Discovery and SAR of novel tetrahydropyrrolo[3,4-c]pyrazoles as inhibitors of the N-type calcium channel

Article abstract of DOI:10.1016/j.bmcl.2014.03.062

A novel series of substituted tetrahydropyrrolo[3,4-c]pyrazoles were investigated as blockers of the N-type calcium channel (Ca_v2.2 channels), a chronic pain target.

Full text of DOI:10.1016/j.bmcl.2014.03.062

Bioorganic & Medicinal Chemistry Letters 24 (2014) 2053–2056
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery and SAR of novel tetrahydropyrrolo[3,4-c]pyrazoles as
inhibitors of the N-type calcium channel
⇑
Michael P. Winters , Nalin Subasinghe, Mark Wall, Edward Beck, Michael R. Brandt, Michael F. A. Finley,
Yi Liu, Mary Lou Lubin, Michael P. Neeper, Ning Qin, Christopher M. Flores, Zhihua Sui
Janssen Research and Development, LLC, 1400 McKean Rd., Spring House, PA 19477, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 27 March 2014
A novel series of substituted tetrahydropyrrolo[3,4-c]pyrazoles were investigated as blockers of the
N-type calcium channel (Cav2.2 channels), a chronic pain target.
Ó 2014 Elsevier Ltd. All rights reserved.
Keywords:
Tetrahydropyrrolo[3,4-c]pyrazoles
N-type calcium channel
Cav2.2 blockers
Pain
Voltage-dependent Ca2+ channels play important roles in many
critical biological processes.1 One particular type, the N-type cal-
cium channel (Cav2.2 channel), is located mainly at pre-synaptic
nerve terminals and has been implicated in the neurotransmission
of pain.2 N-type knockout mice show moderated response to pain
sensation, and show reduced neuropathic pain symptoms com-
pared to normal mice.3 The cone snail toxin ziconotide (PrialtÒ)
is a potent blocker of N-type calcium channels and is used for se-
vere chronic pain. It must be administered intrathecally, and has
a side-effect profile that gives it a very narrow therapeutic win-
dow.4 Due to these features, we feel that N-type calcium channel
is an attractive target for an oral therapeutic for the treatment of
pain, in particular neuropathic pain. Other groups have also been
pursuing this target with varying degrees of success, with one
compound, CNV-2197944 (structure not disclosed), currently in
clinical development for the treatment of chronic pain (Fig. 1).5
One hit from a screen of our compound collection was pyrazole
1, a 100 nM inhibitor of N-type calcium channel (Fig. 2). The side-
chain of 1 was rigidified by forming a fused cyclopentane ring on
the pyrazole scaffold 2 that improved potency by lessening the
rotational degrees of freedom of the sidechain.6 It was hypothe-
sized that modifying 2 by inserting a nitrogen into the cyclopentyl
ring giving tetrahydropyrrolo[3,4-c]pyrazoles 3 should enhance
the lipophilicity of the compounds and may impact other proper-
ties such as rat liver microsome (RLM) stability. These compounds
were designed using a 3+2 dipolar cycloaddition reaction of a nit-
rilimine and an alkyne tethered together by an alkylamino chain.
O2
S
F3C
O
S
OMe
N
N
N
H
N
NH
O
BocHN
O
Ono
Convergence/GSK
Z160 (NMED-160)
Figure 1. Some Representative N-Type Calcium Channel Blockers.
OMe
OMe
OMe
N
N
N
N
N
N
Cl
Cl
Cl
N
HO
HO
OH
3
2
1
2,3,5-substitued
pyrazole
2,3,5-substitued
2,3,5-substituted
tetrahydropyrrolo[3,4-
c]pyrazoles
tetrahydrocyclopenta[c]
pyrazole
Figure 2. Initial pyrazole hit and proposed bicyclic pyrazoles.
The general synthetic schemes for the synthesis of 2,3,5-tetra-
hydropyrrolo[3,4-c]pyrazoles are described below (Scheme 1).7
Treatment of alkyne 4 with bromine and KOH gives alkynyl bro-
mide 5.8 The hydrazide, generated by displacement of the mixed
anhydride with an arylhydrazine, is treated with polystyrene-
⇑
Corresponding author. Tel.: +1 215 628 7843; fax: +1 215 540 4611.
E-mail address: mwinter4@its.jnj.com (M.P. Winters).
http://dx.doi.org/10.1016/j.bmcl.2014.03.062
0960-894X/Ó 2014 Elsevier Ltd. All rights reserved.

