The discovery of aminopyrazines as novel, potent Nav1.7 antagonists: Hit-to-lead identification and SAR

Article abstract of DOI:10.1016/j.bmcl.2012.01.023

Herein the discovery of a novel class of aminoheterocyclic Na _v1.7 antagonists is reported. Hit compound 1 was potent but suffered from poor pharmacokinetics and selectivity. The compact structure of 1 offered a modular synthetic strategy towards a broad structure-activity relationship analysis. This analysis led to the identification of aminopyrazine 41, which had vastly improved hERG selectivity and pharmacokinetic properties.

Full text of DOI:10.1016/j.bmcl.2012.01.023

Bioorganic & Medicinal Chemistry Letters 22 (2012) 2033–2042
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
The discovery of aminopyrazines as novel, potent Nav1.7 antagonists:
Hit-to-lead identification and SAR
Howard Bregman a, , Hanh Nho Nguyen a, Elma Feric b, Joseph Ligutti d, Dong Liu d, Jeff S. McDermott b,
⇑
Ben Wilenkin b, Anruo Zou d, Liyue Huang c, Xingwen Li c, Stefan I. McDonough b, Erin F. DiMauro a
a Department of Chemistry Research and Discovery, Amgen Inc., 360 Binney St., Cambridge, MA 02142, USA
b Department of Neuroscience, Amgen Inc., 360 Binney St., Cambridge, MA 02142, USA
c Department of Pharmacokinetics and Drug Metabolism, 360 Binney St., Cambridge, MA 02142, USA
d Department of Neuroscience, Amgen Inc., One Amgen Center Drive, Thousand Oaks, CA 91320, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Herein the discovery of a novel class of aminoheterocyclic Nav1.7 antagonists is reported. Hit compound 1
was potent but suffered from poor pharmacokinetics and selectivity. The compact structure of 1 offered a
modular synthetic strategy towards a broad structure–activity relationship analysis. This analysis led to
the identification of aminopyrazine 41, which had vastly improved hERG selectivity and pharmacokinetic
properties.
Received 8 December 2011
Revised 6 January 2012
Accepted 9 January 2012
Available online 18 January 2012
Ó 2012 Elsevier Ltd. All rights reserved.
Keywords:
Nav1.7
Analgesia
Sodium channels
Aminopyrazine
The treatment of chronic pain represents a significant unmet
medical need in society, providing an impetus for effort toward
the identification of novel and efficacious medicines. Compelling
human genetic data has implicated Nav1.7 as a key mediator in
nociception, sparking broad efforts aimed at Nav1.7 antagonism
across the scientific community.1–3 One of the paramount chal-
lenges is the identification of chemical matter which possesses
selectivity over physiologically important ion channels such as hu-
man Ether-a-go-go Related Gene (hERG) and subtype selectivity
over similar channels within the Nav family, such as the cardiac
channel Nav1.5.
citable neurons. Accordingly, compound potency was assessed
using the PX system on hNav1.7 with an average 20% contribu-
tion from inactivated channels.
