Phthalazine PDE IV inhibitors: Conformational study of some 6-methoxy-1,4-disubstituted derivatives

Article abstract of DOI:10.1016/j.bmc.2005.04.057

This report describes the detailed conformational analysis and synthesis of a series of phthalazine phosphodiesterase-type (IV) (PDE IV) inhibitors bearing either mono- or dichloro 4-methylenepyridine at the 'down' position of the phthalazine nucleus or different heterocycles at the 'top' position 4 of the phthalazine moiety. Both the mono- and dichloro 4-methylenepyridine units linked at carbon C₁ of the phthalazine nucleus show identical conformational behaviour with the substituent preferentially oriented towards the external part of the molecule and the pyridine plane almost orthogonal to that of phthalazine ring. The heterocyclic five-membered rings linked at carbon C₄ of phthalazine show quite different conformational behaviour. The 1,3-thiazole ring exists in a well-defined conformation almost coplanar with respect to the phthalazine nucleus while the 1,2,4-triazole and the 1,3-diazole heterocycles show a great conformational freedom with large torsion angles. Compound 3 bearing the thiazole ring at C₄ displays the major biological activity, thus suggesting that a planar and rather rigid conformation of the pentacycle should favour the PDE IV inhibition capacity of this class of compounds.

Full text of DOI:10.1016/j.bmc.2005.04.057

Bioorganic & Medicinal Chemistry 13 (2005) 4425–4433
Phthalazine PDE IV inhibitors: Conformational study of some
6
-methoxy-1,4-disubstituted derivatives
a,
Thomas Haack, Raimondo Fattori, Mauro Napoletano, Franco Pellacini,
a
a
a
*
b
Giovanni Fronza, Giuseppina Raffaini and Fabio Ganazzoli
c
c
a
Inpharzam Ricerche, Zambon Group, Via ai S o¨ i, 6807 Taverne, Switzerland
b
CNR, Istituto di Chimica del Riconoscimento Molecolare, Dipartimento di Chimica, Materiali e Ingegneria Chimica ‘G. Natta’,
Via Mancinelli 7, 20131 Milano, Italy
Politecnico di Milano, Dipartimento di Chimica, Materiali e Ingegneria Chimica ‘G. Natta’, Via Mancinelli 7, 20131 Milano, Italy
c
Received 26 January 2005; accepted 19 April 2005
Available online 31 May 2005
Abstract—This report describes the detailed conformational analysis and synthesis of a series of phthalazine phosphodiesterase-type
IV) (PDE IV) inhibitors bearing either mono- or dichloro 4-methylenepyridine at the ꢀdownꢁ position of the phthalazine nucleus or
different heterocycles at the ꢀtopꢁ position 4 of the phthalazine moiety. Both the mono- and dichloro 4-methylenepyridine units
linked at carbon C of the phthalazine nucleus show identical conformational behaviour with the substituent preferentially oriented
towards the external part of the molecule and the pyridine plane almost orthogonal to that of phthalazine ring. The heterocyclic five-
membered rings linked at carbon C of phthalazine show quite different conformational behaviour. The 1,3-thiazole ring exists in a
well-defined conformation almost coplanar with respect to the phthalazine nucleus while the 1,2,4-triazole and the 1,3-diazole het-
erocycles show a great conformational freedom with large torsion angles. Compound 3 bearing the thiazole ring at C displays the
(
1
4
4
major biological activity, thus suggesting that a planar and rather rigid conformation of the pentacycle should favour the PDE IV
inhibition capacity of this class of compounds.
ꢀ
2005 Elsevier Ltd. All rights reserved.
1. Introduction
phosphodiesterase highly expressed in inflammatory
1
cells and in the airway smooth muscle. The observation
The phosphodiesterases (PDEs) are involved in intracel-
lular degradation of cyclic adenosine monophosphate
that an elevation of cAMP in these cells can suppress
inflammatory effects and can induce muscle relaxation
has stimulated great interest in developing selective
PDE IV inhibitors as therapeutic agents for asthma or
chronic obstructive pulmonary disease (COPD) and
(
cAMP) and cyclic guanosine monophosphate (cGMP)
0
to form their corresponding 5 -monophosphates. Phos-
phodiesterase type IV (PDE IV) is a cAMP-specific
R
O
O
O
O
N
N
O
O
O
Cl
Cl
N
O
NC
O
N
N
CO H
N
2
Cl
Cl
Rolipram
Ariflo
Piclamilast
Phthalzine based
inhibitors
Keywords: PDE IV inhibitors; Conformational analysis; NMR; Molecular dynamics.
*
Corresponding author at present address: Chemistry Department, Inpharzam Ricerche SA, Via ai S o¨ i, CH-6807 Taverne, Switzerland. Tel.: +41 91
35 45 18; fax: +41 91 935 45 12; e-mail: thomas.haack@zambongroup.com
9
0
968-0896/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.04.057

4426
T. Haack et al. / Bioorg. Med. Chem. 13 (2005) 4425–4433
2
other inflammatory diseases. Many inhibitors have
3
been reported that are under clinical investigation.
compounds seemed rather difficult since polar substitu-
ents such as tertiary amines apolar groups such as phe-
nyl residues showed comparable activity. Very recently
5
The archetypal PDE IV inhibitor Rolipram was fol-
lowed by several other compounds, most notably Ariflo
and Piclamilast, replacing the pyrrolidinone of Roli-
pram with another functionality such as the dichloro-
pyridine unit. However, the clinical utility of the
pioneer compounds has been limited by side effects such
a comprehensive effort has been made to establish a
three-dimensional quantitative structure–activity rela-
tionship of the homogeneous class of the phthalazine
PDE IV inhibitors using the Comparative Molecular
7
Field Analysis (CoMFA) method. This method pro-
3
as nausea, emesis and increased gastric acid secretion.
vides a correlation between both the electrostatic and
steric field variations of the investigated molecules and
their biological activity, allowing in principle the design
of new and more potent compounds.
We showed recently that the use of the phthalazine nu-
cleus as a conformationally constrained analogue of Pic-
4
–6
lamilast was compatible with inhibitory potency. This
finding confirms our hypothesis that the carbonyl func-
tion of Rolipram and Piclamilast could be replaced by a
p-bond of an aromatic ring and that a planar dihedral
angle between the phenyl ring and the linker region of
In the present contribution we report the results of a de-
tailed conformational analysis of the phthalazine deriv-
atives 2 to 5 representative of this class of novel PDE IV
inhibitors. Compounds 2, 3 and 5, already disclosed in a
4
5
0
0
0
Rolipram-like analogues is allowed. Nevertheless, these
prior publication, show the identical 3 ,5 -dichloro-4 -
methylenepyridine substituent in position 1 of the
phthalazine ring, while they differ for the five-membered
ring heterocyclic substituents in position 4. The three
compounds are characterised by very different biological
activity in vitro with IC50 values ranging from 241 nM
phthalazine derivatives showed a decreased potency
compared to open amide, probably due to unfavourable
interaction between the cyclopentyloxy and the peri-hy-
drogen in position 4 of the phthalazine nucleus (1).