2054
M. P. Winters et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2053–2056
supported Ph3P and CCl4 to give hydrazonyl chloride 6.7 Upon
treatment with TEA in toluene at 100 °C, 6 undergoes an intramo-
lecular 3+2 cycloaddition of the nitrilimine on the alkyne to give
the desired 3-bromotetrahydropyrrolo[3,4-c]pyrazole 7.9 The
3-aryl/heteroaryl substituent is installed by Suzuki coupling of
the 3-bromopyrazole with aryl or heteroaryl boronic acids, and
the Boc is removed by treatment with TFA to give key intermediate
8.10
Table 1
SAR at the 5-position of the pyrrolopyrazole
R2
Cl
N
N
N
R1
The next scheme shows the functionalization of the tetrahydro-
pyrrole nitrogen (Scheme 2). Treatment with sulfonyl chlorides or
acid chlorides in the presence of base gives the sulfonamides 9 or
amides 10, respectively. Derivatives with aryls or heteroaryls di-
rectly attached to the nitrogen were made by Buchwald/Hartwig
couplings 11. Reductive amination with aldehydes gives the ali-
phatic and benzylic derivatives 12.
Our initial efforts focused on optimizing the groups at
2-, 3-, and 5-positions of the bicyclic heterocycle. Solubility was
low for our initial compounds, so basic groups, such as pyridyl,
were installed in an attempt to increase solubility while maintain-
ing potency in the Functional Drug Screening System (FDSS)
assay.11 The initial hits also suffered from poor RLM stability.
RLM stability was important because stable compounds were nec-
essary to show activity in the initial in vivo model, the rat Com-
plete Freund’s Adjuvant (CFA) radiant heat model.
Compd no.
R1
R2
FDSS IC50 (nm)
RLMb (%)
a
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Et
4-Cl-Ph
pyrid-2-yl
CH2-4-Cl-Ph
CH2-pyrid-4-yl
CONMe2
Boc
COCF3
SO2Me
SO2iPr
SO2-cyclohexyl
SO2NMe2
SO2-morpholine
SO2-piperidine
SO2CH2CH2NHMs
SO2Bn
OMe
Et
Et
Et
Et
OMe
OMe
OCF3
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
61%
34%
36%
1
a
NT
18
18
10
10
NT
100
3
a
300
280
170
73
180
21
10
16
2
12
13
5
14
13
1
9.0
1.8
4.0
9.0
27
a
Substitution on the nitrogen at the 5-position of the pyrrolopy-
razole gave a range of potencies and was the most tolerant of di-
verse substitution (Table 1). Simple alkyl groups, such as ethyl
% Inhibition at 1
% Remaining after 10 min incubation with rat liver microsomes.
lM.
b
13, were not favored. Direct aryl and heteroaryl substitution as
well as benzyl substitution were also not tolerated, presumably
due to the basicity of the nitrogen in the ring. Amides, ureas and
carbamates are somewhat more potent, as evidenced by 18, 19
and 20. The most fruitful substitutions in this series were the
5-position sulfonamides and sulfamides. Methanesulfonamide 21
exhibited relatively high potency, and that of isopropyl sulfon-
amide 22 was 2-fold higher. However, cyclic sulfamides, such as
25 and 26 were the most potent compounds made in this sub-ser-
ies. Longer chain sulfonamides and sulfamides exhibited reduced
potency. RLM stability remained poor for all of the potent com-
pounds in this series; and only the weakly active trifluoroacetyl
(20), wherein R2 is OCF3, had good RLM stability.
Next, a number of substituents on the aromatic ring at the 2-po-
sition of the pyrazole were explored (Table 2). Metabolite identifi-
cation studies on 22 showed that loss of the methyl from the
o-methoxy was the major metabolite, so a number of substituents
on that ring were explored in an attempt to mitigate this issue.
Methoxy 22 and ethoxy 29 were both quite potent; however, the
more electron-withdrawing trifluoromethoxy substituent 30, lost
some potency. Contrary to our hypothesis, the ethyl substituent
31 did not afford greater RLM stability and exhibited diminished
potency as well. Similarly, the dimethylamino substituent 32 also
did not improve the RLM stability, although potency was restored.
