Discovery of pyrrolo-benzo-1,4-diazines as potent Nav1.7 sodium channel blockers

Article abstract of DOI:10.1016/j.bmcl.2014.07.060

A series of pyrrolo-benzo-1,4-diazine analogs have been synthesized and displayed potent Na_v1.7 inhibitory activity and moderate selectivity over Na_v1.5. The syntheses, structure-activity relationships, and selected pharmacokinetic data of these analogs are described. Compound 41 displayed anti-nociceptive efficacy in the rat CFA pain model at 100 mpk oral dosing.

Full text of DOI:10.1016/j.bmcl.2014.07.060

Bioorganic & Medicinal Chemistry Letters 24 (2014) 4110–4113
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of pyrrolo-benzo-1,4-diazines as potent Nav1.7 sodium
channel blockers
Ginny D. Ho a, Deen Tulshian a, Ana Bercovici a, Zheng Tan a, Jennifer Hanisak a, Stephanie Brumfield a,
Julius Matasi a, Charles R. Heap b, William G. Earley b, Brandy Courneya b, R. Jason Herr b, Xiaoping Zhou c,
Terry Bridal d, Diane Rindgen e, Steve Sorota d, Shu-Wei Yang a,
⇑
a Discovery Chemistry, Merck Research Laboratory, Kenilworth, NJ 07033, United States
b Department of Medicinal Chemistry, AMRI Global, Albany, NY 12203, United States
c In Vivo Pharmacology Group, Merck Research Laboratory, Kenilworth, NJ 07033, United States
d Cardiorenal Group, Merck Research Laboratory, Kenilworth, NJ 07033, United States
e Pharmacokinetics, Pharmacodynamics and Drug Metabolism, Merck Research Laboratory, Kenilworth, NJ 07033, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of pyrrolo-benzo-1,4-diazine analogs have been synthesized and displayed potent Nav1.7
inhibitory activity and moderate selectivity over Nav1.5. The syntheses, structure–activity relationships,
and selected pharmacokinetic data of these analogs are described. Compound 41 displayed
anti-nociceptive efficacy in the rat CFA pain model at 100 mpk oral dosing.
Received 22 April 2014
Revised 17 July 2014
Accepted 21 July 2014
Available online 29 July 2014
Ó 2014 Elsevier Ltd. All rights reserved.
Keywords:
Nav1.7
Nav1.5
Pain
Sodium channel blockers
Pyrrolo-benzo-1,4-diazine
Voltage-gated sodium channels (VGSCs), a family of nine
sodium channel subtypes, generate and proliferate action poten-
tials in excitable cells, and are crucial for muscle and nerve func-
tion.1 Although most compounds identified as sodium channel
blockers do not display selectivity against all channel subtypes, a
few blockers have been shown to produce sufficient therapeutic
selectivity and window for the treatment of action potential mala-
dies such as epilepsy (phenytoin), pain (lidocaine), and certain car-
diac arrhythmia (mexiletine).
Nav1.7 is a VGSC that is highly expressed in nociceptive neurons
and sympathetic ganglia.2 It plays a critical role in the generation
and conduction of electrical signaling in small and medium-
diameter sensory afferent nerves.3 Knockout mice lacking Nav1.7
in Nav1.8 expressing neurons displayed reduced pain response to
mechanical or inflammatory pain.4 Humans lacking Nav1.7 exhibit
a profound loss of all pain sensation.5 Consistent with the human
phenotype, a more complete loss of pain sensation was observed
in a pan-neuronal Nav1.7 knockout mouse.4 In contrast to the
human and mouse knockout phenotype gain of function mutations
in Nav1.7 causes pain syndromes in humans, either primary eryth-
romelalgia or paroxysmal extreme pain syndrome.6,7 On balance
these observations suggest that Nav1.7 provides an excellent ther-
apeutic target for pain relief.
