From arylureas to biarylamides to aminoquinazolines: Discovery of a novel, potent TRPV1 antagonist

Article abstract of DOI:10.1016/j.bmcl.2006.07.010

Bioisosteric replacement of piperazine with an aryl ring in lead VR1 antagonist 1 led to the biarylamide series. The development of B-ring SAR led to the conformationally constrained analog 70. The resulting aminoquinazoline 70 represents a novel VR1 anta

Full text of DOI:10.1016/j.bmcl.2006.07.010

Bioorganic & Medicinal Chemistry Letters 16 (2006) 5217–5221
From arylureas to biarylamides to aminoquinazolines:
Discovery of a novel, potent TRPV1 antagonist
Xiaozhang Zheng,a Kevin J. Hodgetts,a Harry Brielmann,a Alan Hutchison,a
Frank Burkamp,b A. Brian Jones,b Peter Blurton,b Robert Clarkson,b
Jayaraman Chandrasekhar,a Rajagopal Bakthavatchalam,a
a
a
a
a,
*
´
Stephane De Lombaert, Marci Crandall, Daniel Cortright and Charles A. Blum
aNeurogen Corporation, 35 N. E. Industrial Road, Branford, CT 06405, USA
bMerck Sharp and Dohme, The Neuroscience Research Centre, Terlings Park, Harlow, Essex CM20 2QR, UK
Received 19 May 2006; revised 4 July 2006; accepted 5 July 2006
Available online 25 July 2006
Abstract—Bioisosteric replacement of piperazine with an aryl ring in lead VR1 antagonist 1 led to the biarylamide series. The devel-
opment of B-ring SAR led to the conformationally constrained analog 70. The resulting aminoquinazoline 70 represents a novel
VR1 antagonist with improved in vitro potency and oral bioavailability vs the analogous compounds from the lead series.
Ó 2006 Elsevier Ltd. All rights reserved.
CF3
CF3
TRPV1 (VR1 receptor) is a member of the transient
receptor potential family of ion channels and is localized
on primary afferent neurons as well as sensory neurons
inervating the bladder and gut. Activation of VR1 on
sensory neurons by chemical stimulants including capsa-
icin and resiniferatoxin, as well as low pH and heat,
leads to an influx of calcium and sodium ions through
the channel, causing depolarization of the cell and trans-
mission of painful stimuli. It is now generally hypothe-
sized that direct blockade of the channel should
provide a clinically effective therapeutic method to treat
pain.
C
C
HN
HN
CF3
A
N
CF3
A
O
O
B
B
N
N
N
1
2
hVR1 =17 nM
hVR1 = 14 nM
Figure 1. Bioisosteric replacement of the piperazine by a phenyl ring.
arylureas have proven to be valuable tools to explore
pre-clinical VR1 pharmacology,4 although the commer-
cial potential of the best characterized members of the
series has generally been limited due to a variety of
issues including poor aqueous solubility, poor oral
exposure, and metabolic/chemical instability. During
our own SAR exploration of the arylurea series it was
discovered that bioisosteric replacement of the pipera-
zine B-ring with an aryl-ring was feasible (e.g., 1 and
2, Fig. 1) with minimal loss of potency at the VR1 recep-
tor.5 The resulting arylcarboxamides offered several pos-
sible advantages, among them: (1) increased metabolic/
chemical stability; (2) more direct access to under-
explored regions of the arylurea pharmacophore via
substitution on the B-ring; (3) a springboard from which
to develop novel second generation VR1 antagonists.
Therefore, the identification of selective and potent
antagonists of VR1 has become a focus of attention
within the pharmaceutical industry.1
Among the earliest non-vanilloid VR1 antagonists to
appear in the patent literature were the arylureas e.g.,
1 (Fig. 1) which we originally discovered several years
ago through screening of our compound library.2 Sever-
al other groups have independently reported on the
SAR of these and related VR1 antagonists.3 The
Keywords: VR1; TRPV1; Pain; Bioisostere; Quinazoline; Ureas;
Antagonist.
