2-Arylimino-5,6-dihydro-4H-1,3-thiazines as a new class of cannabinoid receptor agonists. Part 1: Discovery of CB2 receptor selective compounds

Article abstract of DOI:10.1016/j.bmcl.2007.04.093

2-Arylimino-5,6-dihydro-4H-1,3-thiazines have been identified as a novel class of cannabinoid agonists. A lead structure with moderate activity was discovered through a high throughput screening assay. Structure-activity relationships led to the discovery of potent agonists of CB₂ receptor. The most potent compound 13 displays K_i values of >5000 and 9 nM to CB₁ and CB₂ receptors, respectively.

Full text of DOI:10.1016/j.bmcl.2007.04.093

Bioorganic & Medicinal Chemistry Letters 17 (2007) 4030–4034
2-Arylimino-5,6-dihydro-4H-1,3-thiazines as a new class
of cannabinoid receptor agonists. Part 1: Discovery of CB2
receptor selective compounds
Hiroyuki Kai,a,* Yasuhide Morioka,b Takami Murashi,a Koichi Morita,a
Satomi Shinonome,a Hitoshi Nakazato,a Keiko Kawamoto,a Kohji Hanasaki,a
Fumiyo Takahashi,a Shin-ichi Mihara,a Tohko Arai,b Kohji Abe,b Hiroshi Okabe,b
Takahiko Baba,b Takayoshi Yoshikawaa and Hideyuki Takenakaa
aShionogi Research Laboratories, Shionogi & Co., Ltd., 12-4, Sagisu, 5-chome, Fukushima-ku, Osaka 553-0002, Japan
bDevelopmental Research Laboratories, Shionogi & Co., Ltd., 1-1, Futaba-cho 3-chome, Toyonaka, Osaka 561-0825, Japan
Received 13 February 2007; revised 19 April 2007; accepted 25 April 2007
Available online 30 April 2007
Abstract—2-Arylimino-5,6-dihydro-4H-1,3-thiazines have been identified as a novel class of cannabinoid agonists. A lead structure
with moderate activity was discovered through a high throughput screening assay. Structure–activity relationships led to the discov-
ery of potent agonists of CB2 receptor. The most potent compound 13 displays Ki values of >5000 and 9 nM to CB1 and CB2 recep-
tors, respectively.
ꢀ 2007 Elsevier Ltd. All rights reserved.
The pharmacological activity of cannabinoids is
known to be mediated by two cannabinoid receptors
(CB1 and CB2).1 The CB1 receptor discovered in
1990 was mainly distributed in components of the cen-
tral nervous system, such as the brain.2 Agonists of
this receptor were found to suppress the release of
neurotransmitters and to cause some central actions
such as euphoria or analgesia. The CB2 receptor dis-
covered in 1993 was mainly distributed in immune tis-
sues, such as the spleen and tonsil.3 It has been
reported that both agonists and antagonists of this
receptor have anti-inflammatory activity, and only
agonists, but not antagonists, have analgesic activity.
Therefore, selective agonists4 to the CB2 receptor were
expected to be anti-inflammatory agents and analgesic
agents that would not cause side effects associated
with CB1 receptors, such as euphoria or drug
dependence.
To discover novel cannabinoid receptor (CBR) agonists
with different pharmacophores from conventional can-
nabinoids,5 we screened our compound libraries. High-
throughput screening assay resulted in the discovery of
compound 1, which exhibited modest activities to
CBR (binding affinity, CB2 Ki 428 nM; CB1 Ki
>5000 nM). In attempts to improve the binding affinity
of compound 1, a series of 2-arylimino-5,6-dihydro-4H-
1,3-thiazines have been synthesized with various substit-
uents to find the optimal structure exhibiting the stron-
gest binding affinity. In this paper, we describe the
synthesis and the structure–activity relationships of 2-
phenylimino-5,6-dihydro-4H-1,3-thiazines as cannabi-
noid receptor agonists (Fig. 1).
The 2-phenylimino-5,6-dihydro-4H-1,3-thiazines 5 were
prepared as outlined in Scheme 1.6 Phenylthioureas 4
Keywords: 1,3-Thiazine; Cannabinoid receptor agonist; CB2 receptor
selective; CB1; CB2; Pain; Structure–activity relationship.
*
Corresponding author. Tel.: +81 6 6458 5861; fax: +81 6 6458
0987; e-mail: hiroyuki.kai@shionogi.co.jp
Figure 1. Lead structure found in a high throughput screening assay.
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.04.093