In our Nav1.7 program we screened our compound collection
using veratridine-modified human Nav1.7 and Nav1.5 channels
and a voltage-sensitive dye in FLIPR assays. Nav1.5 channel is the
major VGSC in the heart. Block of Nav1.5 is undesirable because
it is responsible for the cardiac action potential initiation and con-
duction. The project goal was to identify Nav1.7 blockers with rea-
sonable selectivity against Nav1.5 channels in vitro, and most
importantly in vivo. During high-throughput screening (HTS) of
our compound collection, some analogs with the pyrrolo-benzo-
1,4-diazine skeleton were identified to display Nav1.7 inhibitory
activity and modest selectivity against human Nav1.5. Compound
1, the initial lead, displayed potent Nav1.7 inhibitory activity
(IC50 160 nM) with ꢀ50-fold selectivity over Nav1.5 in the FLIPR
screen, shown in Figure 1. This apparent selectivity decreased to
ꢀ10-fold when whole cell sodium currents were measured in man-
ual voltage clamp (VC) with 50 ms depolarizations at 10 Hz from a
holding potential of ꢁ120 mV. However 1 did not exhibit high
⇑
Corresponding author. Tel.: +1 908 740 2589.
E-mail address: shu-wei.yang@merck.com (S.-W. Yang).
http://dx.doi.org/10.1016/j.bmcl.2014.07.060
0960-894X/Ó 2014 Elsevier Ltd. All rights reserved.

G. D. Ho et al. / Bioorg. Med. Chem. Lett. 24 (2014) 4110–4113
4111
O
S
O
R2
H
N
N
R2
H
N
N
H
N
N
O
O
R1
R1
N
N
8
N
9
1
b
a
NaV1.7 IC50
NaV1.5 IC50
rat-AUC6h
:
:
160 nM
8092 nM
788 nM.h
5
:
c
d
Figure 1. Initial HTS lead compound 1, and its biological data.
O
R2
O
H
N
N
N
H
H
N
N
R2
plasma exposure in rat based on area under curve (AUC) data
(AUC6h: 788 nM h).
R1
R1
N
10
N
Our strategy was to quickly identify Nav1.7 blockers which
exhibited in vivo efficacy for proof of concept. Therefore achieving
high oral plasma exposure was considered as an important goal.
Rat pharmacokinetic (PK) AUC values of the active compounds
were obtained to guide the compound selection for further
in vivo study. In this communication, we describe syntheses and
structure activity relationships (SAR) of the pyrrolo-benzo-1,4-dia-
zine series versus Nav1.5, Nav1.7, and plasma exposure. The PK and
in vivo data of selected compounds are described.
The synthetic route to the key intermediate 5 is summarized in
Scheme 1. The reaction of nitro-aniline 2 with 2,5-dimethoxytetra-
hydrofuran (2,5-DMTHF) in acetic acid, followed by reduction of
the nitro group using BiCl3–NaBH4 yielded pyrrolo-aniline inter-
mediates 3.8 The pyrrole intermediates (3) were then cyclized with
piperidin-4-one or N-substituted piperidin-4-one derivatives to
form 50H-spiro[piperidine-4,40-pyrrolo[1,2-a]quinoxaline] cores
(4–7). Intermediates 5 were the major intermediates for further
diverse modification. Cyclization of pyrrolo-anilines 3 and N-
substituted piperidin-4-one derivatives allowed us to directly
obtain certain desired targets, such as carbamates (4), amides (7),
and amines (6). However this strategy showed restriction due to
limited commercially available substituted piperidin-4-ones. The
late stage intermediates 5 were the subjects for versatile and fea-
sible structure modification. In general the yields of the cyclization
(reaction c in Scheme 1) to form 4–7 were excellent (>80%).
Conversion of intermediates 5 to sulfonamides, N-substituted
amines, amides, and ureas is shown in Scheme 2. Standard proce-
dures were used to prepare these intermediates. The commercially
7
Scheme 2. Reagents and conditions: (a) R2-Ph-SO2Cl, Et3N, CH2Cl2, rt (ꢀ30–50%);
(b) R2-benzyl bromide, DIPEA, DMF (ꢀ60–70%); (c) R2-PhNCO, Et3N, CH3CN, 0 °C to
rt (ꢀ50–60%); (d) R2CO2H, EDC, HOAt, DMAP, DIPEA, CH2Cl2 (ꢀ50–85%).
available sulfonyl chlorides, isocyanates, and substituted benzyl
bromides were reacted with 5 under basic conditions to form sul-
fonamides (8), ureas (10), and benzyl amines (9), respectively.
