Phenyl isoxazole voltage-gated sodium channel blockers: Structure and activity relationship

Article abstract of DOI:10.1016/j.bmcl.2011.05.041

Blocking of certain sodium channels is considered to be an attractive mechanism to treat chronic pain conditions. Phenyl isoxazole carbamate 1 was identified as a potent and selective Na_V1.7 blocker. Structural analogues of 1, both carbamates,

Full text of DOI:10.1016/j.bmcl.2011.05.041

Bioorganic & Medicinal Chemistry Letters 21 (2011) 3871–3876
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Phenyl isoxazole voltage-gated sodium channel blockers: Structure
and activity relationship
Istvan Macsari a, , Lars Sandberg a, Yevgeni Besidski a, Ylva Gravenfors a, Tobias Ginman a, Johan Bylund b,
⇑
Tjerk Bueters b, Anders B. Eriksson c, Per-Eric Lund c, Elisabet Venyike c, Per I. Arvidsson d,e,f
a Medicinal Chemistry, CNSP iMed Science, AstraZeneca R&D, Innovative Medicines, SE-15185 Södertälje, Sweden
b DMPK, CNSP iMed Science, AstraZeneca R&D, Innovative Medicines, SE-15185 Södertälje, Sweden
c Neuroscience, CNSP iMed Science, AstraZeneca R&D, Innovative Medicines, SE-15185 Södertälje, Sweden
d CNSP iMed Project Management, AstraZeneca R&D, Innovative Medicines, SE-15185 Södertälje, Sweden
e Organic Pharmaceutical Chemistry, Department of Medicinal Chemistry, Uppsala Biomedical Centre, Uppsala University, Box 574, SE-75123 Uppsala, Sweden
f School of Pharmacy and Pharmacology, University of KwaZulu-Natal, Durban 4001, South Africa
a r t i c l e i n f o
a b s t r a c t
Article history:
Blocking of certain sodium channels is considered to be an attractive mechanism to treat chronic pain
conditions. Phenyl isoxazole carbamate 1 was identified as a potent and selective NaV1.7 blocker. Struc-
tural analogues of 1, both carbamates, ureas and amides, were proven to be useful in establishing the
structure–activity relationship and improving ADME related properties. Amide 24 showed a good overall
in vitro profile, that translated well to rat in vivo PK.
Received 20 April 2011
Revised 11 May 2011
Accepted 12 May 2011
Available online 18 May 2011
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Sodium channel blockers
NaV1.7 blocker
Neuropathic pain
Phenyl isoxazoles
Voltage-gated channel
Millions of people worldwide are diagnosed with chronic
neuropathic pain. One of the main reason behind the disease is a
malfunction in pain signal transmission in the peripheral or central
nervous system. Voltage-gated sodium channels play a key role in
controlling the transmission of the pain signal. It was reported that
inherited mutation of the NaV1.7 channel in humans is responsible
for congenital insensibility for pain stimuli.1,2 Moreover, eryther-
malgia and familial rectal pain, two inheritable pain conditions in
humans, are also connected to NaV1.7 mutations.3 Unfortunately,
sodium channel blockers currently available in the clinic are only
partially efficacious for the treatment of neuropathic pain and
are often associated with tolerability limitations. It is widely
believed that identification of more potent and subtype selective
sodium channel blockers would provide a step-change for treat-
ment of neuropathic pain.4
NaV1.7 pIC50 7.55
NaV1.5 pIC50 5.26
clogP 4.48
LLE 3.07
hERG pIC50 <4.58
O
Cl
O
N
O
N
H
Cl
Cl
F
Solubility <1µM
RLM Clint NV
HLM Clint NV
1
Figure 1. Profile of phenyl isoxazole hit 1. LLE, lipophilic ligand efficiency is
calculated as NaV1.7 pIC50–c log P. NV, no value, the substance could not be
detected.
sodium-channel is widely expressed in heart muscle and inhibition
leads to ventricular arrhythmia.7,8 Furthermore, the compound had
low hERG affinity. However 1 was associated with very poor solu-
bility and metabolic instability. These early findings prompted us
to conduct further investigations.
In order to establish a structure–activity relationship within the
series close analogues were designed and synthesized. The results
are summarized in Table 1. We focused our attention first to probe
the right hand side of the molecule (dihalo-benzylalcohol part).
Monosubstitution at ortho position was well tolerated, that is, 2.
In 3, the combination of two chlorine atoms led to decreased
During the course of our NaV1.7 blocker program we have iden-
tified a series of phenyl isoxazoles, represented by compound 1
(Fig. 1). The compound showed high in vitro NaV1.7 potency and
good selectivity against NaV1.5, as measured in whole-cell voltage
clamp electrophysiology assays.5,6 Very high NaV1.5 selectivity is
an absolute requirement for the progression of a compound, as this
⇑
Corresponding author. Tel.: +46 8 553 24322; fax: +46 8 553 25646.
E-mail address: istvan.macsari@astrazeneca.com (I. Macsari).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.05.041