*
Corresponding author. Tel.: +1 203 315 3073; fax: +1 203 481
5290; e-mail: cblum@nrgn.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.07.010

5218
X. Zheng et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5217–5221
Table 2. Biarylamides with A-ring 3-position variationsa
This letter highlights the pertinent SAR of the A-, B-,
and C-rings of the biarylamides with emphasis on the
B-ring and how this related to the discovery of a series
of aminoquinazolines with potent VR1 antagonist activ-
ity. The compounds were assayed for their ability to
inhibit capsaicin activation of human VR1 receptors
using fluorometric imaging plate reader (FLIPR) tech-
nology as well as their ability to inhibit the rat VR1
receptor upon activation at low pH (5.0–5.5).8 The data
are presented in Tables 1–3. The compounds in the
study were conveniently prepared by the route outlined
in Scheme 1.
HN
R
O
3
4
5
A
N
Compound
R
hVR1-capb rVR1-pHc
(nM) (nM)
24
25
21
22
11
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
H
F
11
8
In general, the SAR with respect to simple substitutions
on the A- and C-rings tracked well with the arylurea ser-
ies. For example, the preferred A-ring was a pyridin-2-yl
linkage with a 3-substituent (e.g., 3-CF3, 3-Me or 3-Cl).
Substitution at other positions of the A-ring or replace-
ment of the A-ring with a variety of five and six mem-
bered heterocycles resulted in a significant loss of
potency (data not shown). The optimal substitution on
the C-ring incorporated a lipophilic group (e.g., CF3
3
13
1
10
2
Me
Cl
8
CF3
7
62
4
O-propyl
CN
NO2
20
13
5
30
5
NH2
123
158
190
1106
49
325
37
44
290
5
NH–propyl
NH–cyclopentyl
NH(C@O)Me
NHOH
t
or butyl) at the para-position of the aniline (2 and 11,
Table 1). Smaller alkyl groups (12 or 18) or more-hydro-
philic groups (9 or 13) led to a degradation of VR1
antagonist potency. A pyridine C-ring was tolerated as
long as the requisite lipophilic group was present in
the para-position (see 10 vs 19). The data illustrate the
effectiveness of these compounds in blocking the VR1
receptor irrespective of species (human or rat) or modal-
ity of activation (capsaicin or low pH). While the simple
NHSO2Me
NMeSO2Me
NHSO2Et
152
39
35
9
49
4
0.4
N(SO2Me)2
NHSO2–n-butyl
NHSO2CH2Ph
NHSO2Ph
0.8
162
377
141
159
166
88
91
168
CH2OH
CO2H
>4000
>4000
746
560
>4000
234
334
639
196
694
34
C(@N)NH2
CH2N(Me)2
CH2N(Et)2
Table 1. Biarylamides with A- and C-ring variationsa
1412
4303
1527
3669
387
Ar
HN
CH2NH–propyl
CH2–pyrrolidin-1-yl
Imidazol-2-yl
R
A
O
2-Me–1,3,4-oxadiazol-5-yl
1,3,4-Oxadiazol-2-yl
Tetrazol-5-yl
91
>4000
59
>4000
N
a Values are means of 2–4 experiments, (nd = not determined).
b Values are IC50s at human VR1 receptor activated by capsaicin.
c Rat VR1 receptor activated by low pH (5.0–5.5).
Compound
R
Ar
hVR1-capb rVR1-pHc
(nM) (nM)
1
8
Figure 1
CF3 Ph
CF3 4-CF3-Ph
17 11
nd
26
>4000
14
biarylamides maintained much of the in vitro potency of
the corresponding arylureas (e.g., 2 vs 1), the com-
pounds also retained the less desirable drug-like proper-
ties of the corresponding ureas, most notably poor
aqueous solubility. For example, the most potent exam-
ple, 11 (hVR1 = 6 nM), exhibited a bioavailability in
rats of only 2%, as well as poor aqueous solubility (be-
low limit of quantitation, BLQ) in our high throughput
solubility (HTSol) assay.6
2
9
CF3 4-CN–Ph
417
24
1299
65
10
11
12
13
14
15
16
17
18
19
20
21
22
23
CF3 5-CF3–pyridin-2-yl
CF3 4-tert-Butyl–Ph
Me 4-i-Propyl–Ph
Me 4-OMe–Ph
Me 4-F–Ph
6
73
4
43
>4000
>4000
1502
974
189
2585
>4000
65
>4000
>4000
2024
896
366
>4000
>4000
81
Me 4-Cl–Ph
Me 4-OCF3–Ph
Me 4-Br–Ph
Me 4-Me–Ph
Utilizing our knowledge of arylurea SAR, we focused
our efforts to solubilize these compounds on the 3-posi-
tion of the A-ring, a position more tolerant of substitu-
tion (Table 2).
