Discovery and development of a novel class of phenoxyacetyl amides as highly potent TRPM8 agonists for use as cooling agents

Article abstract of DOI:10.1016/j.bmcl.2017.04.003

The paper presents the activity trends for a novel series of phenoxyacetyl amides as human TRPM8 receptor agonists. This series encompasses in vitro activity values ranging from the micromolar to the picomolar levels. Sensory evaluation of these molecules highlights their relevance as cooling agents for oral applications. The positive outcome of the complete evaluation of N-(1H-pyrazol-3-yl)-N-(thiophen-2-ylmethyl)-2-(p-tolyloxy)acetamide resulted in its approval for Generally Recognized As Safe (GRAS) status by the Flavor & Extract Manufacturer Association (FEMA) as FEMA 4809.

Full text of DOI:10.1016/j.bmcl.2017.04.003

Accepted Manuscript
Discovery and development of a novel class of phenoxyacetyl amides as highly
potent TRPM8 agonists for use as cooling agents.
Alain Noncovich, Chad Priest, Jane Ung, Andrew P. Patron, Guy Servant, Paul
Brust, Nicole Servant, Nathan Faber, Hanghui Liu, Nicole S. Gonsalves, Tanya
L. Ditschun
PII:
S0960-894X(17)30366-9
DOI:
http://dx.doi.org/10.1016/j.bmcl.2017.04.003
BMCL 24856
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
1 December 2016
28 March 2017
1 April 2017
Please cite this article as: Noncovich, A., Priest, C., Ung, J., Patron, A.P., Servant, G., Brust, P., Servant, N., Faber,
N., Liu, H., Gonsalves, N.S., Ditschun, T.L., Discovery and development of a novel class of phenoxyacetyl amides
as highly potent TRPM8 agonists for use as cooling agents., Bioorganic & Medicinal Chemistry Letters (2017), doi:
http://dx.doi.org/10.1016/j.bmcl.2017.04.003
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Discovery and development of a novel class of phenoxyacetyl amides as highly potent
TRPM8 agonists for use as cooling agents.
Alain Noncovich, Chad Priest,* Jane Ung, Andrew P. Patron, Guy Servant, Paul Brust,
Nicole Servant, Nathan Faber, Hanghui Liu, Nicole S. Gonsalves and Tanya L. Ditschun.
Senomyx Inc., 4767 Nexus Centre Dr, San Diego, CA, 92121
*Corresponding author. Tel.: +1 858 646 8376; fax: +1 858 404 0752; e-mail: chad.priest@senomyx.com
The perception of temperature is regulated by a sophisticated biological system centered on a
series of temperature sensitive ion channels collectively known as thermosensitive transient
receptor potential (thermo-TRP) ion channels.1 These thermo-TRP channels can be gated either
by temperature or by a ligand triggering the response generally known as chemesthesis, a
phenomena including the perception of pain, touch and temperature. Chemesthesis occurs when
the channel is opened in response to a ligand, resulting in a sensation of temperature that has no
actual physical basis but it is exclusively associated to a chemical event, namely a small
molecule-macromolecule interaction.2
Figure 1. Schematic representation of the TRPM8 channel.13
TRPM8 is the ion channel responsible for the cool perception, which is triggered either by
temperatures below 25 °C or by ligands such as (-)-menthol (1) (Figure 2). The structures and
key features of TRPM8 are depicted in Figure 1. Wilkinson Sword Ltd developed the first
synthetic analogs of menthol in the 1970’s for use in personal hygiene products,3 when more

