3-oxoisoindoline-1-carboxamides: Potent, state-dependent blockers of voltage-gated sodium channel nav1.7 with efficacy in rat pain models

Article abstract of DOI:10.1021/jm300623u

The voltage-gated sodium channel Na_V1.7 is believed to be a critical mediator of pain sensation based on clinical genetic studies and pharmacological results. Clinical utility of nonselective sodium channel blockers is limited due to serious adverse drug effects. Here, we present the optimization, structure-activity relationships, and in vitro and in vivo characterization of a novel series of Na_V1.7 inhibitors based on the oxoisoindoline core. Extensive studies with focus on optimization of Na _V1.7 potency, selectivity over Na_V1.5, and metabolic stability properties produced several interesting oxoisoindoline carboxamides (16A, 26B, 28, 51, 60, and 62) that were further characterized. The oxoisoindoline carboxamides interacted with the local anesthetics binding site. In spite of this, several compounds showed functional selectivity versus Na _V1.5 of more than 100-fold. This appeared to be a combination of subtype and state-dependent selectivity. Compound 28 showed concentration- dependent inhibition of nerve injury-induced ectopic in an ex vivo DRG preparation from SNL rats. Compounds 16A and 26B demonstrated concentration-dependent efficacy in preclinical behavioral pain models. The oxoisoindoline carboxamides series described here may be valuable for further investigations for pain therapeutics.

Full text of DOI:10.1021/jm300623u

Article
pubs.acs.org/jmc
3‑Oxoisoindoline-1-carboxamides: Potent, State-Dependent Blockers
of Voltage-Gated Sodium Channel NaV1.7 with Efficacy in Rat Pain
Models
Istvan Macsari,*,† Yevgeni Besidski,† Gabor Csjernyik,† Linda I. Nilsson,† Lars Sandberg,† Ulrika Yngve,†
Kristofer Åhlin,† Tjerk Bueters,‡ Anders B. Eriksson,∥ Per-Eric Lund,§ Elisabet Venyike,§ Sandra Oerther,§
Karin Hygge Blakeman,§ Lei Luo,§ and Per I. Arvidsson∥,⊥,#
‡
§
∥
†Medicinal Chemistry, DMPK, Neuroscience, Project Management, CNSP iMed, AstraZeneca R&D Sodertalje, SE-151 85
̈
̈
Sodertalje, Sweden
̈
̈
⊥Organic Pharmaceutical Chemistry, Department of Medicinal Chemistry, Uppsala Biomedical Centre, Uppsala University, Box 574,
SE-751 23 Uppsala, Sweden
#School of Pharmacy and Pharmacology, Westville Campus, University of KwaZulu-Natal, Private Bag X54001, Durban 4000, South
Africa
S
* Supporting Information
ABSTRACT: The voltage-gated sodium channel NaV1.7 is
believed to be a critical mediator of pain sensation based on
clinical genetic studies and pharmacological results. Clinical
utility of nonselective sodium channel blockers is limited due
to serious adverse drug effects. Here, we present the
optimization, structure−activity relationships, and in vitro
and in vivo characterization of a novel series of NaV1.7
inhibitors based on the oxoisoindoline core. Extensive studies
with focus on optimization of NaV1.7 potency, selectivity over
NaV1.5, and metabolic stability properties produced several
interesting oxoisoindoline carboxamides (16A, 26B, 28, 51,
60, and 62) that were further characterized. The oxoisoindo-
line carboxamides interacted with the local anesthetics binding site. In spite of this, several compounds showed functional
selectivity versus NaV1.5 of more than 100-fold. This appeared to be a combination of subtype and state-dependent selectivity.
Compound 28 showed concentration-dependent inhibition of nerve injury-induced ectopic in an ex vivo DRG preparation from
SNL rats. Compounds 16A and 26B demonstrated concentration-dependent efficacy in preclinical behavioral pain models. The
oxoisoindoline carboxamides series described here may be valuable for further investigations for pain therapeutics.
other sodium channels, i.e., NaV1.1−1.9, that influence basal
functions in the brain, heart, or muscle. Consequently, much
effort has been focused on identifying potent and selective
NaV1.7 blockers, which do not inhibit other members of the
voltage-gated sodium channel family.10−17 There are several
ways to obtain selectivity between NaV1.7 and other sodium
channels. True subtype selectivity for NaV1.7 could be achieved
by identifying a unique binding site at NaV1.7 that is not present
within other NaV channels. This is challenging, indeed, because
all sodium channel are very similar. An alternative is to develop
state-dependent ligands that preferentially bind to the inactivated
state of the sodium channels and exploit the difference in kinetic
states of different sodium channels in different tissues. It is
believed that in hyperexcitable neurons that are an underlying
cause for chronic pain, most NaV1.7 channels are inactivated.18,19
INTRODUCTION
■
The voltage-gated sodium channel NaV1.7 represents an
extraordinarily interesting drug target for the treatment of pain.
Clinical genetic studies by several groups have established strong
genetic links between mutations in the gene SCN9A, coding for
NaV1.7 and pain related conditions.1 Loss of function truncation
mutations of SCN9A result in congenital insensibility for pain
stimuli, without additional major physiological abnormalities.2−5
In contrast, gain of function mutations that increase channel
opening are known to cause erythermalgia and familial rectal
pain, two human inheritable pain conditions.6−8 Additional
support for the potential of voltage-gated sodium channel
modulators as pain treatments comes from the notion that a
number of clinically used agents developed for other indications,
such as mexiletine, lamotrigine, and carbamazepine, showed
analgesic activity through nonselective blockade of sodium
channels.9 However, these drugs give rise to serious adverse
events in the clinical setting, presumably due to the inhibition of
Received: May 4, 2012
Published: July 9, 2012
© 2012 American Chemical Society
6866
dx.doi.org/10.1021/jm300623u | J. Med. Chem. 2012, 55, 6866−6880

Journal of Medicinal Chemistry
Article
a
Scheme 1. Preparation of Oxoisoindoline Carboxamides via Ugi Reaction (Method 1)
a
Reagents: (a) MeOH, rt; (b) borane, THF, 60 °C; (c) phenylformate, DCM, 0 °C; (d) POCl3, DIPEA, DCM, −15 °C. R1 = H, Me; R2, R3 =
varied; R4 = H, F, Me, OMe; R5 = H, F. (e) BBr3, DCM, −78 °C for R4 = OH.
A combination of subtype selectivity and state dependency could
confer compounds that possess sufficient functional selectivity in
vivo in order to avoid the adverse effects of the current clinically
available nonselective sodium blockers. Herein we present the
optimization, structure−activity relationship (SAR), and in vitro
and in vivo characterization of a novel series of NaV1.7 inhibitors
based on the oxoisoindoline core.
Scheme 2. Preparation of Oxoisoindoline Carboxamides via
Amide Coupling (Method 2)
a
CHEMISTRY
■
The synthesis of oxoisoindoline carboxamides (4) is based upon
the three component Ugi reaction and outlined in Scheme 1.
Acid 1, amine 2, and isocyanide 3 were mixed in methanol and
stirred at room temperature overnight to afford the desired
products.20−23 The method was general, and a wide range of
amines and isocyanides could be used in a library fashion to
prepare the desired molecules. The commercial availability of
functionalized aryl isocyanides is limited, therefore a short
sequence starting from the appropriate aryl nitrile (5) was
devised. The amines 6 were prepared by reduction of 5 and then
subjected to formylation and water elimination reactions in order
to obtain the functionalized isocyanides 3.
The synthesis of substituted isooxindoles was also achieved by
the above-mentioned Ugi reaction (Scheme 1) employing 4- or
7-substituted isobenzofuranones (7), which were prepared by
published methods.24,25 Fluoro, methyl, and methoxy groups
were introduced using this approach. Moreover, demethylation
of methoxy moiety afforded the phenol derivative of 4.
An alternative method to the Ugi reaction was also developed,
as depicted in Scheme 2.20−23 The diester 8 was brominated in
the benzylic position and then reacted with the appropriate
amine 2 to form the oxoisoindoline ring, followed by hydrolysis
of the ester intermediate.26 The acid 9 was reacted with benzyl
amine 6 to afford the product amide 4. This method made it
possible to avoid the use of poisonous isocyanide reagent. On the
other hand, acid 9 proved to be very unreactive in the amidation
reaction and only via the very sensitive acid fluoride could the
coupling be achieved.
a
Reagents: (a) NBS, AIBN, CCl4, reflux; (b) 2, TEA, acetonitrile, 0
°C; (c) 1 M NaOH, MeOH, rt; (d) 6, fluoro-N,N,N′,N′-
tetramethylformamidinium hexafluorophosphate, TEA, DMF, 45 °C.
R5 = H, OMe. (e) BBr3, DCM, −78 °C for R5 = OH.
BIOLOGICAL EVALUATION
■
Oxoisoindoline carboxamides were initially identified as blockers
of the human NaV1.7 channel, recombinantly expressed in HEK
293 cells, using a Li flux atomic absorption spectroscopy (AAS)
assay.27 Potency of identified and synthesized compounds was
determined by automated perforated whole-cell patch clamp
using the automated IonWorks instrument, mainly with a steady-
state protocol (Vhold −65 mV), which is physiologically relevant
to the conditions in C-type dorsal root ganglion neurons.28,29
Analogues were further examined for their selectivity mainly over
the human NaV1.5 channel using a physiologically relevant assay
with a Vhold of −90 mV.30,31 In addition, selectivity over the
hERG channel as well as analysis of state-dependent properties of
the compounds on NaV1.7 was determined. Mode of action
studies were conducted on a point-mutated channel where one of
the key amino acid residues of the local anesthetic binding site
was changed (F1737A). Furthermore, key compounds were
tested against a broader panel of mainly CV and CNS liability
targets. Finally, on the basis of their potency, selectivity, and
pharmacokinetic profiles, compounds were evaluated for
analgesic efficacy using ex vivo and in vivo models of
inflammatory and neuropathic pain. These studies include ex
vivo effects of L5/L6 spinal nerve ligation-induced (Chung)
ectopic activity, in vivo behavioral effects in the formalin model of
6867
dx.doi.org/10.1021/jm300623u | J. Med. Chem. 2012, 55, 6866−6880