Substituted amides were prepared under EDC–HOAt coupling con-
dition. The amides shown in Tables 1, 2, 4 and 5 were prepared
from intermediates 5 according to this condition.
Target compounds were tested for their inhibitory activities at
the cloned human VGSCs (Nav1.7 or Nav1.5) expressed in HEK
293 cells by measuring fluorescence from a membrane potential
dye. After 60 min of compound and dye loading, cell depolarization
was evoked on the FLIPR by the addition of veratridine. Maximum
veratridine-induced increase in fluorescence intensity from base-
line was used to plot concentration-effect curves and generate
IC50 values. The rat PK assay was carried out with oral dosing at
10 mpk, and the AUC values were measured over a 0–6 h duration.
The initial SAR study focused on analogs without substitution on
the left-hand-side (LHS) phenyl ring and with varying substitution
on the right-hand-side (RHS) phenyl ring (Table 1). Compounds 11
and 12 with no or cyano substitution, respectively, on the RHS
phenyl ring did not show activity up to 3 lM. Compound 13 with
3-N,N-dimethyl substitution displayed reasonable Nav1.7 inhibi-
tory activity (IC50 381 nM), and poor selectivity against Nav1.5
(ꢀ11-fold). The 3,5-di-methyl substitution (14) however greatly
improved Nav1.7 inhibitory activity (IC50 47 nM) and selectivity
over Nav1.5 (379-fold). Unfortunately 14 did not show detectable
plasma exposure in the rat PK study (AUC6h 0 nM h). Although
H
N
NCbz
R1
N
4-cyano analog 12 was inactive up to 3 lM in the Nav1.7 FLIPR
O
NCbz
assay, it showed some plasma exposure (AUC6h 1011 nM h) in the
rat PK study. Hence additional analogs with 4-cyano-phenyl moiety
were prepared for further SAR study.
4
c
d
HCl
NH
H
O
NH
NO2
NH2
NH2
N
N
a, b
R1
R1
Table 1
R1
SAR of the analogs with the RHS substitution
c
N
O
2
3
5
O
R3
H
N
N
N
O
R
R2
4
N
3
c
N
c
O
R3
O
N
R2
H
N
a
a
R
Nav1.7 IC50 (nM) Nav1.5 IC50 (nM) AUC6h (nM h)
H
N
N
R1
11
H
>3000
>3000
381
ND
ND
4384
17810
ND
1011
633
0
R1
N
12 4-CN
13 3-N(CH3)2
14 3,5-Di-(CH3)2 47
N
6
7
Scheme 1. Reagents and conditions: (a) 2,5-DMTHF, AcOH, reflux (ꢀ70–90%); (b)
BiCl3, NaBH4, EtOH (ꢀ50–80%); (c) cat. maleic acid, EtOH, 80 °C (>80%); (d) H2
(1 atm), Pd/C, EtOH (>90%).
a
IC50 represent single determinations performed with four replicates per con-
centration point.

4112
G. D. Ho et al. / Bioorg. Med. Chem. Lett. 24 (2014) 4110–4113
Table 2
Table 4
SAR of the 7- or 8-substituted analogs
SAR of the RHS mono-substituted analogs
O
O
H
N
N
H
N
R
N
H3CO
7
8
4
R
CN
N
N
a
a
R
Nav1.7 IC50 (nM)
Nav1.5 IC50 (nM)
AUC6h (nM h)
a
a
R
Nav1.7 IC50 (nM)
Nav1.5 IC50 (nM)
AUC6h (nM h)
28
29
30
31
32
33
34
35
36
37
2-OCH3
2-Br
3-OCH3
3-CH3
3-Cl
436
414
280
321
48
201
437
312
94
4891
7601
4325
7172
4418
9716
8033
6943
3949
947
0
ND
0
0
0
783
91493
17199
1960
9436
15
16
17
18
19
20
21
7-OCH3
7-OCF3
7-CF3
7-Cl
8-F
8-Cl
613
81
13770
1452
4871
8289
9291
8144
9174
51081
11658
15202
2897
11982
26066
23416
311
404
478
563
344
3-CN
4-OCH3
4-CH3
4-Cl
8-Br
a
IC50 represent single determinations performed with four replicates per con-
centration point.