Hit molecule (1) was a moderately potent, state-dependent
hNav1.7 inhibitor (hNav1.7 manual IC50-partially inactivated/
closed = 0.3/2.9
lM) with low intrinsic clearance in human and
rat liver microsomes (HLM, RLM intrinsic clearance = 25, 38
lL/
min/mg) (Fig. 1). The structure of 1 was compact and amenable
to rapid SAR evaluation. Although 1 exhibited low intrinsic clear-
ance in liver microsomes, it suffered from poor solubility6 and
was rapidly metabolized in rat plasma resulting in extremely
high clearance.7 In addition, compound 1 lacked selectivity over
the cardiac channels hNav1.5 and hERG. We embarked on SAR
investigations with the goal to improve potency, selectivity and
pharmacokinetic properties. For this effort, we routinely em-
ployed PX for assessment of potency on hNav1.7 and hNav1.5,
dofetilide displacement for assessment of binding to hERG, and
in vitro metabolic stability assays in liver microsomes and rat
plasma to screen new molecules. The best new leads were fur-
ther examined with additional electrophysiology (manual8 on
hNav1.7 and hNav1.5, and PX on hERG) and dosed in vivo to ob-
tain rat PK data.5
A
high throughput screening campaign utilizing the Ion-
WorksÒ Quattro automated electrophysiology platform4,5 yielded
amino-triazine (1). We utilized the IonWorksÒ Quattro (IWQ)
and PatchXpressÒ (PX) planar patch-clamp automated electro-
physiology platforms for screening and hit-to-lead assessments,
respectively.5 These assays enabled receptor–ligand interactions
to be quantified directly, with a readout of functional inhibition,
and allowed compound potency to be determined against de-
fined gating states of sodium channels. The vast majority of so-
dium channel antagonists used clinically exert preferential
inhibition of channel inactivated states, a property thought to
confer functional selectivity for the sodium channels of hyperex-
The exceedingly high in vivo clearance (>10 L/h/kg) of 1 was
found to be mediated in large part by plasma instability; thus
our initial efforts were geared towards overcoming this liability.
To understand the source of instability, compound 1 was analyzed
⇑
Corresponding author.
E-mail address: hbregman@amgen.com (H. Bregman).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2012.01.023

2034
H. Bregman et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2033–2042
N
N
HN
O
N
NH2
Cl
2.2a
0.3 / [2.9]a
3.2a
hNav1.7 PX IC50 (μM)
hNav1.7 Manual IC50 (μM) 20% inactivated / closed
hNav1.5 PX IC50 (μM)
0.2 / [> 10]a
2.3 , 1.5
25 , 38
hNav1.5 Manual IC50 (μM) 20% inactivated / closed
hERG dofetilide binding Ki derived , PX IC50 (μM)
HLM , RLM intrinsic clearance (μL/min/mg)
Binding efficiency (kcal/NHA)
0.35b
7 , 1 , 6
4.6 , 86
Symyx solubility - SIF, PBS (pH 7.4), 0.01 N HCl (μg/mL)
clogP , PSA
aElectrophysiology IC50 of partially inactivated channels is shown.5 Where applicable, bracket indicates IC50 against fully noninactivated channels
determined with manual electrophysiology. Inhibitory activity represents an average of at least two determinations. bBinding efficiency = RT
[log(hNav1.7 PX IC50)]*(1x10-6)/(number of heavy atoms)
Figure 1. Profile of hit compound 1.
after various incubation times in rat plasma by mass spectrome-
try.7 After 3 h of incubation in fresh rat plasma near complete con-
pyrimidines, we decided to explore pyridine cores. Pyridines 7
and 8 were greater than tenfold less potent than pyrimidine 4. Pyr-
idine 9 imparted modest Nav1.5 selectivity (fourfold) but also was
version to
a single mono-N-oxidized product was observed.
Towards improving plasma stability, the primary aniline group
was methylated (2) or removed (3), leading to a substantial loss
of potency in both cases (Table 1).
Alternatively, we looked towards modification of the triazine
core as a means of modulating the electronics and nucleophilicity
of the amine functionalities in hopes of reducing in vivo clearance
while retaining potency (see Table 2). Pyrimidines 4 and 5 were
prepared, each with one nitrogen removed from the core relative
to 1 at a position proximal to the primary aniline. Compound 4
very active against hERG (IC50 = 0.16 lM) and 10 suffered from
high intrinsic clearance in liver microsomes.