Substitution with methyl and fluoro at the 4-position reduced po-
tency somewhat; however, reversing the methyl and methoxy 35
led to a much more dramatic loss of potency. Fused rings 36 and
37 did not enhance the potency. Attempts to adjust the poor
RLM stability by making numerous changes to this ring did not re-
sult in any improvement except for 34, the least potent compound
in this series. Collectively, these data support the hypothesis that
larger ortho substituents, such as methoxy, are necessary for inhib-
itory activity, possibly due to the angle of the aromatic ring to the
pyrazole. An electron-rich ring also seems important for N-type
inhibitory activity; however, this SAR runs counter to RLM
stability.
Br
Boc
a
b
N
O
Boc
N
O
OH
OH
4
5
R1
Br
R1
N
N
R2
Br
c
N
d
N
Boc
N
N
Boc
7
N
H
N
R1
N
H
Cl
8
6
Scheme 1. Reagents and conditions: (a) Br2, KOH; (b) (i) i-BuOCOCl, NMM,
R1NHNH2, DCM; (ii) PS-Ph3P, CCl4, ACN, 50 °C; (c) TEA, toluene, 100 °C; (d) (i)
R2B(OH)2, Pd(OAc)2, (o-tol)3P, Na2CO3; (ii) TFA.
R1
N
R1
N
R2
R2
N
N
N
N
SO2R3
R3
d
a
9
12
R1
N
R2
N
N
H
8
b
c
R1
R1
R2
N
R2
N
N
N
N
N
COR3
R3
10
11
Substitution on the 3-aryl ring did not improve potency over
the 4-chlorophenyl starting point 22 (Table 3), and removing the
aryl completely 38 or substituting bromine 39 resulted in a
Scheme 2. Reagents and conditions: (a) R3SO2Cl, TEA; (b) R3COCl, TEA or R3COOH,
EDC; (c) R3Br or R3I, Pd2dba3, BINAP, Cs2CO3, toluene, 100 °C (d) R3CHO,
NaBH(OAc)3.

M. P. Winters et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2053–2056
2055
Table 2
Table 4
SAR of 2-aryl analogs
Patch clamp data for selected compounds
R3
N
Compd no.
FDSS IC50
(nm)
QPatch
QPatch Hi
Freqa
QPatch ratio
Lo/Hi (%)
R2
R1
Lo Freqa
22
25
26
27
29
32
35
36
40
10
1.8
4
9
3
5.1
21
10
16
11
28
20
0
42
26
0
47
51
45
0
50
47
0
23
55
44
0
84
55
0
Cl
N
N
SO2
R4
Compd no.
R1
R2
R3
R4
FDSS IC50 (nm)
RLMa (%)
0
14
7
10
0
140
22
29
30
31
32
33
34
35
36
37
OMe
OEt
OCF3
Et
NMe2
OMe
Me
OMe
–OCH2CH2–
–OCH2O–
H
H
H
H
H
H
H
H
H
H
H
H
i-Pr
i-Pr
i-Pr
Me
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
10
3.0
27
85
5.1
32
190
21
10
20
2
a
NT
NT
13
4
13
76
23
1
% Inhibition at 0.1 lM.
H
Me
OMe
F
H
H
To assess selectivity, compound 31 was examined in an L-type
Calcium Channel assay and was found to have only 12% inhibition
of L-type at 1 M and 48% inhibition at 5 M concentration of com-
pound. No other compounds from this series were assayed versus
L-type, but this single data point does suggest that the series may
be selective for N-type over L-type.14
l
l
7
a
% Compound remaining after 10 min incubation with RLM.
In conclusion, the potency of a straight chain pyrazole hit for N-
type calcium channel was improved by modifying the scaffold to a
Table 3
more rigid pyrrolo[3,4-c]pyrazole.
A number of compounds
SAR at the 3-position of the pyrrolopyrazole
showed state-dependent inhibition in patch clamp assays, but sta-
bility in rat liver microsomes remains an issue. Some of the lear-
nings from this attempt to make oral N-type calcium channel
inhibitors have been incorporated into efforts on other series that
we hope will lead to improved therapeutics in the future.
OMe
R1
N
N
N
SO2
Acknowledgments
The authors gratefully acknowledge the ADME and analytical
teams in Spring House for providing the human and rat liver
metabolism data, and Renee DesJarlais for initial modeling
discussions.
Compd no.