Journal of Medicinal Chemistry
Article
peripherally driven spontaneous pain, and the carrageenan-
induced monoarthritis model of chronic inflammatory pain.
20 was tolerated but did not significantly influence the potency
(cf. 19). The phenyl ring was replaced by a pyridine ring in
compounds 21−23. This type of moiety was successfully
incorporated in other in-house series of NaV1.7 channel
blockers.14 Unfortunately, the potency was lower compared to
the corresponding phenyl analogues (21 and 22, cf. 16 and 19).
On the other hand, the selectivity over NaV1.5 was improved.
These pyridine-containing compounds also showed an improved
in vitro metabolic stability profile compared to the other
analogues presented in Table 1, presumably, at least partly, due
to lower lipophilicity. Particularly, the metabolic stability in
human liver microsomes for most other analogues was too high
and required additional optimization in order to meet the desired
profile of once or twice daily dosing in human.
RESULTS AND DISCUSSION
■
During the course of our NaV1.7 blocker program, we identified a
series of oxoisoindoline carboxamides from a targeted library
represented by compound 10 (Figure 1). Analysis of the library
In general, the meta and para substituted benzyl amides were
equally well tolerated (11, 12, 16, and 17). Small groups such as
trifluoromethyl (14) and trifluoromethoxy (16, 17) had the
highest positive influence on the NaV1.7 potency. Alkoxy (15) or
trifluoroalkoxy groups, especially in case of pyridines 21−23,
helped to lower the NaV1.5 potency. The solubility was generally
good for most of the benzyl amides, the exception being 12 and
14, where the solubility dropped below 10 μM. The analyzed
compounds blocked the hERG channel with potencies between
7 and 20 μM.
Compound 16 showed an attractive profile due to good
NaV1.7 potency and in vitro microsomal stability in rat compared
to close analogues (Table 1). Therefore, we decided to study this
compound further. The enantiomers were separated and tested;
however, the absolute stereochemistry could not be determined,
and instead the first eluting enantiomer from the chiral HPLC
column was labeled A and the second eluting enantiomer was
labeled B. We were pleased to see that enantiomer 1 (16A)
showed 56 nM potency against NaV1.7 (Table 1, cf. 16). For the
less active enantiomer 2 (16B), a potency value of 1.4 μM was
determined. The pharmacokinetic profile of 16A in rat is
summarized in Table 2. Following intravenous administration,
the plasma concentration of 16A increased slightly from 6 to 24 h
due to redistribution or reabsorption. This explained the large
volume of distribution. The clearance was approximately 40% of
plasma liver blood flow, which correlated well to the observed
metabolic stability in vitro. The oral pharmacokinetics
demonstrated rapid absorption, good exposure, and bioavail-
ability. Encouraged by the favorable pharmacokinetic properties
in rodents, and with the high hERG affinity and low human liver
microsome stability in mind, 16A appeared a good preclinical
tool compound. It was also used as a starting point for further
optimizations.
Figure 1. Profile of oxoisoindoline carboxamide hit 10. LLE, lipophilic
ligand efficiency, is calculated as NaV1.7 pIC50 − clogP.
showed that close analogues to 10 with medium NaV1.7 potency
(IC50 < 0.5 μM) all possessed an ortho disubstituted aniline group
regardless of the side chain. Those compounds showed very poor
solubility, typically around or below 1 μM. We were particularly
pleased to see that 10, which does not have a substituent in the
ortho position, was associated with improved solubility. However,
10 showed modest in vitro NaV1.7 potency and no selectivity
over NaV1.5, as measured in whole-cell voltage clamp electro-
physiology assays.31 Good NaV1.5 selectivity is an absolute
requirement for further development of a potential drug
candidate, as this sodium channel is widely expressed in the
heart muscle and its inhibition leads to ventricular arrhyth-
mia.32,33 These initial findings prompted us to conduct further
investigations aimed to identify compounds with higher NaV1.7
potency and increased selectivity toward NaV1.5. CNS exposure
was not in the scope of the optimization because the expression
of the NaV1.7 protein suggests a peripheral mechanism of action.
In our first attempt to optimize the series, we focused our
attention on replacing the aniline group, which could become a
toxicity liability if the amide function is hydrolyzed in vivo.
Because the dialkyl substituted aniline moiety in 10 is well
tolerated, replacement with a benzyl amine moiety should
potentially be possible due to similar steric demand. We initially
decided to design and prepare analogues to 10 with variations of
the amide part. The results of 13 synthesized analogues are
summarized in Table 1.
Next, we decided to probe the effects of the side chain
modifications on the SAR. The generated results are summarized
in Table 3. Replacement of the pyridine ring of 16 by
tetrahydropyran (24) and morpholine (25) mainly lowered the
NaV1.5 channel potency. Surprisingly, the removal of the chain in
26 gave an equipotent compound to 24. The enantiomers of 26
showed different blocking activity on the 1.7 channel, as we have
previously seen for 16 (Table 1). Furthermore, the contraction of
the ring to tetrahydrofuran (27) was well tolerated. When the
oxygen atom of 26 was replaced by a −CF2 group in 28, the
NaV1.7 potency increased. It is interesting to note the ∼70-fold
difference in potency against the NaV1.7 channel between the
two enantiomers (28A,B). A cyano group (29) or carbamate
function (30) were not beneficial in this position. The
introduction of the smaller azetidine and cyclobutyl rings in
31−33 retained NaV1.7 potency, cf. 28. When the ring was
Our hypothesis proved to be correct as compound 11 retained
NaV1.7 potency in the steady state protocol (vide infra) and even
increased the selectivity somewhat over NaV1.5. The isomeric 12
was as potent as 11, but the bromo atom in the ortho position
(13) was not tolerated. The trifluoromethyl group (14) helped
to further increase the NaV1.7 potency (cf. 10), whereas the
NaV1.5 blocking activity was retained. The electron donating
methoxy group (15) did not influence NaV1.7 potency but
considerably lowered the NaV1.5 activity, thereby improving
selectivity. In 16 and 17, the introduction of trifluoromethoxy
group increased the NaV1.7 potency and, similarly to 14, did not
considerably influence the potency to the NaV1.5 channel. The
replacement of the trifluoromethyl group with the trifluoroethyl
functionality in 18 gave a similarly potent compound as 14.
Further elongation of the chain in 19 helped only to lower the
NaV1.5 potency (cf. 16). The introduction of the methyl group in
6868
dx.doi.org/10.1021/jm300623u | J. Med. Chem. 2012, 55, 6866−6880

Journal of Medicinal Chemistry
Article
Table 1. The Effect of SAR around the Variation of Amide Part
a
b
c
Vhold −65 mV. Vhold −90 mV. Dried DMSO solubility, ref 34. ND, not determined.
replaced by short aliphatic chains in 34−37, somewhat decreased
NaV1.7 potencies, cf. 16, were observed.
Unfortunately, the NaV1.7 potency was negatively influenced
by the introduction of smaller aliphatic chains and rings, the
exceptions being 28, 31 and 32. On the positive note, the NaV1.5
affinity was significantly lowered in most cases, which led to an
increased selectivity window (up to 100 times). Furthermore, the
solubility improved with a few exceptions (28, 32, 33, and 35)
and the hERG potency was lowered. The good metabolic
stability in rat was also retained and in human significantly
6869
dx.doi.org/10.1021/jm300623u | J. Med. Chem. 2012, 55, 6866−6880

Journal of Medicinal Chemistry
Article
high selectivity over the NaV1.5 channel and good metabolic
stability measured in rat liver microsomes.
Table 2. Rat in Vivo PK for 16A, 26B, 28, 51, 60, and 62
16A
26B
28
51
60
62
The oxoisoindoline carboxamides contain a chiral center and
are prone to racemization due to the acidity of the proton.
Therefore, a methyl group was introduced into the quaternary
center (52, in Table 5) and the enantiomers were separated by
chiral HPLC. We found that the enantiomers showed different
potency for the NaV1.7 channel similarly to 16, 26, and 28 (IC50
= 3.6 μM for 52A and IC50 = 0.17 μM for 52B). The absolute
configurations were not determined. In comparison to 16, the
racemic 52 showed decreased potency but increased selectivity
against NaV1.5. Unfortunately, both solubility and metabolic
stability were dramatically decreased.
a
AUC (h·μmol/L)
1.7
7.6
0.8
29
1.7
0.8
1.0
30
2.4
5.2
4.1
21
1.4
1.1
0.4
37
2.1
0.7
0.4
29
1.6
0.9
0.4
32
a
Vss (L/kg)
a
t1/2 (h)
a
CL (mL/min/kg)
b
AUC (h·μmol/L)
2.1
0.6
0.5
5.9
37
1.1
0.6
0.3
1.9
20
6.5
0.5
3.7
5.9
80
2.1
0.8
0.3
2.3
44
0.4
0.09
0.4
8.1
6
0.2
0.1
0.3
1.2
4
b
Cmax (μmol/L)
b
tmax (h)
b
t1/2 (h)
b
F
(%)
fu (% free)
4.5
13.2
3.0
9.2
18.0
13.0
a
b
Intravenous (iv), dose 3 μmol/kg. Per oral (po), dose 10 μmol/kg.
To gain further insight to the structure−activity relationship in
the series, different functional groups were introduced on the
oxoisoindoline core (see Table 5). Replacement of the proton in
the 7 position (R4) with fluoride atom (53), methyl (54), and
methoxy (55) functionalities were well tolerated. However, a
hydroxyl group (56) lowered the NaV1.7 potency. Among the
modifications in the 4 position (R5, 57−59), all except the
methoxy functional group was tolerated (58). The NaV1.5
potency was decreased in all cases, except in 53 and 59.
Unfortunately, the solubility was very poor for all compounds,
with exceptions for 53, 57, and 59. Furthermore, the affinity to
the hERG channel increased somewhat with only one exception
(57) and the metabolic stability was lower throughout the series.
In summary, the fluoro atom in the 7 and the hydroxyl group in
the 4 position only led to compounds which showed similar
properties as 16 but provided no clear advantages.
At this point, we turned our attention again on compound 51,
which showed an attractive profile with high selectivity over
NaV1.5 channel and good metabolic stability measured in rat liver
microsomes (Table 4). We decided to design and prepare a few
analogues to 51 to improve selectivity against the hERG channel
among others. The generated results are shown in Table 6. In 60,
the trifluoromethoxy group in the meta position (R7) was
tolerated as previously seen for 17 (Table 1). Furthermore,
disubstituted 61 and 62 also retained the good potency. The
hydroxyl group in R5 (63) slightly increased the NaV1.7 activity,
as seen in 59 (see Table 5). The selectivity over NaV1.5 was
retained for all compounds. The solubility increased in case of 60
and was very poor in 63. Compounds 60 and 62 showed
significantly lower potency for the hERG channel than 51.
The overall profiles of 60 and 62 were promising, and rat
pharmacokinetics were determined together with 51 (Table 2).
When administered intravenously, all three compounds showed
very similar profiles. The oral PK profile of 51 was similar to 16A.
However, in the cases of 60 and 62, low exposure and
bioavailability were measured. Surprisingly, the good in vitro
metabolic stability in rat for 60 and 62 did not translate well to
good overall in vivo PK data. The low bioavailability of 60 and 62
could not be explained by systemic metabolism or poor
permeability. The clearance after iv administration was
approximately 40% of plasma liver blood flow, similarly to 16
or 51. Also, in vitro permeability determined in CaCo2 cells was
excellent for these compounds, 50 × 10−6 and 39 × 10−6 cm/s for
60 and 62, respectively. Possible explanations are presystemic
metabolism, instability, or precipitation in the rat gut. Although
these compounds might still have acceptable pharmacokinetics
in human based on scaling, poor rodent PK poses a great hurdle
in the further development of such compounds.
improved. This suggested that the ethylpyridine moiety was
associated with a metabolic soft-spot, particularly in human
microsomes. Subsequent biotransformation studies showed that
this moiety was indeed oxidized at several locations. In summary,
compound 26B and 28 showed the most attractive overall
profile. The pharmacokinetics of 26B and 28 in rat are
summarized in Table 2. Particularly, the long-lasting exposure
and high bioavailability of 28 after 10 μmol/kg po was of interest.
To allow administration of larger doses, however, the
formulation had to be changed from a solution-based to a
suspension-based one. The total AUC in rat was reduced 3-fold
to 2 h·μmol/L and the Cmax 2-fold to 0.3 μmol/L after 30 μmol/
kg per oral dosing with a suspension-based formulation
compared to the 10 μmol/kg dose given as a solution,
presumably caused by the low solubility of 28. This impeded
additional in vivo studies with 28. Therefore, 26B, which had
significantly higher solubility than 28 and similar PK profile as
16A, was chosen for further studies (vide infra).
At this point in our investigation, we turned our attention to
design and prepare close analogues to the ethyl pyridine side
chain of 16 to investigate primarily if the human metabolic
stability could be improved. The generated results are
summarized in Table 4. The isomeric pyridines (38 and 39, cf.
16) were tolerated, 39 being the least active. The introduction of
small substituents on the ring, such as fluoride (40) and methyl
(41), retained the NaV1.7 potency. In the case of compounds
42−44, the chain length was varied; elongation was well
tolerated (42 and 43) but not shortening (44). When a methyl
substituent was introduced on the chain, the NaV1.7 potency was
retained in 45 and 46, however, decreased by 10 times in case of
47. On the positive note, difluoro substitution of the chain (48)
increased the NaV1.7 potency. Replacement of the pyridine ring
with a pyrimidine (49) and pyridazine (50) decreased the NaV1.7
blocking activity. The para fluoro substitution in 51 increased the
potency, cf. 49.
In general, the good NaV1.7 potency was retained throughout
the subseries. Furthermore, the NaV1.5 potency was decreased in
all cases, except for 38, 40, 41, and 48. The solubility was
improved for most compounds, the exceptions being 40−44, 46,
and 47. A few possibilities were identified that improved the
metabolic stability in human microsomes. Removal (44) or
disubstitution (47−48) of the α-C carbon in the ethyl chain
diminished oxidative metabolites. Also, replacement of pyridine
by pyrimidine reduced oxidative metabolism. However, the
potency for the hERG channel was retained, or slightly increased,
for all compounds with the exception of 49. We would like to
highlight compound 51, which showed an attractive profile with
Of the 23 compounds with a NaV1.7 potency below 0.2 μM,
seven compounds (28A and B, 43, 46, 52B, 55, 60, and 61)
6870
dx.doi.org/10.1021/jm300623u | J. Med. Chem. 2012, 55, 6866−6880