efficacious cooling agents were needed to improve on the relatively weak potency of (-)-
menthol, as well as its undesired physicochemical properties (poor water solubility and
volatility). The result of this work brought to the market WS-3 (2), still a widely used coolant,
that is 1.5 to 5 times more potent than menthol in producing a cooling sensation in humans.4 A
review of the patent literature suggests that the search for coolants of improved properties has
remained a very lively field for the past several decades.5 However, the majority of this work
has still employed a menthane moiety as the privileged scaffold in the search for more potent
cooling agents, with the exception of its replacement with alternative bulky and highly lipophilic
groups, such as the 2-isopropyl-2,3-dimethylbutanoate found in WS-23 (3)3.
H
H
N
N
O
4
OH
OH
O
O
2
1
3
WS-3
FEMA 3455
(1975)
(-)-menthol
FEMA 2665
(1965)
Coolact 5
FEMA 4154
(2005)
WS-23
FEMA 3455
(1996)
HO
O
O
H
H
HN
N
N
N
O
O2N
O
OH
O
CN
5
Cubebol
FEMA 4497
(2009)
6
7
8
Evercool 180
FEMA 4496
(2009)
Questice
FEMA 4155
(2005)
icilin
not GRAS approved
Figure 2. Sample of TRPM8 known agonists, with commercial name, FEMA number and year
of granted GRAS status, when available.
The lead compound 96 for the work described in this publication emerged from a high-
throughput screening (HTS) campaign of TRPM8 on the Senomyx corporate library (the
employed screening platform has been described elsewhere).7 This compound showed high
potency in our cell-based, functional assay as a TRPM8 activator, and was a novel structural
chemotype with significant potential for chemical modification to improve on activity and
physicochemical properties (see Table 2). The details of the preparation of these compounds
have already been previously reported.6 Scheme 1 depicts the synthetic routes that were utilized
in the preparation of analogs of compound 9.

R3
HN
R4
, iii
O
O
O
R3
ii
i
R3
O
O
O
+
HN
R4
Ar1
Ar1
Ar1
N
R4
OH
Cl
R1 R2
R1 R2
R1 R2
O
O
R3
iv
v
R3
OH
X
X
+
HN
R4
+
Ar1
LG
N
R4
R1 R2
R1 R2
X = Cl or Br
X = Cl or Br
Scheme 1.
General synthetic approaches for the synthesis of the described
phenoxyacetylamides. Reagents and conditions: (i) oxalyl chloride, DMF (cat.), DCM, 0 °C to rt;
(ii) TEA, DCM, 0 °C to rt; (iii) EDCI, HOBt, DCM, DMF, MW, 100 °C, 10 min; (iv) NaHCO3,
DMF, rt, 18 h; (v) K2CO3, DMF, MW, 120 °C, 10 min.
Initially we explored the possibility of replacing the 2-pyridyl fragment. This fragment of the
molecule was amenable to exchange with simple alkyl groups, as well as with other aromatic and
heteroaromatic fragments. In general, these substitutions lead to agonists with improved potency,
as summarized in Tables 1 and 2. Since the absolute EC50 values in our cell-based assay were
found to be subject to cell batch-to-batch fluctuations, we also compared EC50 ratio to that of
WS-3 as a reference in each screening plate. The effectiveness of these compounds as cooling
agents was assessed by determining a concentration that would deliver an equivalent cooling
sensation to that of 45 µM of WS-3 as judged by a group of trained panelists.8
The optimal size for maximum potency for the alkyl chain was found to be the ethyl (12),
although Table 1 shows four analogs with similar levels of activity (13, 16, 17, 26). Among the
heterocyclic fragments studied (Table 2), the oxazoles (32, 33) were found to be among the most
potent, but proved to be hydrolytically unstable.11 While 3 and 4-pyrazolyl amides (35, 36) also
exhibited similar high potency, only 35 resulted in a good in-vitro to in-vivo sensory correlation.
The N-methylated pyrazolyl amides showed significantly reduced activity, while C-methylated
pyrazolyl amides were well tolerated. Because of the ease of synthesis, good in-vitro to in-vivo
sensory8 correlation, and high levels of activity for the compound containing the ethyl fragment
(12), the phenoxyacetyl portion was optimized while keeping this residue constant.

TRPM8 Agonist Activity 7
Sensory data 8
Compound
R
EC50 (nM)
WS-3 EC50 Ratio
10
11
12
13
14
15
16
17
18
19
H
25,000
400
8
0.1
15
Methyl
Ethyl
620
149
66
5 μM cmp ~ 45 μM WS-3
5 μM cmp < 45 μM WS-3
Propyl
i -Propyl
c-Hexyl
Allyl
40
70
40,000
20
0.1
365
577
2
5 μM cmp ~ 45 μM WS-3
5 μM cmp ~ 45 μM WS-3
c-Propyl
s-Butyl
t-Butyl
8
2,000
11,000
0.4
20
21
22
23
23,000
900
0.3
5
3,000
400
1
11
24
25
i-Butyl
900
3
OH
8,000
0.4
26
10
317
Table 1. TRPM8 activity for N-alkyl analogs.