ꢀ
7
for 2 to 14 nM for 3 and to 17% at 10 M for 5. Com-
pounds 2 and 3 were also included in the data base of
the above mentioned CoMFA analysis and their mark-
N
N
7
O
H
N
edly different activities were explained in terms of elec-
trostatic field variations. In fact the model shows that
the occurrence of electronegative atoms in the internal
O
O
N
N
N
N
Cl
Cl
00
00
00
positions 1 , 2 or 5 of the pentacyclic ring induces
an activity enhancement of the molecule, while the
inhibitory potency decreases when the electronegative
N
N
Cl
Cl
00
atoms are located in the external positions 3 and (or)
1
2
00
4
. In this work we use a combination of computational
Subsequently, we were able to show that the catechol
substitution pattern of Rolipram was not necessary for
PDE IV potency if the cyclopentyloxy group is replaced
by an appropriate group positioned at the heteroaro-
methods and NMR experiments to describe in detail the
conformational behaviour of the selected molecules with
the aim of determining the preferential position or mo-
tion of the electronegative atoms of the five-membered
ring and eventually their correlation with the biological
activity. For compound 2, a complete single crystal
structure was also obtained, thus providing a relation
between its conformational properties in the solid state
and in solution. In addition, a newly synthesised com-
pound 4 bearing a monochloro-substituted methylene-
pyridine ring at carbon C-1 of phthalazine has also
been examined. This compound shows a much lower
PDE IV inhibition potency than the corresponding
dichloroderivative 3 (IC50 of ca. 600 and 14 nM, respec-
tively) and thus it may be worth investigating whether
the structural variation is also accompanied by signifi-
cant differences in the conformational behaviour.
6
matic ring. This finding finally leads to the observation
that several heterocycles as cyclopentyloxy surrogates at
position 4 of the phthalazine moiety are allowed. Most
notably, derivative 2 with a triazole heterocycle was de-
void of the common emetic side effect normally observed
for the catechol rolipram-like compound class while
maintaining good potency (241 nM) and an acceptable
pharmacokinetic and safety profile.
In general, potential PDE IV inhibitors based on the
phthalazine ring as scaffold showed activities from the
4
–6
low nanomolar to micromolar range. The rationalisa-
tion of the structure–activity relationship of this class of
3
'
3
''
N 4''
5''
3''
4''
3''
4''
N
4
''
2''
5''
2
'' N
1
2'' N
1''
S5''
2'' N
1''
S5''
N
1'' N
''
5
4
O
O
O
O
N
N
N
N
N
N
N
N
1
0
7
9
Cl
Cl
Cl
8
1
3
'
1
1
2'
4
'
N
N
N
N
Cl 5'
Cl
Cl
Cl
1
'
6
'
2
3
4
5

T. Haack et al. / Bioorg. Med. Chem. 13 (2005) 4425–4433
4427
2
. Results and discussion
prior to phthalazine formation under strong base condi-
tions forming hemiketals 9. Compound 9a was yielded
as a (7:2) mixture of E/Z diastereoisomers, whereas 9b
was pure E. The final phthalazines 3 and 4 were ob-
tained after hydrazine cyclisation under acid catalysis.
Crystallisation from an ethanol/DMF mixture yielded
4 as a light yellow solid, whereas 3 was a white solid.
2
.1. Chemistry
The syntheses of compounds 2, 3 and 5 bearing the
5
dichloropyridine unit have been reported recently but
not of the monochloro compound 4, presented as a sub-
stance without a synthetic scheme during the European
8
chemistry symposium 2002. Briefly, furanonecarboxylic
acid 6 was prepared in two steps and in high yield from
commercially available m-anisic acid by using an old
Otherwise, imidazole or triazole heterocycles of deriva-
tives 2 and 5, respectively, were introduced after phthal-
azine formation, as outlined in Scheme 2. Briefly, key
intermediate 8b was cyclised by acid-catalysed hydrazi-
9
procedure. The key intermediates, the methylene fura-
nones 8, were obtained in high yields by acetic anhy-
nolysis (10) and chlorinated by POCl to yield the corre-
3
dride-induced condensation of
1
6
with pyridine
sponding chloro derivative 11, which subsequently was
treated with the preformed heterocycle anion to yield
the final products 2 and 5.
0
aldehydes 7a and 7b, as outlined in Scheme 1. The thi-
azole heterocycle of derivatives 3 and 4 was introduced
O
O
H
O
O
O
O
(
a)
X
(
b)
O
+
Cl
X
N
CO H
N
2
Cl
6
7a X=H
b X=Cl
8a X=H
8b X=Cl
7
N
S
S
N
HO
O
O
(
c)
N
N
O
X
X
N
N
Cl
Cl
9
a X=H
4 X=H
3 X=Cl
9b X=Cl
Scheme 1. Synthesis of thiazole compounds 3 and 4. Reagents and conditions: (a) acetic anhydride, toluene, reflux, 10 h, (70% X@H, 99% X@Cl); (b)
2-bromothiazole, LDA, ꢀ78 ꢁC, 2 h, ꢀ65 ꢁC, 27% (X@H), 38% (X@Cl); (c) hydrazine hydrate, MeOH, AcOH, reflux, 6 h, 70% (X@H), 60% (X@Cl).
O
O
Cl
O
O
O
Cl
N
N
O
N
N
(
a)
(b)
Cl
N
Cl
Cl
N
Cl
N
Cl
8b
11
10
N
X
N
O
(
c)
N
N
Cl
N
Cl
2
5
X=N
X=C
Scheme 2. Synthesis of triazole compound 2 and imidazole-containing compound 5. Reagents and conditions: (a) hydrazine hydrate, MeOH, AcOH,
reflux, 2 h, 99%; (b) POCl
, reflux, 4 h, 94%; (c) imidazole or triazole, NaH, DMF, 100 ꢁC, 20 h, 53% 2, 50% 5.
3

4428
T. Haack et al. / Bioorg. Med. Chem. 13 (2005) 4425–4433
a
2
2
.2. NMR and simulation methods
Table 2. Dihedral angles
Dihedral angle
2
3
4
5
.2.1. Conformational behaviour of the chloropyridine
U
W
P
60.9ꢁ
125.7ꢁ
86.6ꢁ
61.2ꢁ
ꢀ66.2ꢁ
129.6ꢁ
16.8ꢁ
ꢀ61.0ꢁ
125.6ꢁ
ꢀ113.5ꢁ
unit. The phthalazine derivatives 2–5 here examined dis-
play only two regions of potential flexibility: that of the
chlorosubstituted pyridine residue (rotation about C1/
0
C11 and C /C bonds) and that of the five-membered
ring (rotation about the bond C /pentacycle).