Journal of Medicinal Chemistry
Article
Table 3. SAR of Replacements of Ethyl Pyridine Chain
a
b
c
Vhold −65 mV. Vhold −90 mV. Dried DMSO solubility, ref 34. ND, not determined.
display functional selectivity over NaV1.5 by 100-fold or more
(Table 3−6). This is a significant improvement of NaV1.5
selectivity compared to clinically used compounds. Two well
characterized and potent NaV1.7 blockers, tetracaine and
amitriptyline, were used for comparison.35 They displayed only
2−5-fold difference in potency between NaV1.7 and 1.5 (Table
6871
dx.doi.org/10.1021/jm300623u | J. Med. Chem. 2012, 55, 6866−6880

Journal of Medicinal Chemistry
Article
Table 4. SAR of Heterocyclic Ring Variations
a
b
c
Vhold −65 mV. Vhold −90 mV. Dried DMSO solubility, ref 34. ND, not determined.
7). The functional selectivity is to the largest extent achieved
through increased state-dependent properties, e.g., compounds
43, 52B, and 55 show more than 100-fold difference between the
resting-state (Vhold −90 mV) and the steady-state (Vhold −65
mV) protocols, reflecting preferential affinity for the inactivated
versus resting state of the NaV1.7 channel. This is largely in
agreement with what has been reported for other recently
described NaV1.7 blockers.12 Although state-dependent proper-
ties have also been observed with clinically used sodium channel
blockers such as local anesthetics, anticonvulsants, and
6872
dx.doi.org/10.1021/jm300623u | J. Med. Chem. 2012, 55, 6866−6880

Journal of Medicinal Chemistry
Article
Table 5. The Effect on SAR around Core Modifications
a
b
b
c
NaV1.7 IC50
NaV1.7 IC50
NaV1.5 IC50
hERG IC50
(μM)
solubility
(μM)
RLM/HLM Clint
(μL/min/mg)
entry
R3
R4
R5
(μM)
(μM)
(μM)
16
H
H
H
0.16
0.88
3.6
0.41
ND
ND
21
1.2
>31
27
7.8
ND
ND
11
13
<1
20
20
17
1
24/292
57/>500
ND/ND
ND/ND
10/150
59/190
37/97
52
Me
Me
Me
H
H
H
52A
52B
53
H
H
H
H
0.17
0.087
0.051
0.16
0.40
0.29
3.1
25
F
H
0.28
0.18
22
0.81
4.7
16
5.6
54
H
Me
OMe
OH
H
H
5.6
55
H
H
3.6
4
56
H
H
>33
1.7
>33
17
ND
8.2
<1
26
6
23/26
57
H
F
16/170
<10/50
10/59
58
H
H
OMe
OH
ND
ND
8.5
0.94
ND
ND
59
H
H
0.064
15
a
b
c
Vhold −65 mV. Vhold −90 mV Dried DMSO solubility, ref 34. ND, not determined.
Table 6. SAR of Analogues of 51
a
b
b
c
NaV1.7 IC50
NaV1.7 IC50
NaV1.5 IC50
hERG IC50
(μM)
solubility
(μM)
RLM/HLM Clint
(μL/min/mg)
entry
R5
R6
R7
(μM)
(μM)
(μM)
51
60
61
62
63
H
OCF3
H
0.23
0.13
0.14
0.22
0.14
ND
0.78
ND
0.99
ND
19
>16
17
5.5
24
25
61
27
14
2
<10/53
28/180
110/220
<10/<10
36/42
H
H
F
OCF3
OCF3
CF3
H
ND
>33
ND
H
F
>25
>10
OH
OCF3
H
a
b
c
Vhold −65 mV. Vhold −90 mV. Dried DMSO solubility, ref 34. ND, not determined.
the properties of 16A, 26B, 28B, 60, and 62 as well as tetracaine
and amitriptyline. Interestingly, the most NaV1.5 selective
compounds (28B, 60, and 62) showed no inhibition of NaV1.2
at 33 μM. For comparison, tetracaine and amitriptyline inhibited
NaV1.2 with IC50 of 1.1 and 3.1 μM, respectively. In depth
electrophysiological studies are required to understand the
mechanism of action of the described compounds as well as the
absolute selectivities versus the different sodium channels.
In vitro selectivities for compound 16A and 26B were
determined in functional or radioligand bindings assays. In
functional ion channel assays, using either IonWorks or FLIPR,
compound 26B was inactive at 10 μM on the potassium channels
KV1.5, KCNQ2, at 33 μM on hERG, hIKs, and at 31.6 μM on the
calcium channel CaV3.2. The more NaV1.7 potent compound,
16A, blocked the potassium channels: hERG, KV1.5, and
KCNQ2 with IC50s of 4.6, 1.5, and 1.6 μM, respectively, but
was inactive on IKs at 33 μM. In addition, this compound
inhibited at 31.6 μM the calcium channel CaV3.2 by 44%. Both
compounds were inactive on the ligand-gated channels TRPA1
(at 11 μM) and glycine receptor α1 (at 33 μM). As determined
by radioligand binding, no activity was detected at 10 μM for a
Table 7. In Vitro Profile of 16A, 26B, 28B, 60, 62,
Amitriptyline, and Tetracaine
a
b
a
b
a
NaV1.7
IC50
(μM)
NaV1.7
IC50
(μM)
NaV1.7
NaV1.5
IC50
(μM)
NaV1.2
IC50
(F1737A)
IC50 (μM)
entry
16A
(μM)
0.053
0.41
0.092
0.13
0.23
1.0
0.083
3.8
10
>33
>10
ND
33
0.73
>30
>31
>16
>25
1.6
8.1
>33
>33
>33
>33
3.1
26B
28B
1.0
60
0.78
0.99
7.0
62
amitriptyline
tetracaine
12
0.11
0.62
33
0.53
1.1
a
b
Vhold −65 mV. Vhold −90 mV. ND, not determined.
antiarrhythmics, the absolute differences in the potency between
compounds such as tetracaine and amitriptyline are less
pronounced than that for the described oxoisoindoline
carboxamides (Table 1, 3−7).
Selected compounds were also evaluated for their activity on
one of the major brain sodium channels, NaV1.2. Table 7 shows
6873
dx.doi.org/10.1021/jm300623u | J. Med. Chem. 2012, 55, 6866−6880