Me Pyridin-2-yl
Me 4-CF3–Ph
Me 4-tert-Butyl–Ph
11
9
Cl
Cl
4-tert-Butyl–Ph
4-CF3–Ph
8
58
2
nd
a Values are means of 2–4 experiments, (nd = not determined).
b Values are IC50s at human VR1 receptor activated by capsaicin.
c Rat VR1 receptor activated by low pH (5.0–5.5).
Among the neutral hydrophilic substituents, several sul-
fonamides (35–37, and 40) exhibited a modest improve-
ment in aqueous solubility (HTSol 1–20 lg/mL) versus

X. Zheng et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5217–5221
5219
Table 3. Biarylamides with B-ring variationsa
R'
CO2
H
+
R'
CF3
HN
a
(HO)2B
H2N
O
HN
3
4
(HO)2
B
R
O
5
B
2
3
R
R = NO2
R = NH2
c
N
b
Cl
N
6
d
R = NHSO2R2
Compound
R
B-ring (substituents) hVR1-capb rVR1-pHc
(nM) (nM)
R = NH2
R = NHalkyl
R'
e
1
2
Figure 1
CF3
17 11
HN
R = CN
f
g
R = CHO
14
26
R
O
R = CH2 NR3 R4
52
53
54
55
56
57
58
59
60
CF3 2-Me
CF3 3-F
257
131
192
348
R = CHO
h
N
7
R = CH2 OH
CF3 3-Cl
CF3 3-OH
>4000
43
93
2403
36
Scheme 1. General synthesis (a) BOP, CH2Cl2, (b) 2 M Na2CO3,
Pd(PPh3)4, DME, 80 °C, (c) SnCl2 Æ 2H2O, EtOAc, reflux, (d) MsCl,
TEA, CH2Cl2, (e) alkyl-CHO, NaBH(OAc)3, HOAc, DCE, (f)
DIBAL, CH2Cl2, À78 °C, (g) NHR3R4, NaBH(OAc)3, DCE, HOAc,
and (h) NaBH4, EtOH.
CF3 3-NH2
CF3 3-NO2
CF3 3-OMe
CF3 3-N(SO2Me)2
CF3 3-NHSO2Me
103
222
392
>4000
>4000
1015
1044
>4000
722
Exploration of the central ring (Table 3) was of primary
importance since this region was less accessible synthet-
ically in the arylurea series and largely unexplored. The
introduction of small lipophilic groups was investigated
first. At the 2-position, methyl substitution gave a 4- to
18-fold decrease in potency compared to the parent ana-
logs (52 vs 2 and 66 vs 65). At the 3-position, introduc-
tion of a chloro- (54) or methyl-substituent (67) resulted
in an even more dramatic loss in potency. A similar
trend was observed for the 2-pyridyl analogs 61 and
63 (Table 3). Conformational analysis of 2 using the
MMFFs force field7 suggests that the central phenyl ring
is twisted with respect to the amide bond by 35° in the
minimized structure, although enforcing planarity of
the benzamide requires only 1.2 kcal/mol. By making
the planar benzamide unit energetically inaccessible
through 3-Cl substitution, a large reduction in VR1
potency was observed. Conversely, introduction of a
small hydrogen bond donor at the 3-position gave com-
pounds only slightly less potent at VR1 than the parent
2. For example, the 3-OH (55) and 3-NH2 (56) analogs
are capable of forming a hydrogen bond with the amide
carbonyl thereby enforcing planarity of the benzamide
unit. Analogs, such as the 3-OMe (58, >4000 nM) or
3-NO2 (57, 392 nM) which contain H-bond acceptor
groups only and which were capable of forming
H-bonding interactions with the aniline NH, were
61
62
63
64
65
CF3
60
>4000
259
56
764
263
60
N
CF3
HN
O
CF3
N
N
CF3
42
N
Me
64
79
66
67
Me 2-Me
Me 3-Me
233
>4000
190
>4000
68
Me
624
583
N
69
Me
1779
1583
N
a Values are means of 2–4 experiments, (nd = not determined).