4-N(CH3)2
56
a
IC50 represent single determinations performed with four replicates per con-
centration point.
Table 3
SAR of the analogs with different linkers
R
Table 5
H
N
N
O
SAR of the RHS di-substituted analogs
O
N
H
N
H3CO
N
R1, R2
4
2
a
a
R
Nav1.7 IC50
(nM)
Nav1.5 IC50
(nM)
AUC6h
(nM h)
3
N
22 CO2-Bn
23 Bn
24 4-CN-Bn
25 3-CN-Ph
26 CONH-(4-CN-
Ph)
1500
17
562
9279
8732
6882
455
4577
12875
>30000
13108
370
ND
ND
ND
a
a
R1,R2
Nav1.7 IC50
(nM)
Nav1.5 IC50
(nM)
AUC6h
(nM h)
38 3-OCH3, 4-CN
39 3-Cl, 4-N(CH3)2
40 3-Cl, 4-OCH3
41 3-N(CH3)2, 4-
OCH3
439
9
37
9
7240
572
934
351
26641
1271
6809
27 SO2-(4-CN-Ph)
6966
ND
ND
20271
a
IC50 represent single determinations performed with four replicates per con-
42 3-OCH3, 4-
N(CH3)2
87
1329
ND
centration point.
a
IC50 represent single determinations performed with four replicates per con-
Substitution on the LHS phenyl ring was carried out mainly on
7- or 8-position due to their potential of being metabolized. A ser-
ies of compounds with different substitutions on C-7 and C-8 were
prepared. The data of selected compounds (15–21) are summa-
rized in Table 2. These compounds possessed similar inhibitory
and selectivity profiles (Nav1.7/Nav1.5: between 15 and 27 folds).
Compound 16 with 7-OCF3 substitution displayed most potent
Nav1.7 blocking activity (IC50 81 nM). The strategy of blocking
potential metabolic spots was very successful. All these com-
pounds showed moderate to great plasma exposure in the rat PK
centration point.
did not improve inhibitory activity and selectivity over Nav1.5, and
failed to show detectable plasma exposure in the rat PK study.
Among four 3-substituted analogs (30–33), 3-chloro substitution
(32) improved Nav1.7 inhibitory activity (IC50 48 nM) and selectiv-
ity over Nav1.5 (92-fold). However, only the 3-cyano analog (33)
displayed some oral plasma exposure (AUC6h 783 nM h), and 30–
32 were not detectable in the rat plasma after oral dosing at
10 mpk. Compared to 3-substituted analogs, 4-substituted analogs
(34–37) displayed similar Nav1.7 and Nav1.5 profiles, whereas the
plasma exposures of these compounds significantly improved.
Compound 34 with 4-OCH3 substitution exhibited outstanding
study with AUC6h values between 2.9 and 51 lM h. Among them,
the 7-methoxyl analog (15) exhibited the highest plasma exposure
in the rat oral PK study (AUC6h 51081 nM h). In addition a few 6- or
9-substituted analogs were prepared. However, all of them showed
poor selectivity over Nav1.5 (<6-fold, data not shown), and were
not further explored.
In parallel the linkers between the core and the RHS phenyl ring
were examined. The data of some selected compounds (22–27) are
listed in Table 3. Among these representatives, carbamate 22, urea
26, sulfonamide 27, and N-phenyl analog 25 all showed poor
Nav1.7 activity (>1500 nM) and selectivity over Nav1.5 (<5-fold).
The benzyl analog 23 showed excellent activity in the Nav1.7 FLIPR
assay (IC50 17 nM), but failed to display adequate oral plasma
exposure (AUC6h 370 nM h). From this SAR information we further
focused on compounds with the amide linkage, the LHS 7-OCH3
substitution, and varying substitution on the RHS phenyl ring.
The data of ten selected mono-substituted analogs (28–37) are
summarized in Table 4. Substitution with 2-OCH3 (28) or 2-Br (29)
AUC6h value (ꢀ91
lM h). From these SAR data, 4-position of the
RHS phenyl ring appeared to be the metabolic hot spot in the
amide series. Blocking this position greatly enhanced the oral PK
plasma exposure. Although 34 showed great metabolic stability,
the potency and the selectivity could be further improved. Hence
we shifted our focus to the bi-substituted analogs.