R1
FDSS IC50 (nm)
10
RLMb (%)
22
38
39
40
41
42
43
44
45
46
4-Cl-Ph
H
Br
4-Cl-pyrid-3-yl
4-MeO-pyrid-3-yl
4-i-PrO-pyrid-3-yl
4-Me- pyrid-3-yl
Quinolin-3-yl
Indol-5-yl
2
8
1
1
1
2
NT
1
21
4
a
33%
30%
a
16
43
130
46%
References and notes
a
100
34%
45%
1. (a) Lnenicka, G.; Hong, S. Mol. Neurobiol. 1997, 14, 37; (b) Yamamoto, T.;
Takahara, A. Curr. Top. Med. Chem. 2009, 9, 377; (c) Lee, S. Curr. Neuropharmacol.
2013, 11, 606.
a
a
3-Cl-pyrid-4-yl
2. (a) Schroeder, C.; Doering, C.; Zamponi, G.; Lewis, R. Med. Chem. 2006, 2, 535;
(b) Elmslie, K. J. Neurosci. Res. 2004, 75, 733; (c) Belardetti, F. Future Med. Chem.
2010, 2, 791; (d) McGivern, J. In Peripheral Receptor Targets for Analgesia: Novel
Approaches to Pain Management; Cairns, B., Ed.; John Wiley & Sons, Hoboken, NJ
2009, pp 111.
a
% Inhibition at 0.33
% Remaining after 10 min incubation with rat liver microsomes.
l
M.
b
3. McGivern, J.; McDonough, S. Curr. Drug Targets CNS Neurol. Disord. 2004, 3, 457.
4. (a) Prommer, E. Drugs Today 2006, 42, 369; (b) Miljanich, G. Curr. Med. Chem.
2004, 11, 3029.
dramatic loss in potency. Installing a 3-pyridyl 40 in an attempt to
increase solubility resulted in a slight loss in potency, but changing
the 4-substituent to oxygen or methyl on the 3-pyridyl ring also
was not favored. Fused heterocycles, such as 3-quinoline 44 and
5-indole 45 also did not exhibit improved potency nor did 3-posi-
tion substitution, such as 46. RLM stability was not affected by the
changes made at the 3-position and remained poor for all
compounds.
A number of the more potent compounds were selected for fol-
low-up by screening in patch clamp assays (Table 4).12 In general,
the more potent compounds in the FDSS assay were more potent in
the patch clamp assay with a few exceptions. Compound 22, mod-
erately potent in the FDSS assay, had greater than expected po-
tency in the patch clamp assay; 35, 27 and 40, with similar
potency to 22, had very little potency in the QPatch assay. It has
been postulated that relatively greater inhibition at high vs. low
frequency (i.e., use- or state-dependent inhibition) might result
in an improved therapeutic index.13 Compounds such as 22, 25,
26 and 32 all showed some degree of state-dependent inhibition
leading to an increased interest in this series.
5. (a) Yamamoto, T.; Ohno, S.; Niwa, S.; Tokumasu, M.; Hagihara, M.; Koganei, H.;
Fujita, S.; Takeda, T.; Saitou, Y.; Iwayama, S.; Takahara, A.; Iwata, S.; Shoji, M.
Bioorg. Med. Chem. Lett. 2011, 21, 3317; (b) Tyagarajan, S.; Chakravarty, P.; Park,
M.; Zhou, B.; Herrington, J.; Ratliff, K.; Bugianesi, R.; Williams, B.; Haedo, R.;
Swensen, A.; Warren, V.; Smith, M.; Garcia, M.; Kaczorowski, G.; McManus, O.;
Lyons, K.; Li, X.; Madeira, M.; Karanam, B.; Green, M.; Forrest, M.; Abbadie, C.;
McGowan, E.; Mistry, S.; Jochnowitz, N.; Duffy, J. Bioorg. Med. Chem. Lett. 2011,
21, 869; (c) Pajouhesh, H.; Feng, Z.; Zhang, L.; Pajouhesh, H.; Jiang, X.;
Hendricson, A.; Dong, H.; Tringham, E.; Ding, Y.; Vanderah, T.; Porreca, F.;
Belardetti, F.; Zamponi, G.; Mitscher, L.; Snutch, T. Bioorg. Med. Chem. Lett. 2012,
22, 4153; (d) Beebe, X.; Yeung, C.; Darczak, D.; Shekhar, S.; Vortherms, T.;
Miller, L.; Milicic, I.; Swensen, A.; Zhu, C.; Banfor, P.; Wetter, J.; Marsh, K.; Jarvis,
M.; Scott, V.; Schrimpf, M.; Lee, C.-H. Bioorg. Med. Chem. Lett. 2013, 23, 4857.