4
ꢀ125.3ꢁ
ꢀ16.3ꢁ
a
U = C
9
–C
1
–C11–C
4
0
; W = C
1
–C11–C
4
0
–C
5
0
4 1 2
; P = C10–C –X
00–X 00 U, W
11
4
and P at the absolute energy minimum obtained through the MD
simulations of compounds 2–5 performed in vacuo.
The X-ray study performed on 2 (Fig. 1) showed that
the dichloropyridine unit is orthogonal to the plane of
the phthalazine ring (W = 93.4ꢁ) and oriented towards
an apparent distance, which mainly reflects the distance
between H and the nearest H hydrogen. The data ob-
tained in CDCl and DMSO solutions are collected in
8
11
3
the external nitrogen atom N . The dihedral angle U is
2
Table 1. Within the experimental errors (estimated pre-
˚
close to 180ꢁ, so that the four atoms are nearly planar
and the methylene hydrogens H11 point towards the
benzenic proton H with distances H /H of 2.34 and
cision ±0.1 A) the H –H apparent distance is ca.
2
8 11
˚ ˚
.47 A in chloroform and ca. 2.40 A in DMSO (values
8
11
8
˚
averaged over the four compounds). This result indi-
cates that the conformation of the substituent at the
phthalazine carbon C is the same for compounds 2–5,
2
.48 A.
1
In solution the conformational preference of the dichlo-
ropyridine residue may be investigated by determining
the interproton distances H –H from NOE measure-
including compound 4 also which bears only one chlo-
rine substituent on the pyridine ring. The apparent dis-
tance H –H determined in the solution is similar to
8
11
8
11
ments. The two methylene protons H11a and H11b are
magnetically equivalent and give rise to a singlet in the
spectrum, thus only the mean distance H –H can be
that measured in the solid state for 2 (mean value
˚
.41 A) suggesting that the pyridine ring points towards
2
the external part of the phthalazine nucleus in both
8
11
determined. Such distance is very different from the
arithmetic mean since physically the averaging process
occurs over the inverse sixth power of the internuclear
cases. On the other hand, the angle U, which is close
to 180ꢁ for 2 in the solid state with the atoms N and
2
1
1
C practically eclipsed, is reasonably expected to have
0
4
distances. Therefore, the experimental result provides
values of ca. ±120ꢁ in solution to release the steric
hindrance.
However, the molecular dynamics simulations suggest a
more complicated situation. The U and W torsion angles
for the conformations with the global energy minimum
in vacuo are mentioned in Table 2. In the dielectric med-
ium, the values of U and W are equal to those shown in
the table to within ± 0.5ꢁ, while in water they differ by
up to 4ꢁ, indicating very little solvent effects. The U val-
ues are very close to 60ꢁ, while those of W occur at about
1
20ꢁ. As a result, the dichloropyridine ring is approxi-
mately orthogonal to the phthalazine plane, but ori-
ented towards the internal part of the phthalazine
nucleus, unlike what is suggested above. Correspond-
ingly, the distances H –H are about 2.4–2.5 and
8
11
˚
3
fixed conformation, the average distance probed by the
.8 A for all compounds. If the molecules existed in this
˚
NOE measurements should be 2.72 A, using the expres-
1
1
ꢀ6
ꢀ6
ꢀ6
sion
r
¼ ðr
þ r
Þ=2, a value substan-
H8–H11
H8–H11a
H8–H11b
Figure 1. ORTEP drawing of the X-ray structure of 2. U and W are the
tially different from the experimental one. However,
the energy difference between the internal and the exter-
nal arrangement of the pyridine ring (the global and a
local energy minimum, respectively) amounts to about
9 1 4 1 4 5
dihedral angles defined by C –C –C11–C and by C –C11–C –C ,
0 0 0
respectively.
1
.5 kcal/mol. Moreover, the energy barrier separating
Table 1. Mean values of the apparent distance H
a
8
–H11 determined
˚
these minima is about 2.7 kcal/mol only, so that they
are both visited relatively often at room temperature.
Interestingly, the PDF gH8–H11 ðrÞ, reported in Figure 2
for compound 2 as a typical case, shows that during
the MD simulations the most probable hydrogen sepa-
rations are about 2.4–2.5 and 3.8 A. However, the
broader distribution for the closer hydrogens (filled sym-
bols) should be noted, since it indicates that this distance
˚
can be even smaller than 2 A, thanks to thermal fluctu-
ations. Therefore, we calculated the weighted average
from NOE data
˚
Sample
r
8,11 (A)
r8,11 (A)
(
CDCl
3
)
(DMSO)
2
3
4
5
2.46
2.50
2.45
2.48
2.40
2.40
˚
b
—
2.42
a
˚
Mean of three measurements; estimated precision ±0.1 A.
Not measured due to the low solubility of 4 in DMSO.
b

T. Haack et al. / Bioorg. Med. Chem. 13 (2005) 4425–4433
4429
0
0
0
0
0
.04
.03
.02
.01
.00
Table 3. Steady state NOE enhancements measured on the five-
membered ring hydrogens by irradiation of H of phthalazine ring
H8-H11a
H8-H11b
5
0
0
00
00
00
Sample
H-2 (%)
H-3 (%)
H-4 (%)
H-5 (%)
a
a
2
1.2
1.5
b
b
2
1
.3
.2
2.6
c
c
1.5
ND
a
b
a
a
3
1.5
.0
1.4
b
2
ND
a
4
5
ND
d
7.2
d
5.9
d
ND
0
2
4
6
a
b
c
r
CDCl
DMSO-d
CD OD.
/CD
3
.
6
.
Figure 2. The pair distribution function (PDF) for the distances r
˚
in A) between hydrogen H
3
d
(
8
and the methylene hydrogens H11a and
C
6
D
6
3
OD 1:1.
H11b for compound 2.
The experimental data are collected in Table 3 as per-
cent enhancements of the signals.
value ꢀr H8–H11 as above using the averages ꢀr H8–H11a and
obtained from the appropriate volume integrals
involving the separate PDFs. The resulting value is
ꢀ
r
H8–H11b
The X-ray analysis of 2 shows that in the solid state the
triazole ring is almost coplanar with the nitrogen atom
˚
2
.57 A for compound 2, being marginally larger for
other compounds. Such values, although still larger than
the experimental ones, are now closer to the distances
determined in chloroform.