Journal of Medicinal Chemistry
Article
panel of 20 targets at MDS pharma.36 At 10 μM, compounds 16A
and 26B inhibited the binding of [3H]batrachotoxin (sodium
channel, site 2) to rat brain membranes by 100% and 79%,
respectively, indicating overlapping binding to the local
anesthetic binding site.
of central sensitization.40 It measures spontaneous rather than
evoked pain responses. Clinical conditions triggered by
activation of NaV1.7 are erythromelalgia and paroxysmal extreme
pain disorder. They are considered to be spontaneous pain
disorders and are likely to be peripherally driven.2 In the formalin
model, the first phase (0−5 min) is more peripherally driven than
the second phase (15−35 min). In carrageenan-induced
monoarthritis, a unilateral inflammation is induced and change
in gait pattern is recorded. The reason for using a small and
unilateral joint is that it may reflect the inflammatory process
seen in patients with rheumatoid arthritis, a disease that usually
starts in small joints.41,42
The rat NaV1.7 potency, using a steady-state protocol (Vhold
−65 mV), for compounds 16A and 26B, was determined to be
0.074 and 1.6 μM, respectively. As seen in Figure 3, compound
16A showed dose-dependent efficacy in the formalin model, with
statistically significant analgesic effects in both phase I and II. In
phase I, there was a significant effect at 140 μmol/kg and in phase
II from 30 μmol/kg with dose-relevant trends even at lower
doses. In the carrageenan model (Figure 5), there was a
significant reduction in the pain behavior by 42% at 100 μmol/
kg. Figure 4 shows that compound 26B displayed a dose- and
concentration-dependent inhibition of painful behavior that
reached statistical significance at the highest dose tested (100
μmol/kg) in phase I of the formalin model. The same dose of
26B resulted in a 42% reduction in pain behavior in the
carrageenan model, although this did not reach statistical
significance (p value = 0.068; Figure 6). For both compounds,
no visual side effects have been observed under the time the tests
were performed. Free plasma concentrations of 100 and 400 nM
were achieved at the end of the formalin experiment for
compounds 16A and 26B, respectively, which is 0.7−4-fold the
measured in vitro IC50 values for the rat NaV1.7 channel. The
brain:plasma ratio was 2−5 for compound 16A and 1−2 for
compound 26B, indicating that both compounds have
unrestricted access to the brain. The efficacies seen with
compounds 16A, 26B, and 28B in the different rat ex vivo and
in vivo models could, in addition to inhibition of NaV1.7, be due
to blockade of, e.g., NaV1.3 and/or NaV1.8. Further studies are
required to fully understand the mechanism behind the effects of
the described oxoisoindoline carboxamides.
The binding epitope for local anesthetics is mostly formed by
residues in the sixth transmembrane segments mainly of the third
and fourth domains of the sodium channel peptide sequence.37
We further sought to understand the site of binding by site-
directed mutagenesis of phenylalanine 1737 to alanine in the
NaV1.7 channel followed by functional analysis. The potency of
compounds 16A, 28B, and 62 for the mutated channel was
reduced by ∼100-fold (Table 7). For comparison, the potency of
amitriptyline was reduced by ∼12-fold and tetracaine by >300-
fold, which is in agreement with previous studies.38,39 These
results demonstrate that the oxoisoindoline carboxamides most
likely bind to the local anesthetic site of the NaV1.7 channel.
Prior to efficacy studies in rat models of pain, the potencies of
selected compounds were determined on recombinantly ex-
pressed rat NaV1.7 channels. For the majority of compounds, less
than 3-fold difference between rat and human NaV1.7 was
observed, which is not a surprise because the channels are
evolutionary well conserved. The ability of compound 28B to
inhibit SNL-induced ectopic activity in a rat DRG ex vivo
preparation was investigated. The in vitro potency on the rat
NaV1.7 channel as measured by whole cell electrophysiology
using a steady-state protocol (Vhold −65 mV) was determined to
be 74 nM, which was selected as the lowest test concentration.
Compound 28B caused a progressive inhibition of the ectopic
firing and reached more than 50% inhibition in 6 out of 11
analyzed fibers. The responding fibers displayed a baseline firing
frequency of 9.38 1.96 Hz (mean SEM), which was used for
calculation of the dose-dependent efficacy. At 222 and 740 nM,
compound 28B caused 45.8 13.4% and 82.3 9.6% inhibition
(mean
SEM, n = 6) of the SNL-induced ectopic firing,
respectively (Figure 2). Thus, the estimated EC50 for ex vivo
CONCLUSION
■
In conclusion, a high-throughput screening campaign using a
functional ion-flux assay identified compound 10 as a non-
selective, state-dependent blocker of NaV1.7. An extensive
structure−activity relationship evaluation with focus on
optimization of NaV1.7 potency, selectivity over NaV1.5, and
metabolic stability properties produced several interesting
oxoisoindoline carboxamides that were further characterized. A
potent and selective compound 28B showed concentration-
dependent inhibition of nerve injury-induced ectopic in an ex
vivo DRG preparation. Two compounds, 16A and 26B, with
good PK properties in rat demonstrated dose and exposure
dependent efficacy in preclinical pain models. Compound 26B
showed selectivity over several ion channels and GPCRs. Binding
of the oxoisoindoline carboxamides to the local anesthetics site of
the NaV1.7 channel was further demonstrated. Despite this,
several compounds showed functional selectivity over NaV1.5 by
more than 100-fold. The selectivity is likely achieved in
combination of the state-dependent properties of the com-
pounds with interactions to nonconserved amino acid residues
outside of the local anesthetics site. The oxoisoindoline
Figure 2. Concentration-dependent effects of compound 28B on
ectopic firing from injured DRG neurons. Buffer or compound 28B at
74, 222, and 740 nM were perfused cumulatively on DRGs excised from
SNL rats at 15 min for each concentration. The effects during the last 3
min were analyzed and expressed as % inhibition (mean SEM, n = 6).
inhibition of SNL-induced ectopic is ∼3-fold for the in vitro
NaV1.7 potency of the compound. Unfortunately, due to low
solubility of 28B, further in vivo analysis of this compound was
stopped.
On the basis of the physicochemical properties and on the in
vitro potency on the NaV1.7 channel, both compounds 16A and
26B were selected to assess their in vivo efficacy in the rat
formalin model and carrageenan monoarthritis model. The
formalin model is a peripherally driven model with components
6874
dx.doi.org/10.1021/jm300623u | J. Med. Chem. 2012, 55, 6866−6880

Journal of Medicinal Chemistry
Article
Figure 5. Dose response effect of 16A in the carrageenan monoarthritis
model. The compound was administered per oral gavage 2 h 45 min post
carrageen injection in the ankle joint of the rats. Behavioral
measurement was performed, 15 min post compound administration,
on a computerized set up developed internally called Paw Print. The
effect is express as the percentage in comparison to vehicle treated group
(n = 9 per group) and free plasma concentration is expressed as mean
SEM (n = 9 per group). One-way ANOVA was used as statistic in
comparison to vehicle group per phase, followed by followed by
unpaired one-tailed t test.*p < 0.05.
Figure 3. Dose response effect of 16A in the rat formalin model.
Compound is administered per oral gavage 15 min before formalin (100
μL of 2% solution) injection into the dorsal paw of Sprague−Dawley
rats. Dose response effect is represented for both phases of the formalin
test (phase I, 0−5 min; phase II, 15−35 min post formalin injection).
The effect is expressed as the percentage in comparison to vehicle
treated group (n = 9 per group) and free plasma concentration is
expressed as mean SEM (n = 9 per group). Pairwise comparisons are
used as t tests (Bonferroni) within a one-Way ANOVA model per phase.
No adjustment is performed for multiple comparisons. *p < 0.05, **p >
0.01, ***p > 0.001.
Figure 6. Dose response effect of 26B in the carrageenan monoarthritis
model. The compound was administered per oral gavage 2 h 45 min post
carrageen injection in the ankle joint of the rats. Behavioral
measurement was performed, 15 min post compound administration,
on a computerized set up developed internally called Paw Print. The
effect is express as the percentage in comparison to vehicle treated group
(n = 9 per group) and free plasma concentration is expressed as mean
SEM (n = 9 per group). One-way ANOVA was used as statistic in
comparison to vehicle group per phase, followed by followed by
unpaired one-tailed t test. p = 0.068.
Figure 4. Dose response effect of 26B in the rat formalin model.
Compound is administered per oral gavage 15 min before formalin (100
μL of 2% solution) injection into the dorsal paw of Sprague−Dawley
rats. Dose response effect is represented for both phases of the formalin
test (phase I, 0−5 min; phase II, 15−35 min post formalin injection).
The effect is expressed as the percentage in comparison to vehicle
treated group (n = 9 per group), and free plasma concentration is
expressed as mean SEM (n = 9 per group). Pairwise comparisons are
used as t tests (Bonferroni) within a one-way ANOVA model per phase.
No adjustment is performed for multiple comparisons. *p < 0.05.
been described before.43 The cDNA encoding the rat NaV1.7 α-subunit
(Genbank U79568) inserted into the pIRES-Neo2 vector was used to
transfect HEK293 cells using the calcium phosphate method. Cells were
selected with 400 μg/mL of G418, and single colonies were picked and
characterized by patch-clamp electrophysiology. The cells were kindly
donated by Mohamed Chahine (Department of Medicine and Laval
Hospital, Laval University, QC, Canada). The human NaV1.7 (F1737A)
mutated cell line (clone 4) was generated by Lipofectamine 2000-
mediated transfection of HEK293 cells with a point mutated expression
construct briefly generated as follows. Two 152-bp long oligonucleo-
tides (5′- GGGATGGATTGCTAGCACCTATTCTTAACAG-
TAAGCCACCCGACTGTGACCCAAAAAAAGTTCATCCTG-
carboxamides series described here, including compound 26B,
may be valuable for further investigations for pain therapeutics.
EXPERIMENTAL SECTION
■
Cell Lines and Culture Conditions. The recombinant HEK293
cell line expressing the human NaV1.7 α-subunit and the Chinese
hamster ovary (CHO) K1 cell line expressing the human NaV1.2 α-
subunit were obtained from Millipore. The human NaV1.5 α-subunit
and hERG expressing Chinese hamster ovary (CHO) K1 cell lines have
6875
dx.doi.org/10.1021/jm300623u | J. Med. Chem. 2012, 55, 6866−6880