Additional analogs were prepared in which the resulting core
had two nitrogen atoms present, as in pyridazines 11 and 12 and
pyrazine 13. Pyridazine derivatives (11 and 12) suffered from a
greater than tenfold reduction in potency relative to 4. Pyrazine
13 possessed retained potency relative to 4, and a fivefold selectiv-
ity over hERG, representing the first compound identified with any
observed hERG selectivity. We also investigated replacement of
two core carbon atoms with a sulfur atom. All three corresponding
regioisomeric thiadiazoles (14–16) suffered from tenfold reduced
potency relative to 4 and 13. From our exploration of triazine core
replacements we identified two new leads, pyrimidine 4, which
possessed improved potency (fivefold) and reduced rat IV clear-
ance (2.0 L/h/kg) relative to hit 1; and pyrazine 13 with improved
potency (fourfold), and hERG selectivity (fivefold) as compared to
1. Notably, both 4 and 13 were stable in rat plasma.
had improved potency relative to 1 (0.44 vs 2.2
tially reduced rat IV clearance (CL = 2.0 L/h/kg), and pyrimidine 5
had greatly reduced potency (IC50 >10 M). Pyrimidine 6 was also
lM) and substan-
l
prepared, which represents lead 1 with a core devoid of the most
distal nitrogen relative to the primary aniline. Although the po-
tency of 6 was retained relative to 1, it was sevenfold less potent
than pyrimidine 4. Somewhat encouraged by the data for
Initial follow-up SAR efforts on pyrimidine lead 4 were aimed
towards augmenting hERG selectivity while retaining or improving
potency and PK. We began with an evaluation of the linker region
with the 4-amino pyrimidine core and para-chloro phenyl aryl tail
group locked (Table 3). Replacement of the aniline NH with an oxy-
gen (17) or the bottom ether with an amine (18) led to a loss of
activity. Substitution of the bottom ether with a methyl amine
(19) was better tolerated but still led to reduced potency relative
to ether-linked lead 4. Introduction of a methyl group at 3-position
(20) or 2-position (21) yielded a >25-fold and fivefold decrease in
activity, respectively. With no improved linker replacements iden-
tified we performed subsequent SAR assessments with the unsub-
stituted anilino-ether moiety locked as in lead 4.
Table 1
Primary aniline substitution or removal
N
N
HN
O
N
R
Cl
a
Compound
R
hNav1.7 PX IC50 (lM)
1
2
3
NH2
NHCH3
H
2.2
>10
>10
An initial study of positional substituent preference identified
the para position as favorable. The chlorine substituent was moved
from the para position (4) to the meta (22) and ortho (23) positions
leading to a reduction in potency (fourfold and eightfold, respec-
tively). Therefore, further aryl group SAR was performed at the
a
Electrophysiology IC50 of partially inactivated channels is shown. Inhibitory
activity represents an average of at least two determinations.5

Table 2
SAR of core modifications of 1
core
HN
O
NH2
Cl
Compound
core
N
hNav1.7 PX IC50
2.2
(l
M)
hNav1.5 PX IC50
(lM)
hERG Ki-derived (
lM)
HLM CLint
25
(
ll/min/mg)
RLM CLint
38
(l
l/min/mg)
Rat plasma t1/2 (min)
Rat IV CL (L/(h kg))b
>10
a
a
N
1
3.2
2.3
<30
N
N
N
4
5
6
7
8
0.44
13.5
3.1
0.50
—
0.43
—
104
<14
206
18
91
32
>500
—
2.0
—
N
N
N
N
—
0.32
—
41
—
—
N
6.1
—
127
238
—
—
8.8
—
—
100
—
—
N
N
9
10
11
2.9
2.0
7.3
11.8
1.4
—
0.16
0.05
—
36
14
20
88
>399
>399
—
—
—
—
—
—
N
N
N
N
N
12
13
5.5
—
—
<14
183
>399
151
—
—
N
N
0.62
0.72
2.9
>500
8.2
S
N
S
14
15
16
5.7
5.3
—
—
—
—
—
—
<14
16
<14
37
—
—
—
—
—
—
N
N
S
N
N
N
11.3
<14
<14
a
Electrophysiology IC50 of partially inactivated channels is shown. Inhibitory activity represents an average of at least two determinations.5
Male Sprague-Dawley rats were dosed intravenously at 0.5 mg/kg in 0.5 mL/kg of 100% DMSO.