TRPM8 Agonist Activity 7
Sensory data 8
Compound
R
EC50 (nM)
WS-3 EC50 Ratio
9
2-Pyridyl
3-Pyridyl
4-Pyridyl
Pyrazinyl
2-Pyrimidinyl
Phenyl
200
198
25
24
77
3
15 μM cmp ~ 45 μM WS-3
15 μM cmp ~ 45 μM WS-3
27
28
29
30
31
52
1,235
3,700
23
1
197
15 μM cmp ~ 45 μM WS-3
32
33
34
35
36
37
38
39
40
41
42
2
0.3
3,444
119,614
2
2,240
0.2
35,810
3,820
14
3 μM cmp ~ 45 μM WS-3
5 μM cmp < 45 μM WS-3
1.4
5 μM cmp ~ 0 μM WS-3*
*significant bitterness
422
63,000
1,786
0.3
0.1
3
14,725
10,731
13
0.4
5 μM cmp < 45 μM WS-3
855
Table 2. TRPM8 activity for N-aryl analogs.

Systematic modification of the aromatic ring substituents identified the para position as optimal
for high levels of activity, as exemplified by compounds 12, 44, and 45 in Table 3. While other
substituents were also compatible with good activity, the 4-methyl analog was still the most
potent. Bridging the 3- and 4-position of the aromatic ring to form a 5-membered carbocyclic
ring (e.g., 56), resulted in a further improvement in activity.
TRPM8 Agonist Activity 7
R1
R2
R3
Sensory data 8
Compound
EC50 (nM)
1,400
192
WS-3 EC50 Ratio
43
44
45
12
46
47
48
49
50
51
52
53
54
55
56
57
58
4
18
42
620
66
3
Me
Me
97
5 μM cmp ~ 45 μM WS-3
5 μM cmp ~ 45 μM WS-3
5 μM cmp ~ 45 μM WS-3
Me
Et
8
67
Pro
1,813
397
i -Pro
CF3
10
18
32
0.2
2
236
MeO
CH2OH
CN
120
15 μM cmp ~ 45 μM WS-3
21,870
1,985
inactive
422
COOH
F
10
93
Cl
39
-CH2CH2CH2-
0.8
7,684
7
3 μM cmp ~ 45 μM WS-3
5 μM cmp ~ 45 μM WS-3
-CH2(CH2)2CH2-
-CH2OCH2-
689
61
58
Table 3. TRPM8 activity for O-aryl analogs.
The modification of the -OCH2- linking group between the phenyl ring and the amide carbonyl
was also evaluated. Activity was significantly reduced by either mono- or di-methylation of the
phenoxyacetic acid moiety as shown in examples 59 and 60 in Table 4. Switching to a double
bond gave a boost in activity (61), while analogs with an ethylene linkage (62) showed reduced
activity. Fusing the -OCH2- linkage with the neighboring aromatic ring resulted in a constrained
benzofuran analog 65 which still exhibited significant activity on TRPM8. Other constrained
analogs involving this linkage were inactive (69, 70).

TRPM8 Agonist Activity 7
EC50 (nM) WS-3 EC50 Ratio
Sensory data 8
Compound
Structure
12
59
8
620
4
5 μM cmp ~ 45 μM WS-3
889
60
61
62
63
64
65
66
67
68
10,988
2
0.3
2,345
40
3 μM cmp ~ 45 μM WS-3
5 μM cmp < 45 μM WS-3
93
52,023
inactive
120
0.1
42
2
10 μM cmp ~ 45 μM WS-3
2,005
inactive
1,603
3
69
70
weak
inactive
Table 4. TRPM8 activity for different linkers.
Replacement of the thiophene moiety was also explored; the results of these efforts are
summarized in Table 5. It was not possible to find a fragment with improved activity, however
the corresponding 3-thiophenyl and 2- and 3-furanyl analogs (75, 76, 77) maintained comparable
potency both in vitro and/or in vivo.