N200 oriented towards the proton H of the phthalazine
5
nucleus (dihedral angle P = C –C –N 00–N200 of ꢀ1.2ꢁ
1
0
4
1
˚
and distance H /N 00 of 2.24 A) (Fig. 1). In solution such
5
2
conformation no longer exists. In fact both H300 and H500
protons of the pentacycle display weak NOEs with
enhancements in the range 1.2–2.6% upon irradiation
In conclusion, the MD simulations show for all com-
pounds that in the most stable conformation the pyri-
dine ring is oriented towards the internal part of the
phthalazine nucleus, but indicate at the same time that
frequent transitions occur towards the external confor-
mation. The NOE data are roughly in agreement with
the MD calculations, suggesting however that the lat-
ter conformation is more populated than predicted
by the simulations. It is worth noting that compounds
of H (Table 3). The experiment was carried out in dif-
5
ferent solvents to test whether significant variations oc-
cur depending on the dielectric constant of the
medium. For each solvent, H300 and H500 display similar
enhancements suggesting that the triazole ring in solu-
tion is strongly rotated with respect to the phthalazine
plane. Therefore, the distance H –H 00 decreases while
5
5
3
and 4, which only differ for the number of chlorine
the distance H –H 00 decreases while the distance H5–
5 5
atoms on the pyridine ring, show a strictly similar con-
formational behaviour. Thus, the markedly different
activity between these two PDE IV inhibitors (IC50
of 14 and ca. 600 nM, respectively) depends exclusively
on the different substitution pattern of the pyridine
ring. This is an interesting finding, making plausible
a possible interaction between one of the two chlorine
atoms with the zinc metal ion found in the active
site of PDEIV. Thus, the role of disubstitution could
be viewed as enhancement of the local chlorine
concentration.
H300 increases compared to those in the coplanar confor-
mation, so that a NOE is observed on both nuclei. The
MD calculations confirm the qualitative conclusions in-
ferred from NOE data in all the simulation environ-
ments. The PDF is the best way to represent the
distribution of the internuclear distances. These func-
tions calculated for the distances H –H 00 and H –H 00
5
3
5
5
of 2 are reported in Figure 3a for the simulations in va-
cuo. The distribution of distance H –H 00 is rather broad
5
5
˚
with a maximum probability at about 3.2 A, while the
distribution of distance H –H 00 is somewhat sharper
5
3
˚
with a maximum probability centred around 4.5 A.
The P dihedral angle for the global minimum energy
conformation of 2 (see Table 3) indicates that the penta-
cycle is orthogonal to the phthalazine plane, indepen-
dent of the simulation environment.
2
.2.2. Conformational properties of the five-membered
rings. The five-membered ring attached to carbon C of
4
phthalazine has a single conformational degree of free-
dom, the rotation around the bond connecting C to
4
the pentacycle. Such rotation moves about in the space
the electronegative heteroatoms of the ring, thus
strongly influencing the distribution of the electrostatic
fields in this part of the molecule. The conformational
preferences of the pentacycle are dictated by the energy
maxima corresponding to the coplanarity with the
phthalazine ring due to steric interactions with ꢀipsoꢁ
Compounds 3 and 4 have the identical 1,3-thiazole sub-
stituent at carbon C4 and show very similar NOE
enhancements. By irradiation of H only the hydrogen
5
H300 is affected, while H400 does not show any effect indi-
cating that in this case the five-membered ring in solu-
tion adopts a conformation with the nitrogen atom
strongly oriented towards the inner part of the phthal-
azine moiety. The PDF functions calculated for the dis-
tances H –H 00 and H –H 00 (Fig. 3b) show sharp peaks
proton H and to the orthogonality of the rings due to
5
the loss of p conjugation. The steady state NOE
enhancements between the phthalazine hydrogen H5
and the protons located on the pentacycle have been
measured to gain information on the preferential orien-
tation of the ring with respect to the phthalazine plane.
5
3
5
4
˚
with maximum probability at 4.1 A for H –H 00 and
5
3
˚
5.7 A for H –H 00. The P dihedral angle for conforma-
5
4
tion of 3 and 4 with the global minimum energy in vacuo

4430
T. Haack et al. / Bioorg. Med. Chem. 13 (2005) 4425–4433
0
0
0
0
.03
.02
.01
.00
0.03
0.03
0.02
0.01
0.00
H5-H3''
H5-H5''
(a)
H5-H4''
H5-H3''
(b)
H5-H2''
H5-H5''
H5-H4''
(c)
0.02
0.01
0.00
0
2
4
6
0
2
4
6
0
2
4
6
r
r
r
˚
5
Figure 3. The pair distribution functions for the distances r (in A) between proton H and the hydrogens located on the pentacyclic ring for
compounds 2 (a), 3 (b) and 5 (c). The plot for compound 4 is not shown for brevity being the same as for compound 3.
7
is about 16–17ꢁ, indicating that in this case the pentacy-
cle and phthalazine ring are not too far from coplana-
rity. In this case, polar effects and/or hydrogen bonds
with water lead to somewhat larger effects on P with a
difference between the various simulation media being
anyway comprised within ±10ꢁ.
FA in terms of the position of the electronegative nitro-
gen atoms in the five-membered ring: these atoms
produce an increase of the biological activity when lo-
cated in the internal positions while in the external posi-
tions they induce a decrease of the activity. In the
present study it appears that the active and nonactive mol-
ecules differ strongly also in the conformational behav-
iour of the pentacyclic ring. In the most active
compound 3, the thiazole ring has a well-defined confor-
mation with the nitrogen atom oriented towards the ben-
zene moiety and a modest deviation from strict
coplanarity with the phthalazine moiety. On the contrary,
the diazole and triazole rings of the less active molecules 2
and 5 show a much greater rotational freedom with strong
deviations from coplanarity with the phthalazine nucleus,
becoming almost orthogonal. Thus, we suggest that some
coplanarity of heterocycles at position 4 of the phthal-
azine moiety is important for PDE IV activity.
In compound 5, the 1,3-diazole ring is the substituent at
carbon C of the phthalazine nucleus. The irradiation of
4
H produces strong enhancements on protons H 00 and
5
2
H500 (5.9% and 7.2%, respectively), without any effect
on H400. In this case, we can reasonably expect that the
diazole ring makes large-amplitude oscillations about
the C –N 00 bond, which in turn brings both hydrogens
4
1
near H . In fact, the PDF functions show the occurrence
5
of two broad distance distributions with maximum
˚
probability around 3.5 A (Fig. 3c), whereas H 00 is much
4
˚
further away at about 5 A. For this molecule, the pre-
ferred conformation shows in vacuo a P value of
ꢀ
114ꢁ, and in water of ꢀ104ꢁ, much closer to orthogo-
nality than in compounds 3 and 4.
4. Experimental
4
.1. General methods
3
. Conclusion
NMR spectra were run on Bruker ARX 400 or Bruker
Avance 500 spectrometers. Chemical shifts are in d from
internal TMS (tetramethylsilane).
In conclusion, the combined information from the NOE
data of the protons located on the five-membered rings
and the MD calculations indicates that the thiazole sub-
stituent has a well defined conformation, as indicated
also by the sharp peaks of the PDFꢁs of Figure 3b. In
these compounds, the thiazole ring is almost coplanar
with the phthalazine nucleus, with the nitrogen atom lo-
4
.2. NMR
The distance ratio r(H , H )/r(H , H ) of compounds
8
8
11
7
2–5 was determined from the phase-sensitive NOESY
spectra. For each solution the proton relaxation times
T were measured to choose the appropriate mixing
1
and pulse repetition times. The T values range from
1
cated near H . Moreover, in vacuo the rotated confor-
5
mation bringing the S atom close to H represents a
5
local energy minimum for P = ±155ꢁ, which is less sta-
ble than the global one by 4.5 kcal/mol and is separated
by an energy barrier larger than 11 kcal/mol; thus it is
unlikely to be of any significance. On the other hand,
both the diazole and triazole rings exist in a broad dis-
tribution of conformational states with large distribu-
tion of the dihedral angles, and correspondingly with
broad PDFꢁs of Figure 3a and c.