Journal of Medicinal Chemistry
Article
GAAGTTCAGTTGAAGGAGACTGTGGTAACCCATC) and (3′-
TGACTGCAATGTACATGTTCACCACAACCAGGGCGGATAT-
GATGATATAACTAACAAAGTAGAATATTCCAACAGATGGGT-
TACCACAGTCTCCTTCAACTGA) containing the Phe to Ala codon
were annealed and ligated to a Nhe1 and BsrGI digested SCN9A
(database entry NM_002977) in pcDNA3.0. The mutated SCN9A
cDNA was then transferred to pcDNA4.0/TO using the Kpn1and Xho1
restriction sites. The SCN9A inserts of all plasmid constructs were
verified by sequencing of both strands. Cells were selected with Zeocin
at 100 μg/mL, and single colonies were picked and characterized by
Western Blotting and patch-clamp electrophysiology.
The NaV1.7 cells were cultured in DMEM/F12+ Glutamax (Gibco)
and the NaV1.2 cells were cultured in IMDM + L-glutamate (Invitrogen),
supplemented with 10% foetal bovine serum (Hyclone), 1%
nonessential amino acids (Invitrogen), and 400 μg/mL Geneticin
(NaV1.7, and NaV1.2 cells) or 100 μM Zeocine (NaV1.7 [F1737A] cells)
at 37 °C with 5% CO2. Prior to electrophysiological recordings, the cells
were incubated at 30 °C during the last 15−24 h, washed with Mg and
Ca-free PBS, detached with trypsin, centrifuged and resuspended in D−
PBS, and finally added to the IonWorks cell boat.
Electrophysiology. Sodium currents were recorded with the
perforated whole-cell configuration of the patch-clamp from using the
IonWorks Quattro planar patch clamp automated electrophysiology
system (Molecular Devices, Inc.).44
The external recording solution (D-PBS, Gibco) contained (in mM)
NaCl (138), KCl (2.7), KH2PO4 (1.5), Na2HPO4 (8), CaCl2·2H2O
(0.9), MgCl2·6H2O (0.5), and glucose (5.5). The internal recording
solution contained (in mM) K-gluconate (100), KCl (40), MgCl2 (3),
EGTA (3), and HEPES (5). pH was adjusted to 7.25 with KOH, and the
final osmolarity was set at 290 mOsm. Amphotericin B was used as the
perforating agent and added to the internal solution to give a final
concentration of 150 μg/mL.
The compounds were analyzed using two different voltage protocols.
The cells were either clamped to −90 mV, where the majority of
channels are in a resting state, or they were clamped to −65 mV, which is
referred to as the steady state. Comparing the two IonWorks protocols,
as well as with manual patch clamp, we estimated the fraction of
inactivated NaV1.7 channels in the steady-state assay to be 40−50%. The
sodium currents were evoked by a single voltage train consisting of eight
30 ms depolarizing steps to −20 mV (for NaV1.7) and 10 mV (for
NaV1.2 and 1.5) from the holding potential at a frequency of 3 Hz. To
recover a fraction of the channels from slow inactivation with the steady-
state (Vhold −65 mV) protocol, a 20 ms hyperpolarizing step to −100
mV was executed prior to the activation step. The protocol was activated
twice, prescan and postscan compound addition and the degree of
inhibition for each well was assessed by dividing the two values for the
eight pulse.
All compounds were dissolved in 100% dimethylsulfoxide (DMSO)
to a final stock concentration of 10 mM. Half log serial dilutions were
prepared in DMSO, where the final DMSO concentration in the
external recording solution was 0.3%. The NaV1.5 assay has been
described before where a good correlation to manual electrophysiology
was demonstrated for several known sodium channel blockers.31 The
hERG assay using IonWorks HT was performed as previously
described.45
Some IC50 values may be uncertain for compounds with low
solubility, although the buffer conditions used for the functional assays
and the solubility assay are not the same. All compounds were initially
analyzed with an n = 2 at two different test occasions. The compounds
selected for further profiling were analyzed with n = 4−22. Tetracaine
was used as a reference compound present at every test occasion and
showed mean IC50 and SEM values of 0.11 0.002 μM for Nav1.7 (Vhold
−65 mV), 0.62 0.017 μM for Nav1.7 (Vhold −90 mV), 0.53 0.027
μM for NaV1.5 (Vhold −90 mV), and 1.1 0.12 μM for Nav1.2 (Vhold
−65 mV).
Ex Vivo Electrophysiology. Three to four days after the SNL
operation, animals were anesthetized with isoflurane. The L4 and L5
dorsal root ganglia (DRG) with attached dorsal roots and spinal nerve
stumps were excised. The DRG were mounted in an in vitro recording
chamber with one compartment for the DRG and the spinal nerve and
an adjoining compartment for the dorsal root. The volume of DRG
compartment is about 1 mL. The DRG and spinal nerve compartment
was perfused with oxygenated (95% O2−5% CO2) artificial
cerebrospinal fluid (ACSF) (in mM: 126 NaCl, 3 KCl, 2 MgSO4, 26
NaHCO3, 1.25 NaH2PO4, 10 glucose, 2 CaCl2, pH 7.4) at a rate of
1.25−1.5 mL/min. The dorsal root compartment was filled with paraffin
oil. The temperature was kept at 35 0.5 °C through a temperature
controlled water bath.
Single fiber recordings were performed, and ectopic discharges were
recorded from teased dorsal root fascicles. Action potentials were
amplified via a differential amplifier (DAM80, World Precision
Instrument, USA) with a gain of ×1000 and filtered, with lower pass
10 kHz and high pass 10 Hz. Signals were digitized at 20 kHz by a data
acquisition system (Power 1401, Cambridge Electronic Design, UK)
and fed into a personal computer (Spike2, 5.01 software (Cambridge
Electronic Design, UK), Microsoft Windows 2000 operating system), to
allow sampling and offline analysis of the neuronal activity. Neuronal
activity from single spike was monitored online by spike sorting function
of data acquisition system and was further converted into a histogram.
Once stable baseline of ectopic discharge was established and
recorded for at least 5 min, ACSF followed by three concentrations of a
single compound were applied to the DRG compartment for 15 min
each via a pump at a rate of 1.25−1.5 mL/min. Compound 28B was
diluted with ACSF from a 10 mM solution in DMSO prior to each
experiment to final concentrations of 74, 222, and 740 nM.
The effect of the compound on the ectopic firing was analyzed
through calculation of the mean frequency. For each fiber, 5 min
baseline ectopic firing was averaged and used as a control. The ectopic
activities during application of buffer (ACSF) and drugs were averaged
every 3 min and then normalized to the percentage change of mean
control frequency. The magnitude and time profile of inhibition were
calculated. For a particular fiber, the ectopic firing was considered to be
inhibited if the reduction of its firing frequency at the end of perfusion,
i.e., the last 3 min, exceeded 50% of the control frequency. Fibers that
fulfill this criterion were included for further analysis. Data are expressed
as mean standard error of mean (SEM).
In Vivo Pharmacokinetics. Male Sprague−Dawley rats (Scanbur
B&K AB, Sollentuna, Sweden) were housed with up to five animals per
cage and were allowed to acclimatize to the new environment for at least
one week upon arrival. In the conditioned animal facility, room
temperature was kept at 20 2 °C, relative humidity at 60 20%, and a
12 h light−dark cycle was maintained including a 0.5 h dusk or dawn
period (lights on at 6.30 a.m.). Water and standard rodent diet were
freely accessible. All animal handling and experiments within
AstraZeneca were performed in full compliance with authorial and
ethical guidelines.
Rats (∼300 g) received 3 μmol/kg of the test compound
intravenously as a bolus injection (N = 3) or 10 μmol/kg orally via
gavage (N = 3). The formulations consisted of 1.5−5% DMA and 98.5−
95% (HPbCD and 23.5 mg/mL glycerin in water). The content of
HPbCD was adjusted dependent on the physicochemical properties of
the test compounds. Serial blood samples (400 μL) from the tail vein
were collected in microtainer tubes containing EDTA at 0.02, 0.08, 0.33,
0.67, 1, 3, 6, and 24 h following iv administration, or at 0.25, 0.5, 0.75, 1,
1.5, 2.5, 6, and 24 h following po administration. The blood samples
were centrifuged, and plasma was transferred to a new tube. Sample
preparation, bioanalysis with LC-MS/MS, and data analysis has been
described previously.46,47 Plasma protein binding in rat was determined
using equilibrium dialysis.48
In Vivo Pharmacology. The rat formalin test was performed on
male Sprague−Dawley rats from Scanbur B&K Universal, Sollentuna,
Sweden (∼220 g). In brief, the rats were habituated to the experimental
environment for 30 min prior to testing. The animals were randomly
assigned to different experimental groups (n = 9 per group), and
compounds 16A or 26B were administered per oral gavage 15 min
before formalin injection. The formalin test is performed by injecting
100 μL of 2.0% formalin (in saline) subcutaneously on the dorsal side of
the left hind paw. The behavior of the animals was recorded for 35 min
after formalin injection, and time spent licking the injected paw was
analyzed to represent nociceptive behavior. The response to formalin
6876
dx.doi.org/10.1021/jm300623u | J. Med. Chem. 2012, 55, 6866−6880

Journal of Medicinal Chemistry
Article
injection was biphasic: phase I (0−5 min after formalin injection) and
phase II (15−35 min after formalin injection).
Compound 16A was dissolved in 10% hydropropylbetacyclodextrin
in Milli-Q water at doses of 30, 100, and 140 μmol/kg, whereas
compound 26B was dissolved in 0.5% hydropropylmethylcellulose in
0.1% Tween 80 in water at doses 10, 30, and 100 μmol/kg.
positive mode electrospray ionization (ESI+) in full scan and leucine
encephalin (Sigma) was used as the lock mass (m/z 556.2771).
Chromatographic separation was achieved with a 2.3 min gradient from
5 to 95% ACN (0.1% formic acid) over an Acquity UPLC BEH C18 1.7
μm, 2.1 mm × 50 mm column (Waters) maintained at 50 °C and run at a
flow rate of 0.4 mL/min. NMR spectra were recorded on a Varian 400
MHz or on a Bruker 400 MHz and Bruker 500 MHz NMR
spectrometer. The following reference signals were used: TMS δ 0.00,
or the residual solvent signal of DMSO-d6 δ 2.49, CD3OD δ 3.31, or
CDCl3 δ 7.25 (unless otherwise indicated). Resonance multiplicities are
denoted s, d, t, q, m, and br for singlet, doublet, triplet, quartet, multiplet,
and broad, respectively. Unless otherwise stated, chemical shifts are
given in ppm with the solvent as internal standard. Column
chromatography was performed using Merck Silica gel 60 (0.040−
0.063 mm), or employing a Combi Flash Companion system using
RediSep normal-phase flash columns.
Preparation of Isoindoline Carboxamides via Ugi Reaction,
General Procedure A. To the methanol (1 mL) solution of 2-
formylbenzoic acid (1) (0.3 mmol, 1 equiv), the corresponding amine 2
(0.3 mmol, 1 equiv) in methanol (1 mL) was added followed by
isocyanide 3 (0.3 mmol, 1 equiv). The mixture was stirred overnight at
ambient temperature before it was filtered and purified using preparative
liquid chromatography. The fractions containing the product were
pooled, and the acetonitrile was removed in vacuum. The water solution
was extracted with ethyl acetate. The organic layer was dried over
magnesium sulfate and concentrated to yield the title compound as
solid.
The carrageenan-induced monoarthritis test was performed on male
Sprague−Dawley rat (Harlan Laboratories BV, Netherlands; ∼220 g, n
= 9 per group) as previously described.49 In brief, under isoflurane
anesthesia, 50 μL of carrageenan solution (7.5 mg/mL in saline, λ-
carrageenan, Sigma Aldrich, Stockholm, Sweden) was injected into the
ankle joint of the rats, inducing swelling and hyperalgesia. Two hours 45
minutes after the injection, compound 16A or 26B at doses of 0, 30, and
100 μmol/kg were administered per oral gavage. Behavioral measure-
ment was performed, 15 min post compound administration, on a
computerized set up developed internally called Paw Print. The
observation of changes in gait is automatically measured and removes
any operator bias. In brief, the set up consists of a pathway where rats
were trained to go through. The glass floor of the pathway associated to
light allowed detection of a light-print for each rat paw hitting the floor.
The light intensity and the area of the print were dependent on the
pressure applied by the rat paw. Each print on the glass floor was
recorded via a camera, and gait pattern and weight bearing were
extracted from the prints and analyzed by a computer-based system.
Nociception as a result of the induced hyperalgesia was defined by a
guarding index and represent the difference in weight bearing between
the hind paws (in per mill (‰) of the total weight bearing on all limbs).
A guarding index of 0 indicates that there is an equal amount of weight
bearing on the both hind paws, and an increase in the guarding index
indicates a shift of the weight bearing of the affected paw to the
unaffected paw. At the end of the experiment, blood samples were
collected by heart punction under isoflurane anesthesia prior to sacrifice.
Determination of Metabolic Stability. The microsome Clint
values in rat and human were determined from the disappearance of
the parent compound as described previously.50 In short, 1 μM of the
compound was incubated with 0.5 mg/mL of microsomes for 5, 15, 30,
and 60 min in the presence of 1.5 mM NADPH. The reactions were
stopped by the addition of 100 μL of acetonitrile and analyzed on LC-
MS/MS. Control incubations (without microsomes) were performed
for all compounds.
Chemistry: General. All materials were obtained from commercial
supplier, unless otherwise noted, and used without further purification.
Mass spectra were recorded either on Waters Alliance 2795 LC-MS
(ZQ) with ESI, APCI, and APPI in both positive and negative ionization
mode or Agilent GC-MS with chemical ionization (CI, methane as
reactant gas). For separation, a capillary column was used, DB-5MS
(J&W Scientific). A linear temperature gradient was applied. Preparative
chromatography was run on a Waters FractionLynx system with a MS-
triggered fraction collection (ESI). Column; XBridge Prep C8 5 μm
OBD 19 mm × 100 mm, with guard column; XTerra Prep MS C8 10 μm
19 mm × 10 mm cartridge. A gradient from 100% A (95% 0.1 M
ammonium acetate in Milli-Q water and 5% acetonitrile) to 100% B
(100% acetonitrile) was applied for LC separation at a flow rate 25 mL/
min. Purity for final compounds was greater than 95% unless otherwise
noted and measured on one of the following instruments: (A) An
Agilent HP1100 high performance liquid chromatography (HPLC)
system with UV detections at 220, 254, and 290 nm. Phenomenex
Gemini C18 3.0 mm × 50 mm, 3 μm, flow rate of 1.0 mL/min. A linear
gradient was used starting at 100% A (A: 10 mM ammonium acetate in
5% acetonitrile) and ending at 100% B (B: acetonitrile) after 3.5 min.
(B) A Waters Acquity UPLC system with UV detection at 254 nm was
equipped with Waters ZQ mass spectrometer. Waters Acquity UPLC
BEH C8 2.1 mm × 50 mm, 1.7 μm, flow rate 1.0 mL/min. A linear
gradient ranging from 5 to 95% CH3CN in H2O with 0.01 M
ammonium acetate was used. The gradient was completed in 2 min 15 s.
The ZQ mass spectrometer was run with ESI in positive and negative
modes. High resolution mass spectra (HRMS) were recorded on a
Waters Synapt-G2 mass spectrometer (Waters MS Technologies,
Manchester, UK) connected to an Acquity UPLC system with a PDA
detector (Waters Corp., Milford, USA). All analyses were acquired using
1-(Isocyanomethyl)-3-(trifluoromethoxy)benzene (3a). Step 1:
The DCM solution (4 mL) of the commercially available 3-
(trifluoromethoxy)phenyl)methanamine 6a (0.2 g, 1.05 mmol) was
set under N2 atmosphere and cooled down to 0 °C. Thereafter, phenyl
formate (0.117 mL, 1.05 mmol) was added dropwise via syringe and the
mixture was stirred at room temperature for 16 h. The solvent was
removed in vacuo, and the residue was purified on silica column using
heptane:ethyl acetate = 100:0 to 0:100 as gradient, affording N-(3-
(trifluoromethoxy)benzyl)formamide as colorless oil, 120 mg (52%).
1H NMR (500 MHz, CDCl3) δ 8.32 (s, 1 H), 7.38 (dd, J = 8.84, 7.71 Hz,
1 H), 7.25 (d, J = 7.71 Hz, 1 H), 7.19−7.11 (m, 2 H), 5.90 (br s, 1 H),
4.53 (d, J = 6.06 Hz, 2 H). MS (ESI) m/z 220 [M + H].
Step 2: N-(3-(Trifluoromethoxy)benzyl)formamide (0.120 g, 0.55
mmol) was dissolved in dichloromethane (4 mL) and cooled to −15 °C
under N2 atmosphere. N,N-Diisopropylethylamine (0.362 mL, 2.19
mmol) followed by phosphorus oxychloride (0.061 mL, 0.66 mmol)
were added, and the resulting mixture was allowed to slowly reach room
temperature. Then methanol (1.5 mL) was added to quench the
reaction. The mixture was diluted with dichloromethane and washed
twice with satd NaHCO3 solution. The combined organic extracts were
washed with brine, dried over anhydrous sodium sulfate, filtered, and
evaporated to give the product as a brown oil, 110 mg (100%) which was
used without further purification: 1H NMR (500 MHz, CDCl3) δ 7.49−
7.43 (m, 1 H), 7.31 (d, J = 7.57 Hz, 1 H), 7.25−7.21 (m, 2 H), 4.69 (s, 2
H). MS (ESI) m/z 202 [M + H].
1-Fluoro-4-(isocyanomethyl)-2-(trifluoromethyl)benzene (3e).
Step 1: N-(4-Fluoro-3-(trifluoromethyl)benzyl)formamide was pre-
pared following the procedure described for intermediate 3a in step 1,
using (4-fluoro-3-(trifluoromethyl)phenyl)methanamine (446 mg, 2.31
mmol), triethylamine (0.644 mL, 4.62 mmol), and phenyl formate
(0.302 mL, 2.77 mmol). The mixture was stirred under argon at room
temperature for four days. The reaction mixture was purified by silica gel
chromatography using EtOAc in heptane 0−90% as gradient, yielding
1
the product as white solid, 338 mg (66.2%). H NMR (400 MHz,
CDCl3) δ 8.32 (s, 1 H), 7.59−7.45 (m, 2 H), 7.23−7.14 (m, 1 H), 5.93
(br s, 1 H), 4.52 (d, J = 6.06 Hz, 2 H). GC-MS (CI) m/z 221 [M+].
Step 2: Compound 3e was prepared according to the method
described for 3a in step 2 from N-(4-fluoro-3-(trifluoromethyl)benzyl)-
formamide (338 mg, 1.53 mmol), N,N-diisopropylethylamine (1.010
mL, 6.11 mmol), and phosphorus oxychloride (0.171 mL, 1.83 mmol).
Afterward, the reaction mixture was concentrated, the residue was
dissolved in methanol (3 mL), and triethylamine (2.5 mL) was added.
The reaction mixture was stirred for 5 min and concentrated under
6877
dx.doi.org/10.1021/jm300623u | J. Med. Chem. 2012, 55, 6866−6880