TRPM8 Agonist Activity 7
EC50 (nM) WS-3 EC50 Ratio
R1
R2
Sensory data 8
Compound
12
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
2-thiophenyl
5-methyl-2-thiophenyl
2-thiophenyl
2-thiazolyl
8
620
18
5 μM cmp ~ 45 μM WS-3
178
Me
inactive
3,062
1,102
88
2
4
5-thiazolyl
3-thiophenyl
2-furanyl
49
9
5 μM cmp ~ 45 μM WS-3
15 μM cmp ~ 45 μM WS-3
5 μM cmp ~ 45 μM WS-3
489
3-furanyl
136
32
0.3
0.2
2-N-methyl-pyrrolyl
3-pyrrolyl
21,305
21,752
weak
inactive
31,146
1,876
15,955
15,445
inactive
4-pyrazolyl
3-pyrazolyl
1,2,3-triazolyl
Phenyl
0.1
2
2-pyridyl
0.2
0.2
3-pyridyl
4-pyridyl
Table 5. TRPM8 activity for 2-thiophenyl surrogates.
Finally, by combining the various optimized fragments, several extremely potent TRPM8
agonists were synthesized as depicted in Table 6. It was found that in order to reach EC50 values
in the picomolar range, it was essential to include the 3-pyrazolyl fragment (90, 91, and 93).
This unprecedented level of in vitro potency and their non-menthane chemotype make this class
of compounds unique among the various TRPM8 modulators.
It is interesting to note that the 1H-NMR for pyrazole containing analogs, such as compound 35,
generally displayed only one set of signals at room temperature,14 while the same experiment for
compound 12 for example, displayed two sets of signals for the two possible rotamers.15
Compound 70 could be seen as a conformationally constrained analog of 12 (ethyl group trans to
the carbonyl), which exhibited a complete loss of agonistic activity. Similarly, when replacing
the ethyl group for a tert-butyl for analog 19 led to a weakly active compound. It has been
shown that N-tert-butyl substitution locks the conformation of the amide where the alkyl group is

cis to the carbonyl group.16 This observation might suggests that neither of these two rotamers is
the active one for these agonists, but likely one where the N-substituents are not parallel to the
carbonyl group. Further investigation is needed to better understand this aspect for this type of
TRPM8 agonists.
TRPM8 Agonist Activity 7
Sensory data 8
Compound
Structure
EC50 (nM)
WS-3 EC50 Ratio
87
88
89
24
180
5 μM cmp ~ 45 μM WS-3
3 μM cmp ~ 45 μM WS-3
8
650
0.7
6,834
O
N
S
90
91
92
93
94
95
96
97
0.004
0.002
0.6
>1,000,000
>1,000,000
6,336
3 μM cmp ~ 45 μM WS-3
HN
N
0.001
3.7
>1,000,000
1,143
3 μM cmp ~ 45 μM WS-3
5 μM cmp ~ 45 μM WS-3
5 μM cmp ~ 45 μM WS-3
5 μM cmp ~ 45 μM WS-3
4 μM cmp ~ 45 μM WS-3
80
164
1.1
30,944
40
>1,000,000
Table 6. TRPM8 activity for some analogs combining the optimal identified fragments.