0
.7 to 1.0 s of the CH –11 hydrogens to 5–9 s of the
2
most isolated nuclei of the pyridine and five-membered
rings. Thus the NOESY spectra were obtained using a
quite short mixing time of 0.2–0.3 s due to the short
CH –11 T and a relaxation delay of 10–15 s dictated
2
1
by the rather long T of H-7 (2.0–3.4 s). In these condi-
1
tions the cross peak intensities a are proportional to the
ij
cross relaxation rates r , which in turn are related to the
ij
1
1
internuclear distances r by the equation:
As outlined above, compounds 2, 3 and 5, which differ in
the structure of the pentacyclic substituent at carbon C of
the phthalazine nucleus, are characterised by markedly
different activities as PDE IV inhibitors. Such differences
have been explained by Chakraborthi et al. using CoM-
ij
2
4
6
ij
4
r ¼ ð ꢀh c =2r Þs
ij
c
where c is the proton magnetogyric ratio and s the cor-
c
relation time of the internuclear vector H –H . With the
i
j

T. Haack et al. / Bioorg. Med. Chem. 13 (2005) 4425–4433
4431
1
assumption that the correlation time s is equal for all
J = 3.3 Hz, H300), 7.54 (1H, d, J = 3.3 Hz, H400);
H
c
molecular internuclear vectors the distance ratios can
be obtained directly from the ratio of the cross-peak
intensities:
NMR (DMSO-d ): d 9.19 (1H, d, J = 2.6 Hz, H ), 7.73
6
5
(1H, dd, J = 2.6 and 9.2 Hz, H ), 8.33 (1H, d,
J = 9.2 Hz, H ), 4.87 (2H, s, H ), 4.01 (3H, s, OCH ),
7
1
1
8
11
3
8
.44 (1H, d, J = 4.7 Hz, H2 ), 7.32 (1H, d, J = 4.7 Hz,
0
1
=6
r
ij=rki ¼ ðaki=aijÞ
H3 ), 8.65 (1H, s, H6
7.99 (1H, d, J = 3.2 Hz, H400); C NMR (CDCl ): d
156.6 (C ), 149.9 (C ), 106.2 (C ), 163.4 (C ), 125.2
(C ), 126.6 (C ), 126.6 and 121.9 (C , C ), 36.7 (C ),
0 0), 8.20 (1H, d, J = 3.2 Hz, H 00),
3
1
3
This equation, provided rki is known, allows one to
obtain the unknown distance r . The distance r(H ,
H ) estimated with this method is an apparent distance
3
ij
8
1
4
5
6
1
1
7 8 9 10 11
due to the magnetic equivalence of the methylene
protons and mainly reflects the distance between H8
and the nearest H11 hydrogen. For all compounds 2–5
56.7 (OCH ), 149.7 (C ), 145.0 and 131.9 (C4 , C5 ),
0 0 0
3 6
148.1 (C2
0
), 125.4 (C3
0), 144.2 (C300), 122.5 (C400).
1
˚
the H –H distance of 2.41 A as determined by the mod-
7
Compound 5: H NMR (CDCl ): d 7.32 (1H, d,
8
3
elling calculations was chosen as the reference distance.
J = 2.6 Hz, H ), 7.67 (1H, dd, J = 2.6 and 9.2 Hz, H ),
5
7
8.23 (1H, d, J = 9.2 Hz, H ), 4.95 (2H, s, H ), 3.96
(3H, s, OCH ), 8.55 (2H, s, H
8 11
0 0), 8.05 (1H, s, H200),
2 ,6
The steady state NOE difference experiments were car-
ried out by acquiring both on- and off-resonance spectra
using a 75 dB attenuation of the decoupling power. This
experiment was employed to investigate the spatial dis-
position of the five-membered rings with respect to the
3
7.27 (1H, d, J = 1.6 Hz, H 00), 7.49 (1H, d, J = 1.6 Hz,
4
1
H500); H NMR (DMSO-d ): d 7.26 (1H, d, J = 2.6 Hz,
6
H ), 7.85 (1H, dd, J = 2.6 and 9.1 Hz, H ), 8.62 (1H,
d, J = 9.1 Hz, H ), 5.04 (2H, s, H ), 3.97 (3H, s,
5
7
8
11
), 8.32 (1H, s, H200), 7.25
phthalazine moiety. As the T values of the hydrogens
1
OCH ), 8.69 (2H, s, H
3 2 ,6
(1H, d, J = 1.4 Hz, H 00), 7.88 (1H, d, J = 1.4 Hz, H500);
4
0
0
on five-membered rings are rather long (5–9 s), an irra-
diation time of 30–40 s was used to reach with certainty
the steady state and detect also very small NOE
enhancements.
1
3
C NMR (C D /CD OD 1:1): d 156.2 (C ), 150.3
6
6
3
1
(C ), 102.9 (C ), 164.2 (C ), 126.1 (C ), 127.1 (C ),
126.1 and 123.5 (C , C ), 34.4 (C ), 56.0 (OCH ),
4
5
6
7
8
9
10
11
3
147.8 (C2
0
0
), 134.5 (C
0
0
), 143.9 (C4
0
), 138.3 (C200),
,6
3 ,5
The NMR spectra of compounds 2–5 were obtained in
CDCl and DMSO-d solutions.
130.0 (C400), 120.9 (C500).