Journal of Medicinal Chemistry
Article
reduced pressure. The crude compound was purified by silica gel
7.59 (td, J = 7.52, 1.26 Hz, 1 H), 7.48 (t, J = 7.39 Hz, 1 H), 7.19−7.11 (m,
2 H), 7.11−7.03 (m, 2 H), 5.12 (s, 1 H), 4.42 (dd, J = 5.75, 1.71 Hz, 2
H), 4.18−4.03 (m, 1 H), 4.03−3.84 (m, 1 H), 3.37−3.19 (m, 2 H). MS
(ESI) m/z 475 [M + H]. HRMS calcd for C23H18F4N4O3 (M + H)+
475.1393, found 475.1394.
column chromatography using hexane:ethyl acetate = 100:0 to 80:20.
1
The product was obtained as brown liquid, 293 mg (94%). H NMR
(400 MHz, CDCl3) δ 7.65−7.54 (m, 2 H), 7.28 (t, J = 9.22 Hz, 1 H),
4.69 (s, 2 H). GC-MS (CI) m/z 203 [M+].
2 - ( 2 - ( 5 - F l u o r o p y r i m i d i n - 2 - y l ) e t h y l ) - 3 - o x o - N - ( 3 -
(trifluoromethoxy)benzyl)isoindoline-1-carboxamide (60). The title
compound was synthesized according to the general procedure A from
2-formylbenzoic acid (112 mg, 0.74 mmol), 2-(5-fluoropyrimidin-2-
yl)ethanamine (105 mg, 0.74 mmol), and 1-isocyano-3-
(trifluoromethoxy)benzene 3a (139 mg, 0.74 mmol). The reaction
mixture was stirred at 40 °C overnight. The mixture was purified by silica
gel chromatography using EtOAc in heptane 0−100% as gradient. The
product was obtained as semisolid, 187 mg (53%). 1H NMR (400 MHz,
CD3OD) δ 8.64−8.50 (m, 2 H), 7.73 (dt, J = 7.20, 1.20 Hz, 1 H), 7.65−
7.57 (m, 1 H), 7.57−7.50 (m, 2 H), 7.43−7.35 (m, 1 H), 7.26 (ddd, J =
7.83, 1.33, 1.20 Hz, 1 H), 7.22−7.09 (m, 2 H), 5.28 (s, 1 H), 4.50−4.38
(m, 3 H), 3.69 (dt, J = 14.27, 6.38 Hz, 1 H), 3.37−3.33 (m, 1 H), 3.30−
3.23 (m, 1 H). MS (ESI) m/z 475 [M + H]. HRMS calcd for
C23H18F4N4O3 (M + H)+ 475.1393, found 475.1402.
N-(4-Fluoro-3-(trifluoromethyl)benzyl)-2-(2-(5-fluoropyrimidin-2-
yl)ethyl)-3-oxoisoindoline-1-carboxamide (62). The title compound
was prepared according to the modified procedure A, described for 28,
using 2-(5-fluoropyrimidin-2-yl)ethanamine hydrochloride (126 mg,
0.71 mmol), triethylamine (0.123 mL, 0.89 mmol), 2-formylbenzoic
acid (89 mg, 0.59 mmol), and 1-fluoro-4-(isocyanomethyl)-2-
(trifluoromethyl)benzene 3e (0.120 mL, 0.59 mmol). The mixture
was stirred at 45 °C for 48 h. White solid, 41 mg, (14.6%). 1H NMR (500
MHz, CDCl3) δ 8.42 (s, 2 H), 8.21−8.13 (m, 1 H), 7.67 (d, J = 8.35 Hz,
1 H), 7.62−7.55 (m, 2 H), 7.50−7.41 (m, 1 H), 7.36−7.28 (m, 2 H),
7.10−7.01 (m, 1 H), 5.11 (s, 1 H), 4.51 (dd, J = 14.98, 6.31 Hz, 1 H),
4.37 (dd, J = 14.98, 5.67 Hz, 1 H), 4.14−4.05 (m, 1 H), 4.05−3.95 (m, 1
H), 3.32−3.22 (m, 2 H). MS (APP−) m/z 475 [M − H]. HRMS calcd
for C23H17F5N4O2 (M + H)+ 477.1350, found 477.1344.
N-(3,4-Dimethylphenyl)-3-oxo-2-(2-(pyridin-2-yl)ethyl)-
isoindoline-1-carboxamide (10). The title compound was synthesized
according to the general procedure A from 2-formylbenzoic acid (75 mg,
0.50 mmol), 2-(pyridin-2-yl)ethanamine (61.1 mg, 0.50 mmol), and 4-
(isocyano)-1,2-dimethylbenzene (65.6 mg, 0.5 mmol). White solid, 101
mg (52%). 1H NMR (400 MHz, DMSO-d6) δ 10.57 (s, 3 H), 8.47−8.42
(m, 1 H), 7.73−7.66 (m, 2 H), 7.65−7.57 (m, 2 H), 7.55−7.48 (m, 1 H),
7.39 (d, J = 2.02 Hz, 1 H), 7.34 (dd, J = 8.08, 2.27 Hz, 1 H), 7.30 (d, J =
7.58 Hz, 1 H), 7.20 (ddd, J = 7.45, 4.93, 1.26 Hz, 1 H), 7.08 (d, J = 8.08
Hz, 1 H), 5.38 (s, 1 H), 4.29−4.19 (m, 1 H), 3.51−3.42 (m, 1 H), 3.12
(d, J = 7.33 Hz, 1 H), 3.09−3.00 (m, 1 H), 2.19 (s, 3 H), 2.17 (s, 3 H).
MS (ESI) m/z 386 [M + H]. HRMS calcd for C24H23N3O2 (M + H)+
386.1869, found 386.1862.
3-Oxo-2-(2-pyridin-2-ylethyl)-N-[[4-(trifluoromethoxy)phenyl]-
methyl]-1H-isoindole-1-carboxamide (16). The title compound was
synthesized according to the general procedure A from 2-formylbenzoic
acid (1 g, 6.6 mmol), 2-(2-aminoethyl)pyridine (814 mg, 6.6 mmol),
and 1-(isocyanomethyl)-4-trifluoromethoxybenzene (1.34 g, 6.6
mmol). White solid, 1.83 g (60%). 1H NMR (400 MHz, DMSO-d6) δ
9.20 (t, J = 6.00 Hz, 1 H), 8.48−8.42 (m, 1 H), 7.72−7.64 (m, 2 H),
7.63−7.57 (m, 1 H), 7.57−7.48 (m, 2 H), 7.41−7.34 (m, 2 H), 7.34−
7.28 (m, 2 H), 7.26 (d, J = 7.83 Hz, 1 H), 7.24−7.18 (m, 1 H), 5.24 (s, 1
H), 4.36 (d, J = 5.81 Hz, 2 H), 4.26−4.15 (m, 1 H), 3.46−3.36 (m, 1 H),
3.14−3.04 (m, 1 H), 3.04−2.94 (m, 1 H). MS (ESI) m/z 456 [M + H].
HRMS calcd for C24H20F3N3O3 (M + H)+ 456.1535, found 456.1537.
2-(Oxan-4-yl)-3-oxo-N-[[4-(trifluoromethoxy)phenyl]methyl]-1H-
isoindole-1-carboxamide (26). The title compound was synthesized
according to the general procedure A from 2-formylbenzoic acid (45 mg,
0.3 mmol), tetrahydro-2H-pyran-4-amine (30 μL, 0.30 mmol), and 1-
(isocyanomethyl)-4-(trifluoromethoxy)benzene (57 mg, 0.29 mmol).
White solid, 78 mg (60%). 1H NMR (500 MHz, DMSO-d6) δ 9.16 (t, J =
5.83 Hz, 1 H), 7.69 (d, J = 7.25 Hz, 1 H), 7.62−7.57 (m, 1 H), 7.53 (d, J
= 7.57 Hz, 1 H), 7.51−7.47 (m, 1 H), 7.41−7.36 (m, 2 H), 7.35−7.30
(m, 2 H), 5.32 (s, 1 H), 4.34 (d, J = 5.67 Hz, 2 H), 4.16−4.07 (m, 1 H),
3.89 (dd, J = 11.35, 4.10 Hz, 1 H), 3.81 (d, J = 10.72 Hz, 1 H), 3.40−3.30
(m, 2 H), 1.90−1.79 (m, 1 H), 1.73−1.64 (m, 3 H). MS (ESI) m/z 435
[M + H]. HRMS calcd for C22H21F3N2O4 (M + H)+ 435.1532, found
435.1530.
2-(4,4-Difluorocyclohexyl)-3-oxo-N-[4-(trifluoromethoxy)benzyl]-
isoindoline-1-carboxamide (28). The title compound was prepared
according to the general procedure A with minor modification as
follows. 4,4-Difluorocyclohexanaminium chloride (0.429 g, 2.50 mmol)
was dissolved in methanol (10 mL), and 2-formylbenzoic acid (0.375 g,
2.5 mmol) was added. The mixture was set under argon atmosphere and
then the pH was adjusted to ∼6 by addition of triethylamine (1.045 mL,
7.50 mmol). After stirring for 5 min, 1-(isocyanomethyl)-4-
(trifluoromethoxy)benzene (0.478 mL, 2.38 mmol) was added and
the resulting mixture was stirred at ambient temperature for 16 h. The
crude mixture was purified by silica gel chromathography using
heptane:ethyl acetate = 100:0 to 50:50 as gradient. White solid, 430
mg (36.7%). 1H NMR (500 MHz, CDCl3) δ 7.71−7.64 (m, 1 H), 7.64−
7.55 (m, 2 H), 7.52−7.41 (m, 1 H), 7.18−7.01 (m, 4 H), 6.32 (t, J = 6.31
Hz, 1 H), 5.11 (s, 1 H), 4.45 (dd, J = 14.74, 6.38 Hz, 1 H), 4.31 (dd, J =
14.82, 5.83 Hz, 1 H), 4.24−4.09 (m, 1 H), 2.25−2.02 (m, 2 H), 1.97−
1.75 (m, 6 H). MS (ESI) m/z 469 [M + H]. HRMS calcd for
C23H21F5N2O3 (M + H)+ 469.1551, found 469.1549.
2 - ( 2 - ( 5 - F l u o r o p y r i m i d i n - 2 - y l ) e t h y l ) - 3 - o x o - N - ( 4 -
(trifluoromethoxy)benzyl)isoindoline-1-carboxamide (51). The title
compound was prepared according to the modified procedure A,
described for 28, using 2-(5-fluoropyrimidin-2-yl)ethanamine hydro-
chloride (59.2 mg, 0.33 mmol), triethylamine (0.093 mL, 0.67 mmol), 2-
formylbenzoic acid (50 mg, 0.33 mmol), and 1-(isocyanomethyl)-4-
(trifluoromethoxy)benzene (0.067 mL, 0.33 mmol). The mixture was
stirred at 45 °C for 16 h. Yellow semisolid, 18 mg (11%). 1H NMR (400
MHz, CDCl3) δ 8.37 (s, 2 H), 8.17−8.00 (m, 1 H), 7.75−7.63 (m, 2 H),
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental procedures for 10−63 and their precursors.
Additional characterization data of 16, 26, 28, 51, 60, and 62.
This material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*Phone: +46 8 553-24322. E-mail: istvan.macsari@astrazeneca.
com.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank colleagues at the Physical Chemistry Characterization
Team for providing solubility and Fanny Bjarnemark for HRMS
data. The Safety Screening Centre for providing NaV1.5 and
hERG data.
ABBREVIATIONS USED
■
NaV1.7, voltage-gated sodium channel type 7; NaV1.5, voltage-
gated sodium channel type 5; NaV1.2, voltage-gated sodium
channel type 2; SCN9A, sodium channel, voltage-gated, type IX,
R subunit; SAR, structure−activity relationship; DIPEA, N,N-
diisopropylethylamine; THF, tetrahydrofuran; DCM, dichloro-
methane; DMF, N,N-dimethylformamide; NBS, N-bromo-
succinimide; AIBN, 2,2′-azobis(2-methylpropionitrile; hERG,
human ether a-go-go-related gene; LLE, lipophilic ligand
efficiency; PK, pharmacokinetic; Clint, in vitro intrinsic clearance;
RLM, rat liver microsome; HLM, human liver microsome; AUC,
6878
dx.doi.org/10.1021/jm300623u | J. Med. Chem. 2012, 55, 6866−6880