H. Kai et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4030–4034
4031
Scheme 1. Reagents and conditions: (a) CSCl2, Et3N, CH2Cl2, rt 1 h; (b) 3-aminopropanol, CH2Cl2, rt 2–8 h; (c) 35% HClaq, reflux 1–3 h.
were prepared by the reaction of anilines 2 with thio-
phosgene in the presence of triethylamine followed by
reaction with 3-amino-1-propanol. Thiazines 5 were pre-
pared by cyclization of the phenylthioureas 4 with 35%
hydrochloric acid.
First, the ethoxycarbonyl moiety of the thiazine ring of
compound 1 was modified structurally to attain more
potent affinity. Table 1 shows the effects of substituent
(R1) on 3-position of the thiazine ring on binding affin-
ity. The reaction precursor (5a) of 1 was inactive. When
the oxygen atom of the ethoxycarbonyl group of 1 was
replaced by a sulfur atom, the affinity increased (7, 11,
and 12). Alkyl (ethyl, 6; propyl, 16; benzyl, 17) and pro-
pionyl (10) derivatives exhibited extremely weak affinity.
Ethylcarbamoyl (8) and ethylthiocarbamoyl (9) deriva-
tives were inactive. Among the (alkylthio)thiocarbonyl
derivatives (7, 13–15), the (methylthio)thiocarbonyl
derivative (13) was the most active, whereas substituents
larger than the methyl group (ethyl, 7; propyl, 14; iso-
propyl, 15) reduced the affinity. These results suggested
that the physicochemical characteristics or hydrophobic
influence of the sulfur atom into a substituent at 3-posi-
tion might lead to enhancement of the affinity, but
no definitive explanation for this observation is yet
available.
The general methods for synthesis of 3-substituted 2-
phenylimino-5,6-dihydro-4H-1,3-thiazines are shown in
Scheme 2. Compound 1 was prepared by the reaction
of 5a with ethyl chloroformate in the presence of trieth-
ylamine (Method A). Compound 6 was prepared by the
reaction of 5a with iodoethane in the presence of sodium
hydride (Method B). Compound 7 was prepared by the
reaction of 5a with carbon disulfide in the presence of
sodium hydride followed by methylation with iodoe-
thane (Method C). Compound 8 was prepared by the
reaction of 5a with ethyl isocyanate (Method D). Com-
pound 9 was prepared by the reaction of 8 with Lawes-
son’s Reagent (Method E). The thiazine analogues (10–
34) were prepared using method A, B or C.
Human CB2 receptor binding affinities7 of 2-phenylimi-
no-5,6-dihydro-4H-1,3-thiazines are shown in Tables 1–3.
Table 2 shows the effect of substituents on the phenyl-
moiety of 5,5-dimethyl-3-(methylthio)thiocarbonyl-2-
N
N
S
S
6
Method B
O
S
OEt
S
S
SEt
H
N
N
N
N
N
N
N
Method A
Method C
1
5a
7
Method D
S
S
NHEt
O
S
NHEt
N
Method E
N
N
9
8
Scheme 2. Reagents and conditions: (method A) EtOCOCl, Et3N, THF, rt 2 h, 68%; (method B) EtI, NaH, DMF, rt 2 h, 71%; (mehtod C) CS2,
NaH, DMF, 0ꢁC 0.5 h, then EtI, 0ꢁC 1 h, 70%; (method D) EtNCO, THF, rt 1 h, 87%; (method E) Lawessons Reagent, pyridine, 80ꢁC 2 h, 65%.

