pharmacol. 108 (1993) 622–626.
[20] Thompson W.J., Appleman M.M., Biochemistry 10 (1971)
[2] Michel A., Laurent F., Bompart J., Hadj-Kaddour K., Chapat
J.P., Boucard M., Bonnet P.A., Br. J. Pharmacol. 110 (1993)
1031–1036.
[3] Cook S.J., Archer K., Martin A., Buchheit K.H., Fozard J.R.,
Mu¨ller T., Miller A.J., Elliott K.R.P., Poster R.W., Small R.C.,
Br. J. Pharmacol. 114 (1995) 143–151.
311–316.
[21] Rutten B.M., Schoot W.J., Depont J., J. Biochem. Biophys.
Acta 315 (1973) 378–383.
[22] Hooker C.S., Calkins P.Y., Fleish J.H., Blood Vessels 14 (1971)
1–11.
.