Journal of Medicinal Chemistry
Article
state-dependent inhibitor of NaV1.7 with oral efficacy in the formalin
model of persistent pain. J. Med. Chem. 2011, 54, 4427−4445.
(13) Chowdhury, S.; Chafeev, M.; Liu, S.; Sun, J.; Raina, V.; Chui, R.;
Young, W.; Kwan, R.; Fu, J.; Cadieux, J. A. Discovery of XEN907, a
spirooxindole blocker of NaV1.7 for the treatment of pain. Bioorg. Med.
Chem. Lett. 2011, 21, 3676−3681.
(14) Macsari, I.; Sandberg, L.; Besidski, Y.; Gravenfors, Y.; Ginman, T.;
Bylund, J.; Bueters, T.; Eriksson, A. B.; Lund, P.-E.; Venyike, E.;
Arvidsson, P. I. Phenyl isoxazole voltage-gated sodium channel blockers:
structure and activity relationship. Bioorg. Med. Chem. Lett. 2011, 21,
3871−3876.
(15) Nguyen, H. N.; Bregman, H.; Buchanan, J. L.; Du, B.; Feric, E.;
Huang, L.; Li, X.; Ligutti, J.; Liu, D.; Malmberg, A. B.; Matson, D. J.;
McDermott, J. S.; Patel, V. F.; Wilenkin, B.; Zou, A.; McDonough, S. I.
Discovery and optimization of aminopyrimidinones as potent and state-
dependent NaV1.7 antagonists. Bioorg. Med. Chem. Lett. 2012, 22, 1055−
1060.
(16) Bregman, H.; Nguyen, H. N.; Feric, E.; Ligutti, J.; Liu, D.;
McDermott, J. S.; Wilenkin, B.; Zou, A.; Huang, L.; Li, X.; McDonough,
S. I.; DiMauro, E. F. The discovery of aminopyrazines as novel, potent
NaV1.7 antagonists: hit-to-lead identification and SAR. Bioorg. Med.
Chem. Lett. 2012, 22, 2033−2042.
(17) Chakka, N.; Bregman, H.; Du, B.; Nguyen, H. N.; Buchanan, J. L.;
Feric, E.; Ligutti, J.; Liu, D.; McDermott, J. S.; Zou, A.; McDonough, S.
I.; DiMauro, E. F. Discovery and hit-to-lead optimization of
pyrrolopyrimidines as potent, state-dependent NaV1.7 antagonists.
Bioorg. Med. Chem. Lett. 2012, 22, 2052−2062.
(18) Gonzalez, J. E.; Termin, A. P.; Wilson, D. M. Small Molecule
Blockers of Voltage-Gated Sodium Channels. Methods and Principles in
Medicinal Chemistry. In Voltage-Gated Ion Channels as Drug Targets;
Triggle, D. J., Gopalakrishnan, M., Rampe, D., Zheng, W., Mannhold, R.,
Kubinyi, H., Folkers, G., Eds.; Wiley-VCH: Weinheim, Germany, 2006;
Vol. 29, pp 168−192.
(19) Priest, B. T. Future potential and status of selective sodium
channel blockers for the treatment of pain. Curr. Opin. Drug Discovery
Dev. 2009, 12, 682−692.
(20) Arvidsson, P. I.; Besidski, Y.; Csjernyik, G; Lange, T.; Macsari, I.;
Nilsson, L. I. Isoindoline derivatives comprising an additional
heterocyclic group and their use in the treatment of pain disorders.
WO2009145718, 2009.
(21) Åhlin, K.; Arvidsson, P. I.; Besidski, Y.; Nilsson, L. I. Isoindoline
derivatives comprising a cyano group and their use in the treatment of
pain disorders. WO2009145719, 2009.
(22) Arvidsson, P. I.; Besidski, Y.; Csjernyik, G.; Sandberg, L.
Isoindoline derivatives comprising additional heterocyclic groups and
their use in the treatment of pain disorders. WO2009145720, 2009.
(23) Besidski, Y.; Claesson, A.; Csjernyik, G.; Macsari, I.; Nilsson, L. I.
Isoindoline derivatives comprising phenyl groups and their use in the
treatment of pain disorders. WO2009145721, 2009.
(24) Kato, S.; Nonoyama, N.; Tomimoto, K.; Mase, T. Noncryogenic
metalation of aryl bromides bearing proton donating groups: formation
of a stable magnesio-intermediate. Tetrahedron Lett. 2002, 43, 7315−
7317.
area under curve; Vss, volume of distribution; t1/2, terminal half-
life; CL, systemic clearance; Cmax, maximal concentration; Tmax
,
time at which maximal concentration is measured; F, oral
bioavailability; fu, unbound fraction; DRG, dorsal root ganglion;
hCav3.2, T type voltage-dependent calcium channel, α 1H
subunit; hIKs, intermediate-conductance calcium-activated
potassium channel, known as KCa3.1; hKV1.5, voltage-gated
potassium channel type 5; hKCNQ2, potassium voltage-gated
channel, KQT-like subfamily, member 2; hTRPA1, transient
receptor potential cation channel, subfamily A, member 1;
GPCR, G-protein-coupled receptor
REFERENCES
■
(1) Catterall, W. A; Goldin, A. L.; Waxman, S. G. International Union
of Pharmacology. XLVII. Nomenclature and structure− function
relationships of voltage-gated sodium channels. Pharmacol. Rev. 2005,
57, 397−409.
(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. A stop codon mutation in SNC9A causes lack of pain
sensation. Hum. Mol. Genet. 2007, 16, 2114−2121.
(3) 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. An SCN9A channelopathy
causes congenital inability to experience pain. Nature 2006, 444, 894−
898.
(4) Goldberg, Y. P.; MacFarlane, J.; MacDonald, M. L.; Thompson, J.;
Dube, M. P.; Mattice, M.; Fraser, R.; Young, C.; Hossain, S.; Pape, T.;
Payne, B.; Radomski, C.; Donaldson, G.; Ives, E.; Cox, J.;
Younghusband, H. B.; Green, R.; Duff, A.; Boltshauser, E.; Grinspan,
G. A.; Dimon, J. H.; Sibley, B. G.; Andria, G.; Toscano, E.; Kerdraon, J.;
Bowsher, D.; Pimstone, S. N.; Samuels, M. E.; Sherrington, R.; Hayden,
M. R. Loss-of-function mutations in the NaV1.7 gene underlie congenital
indifference to pain in multiple human populations. Clin. Genet. 2007,
71, 311−319.
(5) Staud, R.; Price, D. D.; Janicke, D.; Andrade, E.; Hadjipanayis, A.
G.; Eaton, W. T.; Kaplan, L.; Wallace, M. R. Two novel mutations of
SCN9A (NaV1.7) are associated partial congenital insensivity to pain.
Eur. J. Pain 2011, 15, 223−230.
(6) 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. Mutations in
SCN9A, encoding a sodium channel alpha subunit, in patients with
primary erythermalgia. J. Med. Genet. 2004, 41, 171−174.
(7) Drenth, J. P.; te Morsche, R. H.; Guillet, G.; Taieb, A.; Kirby, R. L.;
Jansen, J. B. SCN9A mutations define primary erythermalgia as a
neuropathic disorder of voltage gated sodium channels. J. Invest.
Dermatol. 2005, 124, 1333−1338.
(8) Fertleman, C. R.; Baker, M. D.; Parker, K. A.; Moffatt, S.; Elmslie, F.
V.; Abrahamsen, B.; Ostman, J.; Klugbauer, N.; Wood, J. N.; Gardiner,
R. M.; Rees, M. SCN9A mutations in paroxysmal extreme pain disorder:
allelic variants underlie distinct channel defects and phenotypes. Neuron
2006, 52, 767−774.
(25) Nguyen, T.-H.; Chau, N. T. T.; Castanet, A.-S.; Nguyen, K. P. P.;
Mortier., J. First General, Direct, and Regioselective Synthesis of
Substituted Methoxybenzoic Acids by Ortho Metalation. J. Org. Chem.
2007, 72, 3419−3429.
(9) Rogers, M.; Tang, L.; Madge, D. J.; Stevens, E. B. The role of
sodium channels in neuropathic pain. Semin. Cell Dev. Biol. 2007, 17,
571−581.
(10) Zuliani, V.; Rivara, M.; Fantini, M.; Constantino, G. Sodium
channel blockers for neuropathic pain. Expert Opin. Ther. Pat. 2010, 20,
755−779.
(11) Kemp, M. I. Structural trends among second-generation voltage-
gated sodium channel blockers. Prog. Med. Chem. 2010, 49, 81−111.
(12) Bregman, H.; Berry, L.; Buchanan, J. L.; Chen, A.; Du, B.; Feric,
E.; Hierl, M.; Huang, L.; Immke, D.; Janosky, B.; Johnson, D.; Li, X.;
Ligutti, J.; Liu, D.; Malmberg, A.; Matson, D.; McDermott, J.; Miu, P.;
Nguyen, H. N.; Patel, V. F.; Waldon, D.; Wilenkin, B.; Zheng, X. M.;
Zou, A.; McDonough, S. I.; DiMauro, E. F. Identification of a potent,
(26) Othman, M.; Decroix, B. Synthesis of phthalimidine-3-
carboxylate and benzopyrroloindolizine from N-(pyrrol-2-yl)-
phthalimidine-3-carboxylate. Synth. Commun. 1996, 26, 2803−2809.
(27) Trivedi, S.; Dekermendjian, K.; Julien, R.; Huang, J.; Lund, P.-E.;
Krupp, J.; Kronqvist, R.; Larsson, O.; Bostwick, R. Cellular HTS assays
for pharmacological characterization of NaV1.7 modulators. Assay Drug
Dev. Technol. 2008, 6, 167−179.
(28) Xu, Z-Q. D.; Zhang, X.; Grillner, S.; Hokfelt, T. Electro-
̈
physiological studies on rat dorsal root ganglion neurons after peripheral
axotomy: Changes in responses to neuropeptides. Proc. Natl. Acad. Sci.
U.S.A. 1997, 94, 13262−13266.
6879
dx.doi.org/10.1021/jm300623u | J. Med. Chem. 2012, 55, 6866−6880