b Values are IC50s at human VR1 receptor activated by capsaicin.
c Rat VR1 receptor activated by low pH (5.0–5.5).
the parent compound 24 (HTSol BLQ) but only double-
digit nanomolar potency at VR1 was attainable. The
bis-sulfonamide 37 (hVR1 = 0.8 nM) exhibited the high-
est potency of any of the biarylamides, however, this
compound rapidly reverted to the mono-sulfonamide
34 upon treatment with nucleophilic amines, making it
unsuitable for further development. Basic amines (e.g.,
44, HTSol 50 lg/mL) or acidic functionalities (e.g., 42,
HTSol 104 lg/mL) offered a much improved solubility
profile but were less potent at VR1 (746 to
>4000 nM). Similarly, the incorporation of small het-
erocycles (48–51) yielded compounds with modest or
weak potency.
CF3
CF3
HN
O
HN
CF3
O
H
CF3
N
B
N
N
N
55
70
hVR1 = 43 nM
rVR1 = 36 nM
hVR1 = 1.1 nM
rVR1 = 1.4 nM
Figure 2. Heterocyclization of the biarylamide to the amino-
quinazoline.

5220
CF3
X. Zheng et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5217–5221
moderately high clearance and relatively long half lives
due to a large volume of distribution.
CF3
Cl
a
+
N
(HO)2B
N
Although 70 was not demonstrably better in terms of its
aqueous solubility (HTSol = BLQ), oral pharmacoki-
netics experiments revealed that 70 significantly outper-
formed 1 with respect to oral bioavailability (99 and
27%, respectively). The lower clearance exhibited by 70
is partially responsible for this improvement in oral
exposure, however, factors such as increased rigidity
and/or fewer rotatable bonds may also contribute.9
71
72
73
b, c
CO2H
NH2
CO2H
NO2
CF3
CF3
d
N
N
In conclusion, bioisosteric replacement of the piperazine
of the urea lead 1 led to the biarylamide series. The
improved VR1 potency of biarylamide analogs with
hydrogen bond donors at the 3-position of the B-ring
suggested that enforcing planarity of the benzamide unit
through interaction with the carbonyl group was a
promising approach. To investigate this possibility the
conformationally constrained analog (70) was prepared.
Aminoquinazoline 70 represents a novel VR1 antagonist
template with improved in vitro potency and oral bio-
availability relative to the corresponding urea or carbox-
amide series compound. While the aminoquinazoline
template retains all the pharmacophore features of the
arylureas and biarylamides, the new template also pro-
vides additional handles for further optimization. Fur-
ther exploration of the SAR and biology of the
aminoquinazoline series will be reported in due course.
75
74
e, f
CF3
Cl
N
N
HN
N
g
CF3
CF3
N
N
N
70
76
Scheme 2. Reagents and conditions: (a) 2 M Na2CO3, Pd(PPh3)4,
DME, 80 °C, 83%; (b) HNO3, 0 °C, 100%; (c) KMnO4, pyridine, H2O,
63%; (d) H2, Pd/C, 92%; (e) HCONH2, 63%; (f) OPCl3, 72%; (g)
4-trifluoromethyl aniline, IPA, 94%.
significantly less potent at VR1. Therefore, although it is
possible to achieve planarity via intramolecular H-bond-
ing of 3-substituents with either the carboxamide N–H
or the carbonyl oxygen, the data in Table 3 suggest that
the latter is preferred from a binding energy perspective.
The reasons for this are not clear but could be due to
severe spatial constraints near the amide N–H.
Acknowledgment
The authors express their appreciation to Lauren Dan-
ner for her work in cell culture and also to Bert Chenard
and Dave Wustrow for reviewing the manuscript.
Incorporation of hetero-atoms into the B-ring was also
examined (e.g., 64, 68, and 69) and led to a diminution
in VR1 potency.
References and notes
To investigate the notion that rigidifying the B-ring ori-
entation through hydrogen bonding interaction with the
carbonyl group (as in 55, Fig. 2) is a promising approach
to enhance VR1 potency, we synthesized the conforma-
tionally constrained biarylamide 70 (Scheme 2, Fig. 2).