Although in the in vitro assay these compounds show a ratio of potencies ranging from hundreds
to more than a 106 times the EC50 of WS-3, their cooling efficacy in sensory evaluation tended to
plateau at an approximate 15-fold improvement over WS-3 in a buffered water solution. For
example, 56 is 7,684 times as potent as WS-3 in vitro, while compound 93 is greater than a 106
times as potent as WS-3 in vitro. When 56 and 93 were tested in sensory at 3 µM, they showed
similar cooling intensity to that of a solution of WS-3 at 45 µM. Since TRPM8 is mainly
expressed in subcutaneous nerve endings connected to the trigeminal nerve,1 one possible
explanation for this lack of correlation with in vitro potency could be differences in the ability to
permeate the tissues of the oral cavity. In an attempt to evaluate this hypothesis, a representative
subset of these compounds was profiled in a Parallel Artificial Membrane Permeability Assay
(PAMPA). These data are shown in Table 7. In general, this subset of compounds showed
relatively poor membrane permeability (log Pe < -5.00), while only 58 exhibited good
permeability in the 2 hour interval (log Pe = -4.65). While these data might account for the
equivalent cooling intensity of 58 in the sensory study compared to 12 (which 10 times more
potent in vitro), the PAPMA data is unable to address the lack of correlation among other pairs
of compounds, such as 56 and 93. This suggests that factors other than membrane permeability
are responsible for the poor correlation between in vitro potency and cooling effect in sensory, or
simply the inadequacy of the PAMPA experiment in simulating permeability through the buccal
epithelia.
PAMPA 2h
TRPM8 Agonist Activity 7
WS-3 Ratio
Sensory data 8
Solubility (μM) 12
Compound
ClogP
Isoequivalent to 45 μM WS-3
transport (%)
log Pe
-5.12
-5.20
-5.82
-4.65
-5.82
-5.55
-5.56
-6.29
12
35
56
58
65
90
93
95
3.6
2.8
4.1
3.2
4.0
3.7
3.3
3.7
545
189
85
620
35,810
7,684
12
10
3
5 μM
3 μM
3 μM
5 μM
10 μM
3 μM
3 μM
5 μM
571
27
58
33
2
42
3.7
26
>1,000,000
>1,000,000
164
5
5
49
1
Table 7. In vitro and sensory TRPM8 activity vs PAMPA.
Several of these potent analogs were evaluated as potential product candidates for commercial
application. Important factors for commercial applications include solubility, sensory profile,
TRPM8 receptor selectivity and lack of toxicity. Solubility was important for formulation
considerations; while these compounds are active at low levels (below 5 µM) many product
formulations require ten times or more higher concentration. Sensory profile considerations
included clean, non-minty taste, long lasting, and temporal profile (i.e., fast verses delayed
onset). When tested in sensory as a coolant by a trained panel,8 a 3 μM aqueous solution of 35
showed an equivalent cooling intensity to that of a 45 μM solution of WS-3. No off-tastes were

noted and 35 was described as a clean coolant by the panelists. In addition, 35 produced a
significantly longer cooling sensation relative to WS-3 as rated by the panelists in an evaluation
of the temporal profile of both coolants. The cooling effect of a 3 μM aqueous solution of 35
was still perceived out to at least 60 minutes while that of a 45 μM aqueous solution of WS-3
was no longer apparent at 25 minutes. Based on our in-house selection criteria, compound 35
was chosen for development. The complete safety assessment for 35 has recently been
completed,9 and as a result of its high margin of safety, 35 has been determined to be the
Generally Recognized As Safe (GRAS) under the conditions of intended use as a flavor
ingredient by the Flavor & Extract Manufacturer Association of the United States (FEMA) and
therefore is available for use in human food in the United States as a “FEMA GRAS” flavor
ingredient. Compound 35 was assigned FEMA GRAS Number 4809 in 2014.10
In summary, we have reported a new class of highly potent TRPM8 agonists, validating their
activity in vitro and as coolants. Compound 35 from the series was developed as a cooling
agent, and was recently granted GRAS status as FEMA 4809.
References and notes.
1. Vriens, J.; Nilius, B.; Voets, T. Nat. Neurosci. 2014, 15(9), 573.
2. Appendino, G.; Minassi, A.; Pagani, A; Ech-Chahad, A. Curr. Pharm. Des. 2008, 14(1),
2.
3. Watson, H. R.; Hems, R.; Roswell, D. G.; Spring, D. J. J. Soc. Cosmet. Chem. 1978, 29,
185.
4. (a)
Erman,
M.
Perfumer
&
Flavorist
2004,
29(8),
34;
(b)
http://www.leffingwell.com/cooler_than_menthol.htm
5. (a) Furrer, S. M. WO Patent 2010/128026; (b) Surburg, H.; Backes, M.; Oertling, H;
Machinek, A.; Loges, H.; Simchen, U.; Subkowski, T.; Bollscheiler, C.; Wittenberg, J.;
Siegel, W. WO Patent 2011/061330; (c) Kazimierski, A.; Dewis, M. L. US Patent
2011/0195032; (d) Komatsuki, Y.; Yamamoto, T.; Aida, T.; Ishida, K. US Patent
2011/0081393.
6. Priest, C.; Noncovich, A.; Patron, A. P.; Ung, J. WO Patent 2012/061698.
7. Servant, G.; Brust, P.; Moyer, B.; Lu, M.; Echeverri, F.; Dahan, D.; Zoller, M.; Williams,
M.; Kimmich, R.; Kamdar, P.; Ditschun, T.; Patron, A. P. US Patent Application
2007/0259354.
8. Compounds are rated by trained panelists on a 16 point line scale (0 = no cooling and 15
= strongest cooling) where 45 µM WS-3 is ranked as a 5 in cool intensity. In most cases
compounds are tested to determine at what concentration the cooling intensity is
equivalent to 45 µM WS-3. In each test, the panelist is presented with a 0 µM control
sample, a 45 µM WS-3 control sample and the experimental compound sample and asked
to rate the cooling intensity of each sample. Panelists are also asked to rate bitterness
and/or additional side tastes. All samples are made with Low Sodium Buffer (LSB) pH