Journal of Medicinal Chemistry
Article
(29) Liu, C.-N.; Michaelis, M.; Amir, R.; Devor, M. Spinal nerve injury
enhances subthreshold membrane potential oscillations in DRG
neurons: relation to neuropathic pain. J. Neurophysiol. 2000, 84, 205−
215.
(30) Harmer, A. R.; Valentin, J.-P.; Pollard, C. E. On the relationship
between block of the cardiac Na+ channel and drug-induced
prolongation of the QRS complex. Br. J. Pharmacol. 2011, 164, 260−
273.
(46) Bueters, T.; Dahlstrom, J.; Kvalvagnaes, K.; Betner, I.; Briem, S.
̈
̊
High-throughput analysis of standardized pharmacokinetic studies in
the rat using sample pooling and UPLC-MS/MS. J. Pharm. Biomed. Anal.
2011, 55, 1120−1126.
(47) Briem, S.; Martinsson, S.; Bueters, T.; Skoglund, E. Combined
approach for high-throughput preparation of plasma samples from
exposure studies. Rapid Commun. Mass Spectrom. 2007, 21, 1965−1972.
́
(48) Borgegard, T.; Jureus, A.; Olsson, F.; Rosqvist, S.; Sabirsh, A.;
Rotticci, D.; Paulsen, K.; Klintenberg, R.; Yan, H.; Waldman, M.;
Stromberg, K.; Nord, J.; Johansson, J.; Regner, A.; Parpal, S.;
Malinowsky, D.; Radesater, A. C.; Li, T.; Singh, R.; Eriksson, H.;
Lundkvist, J. First and second generation γ-secretase modulators
(GSMs) modulate amyloid-β (Aβ) peptide production through
different mechanisms. Alzheimer’s Dis. Res. J. 2012, 3, 29−45.
(49) Finn, A.; Oerther, S. Can L(+)-lactate be used as a marker of
experimentally induced inflammation in rats? Inflammation Res. 2010,
59, 315−321.
(50) Sohlenius-Sternbeck, A. K.; Afzelius, L.; Prusis, P.; Neelissen, J.;
Hoogstraate, J.; Johansson, J.; Floby, E.; Bengtsson, A.; Gissberg, O.;
Sternbeck, J.; Petersson, C. Evaluation of the human prediction of
clearance from hepatocyte and microsome intrinsic clearance for 52
drug compounds. Xenobiotica 2010, 40, 637−649.
(31) Harmer, A. R; Abi-Gerges, N.; Easter, A.; Woods, A.; Lawrence, C.
L.; Small, B. G.; Valentin, J.-P.; Pollard, C. E. Optimisation and
validation of a medium-throughput electrophysiology-based hNaV1.5
assay using IonWorks. J. Pharmacol. Toxicol. Methods 2008, 57, 30−41.
(32) Goldin, A. L. Resurgence of sodium channel research. Annu. Rev.
Physiol. 2001, 63, 871−894.
(33) Tfelt-Hansen, J.; Winkel, B. G.; Grunnet, M.; Jespersen, T.
Inherited cardiac diseases caused by mutations in the NaV1.5 sodium
channel. J. Cardiovasc. Electrophysiol. 2010, 21, 107−115.
(34) Alelyunas, Y. W.; Liu, R.; Pelosi-Kilby, L.; Shen, C. Application of
a dried-DMSO rapid throughput 24 h equilibrium solubility in
advancing discovery candidates. Eur. J. Pharm. Sci. 2009, 37, 172−182
The method was modified by a pre-incubation step for 4 h at a fixed
temperature of 1°C with magnetic stirrers at 500 rpm in order to
generate the most stable solid form..
(35) Dick, I. E.; Brochu, R. M.; Purohit, Y.; Kaczorowski, G. J.; Martin,
W. J.; Priest, B. T. Sodium channel blockade may contribute to the
analgesic efficacy of antidepressants. J. Pain 2007, 8, 315−324.
(36) Assays performed at MDS Pharma Services for compound 16A
and 26B with no inhibition at 10 μM: cholinesterase ACES,
cyclooxygenase COX-1, monoamine oxidase MAO-A, endothelial nitric
oxide synthase eNOS, phosphodiesterase PDE4, insulin receptor
protein tyrosine kinase, adrenergicα1D, adrenergicα2A, adrenergicβ1,
calcium channel L-type benzothiazepine, dihydropyridine sites,
dopamine receptorD2L, estrogen ERα, GABA-A flunitrazepam, hista-
mine receptor H1, muscarinic receptor M2, nicotinic acetylcholine,
opiate receptor μ, serotonin receptor 5-HT2A, and norepinephrine
transporter NET.
(37) Nau, C.; Wang, G. K. Interactions of local anesthetics with
voltage-gated Na+ channels. J. Membr. Biol. 2004, 201, 1−8.
(38) Li, H. L.; Galue, A.; Meadows, L.; Ragsdale, D. S. A molecular
basis for the different local anesthetic affinities of resting versus open and
inactivated states of the sodium channel. Mol. Pharmacol. 1999, 55,
134−141.
(39) Wang, G. K.; Russell, C.; Wang, S.-Y. State-dependent block of
voltage-gated Na+ channels by amitriptyline via the local anesthetic
receptor and its implication for neuropathic pain. Pain 2004, 110, 166−
174.
(40) Lebrun, P.; Manil, J.; Colin, F. Formalin-induced central
sensitisation in the rat: somatosensory evoked potential data. Neurosci.
Lett. 2000, 283, 113−116.
(41) Kortekangas, P.; Peltola, O.; Toivanen, A.; Aro, H. T. Synovial
fluid L-lactic acid in acute arthritis of the adult knee joint. Scand. J.
Rheumatol. 1995, 24, 98−101.
(42) Gerster, J. C.; Gobelet, C. Synovial fluid lactic acid in acute and
chronic pyrophosphate arthropathy and in osteoarthritis. Clin.
Rheumatol. 1988, 7, 197−199.
(43) Persson, F.; Carlsson, L.; Duker, G.; Jacobson, I. Blocking
characteristics of hERG, hNaV1.5, and hKvLQT1 hminK after
administration of the novel anti-arrhythmic compound AZD7009. J.
Cardiovasc. Electrophysiol. 2005, 16, 329−341.
(44) Schroeder, K.; Neagle, B.; Trezise, D. J.; Worley, J. Ionworks HT:
A New High-Throughput Electrophysiology Measurement Platform. J.
Biomol. Screening 2003, 8, 50−64.
(45) Bridgland-Taylor, M. H.; Hargreaves, A. C.; Easter, A.; Orme, A.;
Henthorn, D. C.; Ding, M.; Davis, A. M.; Small, B. G.; Heapy, C. G.; Abi-
Gerges, N.; Persson, F.; Jacobson, I.; Sullivan, M.; Albertson, N.
Optimisation and validation of a medium-throughput electrophysiol-
ogy-based hERG assay using IonWorks HT. J. Pharmacol. Toxicol.
Methods 2006, 54, 189−199.
6880
dx.doi.org/10.1021/jm300623u | J. Med. Chem. 2012, 55, 6866−6880