.3. Computer simulations
All simulations were performed with InsightII/Discov-
3
6
4
1
Compound 2: H NMR (CDCl ): d 8.49 (1H, d,
3
J = 2.6 Hz, H ), 7.68 (1H, dd, J = 2.6 and 9.2 Hz, H ),
.23 (1H, d, J = 9.2 Hz, H ), 4.96 (2H, s, H ), 4.04
5
7
1
2
13
8
er, using the consistent valence force field CVFF
8
11
0 0), 8.27 (1H, s, H300),
(
3H, s, OCH ), 8.55 (2H, s, H
3
with a Morse potential for the bonded atoms. The initial
geometries, generated through the available library frag-
ments, were fully minimised up to an energy gradient
2 ,6
.23 (1H, s, H500); H NMR (DMSO-d ): d 7.96 (1H, d,
1
9
J = 2.6 Hz, H ), 7.87 (1H, dd, J = 2.6 and 9.2 Hz, H ),
6
5
7
ꢀ
3
ꢀ1 ꢀ1
˚
lower than 10 kcal mol A . The molecular dynam-
8
.63 (1H, d, J = 9.2 Hz, H ), 5.07 (2H, s, H ), 3.97
8
11
0 0), 8.49 (1H, s, H300),
(
3H, s, OCH ), 8.69 (2H, s, H
3
ics simulations at 300 K were performed (i) in vacuo, (ii)
into an effective dielectric medium mimicking water
using a distance-dependent dielectric constant and (iii)
in the explicit presence of the solvent, using a large num-
2 ,6
.38 (1H, s, H500); C NMR (DMSO-d ): d 156.8 (C ),
6 1
1
3
9
1
49.4 (C ), 104.7 (C ), 163.4 (C ), 126.0 (C ), 127.8
7
4
5
6
(
(
(
C ), 123.3 and 123.2 (C , C ), 34.4 (C ), 56.9
8
9
10
11
ꢀ
3
OCH ), 147.9 (C
3
0
0
), 133.8 (C ), 143.9 (C4
0 0
3 ,5
0
), 146.3
ber of water molecules at a density of 1 g cm with peri-
odic boundary conditions. The temperature was
controlled through the Berendsen thermostat, and the
dynamical equations were integrated through the Verlet
algorithm with a timestep of 1 fs. After an initial equil-
ibration of 30 ps, the data collection was carried out
for 500 ps, and a large number of instantaneous struc-
tures was optimised in search of the global energy min-
ima. These geometries were subjected to a systematic
dihedral search, first by changing the dihedral angles de-
2 ,6
C300), 154.1 (C500).
1
Compound 3: H NMR (CDCl ): d 9.36 (1H, d,
3
J = 2.6 Hz, H ), 7.61 (1H, dd, J = 2.6 and 9.2 Hz, H ),
.18 (1H, d, J = 9.2 Hz, H ), 4.95 (2H, s, H ), 4.08
8 11
5
7
8
(
3H, s, OCH ), 8.53 (2H, s, H
3
0
0
), 8.06 (1H, d,
H
2 ,6
1
J = 3.3 Hz, H300), 7.50 (1H, d, J = 3.3 Hz, H400);
NMR (DMSO-d ): d 9.22 (1H, d, J = 2.4 Hz, H ), 7.82
6
5
(
J = 9.2 Hz, H ), 5.04 (2H, s, H ), 4.05 (3H, s, OCH ),
1H, dd, J = 2.4 and 9.2 Hz, H ), 8.56 (1H, d,
7
fined by atoms C –C –C –C and C –C –C –C5 (U
0
0
0
8
11
3
9 1 11 4 1 11 4
8
(
.67 (2H, s, H2
0
0
), 8.19 (1H, d, J = 3.3 Hz, H 00), 7.96
and W, respectively), for a total of 1369 conformations,
and then by changing the P dihedral angle around the
bond connecting the five-membered ring to the phthal-
azine nucleus defined by atoms X200–X100–C –C
10
,6
3
1
3
1H, d, J = 3.3 Hz, H400); C NMR (CDCl ): d 155.3
3
(
1
C ), 150.2 (C ), 106.8 (C ), 163.8 (C ), 125.7 (C ),
25.6 (C ), 126.9 and 122.3 (C , C ), 34.7 (C ), 56.7
1
4
5
6
7
8
9
10
11
4
(
(
OCH ), 148.1 (C
3
C300), 122.9 (C400).
0
0
), 134.4 (C ), 143.6 (C4
0 0
3 ,5
0
), 144.7
(X = N or C; see Scheme 1), for a total of 73 conforma-
tions. The geometries sampled in the MD runs were ana-
lysed through the pair distribution function g (r) (or
2 ,6
ij
1
Compound 4: H NMR (CDCl ): d 9.33 (1H, d,
PDF for short). This function gives the probability den-
sity of finding atoms j at a distance comprised between r
and r + dr from atoms i, and is defined as
g (r) = dhN (r)i/q Æ dV(r). Here, dhN (r)iis the average
3
J = 2.6 Hz, H ), 7.50 (1H, dd, J = 2.6 and 9.2 Hz, H ),
7
5
7
.94 (1H, d, J = 9.2 Hz, H ), 4.81 (2H, s, H ), 4.06
8 11
(3H, s, OCH ), 8.31 (1H, d, J = 4.9 Hz, H2
d, J = 4.9 Hz, H3
0
0
), 7.07 (1H,
), 8.09 (1H, d,
3
ij
ij
j
ij
0
), 8.62 (1H, s, H6
number of times the j atoms are comprised in a spherical

4432
T. Haack et al. / Bioorg. Med. Chem. 13 (2005) 4425–4433
shell of thickness dr and volume dV(r) at a distance r
from atoms i, and q is their bulk density. For molecules
perature and filtered to provide pure trans 8 (680 mg,
70%) as a white solid. ESI MS for C H ClNO calcd
+
j
15 10
3
+
1
in vacuo or in a dielectric medium, q is set to 1. In the
(287.70) 287, 288.0 (MH ) found. H NMR (CDCl ): d
j
3
text, we mainly discuss the results obtained in vacuo, be-
cause they better approximate the solvents used in the
NMR measurements, since both CDCl3 and DMSO
are weakly complexating solvents, and anyway cannot
form hydrogen bonds, unlike water. Moreover, CDCl3
has a very low dielectric constant (e = 4.8), hence it is
reasonably approximated by the simulations in vacuo,
whereas DMSO has a larger value (e = 45), though not
as large as water. Therefore, use of the latter solvent
in the simulations corresponds to a limiting case. Any-
way, in all cases the overall conformational features
were independent of the simulation medium, with only
minor quantitative differences.
8.58 (s, 1H), 8.47 (d, 1H, J = 5.5 Hz), 8.11 (d, 1H,
J = 5.5 Hz), 7.74 (d, 1H, J = 8.3 Hz), 7.3 (d, 1H,
J = 8.3 Hz), 6.61 (s, 1H), 3.90 (s, 3H).
4.4.4. 3-(3,5-Dichloro-pyridin-4-ylmethylene)-6-methoxy-
3H-isobenzofuran-1-one (8b). Acetic anhydride (60 mL)
was added to a stirred mixture of 6 (12.4 g, 60 mmol)
and 7b (10.8 g, 61 mmol) in toluene (150 mL). The mix-
ture was heated at reflux for 0.5 h, cooled down to room
temperature and evaporated. The residue was three
times dissolved in 50 mL toluene and evaporated yield-
ing trans 8 (19.2 g, quantitative) as a yellow solid pure
+
enough to be used in the following steps. ESI MS for
C H Cl NO calcd (322.15) 321, 322.0 (MH ) found.
+
1
5
9
2
3
1
4
4
.4. Synthesis
.4.1. Material and methods. H NMR spectra were re-
H NMR (CDCl ): d 8.60 and 8.50 (s, 2H), 7.77–6.20
3
(m, 4H), 3.90 (s, 3H).