b

2036
H. Bregman et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2033–2042
Table 3
Linker region SAR
N
N
NH2
linker
Cl
a
a
Compound
Linker
hNav1.7 PX IC50
(
lM)
hNav1.5 PX IC50
(l
M)
hERG Ki-derived (
l
M)
HLM CLint
104
(ll/min/mg)
RLM CLint (ll/min/mg)
HN
4
0.44
0.50
0.43
91
O
O
17
18
19
20
9.8
5.9
—
—
30
<14
41
72
28
O
HN
—
—
NH
HN
1.8
3.7
0.8
>399
>399
167
N
HN
11.1
2.3
—
—
89
O
HN
21
2.27
—
58
O
a
Electrophysiology IC50 of partially inactivated channels is shown. Inhibitory activity represents an average of at least two determinations.5
para position (Table 4). Replacement of the para-chloro substituent
with fluorine (24) did not offer improvements in potency, selectiv-
ity, or intrinsic clearance. Methyl (25) and cyclopropyl (26) deriv-
atives yielded a modest reduction in potency (ꢀthreefold) with a
more profound drop observed for the larger phenyl derivative
(27). Substitution with both electron withdrawing and electron
donating polar substituents was studied. Electron withdrawing
substitution, such as cyano (28), methyl sulfone (29), carboxamido
(30), reduced Nav1.7 potency. Pyridine analog (31) suffered a loss
in potency. Bulky hydrophobic electron withdrawing groups such
as trifluoromethyl (33) and trifluoromethoxy (34) led to slight
reductions in potency (threefold and twofold, respectively), but
improved microsomal stability. The electron donating dimethyl-
amino (32) substitution was not tolerated. Our SAR evaluation of
the tail aryl group with the pyrimidine core provided no improve-
ment in potency or selectivity, but an understanding that retained
potency and improved microsomal stability can be realized with
the introduction of hydrophobic electron withdrawing groups at
the para position.
We then embarked on a follow up effort to pyrazine lead 13,
which had selectivity over hERG, with the goal of improving
upon its PK properties while retaining or improving potency
and selectivity. Using lessons learned from the pyrimidine series
work, we evaluated the effect of para substitution of the tail aryl
ring, concentrating on substitution which had offered modest
potencies and microsomal stabilities in the pyrimidine series
(Table 5). Removal of the para substituent provided phenyl ana-
log (35), which had retained potency and improved selectivity,
along with low intrinsic clearance in HLM but high in RLM.
Replacement of the chlorine with fluorine (36) provided reten-
tion or modest improvements in potency, selectivity and intrin-
sic clearance in HLM. para Substitution with a methyl group (37)
did not offer any improvements, whereas cyclopropyl derivative
(38) had improved hERG selectivity and microsomal stability but
two-fold reduced potency. Compound (39) substituted with a
small electron withdrawing cyano group suffered a drop in po-
tency. Bulky hydrophobic electron withdrawing groups were
evaluated as well. The trifluoromethyl derivative (40) yielded

H. Bregman et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2033–2042
2037
Table 4
Amino-pyrimidine tail aryl SAR
N
HN
N
NH2
OR
a
a
Compound
R
hNav1.7 PX IC50
(lM)
hNav1.5 PX IC50
(lM)
hERG Ki-derived (
lM)
HLM CLint
104
(ll/min/mg)
RLM CLint
91
(ll/min/mg)
4
0.44
0.50
0.43
Cl
22
23
3.2
1.8
1.7
—
44
93
250
Cl
0.36
1.4
>399
Cl
24
25
1.2
1.6
0.78
0.59
1.9
1.1
177
193
161
F
>399
26
1.5
2.0
0.5
19
61
27
28
29
>30
2.0
—
—
—
71
22
—
48
Ph
1.2
—
1.2
NC
>30
—
<14
MeO2S
30
31
32
>30
—
—
—
—
—
—
—
<14
75
—
<14
190
H2NOC
N
25.7
10.8
Me2N
F3C
33
34
1.3
0.9
<1.0
0.24
0.9
24
27
23
17
0.1
F3CO
a
Electrophysiology IC50 of partially inactivated channels is shown. Inhibitory activity represents an average of at least two determinations.5
retained potency and hERG selectivity (tenfold) with reduced
intrinsic clearance relative to the chloro lead (13). Even greater
overall improvements were realized with trifluoromethoxy deriv-
group (41), significant improvements in microsomal stability were
realized, which translated to moderate rat IV clearance, and im-
proved selectivity over hERG (35-fold) and Nav1.5 (fivefold (PX);
twofold (manual)). Beyond the encouraging properties of lead 41,
the pyrazine series overall represents a promising starting point
towards the development of selective and orally bioavailable
Nav1.7 antagonists. The general trend for selectivity over Nav1.5
and hERG is clear upon visualization of the analogs described
herein binned by core type (Fig. 3). The pyrazine analogs generally
offer a better selectivity profile than pyrimidine analogs or other
structures described. This selectivity enhancement, in concert with
favorable rat pharmacokinetics, mark this series as a favorable
starting point for future efforts.