4032
H. Kai et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4030–4034
Table 1. Lead optimization study of R1 part
phenylimino-5,6-dihydro-4H-1,3-thiazines. When the
isopropyl group at 2-position of compound 13 was
changed to 3- or 4-position removed, the affinity was re-
duced (18–20). Among the 2-substituted derivatives (13,
18–34), the isopropyl derivative (13) was the most active,
whereas bulky groups (t-Bu; 24, Ph; 31, PhO; 32) or elec-
tron-withdrawing groups (F; 25, Cl; 26, Br; 26, CF3; 30)
reduced the affinity. Furthermore, the 2,6-dimethyl
derivative (33) was inactive. These findings suggested
that the presence of a hydrophobic group as well as a
steric effect of the substituent at 2-position on the phenyl
moiety were responsible for increasing affinity.
Compound
R1
Method
h-CB2 Ki (nM)
1
10
11
12
7
COOEt
COEt
A
A
A
A
C
C
C
C
D
E
–
428
2090
48
COSEt
CSOEt
CSSEt
CSSMe
CSSPr
CSS-i-Pr
CONHEt
CSNHEt
H
230
35
13
14
15
8
9
119
Since 5,5-dimethyl-3-(methylthio)thiocarbonyl-2-phenyli-
mino-5,6-dihydro-4H-1,3-thiazines were found to have
favorable affinity for the human CB2 receptor, the di-
methyl moiety of the thiazine ring of compound 13
was modified structurally. Table 3 shows the effects of
substituent (R3) on 4-, 5- or 6-position of the thiazine
ring on binding affinity. Among them, the 5,5-dimethyl
derivative (13) showed the greatest affinity, followed
by the 5-methyl derivative (36). On the other hand,
hydrogen (34), 4-methyl (35) and 6-methyl (37) deriva-
tives were weak or had no activity. These results
revealed that the methyl groups at 5-position of the thi-
azine ring of 3-(methylthio)thiocarbonyl-2-phenylimino-
5,6-dihydro-4H-1,3-thiazines seem to play a significant
role in raising affinity.
275
>5000
>5000
>5000
4880
915
9
5a
6
Et
Pr
B
B
B
16
17
CH2Ph
1360
Table 2. Lead optimization study of R2 part
Based on the findings described above, compound 13
was selected for further evaluation (Table 4). The affin-
ity of compound 13 for CB2 receptors was high, and its
potency (Ki value) did not differ between the two animal
species. Also, compound 13 exhibited extremely weak
affinities against CB1 receptor, and higher CB2 selectiv-
ity than WIN55212-2. Both CB1 and CB2 are coupled
with Gi proteins and known to reduce the cyclic AMP
concentration in cells upon receptor stimulation.8 Com-
pound 13 showed the inhibitory activity of cyclic AMP
production in the assay by using CHO cells expressing
human CB2.9 The compound showed a fast clearance
rate of 83.3 mL/min/kg, a short half life (t1/2 = 0.6 h),
and low oral bioavailability of 18%. Unfortunately,
the pharmacokinetics profile of this compound was
not suitable in a rat in vivo examination.
Compound
R2
CB2 Ki (nM)
7
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
2-i-Pr
3-i-Pr
4-i-Pr
H
9
33
22
316
202
54
2-Me
2-Et
2-Pr
57
2-t-Bu
2-F
2-C1
313
252
722
371
18
2-Br
2-MeO
2-EtO
2-CF3
2-Ph
84
141
989
2920
>5000
2-PhO
2,6-Me2
Compound 13 showed analgesic activity10 when admin-
istered subcutaneously to the pain site, and caused sig-
nificant inhibition of both early and late phases of
formalin-induced licking when co-injected with forma-
lin. These effects were reversed by SR144528 (CB2
antagonist), but not by SR141716A (CB1 antagonist)
(Fig. 2). In addition, SR144528 by itself did not affect
analgesic activity (data not shown). These findings sug-
gest that compound 13 has analgesic activity due to acti-
vation of CB2.
Table 3. Lead optimization study of R3 part
Compound
R3
CB2 Ki (nM)
In conclusion, the structure–activity relationships of a
lead structure found in high throughput screening led
to the discovery of new cannabinoid CB2 selective ago-
nists with approximately 500-fold selectivity in potency.
Compound 13 showed an analgesic effect.11 This novel
series of cannabinoid agonists should be useful for
7
34
35
36
37
5,5-Me2
H
9
1581
>5000
98
4-Me
5-Me
6-Me
>5000