~7.1 and contain 0.1% ethanol. The ratings are then collected and analyzed and
statistically analyzed.
9. Karanewsky, D. S.; Arthur A. J.; Liu, H.; Chi, B.; Markison, S. Toxicol. Rep. 2015, 2,
1291.
10. Cohen, S. M.; Fukushima, S.; Gooderham, N. J.; Hecht, S. S.; Marnett, L. J.; Rietjens, I.
M. C. M.; Smith, R. L. GRAS flavouring substances 27: the 27th publication by the expert
panel of the flavour and extract manufacturers association provides an update on recent
progress in the consideration of flavouring ingredients generally recognized as safe under
the Food Additive Amendment, Food Technol., 2015, 69 (8), 40 (key findings of the
FEMA Expert Panel GRAS determinations for each substance on the GRAS 27 list are
available online: www.femaflavor.org).
11. Compounds 32 and 33 were evaluated for thermal stability by measuring the compound
remaining after heating at 100⁰C for 24 hours in a buffered solution. The solutions were
pH 2.8 in 50 mM citric acid, pH 4.0 in 50 mM citric acid, and pH 7.1 in 50 mM
phosphoric acid.
12. Solubility is the thermodynamic value measured at room temperature in a 50 mM
phosphoric buffer at pH 7.1.
13. Adapted from Bandell, M.; Dubin, A. E.; Petrus, M. J.; Orth, A.; Mathur, J.; Hwang, S.
W.; Patapoutian, A. Nat. Neurosci 2006, 9 (4), 493.
14. 1H NMR (400 MHz, DMSO-d6) δ 2.21 (s, 3H), 4.59 (br s, 2H), 4.98 (br s, 2H), 6.23 (br
s, 1H), 6.67 (br d, J = 8.5 Hz, 2H), 6.91 (br m, 2H), 7.04 (br d, J = 8.1 Hz, 2H), 7.41 (m,
1H), 7.78 (br s, 1H), 12.88 (br s, 1H).
15. 1H NMR (400 MHz, DMSO-d6, T = 80⁰C) δ 1.11 (br s, 3H), 2.24 (s, 3H), 3.37 (q, J = 7.1
Hz, 2H), 4.71 (br s, 2H), 4.76 (s, 2H), 6.82 (d, J = 8.5 Hz, 2H), 6.97 (br s, 1H), 7.04 (br s,
1H), 7.07 (d, J = 8.5 Hz, 2H), 7.40 (br s, 1H).
16. Roy, O.; Caumes, C.; Esvan, Y.; Didierjean, C.; Faure, S.; Taillefumier, C. Org. Lett.
2013, 15 (9), 2246.

Graphical abstract