T
a
8
b
7
l
2
e 1
I. Macsari et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3871–3876
SAR of carbamate sub-series, based on compound 1
Compound
Structure
c log P
NaV1.7 pIC50
NaV1.5 pIC50
O
N
Cl
O
O
N
H
1
4.48
7.55
5.26
Cl
Cl
F
O
N
Cl
O
O
Cl
N
2
3
4.34
5.05
7.62
6.86
5.06
5.58
H
Cl
O
N
Cl
O
O
N
H
Cl
Cl
Cl
O
N
O
N
N
H
O
4
5
3.79
4.34
6.82
7.64
<4.59
5.06
Cl
Cl
O
N
O
N
H
O
Cl
O
Cl
N
O
N
O
N
H
Cl
Cl
O
6
7
8
4.42
5.15
5.84
7.54
6.65
6.52
5.87
N
O
O
N
O
N
Cl
O
<4.54
<4.48
Cl
N
O
O
N
O
Cl
N
O
Cl
Cl
F
O
N
Cl
O
9
5.01
3.87
8.03
6.48
5.14
O
N
H
Cl
Cl
F
O
O
N
Cl
O
10
<4.55
N
O
H
Cl
Cl
Cl
O
N
O
O
N
H
Cl
11
12
4.77
4.16
4.9
ND
Cl
F
O
N
Cl
O
O
N
H
7.14
<4.80
F
Cl
F
O
N
Cl
O
N
H
O
13
3.94
7.98
5.11
F