Selected bi-substituted analogs (38–42) and their data are
summarized in Table 5. To maintain good PK profile, all selected
compounds possessed 4-substitution on the RHS phenyl ring.
Among them, two N,N-dimethylamino analogs (39 and 41) showed
highly potent Nav1.7 inhibitory activity (IC50 9 nM) and moderate
selectivity over Nav1.5 (63- and 39-fold, respectively). Compound
41 with 4-methoxy and 5-N,N-dimethylamino substitutions pos-

G. D. Ho et al. / Bioorg. Med. Chem. Lett. 24 (2014) 4110–4113
4113
Table 6
O
Full rat PK data of 41
H
N
N
O
iv
po
O
N
N
Number of animalsa
Dose (mg/kg)
3
3
17.4
2.9
3
10
65.9
41
AUC
(lM h)
(0–24)
Nav1.7 IC50 9 nM (FLIPR)
Nav1.5 IC50 572 nM (FLIPR)
Nav1.7 IC50 331 nM (VC)
Nav1.5 IC50 10485 nM (VC)
hERG (iw): 3% inh @1 uM
14% inh. @ 10 uM
Terminal half-life (h)
Bioavailability (%)
Mean residence time (MRT) (h)
100
4.7
9.3; 1.3
3.0
Cmax
(l
M); Tmax (h)
Clearance (mL/min/kg)
Vd(ss) (L/kg)
6.5
1.2
Rat AUC6h: 20271 nM.h
brain conc.6h: 257 ng/g
a
Male CD rats were used in this study.
Figure 2. Optimized compound 41, and its in vitro and PK data.
In conclusion, a novel series of potent Nav1.7 inhibitors with
moderate selectivity over Nav1.5 based on the pyrrolo-benzo-1,4-
diazine scaffold were discovered. The initial SAR studies based on
Nav1.7 blocking, selectivity over Nav1.5, and plasma exposure (PK
AUC6h) versus structures were established. In particular, the 7- or
8-substitution on the LHS phenyl ring was very important for the
molecule to maintain good metabolic stability, as well as 4-substi-
tution on the RHS phenyl ring. To quickly achieve proof of concept,
we selected 41 with potent Nav1.7 inhibitory activity, moderate
selectivity over Nav1.5, and excellent PK profile for further
in vivo study. Compound 41 exhibited anti-nociceptive efficacy
in the rat CFA pain model at 100 mpk oral dosing. Further improve-
ments to this scaffold should furnish an improved Nav1.7 inhibitor,
which may be useful as an anti-nociceptive agent.
Rat CFA Thermal - Compound 41
25
Vehicle (contra)
20
15
10
5
Vehicle (ipsi)
100mpk (Contra)
100mpk (Ipsi)
**
0
baseline 1.5 hour
3 hour
Figure 3. CFA inflammatory pain responses at 1.5 and 3 h for compound 41 at
100 mpk PO (N, ipsilateral and d, contralateral) versus vehicle (4, ipsilateral and s,
contralateral) in the rat CFA model.9
Acknowledgments
sessed superior rat plasma exposure (AUC6h ꢀ20
lM h) compared
to 39 (1271 nM h), and was selected for further in vitro and
The authors wish to acknowledge Drs. William Greenlee and
George Njoroge for providing the insight for this program, Dr.
Tze-Ming Chan and his group for structure confirmation of some
analogs, and the DMPK group for acquiring pharmacokinetic data.
in vivo studies.
Compound 41 was further investigated in the Nav1.5 and Nav1.7
voltage clamp (VC) and the Human Ether-a-go-go Related Gene
(hERG) assays (Fig. 2). Standard ruptured whole cell patch clamp
techniques were used to measure direct Na channel blocking
activity. The desired compound should show low effect in these
undesired targets, such as hERG, a potassium ion channel which
regulates heart rhythm. Compound 41 showed IC50 331 nM in
the Nav1.7 VC assay, and ꢀ32-fold selectivity over Nav1.5 (VC). It
did not show a high percentage inhibition in the hERG assay
References and Notes
1. Clare, J. J.; Tate, S. N.; Nobbs, M.; Romanos, M. A. Drug Discovery Today 2000, 5,
506.