ative (41) which was potent on Nav1.7 (0.24
lM), selective over
hERG and exhibited low intrinsic clearance in both HLM and
RLM (27, 36
SIF (109 vs 7
l
l
l/min/mg, respectively) and better solubility in
g/mL for 1) (Fig. 2).
Compound 41 was further assessed in rat PK studies. Gratify-
ingly, the good intrinsic stability in liver microsomes and rat plas-
ma translated to moderate rat IV clearance (0.93 L/h kg), and
reasonable oral exposure (AUCinf
(
l
M Ã h), Cmax
(lM) = 0.76,
0.22). Relative to hit compound 1, structure 41 had two structural
changes. Replacement of the triazine core with a pyrazine yielded
(13), which had improved rat clearance and selectivity (see Ta-
ble 2). By further replacing the para-Cl substituent with an OCF3
Chlorotriazine 43 was obtained via SNAr of aniline 42 with 2,4-
dichlorotriazine under basic conditions and was subsequently

2038
H. Bregman et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2033–2042
Table 5
Amino-pyrazine tail aryl SAR
N
N
NH2
HN
OR
a
a
Compound
R
Nav1.7 PX IC50
(
l
M) Nav1.5 PX IC50
(
l
M) hERG Ki-derived (
lM) or % in h @30
lM
HLM CLint
183
(l
l/min/mg) RLM CLint
(ll/min/mg)
13
0.62
0.72
2.9
151
Cl
35
36
37
0.5
1.8
25%
1.1
2.0
<14
71
172
222
129
0.44
0.88
0.47
0.41
F
>399
38
1.3
3.0
48%
33
70
39
40
3.9
—
—
142
40
216
99
NC
0.26
0.46
2.8
F3C
41
0.24
1.26
33%
27
36
F3CO
a
Electrophysiology IC50 of partially inactivated channels is shown. Inhibitory activity represents an average of at least two determinations.5
N
HN
O
N
NH2
F3CO
0.24
0.80 / [>3]a
1.3a
hNav1.7 PX IC50 (μM)
hNav1.7 Manual IC50 (μM) 20% inactivated / closed
hNav1.5 PX IC50 (μM)
1.5 / [10]a
77.3% inh. @ 30 μM , 8.5
27 , 36
hNav1.5 Manual IC50 (μM) 20% inactivated / closed
hERG dofetilide binding Ki derived , PX IC50 (μM)
HLM , RLM intrinsic clearance (μL/min/mg)
Rat IV CL (L/h-kg) , Vss (L/kg) , t1/2 (h)
Rat PO PK (%F , AUCinf (μM * hr) , Cmax (μM) , t1/2 (h))
Rat plasma protein binding (% bound)
CYP 3A4, 2D6 inh. IC50 (μM)
0.93 , 2.1 , 1.9b
13 , 0.76 , 0.22 , 2.47c
99.9
> 27
Binding efficiency (kcal/NHA)
0.35
109 , 1 , 7
5.6 , 82
Symyx solubility – SIF , PBS (pH 7.4) , 0.01 N HCl (μg/mL)
clogP , PSA
aElectrophysiology IC50 of partially inactivated channels is shown. Where applicable, bracket indicates IC50 against fully noninactivated channels
determined with manual electrophysiology. Inhibitory activity represents an average of at least two determinations. bMale Sprague-Dawley rats
were dosed intravenously at 0.5 mg/kg in 0.5 mL/kg of 100% DMSO. cMale Sprague-Dawley rats were dosed orally at 2.0 mpk in 10 mL/kg of
vehicle (30% hydroxypropyl-β-cyclodextrin, pH 2.2 adjusted with methanesulfonic acid).