6. Winters, M.; Subasinghe, N.; Wall, M.; Beck, E.; Brandt, M.; Finley, M.; Liu, Y;
Lubin, M.; Neeper, M.; Qin, N.; Flores, C.; Sui, Z. Bioorg. Med. Chem. Lett. 2014,
24, 2057.
7. Winters, M.; Teleha, C.; Sui, Z. Tet. Lett. 2014, 55, 2150.
8. Brandsma, L.; Verkruijsse, H. D. Synthesis 1990, 11, 984.
9. Molteni, G. Heterocycles 2004, 63, 1423.
10. The same compounds can be accessed using an arylalkyne in place of the
bromoalkyne if desired. See Ref. 7.
11. In the FDSS assay, functional activity of test compounds are determined by
their inhibitory effects on Ca2+ influx via N-type calcium channel after

2056
M. P. Winters et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2053–2056
depolarizing plasma membrane with 50 mM KCl. Calcium mobilization
responses to KCl depolarization are evaluated by measuring the intensity of
Ca2+ fluorescent signal in the presence of BD Calcium Assay Dye (BD
CO2 at 37 °C. On assay day, growth media is removed and cells are loaded with
BD calcium assay dye (BD Bioscience) for 35 min at 37 °C, under 5% CO2 and
then for 25 min at room temp. Utilizing the FDSS, cells are challenged with test
compounds (at varying concentrations) and intracellular Ca2+ is measured for
5 min prior to the addition of 50 mM KCl for an additional 3 min of
measurement. IC50 values for compounds ran in this assay are determined
from six-point dose-response studies and represent the concentration of
compound required to inhibit 50% of the maximal response. Maximal
fluorescence intensity (FI) achieved upon addition of 50 mM KCl was
exported from the FDSS software and further analyzed using GraphPad Prism
3.02 (Graph Pad Software Inc., CA, U.S.A.). Data is normalized to the maximum
average counts of quadruplicate wells for each data points in the presence of
50 mM KCl and to the minimum average counts in the presence of buffer.
Theoretical curves are generated using nonlinear regression curve-fitting
analysis of either sigmoidal dose response or sigmoidal dose response (variable
slope) and the IC50 values with the best-fit dose curve determined by GraphPad
Prism are reported. Conotoxin was run as a positive control in the assay and
typically showed IC50’s of 10–20 nM.
Biosciences), utilizing
Hamamatsu. The percent inhibition at a concentration of 1
a
Functional Drug Screening System (FDSS) by
M as well as
l
IC50 values will be reported. For this screen, a stable cell line (HEK parent)
expressing Cav2.2 (N-type calcium channel, Genbank accession number
AAO53230) subunits is used. Cav2.2 subunits were expressed in pcDNA3.1
and pBudCE4.1 vectors under selection by 400 lg/ml of G418 and 200 lg/ml of
Zeocin. Cells were clonally isolated, expanded and screened by Western blot
analyses, and further tested for expression of characteristic N-type calcium
currents by whole-cell patch clamp. HEK293 cells stably expressing Cav2.2
subunits are routinely grown as monolayer in Dulbecco’s Modification of
Eagle’s Medium (DMEM; low glucose Cat # 12320-032) supplemented with
10% FBS, 2 mM
L-glutamine, 100 I.U./ml penicillin, 100
lg/ml streptomycin,
400 g/ml G418, and 200
l
l
g/ml Zeocin (Split ratio: 1:5). Cells are maintained
in 5% CO2 at 37 °C. JNJ compounds are prepared as 10 mM stock in DMSO from
neat, if available. Otherwise, the 5 or 10 mM DMSO stock solutions provided
in-house are used. Calcium mobilization responses to KCl depolarization are
evaluated by measuring the intensity of Ca2+ fluorescent signal in the presence
12. Mathes, C.; Friis, S.; Finley, M.; Liu, Y. Comb. Chem. High Thr. Screen. 2009, 12,
78.
of BD Calcium Assay Dye (BD Biosciences), utilizing
Screening System (FDSS) by Hamamatsu. 24 h prior to assay, cells are seeded
in clear-base poly- -lysine coated 384-well plates (BD Biosciences, NJ, USA) at
a density of 5000 cells per well in culture medium and grown overnight in 5%
a
Functional Drug
13. Winquist, R.; Pan, J.; Gribkoff, V. Biochem. Pharmacol. 2005, 70, 489.
14. Only certain compounds were intermittently assayed for selectivity versus L-
type Calcium Channel during this program.
D