2. (a) Klugbauer, N.; Lacinova, L.; Flockerzi, V.; Hofmann, F. EMBO J. 1995, 1084, 14;
(b) Plummer, N. W.; Meisler, M. H. Genomics 1999, 57, 323.
3. (a) Raymond, C. K.; Castle, J.; Garrett-Engele, P.; Armour, C. D.; Kan, Z.;
Tsinoremas, N.; Johnson, J. M. J. Biol. Chem. 2004, 279, 46234; (b) Rush, A. M.;
Dib-Hajj, S. D.; Liu, S.; Cummins, T. R.; Black, J. A.; Waxman, S. G. Proc. Natl. Acad.
Sci. U.S.A. 2006, 103, 8245.
4. (a) Nassar, M. A.; Stirling, L. C.; Forlani, G.; Baker, M. D.; Matthews, E. A.;
Dickenson, A. H.; Wood, J. N. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 12706; (b)
Minett, M. S.; Nassar, M. A.; Clark, A. K.; Passmore, G.; Dickenson, A. H.; Wang, F.;
Malcangio, M.; Wood, J. N. Nat. Commun. 2012, 3, 791.
5. Cox, J. J.; Reimann, F.; Nicholas, A. K.; Thornton, G.; Roberts, E.; Springell, K.;
Karbani, G.; Jafri, H.; Mannan, J.; Raashid, Y.; Al-Gazali, L.; Hamamy, H.; Valente,
E. M.; Gorman, S.; Williams, R.; McHale, D. P.; Wood, J. N.; Gribble, F. M.; Woods,
C. G. Nature 2006, 444, 894.
6. Drenth, J. P.; Michiels, J. J. Int. J. Dermatol. 1994, 33, 393.
7. Dib-Hajj, S. D.; Cummins, T. R.; Black, J. A.; Waxman, S. G. Trends Neurosci. 2007,
30, 555.
(14% at 10 lM).
Compound 41 was evaluated for efficacy in the in vivo rat Com-
plete Freund’s Adjuvant (CFA) induced inflammatory pain model.
The pain inducing substance was injected into one of the hind
paws of the animal to induce an inflammatory response. Thermal
hyperalgesia was then measured two days after CFA treatment
based on paw withdrawal latency (PWL) to determine the pain
relief ability of Nav1.7 blocker 41 versus control at 0 h, 1.5 h and
3 h. The assay was validated with morphine, naproxen and mexil-
etine (oral dosing). As expected, morphine (5 mpk) had effects on
both the ipsilateral and contralateral paws. Mexiletine (60 mpk)
had significant effects on the contralateral but not the ipsilateral
paw. Naproxen (30 mpk) only affected the ipsilateral paw.
Compared to the vehicle, 41 significantly increased PWL at 3 h
following oral dosing at 100 mpk. Compound 41 displayed excel-
lent oral PK profile with 100% bioavailability in 24 h study (Table 6).
Average plasma concentrations of 41 at 3.5 h post dosing within
8. Guillon, J.; Grellier, P.; Labaied, M.; Sonnet, P.; Léger, J. M.; Déprez-Poulain, R.;
Forfar-Bares, I.; Dallemagne, P.; Lemaître, N.; Péhourcq, F.; Rochette, J.;
Sergheraert, C.; Jarry, C. J. Med. Chem. 2004, 1997, 47.
9. In Figure 3. One hind paw of rat subjects (6 rats per group) was treated with
Freud’s Adjuvant to induce an inflammatory pain response. With only vehicle
(D, ipsilateral) treatment, PWL remained the same during the 3 h period under
thermal conditions. With 41 (100 mpk po), PWL significantly increased at 3 h (N
ipsilateral). Therma PWL in the non-inflammed hind paw was also measured,
first with only vehicle administration (s contralateral). Compound 41 (100 mpk
oral dosing) displayed increased thermal PWL in the contralateral paw (d) at 3 h
post dosing. Male CD rats (Charles River Laboratories version of Sprague
Dawley) were used for both PK and PD studies.
the CFA thermal study were 47 lM (rat plasma protein binding
98.4%). The results demonstrated that Nav1.7 blocker 41 was able
to reduce pain sensation.