1
corded as CDCl or DMSO-d solutions on a Varian
3
4.4.5. 3-(3-Chloro-pyridin-4-ylmethylene)-6-methoxy-1-
thiazol-2-yl-1,3-dihydro-isobenzofuran-1-ol (9a). n-BuLi
(2.5 M in hexane, 2.64 mL,7.09 mmol) was added to 2-
6
Gemini 200 MHz spectrometer. Chemical shifts are re-
ported in parts per million (d units) relative to CHCl3
or DMSO as internal standards (in NMR descriptions;
bromothiazole (1.06 g, 6.5 mmol) in Et O (30 mL) under
2
s = singlet,
d = doublet,
m = multiplet, and br = broad peak).
t = triplet,
q = quartet,
argon atmosphere at ꢀ70 ꢁC and stirred for 45 min.
Subsequently, a solution of 8 (1.7 g, 5.91 mmol) in
THF (70 mL) was added keeping the temperature at
ꢀ65 ꢁC. After 2 h the reaction was stopped by quench-
ing with water and extracted with EtOAc. The organic
phase was washed with water, dried (Na SO ) and evap-
ESI mass spectra were recorded with a Finnigan Aqa
mass spectrometer (electron spray, positive ionisation)
connected to a Gilson HPLC.
2
4
orated to dryness. The crude product was purified by
chromatography on silica gel using petroleum ether/
EtOAc (1/1) as eluant to provide 9a (600 mg, 27%) as
a yellow oil and as a mixture of 2:7 of Z:E isomers.
Reactions were followed either by reverse phase HPLC/
MS (Gilson Auto purifier, C18 column Zorbax SBC18
(
3.5 lm, 2.1 · 50 mm, DAD detector) or thin-layer chro-
+
matography conducted on precoated silica gel 60F254
plates (Merck). Chromatography was performed on
ESI MS for C H ClN O S calcd (372.83) 372, 373.0
3
1
8
13
2
+
1
(MH ) found. H NMR (CDCl ): d major product (E):
3
1
00–200 mesh silica gel C-200 (Wako Pure Chemical)
8.32 (s, 1H), 8.20 (d, 1H, J = 5.3 Hz), 7.83 (d, 1H,
J = 8.3Hz), 7.78 (d, 1H, J = 3.4 Hz), 7.40 (d, 1H,
J = 3.4 Hz), 7.03 (m, 2H), 6.85 (d, 1H, J = 2.26 Hz),
using the solvent systems (volume ratios) indicated
below.
5
.15 (s, 1H), 3.84 (s, 3H); d minor product (Z): 8.37 (s,
4
(
.4.2. 3-Chloro-4-formyl-pyridine (7). 3-Chloropyridine
4 g, 35.22 mM) in THF (8 mL) was added to 2 M
1H), 8.00 (d, 1H, J = 5.3 Hz), 7.80 (m, 2H), 7.45 (d,
1H, J = 3.4 Hz), 7.05 (m, 2H), 6.63 (d, 1H,
J = 2.26 Hz), 5.00 (s, 1H), 3.84 (s, 3H).
LDA (19.4 mL, 38.7 mM) in THF (50 mL) under Argon
atmosphere at ꢀ65 ꢁC and stirred for 30 min. Subse-
quently, ethylchloroformate (2.86 g, 38.7 mM) in THF
4.4.6. 3-(3,5-Dichloro-pyridin-4-ylmethylene)-6-methoxy-
1-thiazol-2-yl-1,3-dihydro-isobenzofuran-1-ol (9b). Same
procedure as for 9a except quantities: n-BuLi (2.5 M in
hexane, 14.4 mL, 36 mmol), 2-bromothiazole (3 mL,
(
ꢀ
8 mL) was added and the resulting mixture stirred at
65 ꢁC for 2 h. The reaction was stopped by quenching
the solution with saturated NH Cl and THF at ambient
4
mixture. Work-up of the solution was done by addition
of water, extraction with EtOAc, washing with brine,
drying (Na SO ) and evaporation to dryness. The crude
34.2 mmol) in Et O (17 mL) and the solution of 8
2
(10 g, 31 mmol) in THF (60 mL). The crude product
was purified by chromatography on silica gel using
petroleum ether/EtOAc (6/4) as eluant providing pure
2
4
product was purified by chromatography on silica gel
using petroleum ether/EtOAc (7/3) as eluant to provide
the product (2 g, 40%) as a white solid. ESI MS for
+
E 9b (4.64 g, 38%) as a white solid. ESI MS for
C H Cl N O S calcd (407.28) 406, 407.0 (MH )
+
+
1
8
12
2
2
3
+
1
1
C H ClNO calcd (141.56) 141, 141.9 (MH ) found. H
6
found. H NMR (CDCl ): d 8.42 (s, 2H), 7.78 (d, 1H,
4
3
NMR (CDCl ): d 10.50 (s, 1H), 8.78 (s, 1H), 8.67 (d,
J = 5.3 Hz), 7.88 (d, 1H, J = 8.3 Hz), 7.48 (d, 1H,
J = 3.4 Hz), 7.10 (d, 1H, J = 3.4 Hz), 6.80 (d, 1H,
J = 4.9 Hz), 5.38 (s, 1H), 3.84 (s, 3H).
3
1
H, J = 4.9 Hz), 7.68 (d, 1H, J = 4.9 Hz).
4
.4.3. 3-(3-Chloro-pyridin-4-ylmethylene)-6-methoxy-3H-
isobenzofuran-1-one (8a). Acetic anhydride (1.05 mL)
was added to a stirred mixture of 6 (700 mg, 3.36 mM)
and 7a (524 mg, 3.7 mM) in toluene (5 mL). The mixture
was heated at reflux for 10 h, cooled down to room tem-
4.4.7. 1-(3-Chloro-pyridin-4-ylmethyl)-6-methoxy-4-thia-
zol-2-yl-phthalazine (4). Hydrazine monohydrate was
added to a stirred solution of 9a (100 mg, 0.268 mmol)
and acetic acid (0.077 mL, 1.34 mmol) in MeOH

T. Haack et al. / Bioorg. Med. Chem. 13 (2005) 4425–4433
4433
+
(
1 mL). The mixture was heated at reflux for 6 h, than
(306 mg, 53%) as solid. ESI MS for C H C N O
17 12 12 6
+
1
cooled to room temperature, quenched with water and
extracted with EtOAc. The organic phase was washed
with water, dried (Na SO ) and evaporated to dryness.
calcd (387.23) 386, 387.0 (MH ) found. H NMR: see
experimental part NMR.
2
4
The crude product was purified by silica chromatog-
raphy using DCM/MeOH (98/2) as eluant providing 4
4.4.12. 1-(3,5-Dichloro-pyridin-4-ylmethyl)-6-methoxy-4-
(1,4)imidazol-1-yl-phthalazine (5). Same procedure as 2
starting with NaH (0.120 g, 2.5 mmol, 60% in oil),
1,2,4-triazole (0.23 g, 3.4 mmol) in dry DMF (12 mL)
+
70 mg, 70%) as light yellow solid. ESI MS for
(
C H ClN OS calcd (368.85) 368, 369.0 (MH ) found.