Figure 2. Profile of lead 41.

H. Bregman et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2033–2042
2039
Figure 3. (A) Visualization of Nav1.5/1.7 selectivity (log scale) binned by structural class. (B) Visualization of hERG/Nav1.7 selectivity (log scale) binned by structural class.
NH2
N
N
N
N
HN
O
N
Cl
HN
O
N
R
N
N
(b)
(a)
2
: R = NHMe
N
Cl
Cl
3: R = H
O
Cl
42
43
Cl
Cl
Scheme 1. Synthesis of RHS modified derivatives 2–3. Reagents and conditions: (a) DIEA, DMF, 0 °C; (b) for 2: MeNH2, DMF, 0 °C to rt, 11% (2 steps); for 3: Pd/C, H2, MeOH,
21% (2 steps).

2040
H. Bregman et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2033–2042
NH2
O
X
Y
HN
O
Z
NH2
X
Y
4:
X, Z = N; Y = CH
(a)
5: X, Y = N; Z = CH
6: Y, Z = N; X = CH
Z
NH2
Cl
Cl
42
Cl
Scheme 2. Synthesis of core modified pyrimidine analogs 4–6. Reagents and conditions: (a) HCl, H2O, 100 °C, yield of 4, 5, 6: 39%, 17%, 19%.
X
Z
HN
O
Y
NH2
7:
X = CH; Y = CH; Z = N
(a) (for 7), or
(b) (for 8, 10)
X
Z
8: X = CH; Y = N; Z = CH
10: X = N; Y = CH; Z = CH
Y
NH2
Cl
Cl
N
N
OEt
HN
O
NH2
HN
O
N
O
N
N
H
OEt
(b)
(c)
(d)
NH2
N
H
O
Br
Cl
44
45
9
Cl
Cl
Scheme 3. Synthesis of core modified pyridine analogs 7–10. Reagents and conditions: (a) Compound 42, HCl, MeOH:H2O (1:1), 100 °C, 30%; (b) 42, Pd2(dba)3, BINAP, Cs2CO3,
DMA, 120 °C, 8, 10: 23%, 12%; (c) ethyl chloroformate, pyridine, DCM; (d) KOH, EtOH, 70 °C, 11% (3 steps).
N
NH2
O
N
NH2
HN
O
N
N
N
N
(a)
(b)
NH2
Cl
Cl
Cl
Cl
11
42
Cl
N
N
NH2
O
N
N
HN
O
Cl
HN
NH2
N
N
(c)
(b)
Cl
Cl
Cl
O
12
46
42
Cl
Cl
NH2
O
N
N
13: R = Cl - 27%
35: R = H - 4%
36: R = F - 7%
NH2
N
HN
O
(d)
37:
38:
39:
40:
R = CH3 - 17%
R = cyclopropyl - 5%
R = CN - 8%
N
NH2
Cl
R
R = CF3 - 5%
41: R = OCF3 - 13%
R
Scheme 4. Synthesis of core modified pyridazine analogs 11,12, pyrazine analogs 13, 35–41. Reagents and conditions: (a) NH3 (l), steel bomb, rt, 56%; (b) HCl, MeOH, 85 °C
11, 46: 3%, 35%; (c) NH3 (l), steel bomb, 160 °C, 9%; (d) Pd(OAc)2, BINAP, KOtBu, toluene, 120 °C, yields in scheme.