I. Macsari et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3871–3876
3873
NaV1.7 activity. When both ortho halogens of 1 were replaced by a
sole meta substituent, in 4, the NaV1.7 potency dropped. We were
particularly pleased to see that the replacement of ortho-substi-
tuted phenyl moiety with alkoxy-pyridyl group, 5 and 6, retained
the high potency; however, the NaV1.5 affinity also increased. This
type of moiety was successfully incorporated in other in-house ser-
ies of NaV1.7 channel blockers.9 The H-bond donor ability proved
to be important for the activity, as a more than 10-fold drop in po-
tency was observed when the carbamate nitrogen was methylated
(7 and 8).
The effect of alkyl substituent on the isoxazole ring was probed
by compounds 9–11. The replacement of methyl group with ethyl
led to an increased affinity, cf. 1 and 9. Contrarily, methoxymethyl
(10) and hydrogen (11) as substituents on the isoxazole ring were
not beneficial. Though ortho disubstitution of the phenyl ring, at-
tached to the isoxazole, appeared to be mandatory for potency
(cf. 16 and 17, vide infra), the replacement of chlorine atoms was
tolerated in several analogues. One of such examples is fluoro,
chloro derivative 12, another one is compound 13 where both chlo-
rine atoms were changed for methyl groups. Dimethyl analogue 13
is one of the few examples where a compound with reduced lipo-
philicity retained high NaV1.7 potency, though selectivity against
NaV1.5 did not improve.
The solubility (see Fig. 2) was generally poor for most of the car-
bamates, exception being 4–6, where Àlog S was 5 or lower corre-
sponding to a greater than 10 lM solubility. It seems likely that
lowering of c log P and changed substitution pattern on the right
hand-side together resulted in higher solubility. The metabolic sta-
bility of the carbamates was poor, as indicated by high clearance
values in human liver microsomes (see Fig. 3).
The low solubility of the carbamate sub-series was addressed
by preparing a range of urea analogues. The general synthesis of
Figure 3. HLM Clint versus c log P. Blue: carbamates; yellow: ureas and red:
amides. Clint could not be measured for compounds which are not shown.
carbamate and urea phenyl isoxazoles is outlined in Scheme 1.
The starting materials were either commercially available or could
readily be prepared according to published methods.6 Curtius rear-
rangement of acid A followed by the treatment with an appropriate
alcohol B provided carbamates represented by 1. Urea 14 was pre-
pared by using the corresponding amine B as reagent. Carbamates
were readily N-methylated (8) by employing standard conditions.
In order to prepare compound 16, which represented an N-methyl-
ated urea, a longer route was devised. Acid A was subjected to Cur-
tius rearrangement in tert-butanol providing the N-Boc protected
amino isoxazole which was N-methylated and then deprotected.
This intermediate C was coupled with appropriate amine B to give
compound 16.
The low solubility of the carbamate sub-series was addressed
by preparing a range of urea analogues. The general synthesis of
carbamate and urea phenyl isoxazoles is outlined in Scheme 1.
The starting materials were either commercially available or could
readily be prepared according to published methods.6 Curtius rear-
rangement of acid A followed by the treatment with an appropriate
alcohol B provided carbamates represented by 1. Urea 14 was pre-
pared by using the corresponding amine B as reagent. Carbamates
were readily N-methylated (8) by employing standard conditions.
In order to prepare compound 16, which represented an N-methyl-
ated urea, a longer route was devised. Acid A was subjected to Cur-
tius rearrangement in tert-butanol providing the N-Boc protected
amino isoxazole which was N-methylated and then deprotected.
This intermediate C was coupled with appropriate amine B to give
compound 16.
The SAR for ureas is summarized in Table 2. The introduction of
the urea moiety in 1 and 13 (i.e., 14 and 15) increased the lipophil-
icity and, at the same time, lowered the potency. As previously
Figure 2. Dried DMSO solubility (Àlog S) versus c log P. Blue: carbamates; yellow:
ureas and red: amides.
seen in the carbamate sub-series, a-alkoxy pyridines (5 and 6,

3874
I. Macsari et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3871–3876
O
O
O
Cl
O
Cl
N
N
H
i
Cl
Cl
Ar
X
N
O
N
X
ii
+
+
H
R1
N
O
Cl
Cl
R3
X = O, NH
F
F
1, X = O
14, X = NH
B
8
CO2H
R2
O
O
O
N
A
N
R1
Cl
i, ii, iii
N
N
H
O
N
iv
H
Ar
NH2
+
HO
R2
Cl
N
O
B
16
C
Scheme 1. Reagents and conditions: (i) diphenylphosphorylazide, NEt3, dioxane, heating; (ii) NaH, MeI, THF; (iii) TFA, DCM; (iv) CDI, DCM.
Table 2
SAR of urea sub-series
Compound
Structure
c log P
NaV1.7 pIC50
NaV1.5 pIC50
O
N
Cl
O
N
H
N
H
14
5.4
6.79
5.07
Cl
Cl
F
O
N
O
Cl
N
N
15
16
17
4.85
3.89
3.42
7.42
6.77
5.44
<4.98
5.23
H
H
F
O
N
O
N
N
H
O
Cl
Cl
N
O
O
N
O
N
N
O
<4.48
H
Cl
N
O
O
O
N
O
N
N
18
19
3.34
4.16
5.35
6.68
<4.48
4.70
H
O
N
O
N
O
N
N
H
H
O
N
O
N
O
N
H
Cl
N
H
20
4.6
6.58
<4.48
Cl
N
N
O
N
O
N
H
N
H
21
22
4.05
3.46
5.93
4.9
<4.48
<4.48
N
N
O
N
O
N
H
N
H
Cl
Cl
N
O
N