H NMR: see experimental part NMR.
+
1
8
13
4
1
and 11 (0.6 g, 1.7 mmol), which provide 5 (328 mg,
50%) as solid after chromatographic purification. ESI
+
4.4.8. 1-(3,5-Dichloro-pyridin-4-ylmethyl)-6-methoxy-4-
thiazol-2-yl-phthalazine (3). Same procedure as 4 except
MS for C H Cl N O calcd (386.24) 385, 386.0
18 13 2 5
+
(MH ) found. H NMR: see experimental part NMR.
1
quantities are hydrazine monohydrate (0.048 mL,
1
mmol) 9a (125 mg, 0.302 mmol) and acetic acid
0.077 mL, 1.34 mmol) in MeOH (1.5 mL). The crude
(
product was purified by silica chromatography using
Acknowledgments
DCM/MeOH (98/2) as eluant providing 4 (72 mg,
6
calcd (403.29) 402, 403.0 (MH ) found. H NMR: see
experimental part NMR.
The authors thank Dr. Anthony Linden from the
Institute of Organic Chemistry, University of Z u¨ rich,
Switzerland for X-ray measurements and Mr. C. Riva
for synthetic support. CCDC-258393 contains the sup-
plementary crystallographic data for this paper. These
data can be obtained free of charge via www.ccdc.
cam.ac.uk/conts/retrieving.html (or from the Cambridge
Crystallographic Data Centre, 12 Union Road, Cam-
bridge CB2 1EZ, UK; fax: +44 1223 336033; e-mail:
deposit@ccdc.cam.ac.uk).
+
0%) as a white solid. ESI MS for C H Cl N OS
1
8
12
2
4
+
1
4
2
.4.9. 1-(3,5-Dichloro-pyridin-4-ylmethyl)-7-methoxy-
H-phthalazin-1-one (10). Hydrazine monohydrate
(18.4 mL, 0.378 mol) was added to a stirred suspension
of 8b (84.4 g, 0.126 mol) and acetic acid (15.4 mL) in
MeOH (200 mL) under N gas. The mixture was re-
2
fluxed for 1 h, then left over night at rt and cooled over
ice. The solid was filtered, washed with cold MeOH and
dried in oven at 50 ꢁC under vacuum yielding 33.3 g
References and notes
+
80%) of pure 10. ESI MS for C H Cl N O calcd
(
(
1
5
11
2
3
2
+
336.18) 335, 336.0 (MH ) found. H NMR (CDCl ):
1
1
. Torphy, T. J. Am. J. Respir. Crit. Care Med. 1998, 157,
3
3
1
51; Torphy, T. J. Am. J. Respir. Crit. Care Med. 1998,
57, 351.
. Dyke, H. J.; Montana, J. G. Expert Opin. Invest. Drugs
999, 8, 1301.
. Norman, P. Expert Opin. Ther. Pat. 2002, 12, 93; Dyke, H.
J.; Montana, J. G. Expert. Opin. Invest. Drugs 2002, 11, 1.
. Napoletano, M.; Norcini, G.; Pellacini, F.; Marchini, F.;
Ferlenga, P.; Pradella, L. Bioorg. Med. Chem. Lett. 2000,
10, 2235.
5. Napoletano, M.; Norcini, G.; Pellacini, F.; Marchini, F.;
Morazzoni, G.; Ferlenga, P.; Pradella, L. Bioorg. Med.
Chem. Lett. 2001, 11, 33.
6. Napoletano, M.; Norcini, G.; Pellacini, F.; Marchini, F.;
Morazzoni, G.; Ferlenga, P.; Pradella, L. Bioorg. Med.
Chem. Lett. 2002, 12, 5.
. Chakraborthi, A. K.; Gopalakrishnan, B.; Sobhia, E. S.;
Malde, A. Bioorg. Med. Chem. Lett. 2003, 13, 2473.
. Moriggi, J.-D.; Macecchini, S.; Pippo, L.; Fattori, R.;
Napoletano, M.; Fronza, G.; Raffaini, G.; Ganazzoli, F.;
Haack, T. XVII European Medicinal Chemistry Sympo-
sium, 2002, poster number 243.
d: 12.34 (s, 1H), 8.64 (s, 2H), 8.19–7.54 (m, 3H), 4.58
s, 2H), 3.95 (s, 3H).
(
2
3
4
1
4.4.10. 4-Chloro-1-(3,5-dichloro-pyridin-4-ylmethyl)-6-
methoxy-phthalazine (11). Phosphorus oxychloride
(22.2 mL, 0.230 mol) was added to a solution of 10
(10 g, 25.5 mmol) in acetonitrile (300 mL).The mixture
was refluxed for 3 h, the solution concentrated, taken
up in CH Cl and extracted with Na CO (pH 7–8).
The organic phases were discoloured with charcoal,
2
2
2
3
dried with Na SO and concentrated to yield 11 in stoi-
2
4
+
chiometric amounts. ESI MS for C H Cl N O calcd
1
5
10
3
3
+
1
(
354.63) 353, 354.0 (MH ) found. H NMR (CDCl ):
3
d 8.50 (s, 2H), 8.13–7.54 (m, 3H), 4.88 (s, 2H), 4.04 (s,
7
3
H).
8
4
(
.4.11. 1-(3,5-Dichloro-pyridin-4-ylmethyl)-6-methoxy-4-
1,2,4)triazol-1-yl-phthalazine (2). NaH (0.084 g,
2
1
.1 mmol, 60% in oil) was added to a solution of
,2,4-triazole (0.19 g, 2.8 mmol) in dry DMF (10 mL)
9
. Chakhavarti, N. et al. J. Chem. Soc. 1929, 196.
1
1
0. Bertini, V. et al. Heterocycles 1995, 41, 675.
1. The Nuclear Overhauser Effect in Structural and Confor-
mational Analysis; Neuhaus, D., Williamson, M., Eds.;
VCH: New York, 1989.
2. Accelrys Inc. InsightII 2000; Accelrys Inc.: San Diego,
CA, 2000. See also the URL http://www.accelrys.com/.
13. Dauber-Osguthorpe, P.; Roberts, V. A.; Osguthorpe, D.
J.; Wolff, J.; Genest, M.; Hagler, A. T. Proteins: Struct.,
Funct., Genet. 1988, 4, 31–47.
under N and the mixture was heated to 40 ꢁC for 1 h.
2
The cooled solution was added to 11 (0.5 g, 1.4 mmol),
heated overnight at 80 ꢁC, poured into water (10 vol)
and extracted three times with AcOEt. The organic
phases were discoloured with charcoal, dried with
1
Na SO and evaporated to dryness. The crude material
2
4
was purified by silica chromatography using AcOEt/PE
9/1) as eluant, and triturated in Et O, providing 4
(
2