H. Bregman et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2033–2042
2041
X
Y
NH2
O
HN
NH2
Z
X
Y
(a) for 47, 48
(b) for 49
Cl
NH2
Z
O
14
15
47
48
X = S, Y = N, Z = N:
X = N, Y = S, Z = N:
X = S, Y = N, Z = N:
X = N, Y = S, Z = N:
Cl
Cl
42
X = N, Y = N, Z = S: 16
X = N, Y = N, Z = S: 49
Scheme 5. Synthesis of core modified thiadiazole analogs 14–16. Reagents and conditions: (a) HCl, MeOH, 100 °C, 14, 15: 25%, 26%; (b) Et3N, THF, 80 °C, 33%.
N
OH
O
O
N
NH2
N
(a)
Cl
N
NH2
O
17
50
Cl
Cl
N
NH2
X
R1
R2
HN
N
NH2
R1
R2
N
(b)
Cl
N
NH2
X
Cl
1
R ,R2 = H; X = NH
Cl
1
18:
19:
R ,R2 = H; X = NH
51:
52:
1
R ,R2 = H; X = NCH3
1
R ,R2 = H; X = NCH3
20: R1 = CH3; R2 = H; X = O
21: R1 = H; R2 = CH3; X = O
53: R1 = CH3; R2 = H; X = O
54: R1 = H; R2 = CH3; X = O
Scheme 6. Synthesis of linker modified pyrimidine analogs 17–21. Reagents and conditions: (a) NaH, DMF, 100 °C, 15%; (b) HCl, MeOH, 100 °C, 18, 19, 20, 21: 40%, 23%, 36%,
22%.
22:
R = m-Cl - 28%
23: R = o-Cl - 7%
24: R = p-F - 40%
N
25:
R = p-CH3 - 21%
NH2
O
HN
N
NH2
26: R = p-cyclopropyl - 40%
27: R = p-phenyl - 18%
N
(a)
28:
R = p-CN - 10%
Cl
N
NH2
29: R = p-SO2CH3 - 53%
30: R = p-CONH2 - 6%
O
R
31:
R
R = p-pyrido - 62%
32: R = p-NMe2 - 30%
33: R = p-CF3 - 26%
34:
R = p-OCF3 - 20%
Scheme 7. Synthesis of tail-modified aryl ether pyrimidine analogs 22–34. Reagents and conditions: (a) HCl, MeOH, 100–140 °C (thermal or microwave).
exposed to methyl amine in DMF providing 2 or reduced to afford 3
(Scheme 1).
Pyrimidines 4–6 were obtained via SNAr of the appropriate
chloro-amino pyrimidine precursor with aniline 42 under acidic
conditions9 (Scheme 2).
The 2-amino-4-anilino pyridine analog 7 was obtained via
SNAr with 42 under acidic conditions at elevated temperatures.
The 2-amino-6-anilino and 4-amino-2-anilino pyridines (8, 10)
were synthesized from their respective chloro-amino pyridine
precursors by Buchwald–Hartwig reactions. The 3-amino-5-
anilino pyridine analog (9) was obtained through a three-step
sequence: the protection of commercially available 5-bromo-3-
amino pyridine as an ethyl carbamate (44); subsequent Buch-
wald–Hartwig amination (45) and hydrolytic deprotection
yielded pyridine 910 (Scheme 3).
The preparation of pyridazine analogs 11, 12 proceeded via
two-step protocols. Amination of commercially available 3,5-
dichloropyridazine with ammonia at room temperature provided
5-chloropyridazin-3-amine. Subsequent SNAr with aniline 42 un-
der acidic conditions afforded 11. Reversal of this sequence, but
with amination of 4-chloropyridazine 46 at elevated temperatures
(160 °C) afforded 12. Pyrazine analogs 13, 35–41 were prepared by
Buchwald–Hartwig amination (Scheme 4).
The synthesis of 1,2,4-thiadiazole analogs 14, 15 was accom-
plished via SNAr reactions with aniline 42 under acidic conditions
via their respective amino-chloro-thiadiazole precursors (47, 48).
The 1,3,4-thiadiazole analog (16) was obtained utilizing basic SNAr
conditions with aniline 42 (Scheme 5).