I. Macsari et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3871–3876
3875
O
O
NaV1.7 pIC50 6.75
NaV1.5 pIC50 <4.48
clogP 3.79
LLE 2.96
hERG pIC50 <4.58
N
N
O
O
R1
R1
O
N
i, ii, iii, iv
N
H
CO2H
Cl
OH
F
F
R1
Cl
R1
O
N
F
Solubility 36 µM
RLM Clint 64 µl/min/mg
HLM Clint 116 µl/min/mg
D
A
24
+
Ar
NH2
Figure 4. Profile of phenyl isoxazole 24.
v
B
O
O
It is important to note that a reverse linear relationship between
solubility and c log P could be established. Furthermore, the meta-
bolic stability in human liver microsomes was improved (Fig. 3)
particularly for 17, 19, and 22.
N
N
H
Cl
Cl
N
O
Both the carbamate and urea sub-series were shown to have
reactive metabolite liabilities, as proven by glutathione trapping
experiments. The detailed mechanism for this reactivity will be
outlined in a separate paper. We believe the liability has to do with
the presence of the N–H group at C-4; thus derivatives where this
group is replaced, for example, by a methylene group, would have
decreased tendency to generate reactive metabolites.
The synthesis of amides is outlined in Scheme 2. Acid A, alterna-
tively the corresponding ester, was reduced to alcohol, then con-
verted to a nitrile intermediate. The nitrile group was hydrolyzed
under acidic conditions to yield acid D which, in turn, could be cou-
pled with amine B using standard conditions.
The activity data of the amide sub-series are summarized in Ta-
ble 3. Compounds 23 and 24 showed high NaV1.7 activity, however
in case of 25 the replacement of ortho halogens with methyl groups
decreased the potency as was previously seen for ureas (cf. 21 and
20 in Table 2). Generally, the amides were less active in the NaV1.7
assay than corresponding ureas and carbamates (cf. 26 and 20 in
23
Scheme 2. Reagents and conditions: (i) DIBAL-H, THF; (ii) SOCl2, DCM; (iii) NaCN,
DMSO; (iv) cc. AcOH, 4 M H2SO4, heating; (v) HBTU, DIPEA, DMF.
Table 1) were well tolerated, therefore we focused our attention on
this type of substitution pattern. Mono substitution on the phenyl
group caused a drop in potency in 17. The presence of two ortho
substituents apparently brings the molecule into a more beneficial
conformation for binding to the ion-channel. When the chlorine
atoms of 20 were replaced by methyl groups in 21 the NaV1.7 po-
tency decreased. Introduction of a polar 4-pyridine in 22 resulted
in the loss of activity.
In general the NaV1.7 activity was lowered by the introduction
of an urea moiety; however, as compare to the carbamate sub-ser-
ies, the low NaV1.5 affinity was retained. On a positive note, the
solubility increased when going from carbamates to ureas (Fig. 2).
Table 3
SAR of amide sub-series
Compound
Structure
c log P
NaV1.7 pIC50
NaV1.5 pIC50
O
N
O
N
H
23
3.84
6.9
5.33
Cl
Cl
N
O
O
N
O
N
24
25
3.79
3.23
6.75
5.91
<4.48
<4.48
H
Cl
F
Cl
N
O
F
F
O
N
O
N
H
F
N
O
F
F
O
N
O
N
H
26
27
3.73
2.98
6.07
7.23
<4.48
4.92
Cl
N
Cl
N
O
N
O
N
H
N
O
F
F
O
F