The resulting quinazoline (70, hVR1 = 1.1 nM) was 10-
fold more potent in the human VR1 FLIPR assay than
either the parent urea 1 (17 nM) or the parent carbox-
amide 2 (14 nM). As with compounds 1 and 2, 70 was
devoid of agonist activity (HEK 293 cells expressing
human VR1) at concentrations up to 4 lM.
1. (a) Krause, J. E.; Chenard, B. L.; Cortright, D. N. Curr.
Opin. Invest. Drugs 2005, 6, 48; (b) Rami, H. K.;
Gunthorpe, M. J. Drug Discovery Today: Therapeutic
´
´
Strategies 2004, 1, 97; (c) Lopez-Rodrıguez, M. L.; Viso,
A.; Ortega-Gutierrez, S. Mini-Rev Med. Chem. 2003, 3, 729.
´
2. (a) Hutchison, A.; DeSimone, R. W.; Hodgetts, K.; Krause,
J. E.; White, G. PCT WO 02/08221; (b) Bakthavatchalam,
R.; Hutchison, A.; DeSimone, R. W.; Hodgetts, K.;
Krause, J. E.; White, G. U.S. Patent 6,723,730, 2004.
3. (a) Sun, Q.; Tafesse, L.; Islam, K.; Zhou, X.; Victory, S. F.;
Zhang, C.; Hachicha, M.; Schmid, L. A.; Patel, A.;
Rotshteyn, Y.; Valenzano, K. J.; Kyle, D. J. Bioorg. Med.
Chem. Lett. 2003, 13, 3611; (b) Tafesse, L.; Sun, Q.;
Schmid, L.; Valenzano, K. J.; Rotshteyn, Y.; Su, X.; Kyle,
D. J. Bioorg. Med. Chem. Lett. 2004, 14, 5513; (c) Swanson,
D. M.; Dubin, A. E.; Shah, C.; Nasser, N.; Chang, L.; Dax,
S. L.; Jetter, M.; Breitenbrucher, J. G.; Codd, E. E.; Lee, D.
H.; Zhang, S.-P.; Chaplan, S. R.; Carruthers, N. I. J. Med.
Chem. 2005, 48, 1857.
4. (a) Valenzano, K. J.; Grant, E. R.; Wu, G.; Hachicha, M.;
Schmid, L.; Tafesse, L.; Sun, Q.; Rotshteyn, Y.; Francis, J.;
Limberis, J.; Malik, S.; Whittemore, E. R.; Hodges, D.
J. Pharmacol. Exp. Ther. 2003, 306, 377; (b) Pomonis, J. D.;
Harrison, J. E.; Lilly, M.; Bristol, D. R.; Valenzano, K. J.;
Walker, K. J. Pharmacol. Exp. Ther. 2003, 306, 387.
The rat pharmacokinetic (PK) profiles of the arylurea 1
and the aminoquinazoline 70 are compared in Table 4.
Similar pharmacokinetic profiles were observed for 1
and 70 following intravenous (iv) dosing at 3 mg/kg in
67% PEG/water vehicle. Both compounds exhibited
Table 4. Pharmacokinetic profiles of 1 and 70
Compound
T1/2
CL
(mL/min/kg)
Vd
Tmax
F
(%)
(h)
(L/kg)
(h)
1
70
7.2
8.1
39
23
12
12.5
0.5
0.7
27
99

X. Zheng et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5217–5221
5221
5. The same bioisosteric replacement was recently disclosed
by Park, H.-G. et al. Bioorg. Med. Chem. Lett. 2005, 15,
631.
6. High throughput solubility (HTSol) was assessed by
diluting DMSO stock solutions of the test article into pH
7.4 buffer and shaking for 4 h at room temperature.
Solutions are then filtered and analyzed (HPLC/MS).
7. Energy minimizations were carried out using Maestro 7.0,
Schro¨dinger, <http://www.schro¨dinger.com>.
8. These assays conveniently provided information regarding
species selectivity (rat vs human) as well as the ability of
compounds to block different modes of receptor activation
(capsaicin or low pH). Data generated in human VR1 low
pH or rat VR1 capsaicin assays (data not shown) did not
yield an additional understanding of compound
pharmacology.
9. Veber, D. F.; Johnson, S. R.; Cheng, H.-Y.; Smith, B. R.;
Ward, K. W.; Kopple, K. D. J. Med. Chem. 2002, 45, 2615.