The synthesis of bis-ether analog 17 was accomplished utilizing
an SNAr reaction of 2-chloro-6-aminopyrimidine with the anion of

2042
H. Bregman et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2033–2042
subunit was co-expressed) were voltage clamped at a membrane potential that
produced average 20% fractional inactivation. Voltage was set, and reset as
needed, to produce 20% inactivation, since state-dependent pharmacology is
determined not by the absolute voltage but by fractional channel inactivation,
and the voltage dependence of inactivation varies among cells and can vary
with time when recorded from a single cell. Currents were corrected for
rundown offline with DataXpress software, and dose–response curves were
built from an ensemble of single-concentration tests on many cells. For
experimental details of manual electrophysiology, see Bregman, H. et al. (Ref.
4.)
phenol 50. Pyrimidines 18–21 were obtained via SNAr under acidic
conditions with anilines 51–54 (Scheme 6).
Pyrimidine analogs with modification of the tail aryl (22–34)
were obtained with SNAr of 2-chloro-6-aminopyrimidine with
the appropriate aniline (Scheme 7).
Acknowledgement
6. Tan, H.; Semin, D.; Wacker, M.; Cheetham, J. JALA 2005, 10, 364.
7. The plasma concentration of 1 was below the detection limit after 15 min
following a single bolus intravenous administration at 0.5 mg/kg. Fresh rat
plasma was incubated with (1) (dissolved in DMSO, final concentration in
The authors would like to thank Ramin Vismeh for plasma sta-
bility studies.
plasma was 10
out and mixed with 400
centrifuged and supernatants were analyzed directly using LC/MS/MS. Analysis
m) coupled to a
Thermo Scientific LTQ-Orbitrap XL mass spectrometer (Bremen, Germany)
using a 50 min (5–95% acetonitrile with 0.1% formic acid) gradient in ESI
positive mode. Identification of hydroxylated metabolite was based on exact
mass measurement of protonated molecular ion (<1 ppm accuracy) and
comparing MS/MS fragments to those of parent compound. The typical limit
of detection is 0.5 ng/ml.
l
M) at 37 °C. 200
lL aliquots at 30 min, 1 and 2 h were pipetted
References and notes
lL acetonitrile to precipitate proteins. Samples were
1. Cummins, T. R.; Sheets, P. L.; Waxman, S. G. Pain 2007, 131, 243.
2. England, S.; Rawson, D. Future Med. Chem. 2010, 2, 775.
3. Zuliani, V.; Patel, M. K.; Fantini, M.; Rivara, M. Curr. Top. Med. Chem. 2009, 9,
396.
4. 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.
J. Med. Chem. 2011, 54, 4427.
was performed using a dC18 column (Atlantis, 1  15 mm  5
l
8. Manual electrophysiology verifies the PX value, controls for kinetic information
and used to test compound potency on non-inactivated (closed) channels.
9. Rahman, A. A.; Daoud, M. K.; Dukat, M.; Herrick-Davis, K.; Purohit, A.; Teitler,
M.; do Amaral, A. T.; Malvezzi, A.; Glennon, R. A. Bioorg. Med. Chem. Lett. 2003,
13, 1119.
10. Hasegawa, M.; Nishigaki, N.; Washio, Y.; Kano, K.; Harris, P. A.; Sato, H.; Mori,
I.; West, R. I.; Shibahara, M.; Toyoda, H.; Wang, L.; Nolte, R. T.; Veal, J. M.;
Cheung, M. J. Med. Chem. 2007, 50, 4453.
5. The electrophysiology protocols for IonWorks QuattroÒ (IWQ), PatchXpressÒ
(PX) and manual patch-clamp are described in this paper. IWQ protocol:
hNav1.7 or hNav1.5-expressing HEK 293 cells (no exogenous beta-subunit was
co-expressed) were patched in PPC mode (population patch clamp), and
currents were measured in response to a train of depolarizations that induced
successively greater inactivation. Dose–response curves were built from an
ensemble of wells that were exposed to different single concentrations. PX
protocol: hNav1.7 or hNav1.5-expressing HEK 293 cells (no exogenous beta-