3876
I. Macsari et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3871–3876
Table 4
Rat in vivo PK for 1 and 24
olite issue in both series, amide analogues (23–27) were prepared.
Compound 24 showed the best overall in vitro profile, that trans-
lated well to rat in vivo PK.
Compound
1
24
AUC/dosea (h kg/L)
Vss (L/kg)
0.73 0.04
0.46 0.04
1.6 0.5
0.6 0.2
36 2.5
0.12 0.01
0.40 0.04
0.3 0.1
3.8 0.9
19 1.6
a
Acknowledgment
3
0.6
a
t
(h)
1.4 0.1
45 5.5
0.001 0.0004
0.007 0.002
0.33 0.12
NDc
½
CLa (mL/min/kg)
The authors wish to thank colleagues at Physical Chemistry
Characterization Team for providing solubility data.
AUC/doseb (h kg/L)
b
Cmax
(
l
mol/L)
b
tmax (h)
b
t
(h)
½
Supplementary data
Fb (%)
0.3 0.1
a
b
c
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.05.041.
Intravenous (iv), dose 3
l
mol/kg.
Per oral (po), dose 10
ND: not determined.
lmol/kg.
References and notes
Table 2). Interestingly the least lipophilic among the amides (27)
showed unexpectedly high potency. High selectivity against
NaV1.5 and good solubility (see Fig. 2) were observed in the amide
sub-series. The metabolic stability profile was somewhat im-
proved, and the amide sub-series did not demonstrate the reactive
metabolite potential previously seen in the carbamate and urea
sub-series. Compound 24 showed attractive overall profile (see
Fig. 4) in terms of NaV1.7 potency, selectivity against NaV1.5 and
in vitro metabolic stability. In vivo rat PK data for 1 and 24 are pre-
sented in Table 4.10
The clearance for 24 is reduced compared to compound 1, both
in vitro and in vivo. However, this is not reflected in the terminal iv
half life because the volume of distribution has also been changed
considerably. On the other hand, per oral PK data for 24 showed
that improved solubility and metabolic stability led to greatly in-
creased exposure, Cmax and oral bioavailability and significantly
longer half life than 1.
1. Cox, J. J.; Reimann, F.; Nicholas, A. K.; Thornton, G.; Roberts, E.; Springell, K.;
Karbani, G.; Jafri, H.; Mannan, J.; Raashid, Y.; Al Gazali, L.; Hamamy, H.; Valente,
E. M.; Gorman, S.; Williams, R.; Mchale, D. P.; Wood, J. N.; Gribble, F. M.;
Woods, C. G. Nature 2006, 444, 894.
2. Ahmad, S.; Dahllund, L.; Eriksson, A. B.; Hellgren, D.; Karlsson, U.; Lund, P.-E.;
Meijer, I. A.; Meury, L.; Mills, T.; Moody, A.; Morinville, A.; Morten, J.; O’Donnell,
D.; Raynoschek, C.; Salter, H.; Rouleau, G. A.; Krupp, J. J. Hum. Mol. Genet. 2007,
16, 2114.
3. Yang, Y.; Wang, Y.; Li, S.; Xu, Z.; Li, H.; Ma, L.; Fan, J.; Bu, D.; Liu, B.; Fan, Z.; Wu,
G.; Jin, J.; Ding, B.; Zhu, X.; Shen, Y. J. Med. Genet. 2004, 41, 171.
4. Zuliani, V.; Rivara, M.; Fantini, M.; Constantino, G. Expert Opin. Ther. Patients
2010, 20, 755.
5. Harmer, A. R.; Abi-Gerges, N.; Easter, A.; Woods, A.; Lawrence, C. L.; Small, B. G.;
Valentin, J.-P.; Pollard, C. E. J. Pharm. Toxicol. Methods 2008, 57, 30.
6. Besidski, Y.; Claesson, A.; Macsari, I.; Sandberg, L. PCT Int. Appl.,
WO2009010784, 2009.
7. Goldin, A. L. Annu. Rev. Physiol. 2001, 63, 871.
8. Tfelt-Hansen, J.; Winkel, B. G.; Grunnet, M.; Jespersen, T. J. Cardiovasc.
Electrophysiol. 2010, 21, 107.
9. 6-Alkoxypyridinyl-3-methylamines: Arvidsson, P. I.; Besidski, Y.; Csjernyik, G.;
Sandberg, L. PCT Int. Appl., WO2009145720, 2009. 2-alkoxy-nicotinic acids:
Besidski, Y.; Kers, I.; Nylöf, M.; Sandberg, L.; Skogholm, K. PCT Int. Appl.,
WO2008130320, 2008. Besidski, Y.; Kers, I.; Macsari, I.; Nylöf, M.; Rotticci, D.
PCT Int. Appl., WO2008130321, 2008. Besidski, Y.; Gravenfors, Y.; Macsari, I.
PCT Int. Appl., WO2008130323, 2008.
10. Formulation of compound 1 for both iv and po studies: 5% DMA, 40% PEG, 55%
water. In case of 24 5% DMA, 95% HPbCD were used for both studies. In all cases
fasted, Sprague–Dawley, male rats were used, three animals/administration/
compound.
In summary, we established the structure–activity relationship
in the phenyl isoxazole class of NaV1.7 blockers. We started from
the potent but poorly soluble and metabolically unstable carba-
mates (1–13). By introducing urea moiety (14–22) the NaV1.7 po-
tency was somewhat lowered but the solubility and metabolic
stability were improved. In order to overcome the reactive metab-