H. Kai et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4030–4034
Table 4. In vitro data and pharmacokinetic profiles in rat of selected compounds
4033
Compound
In vitro assay (nM)
Rat PK (iv: 2, po: 30 mg/kg)
h-CB2
(Ki)
h-CB1
(Ki)
m-CB2
(Ki)
m-CB1
(Ki)
cAMP
(CB2, IC50
cAMP
(CB1, IC50
T1/2
CLt
(mL/min/kg)
AUC
(po) (lg-h/mL)
BA
(%)
)
)
(h)
1
13
428
9
>5000
>5000
14
187
9
>5000
2020
7
322
6.1
1.1
>5000
>5000
17
NT
1.1
NT
0.6
83.3
18
WIN55212-2
2.2
6
Figure 2. Analgesic effects of compound 13 reversed by SRI44528 The CB1 selective antagonist SRI41716A (SRI, 0.3 mg/kg) or CB2 selective
antagonist SR144528 (SR2, 0.3 mg/kg) was administered iv 1 min prior to compound 13 administration, compound 13 (500 lg/site) was dissolved in
20 lL of formalin solution (2% in saline), and sc-injected into the dorsal surface of the right hindpaw of mice. *p < 0.05, **p < 0.01 (vs
vehicle + vehicle), ##p < 0.01 (vehicle + compound 13), V: vehicle, 13: compound 13, n = 6–8.
at 4 ꢁC. The supernatants were centrifuged at 100,000g for
60 min at 4 ꢁC. The pelleted membrane fractions were
homogenized in the membrane buffer and stored in a deep
freezer (ꢀ80 ꢁC). The Kd values of [3H]-CP55940 for each
membrane fraction were determined by Scatchard plot
analysis.
characterizing the functions of cannabinoid receptors
and evaluating their potential as a new class of analgesic
drugs.
References and notes
8. Felder, C. C.; Joyce, K. E.; Briley, E. M.; Mansouri, J.;
Mackie, K.; Blond, O.; Lai, Y.; Ma, A. L.; Mitchell, R. L.
Mol. Pharmacol. 1995, 48, 9. 443.
1. Pertwee, R. G. Curr. Med. Chem. 1999, 6, 129.
2. Matsuda, L. A.; Lolait, S. J.; Brownstein, M. J.; Young,
A. C.; Bonner, T. I. Nature 1990, 346, 561.
3. Munro, S.; Thomas, K. L.; Abu-Shaar, M. Nature 1993,
365, 61.
9. Cyclic AMP assay method: The CHO cells expressing
h-CB1 or h-CB2 were seeded into 24-well plates. The cells
were incubated at 37 ꢁC for 20 min with compounds in the
cAMP assay buffer (Hanks’ solution with 20 mM Hepes,
0.1 mM IBMX, 0.2 mM Ro20-1724, and 0.1% BSA). The
cells were stimulated with 4 lM forskolin at 37 ꢁC for
25 min (h-CB1 cells) or 45 min (h-CB2 cells). The cAMP
concentrations in the cells were measured using cAMP kits
(CIS Bio International).
10. Formalin test: Twenty microliter of formalin solution (2%
in saline) was injected subcutaneously into the dorsal
surface of the right hindpaw of mice (ICR). The total
amount time the mouse spent licking in the early phase
(acute pain, 0–5 min) and the late phase (inflammatory
pain, 10–30 min) was measured.
11. Experimental procedure for preparation of compound 13:
To a solution of 2-isopropylaniline (29.4 g, 217.5 mmol)
and triethylamine (44.0 g, 434.8 mmol) in dichlorometh-
ane (200 ml) was added dropwise under ice-cooling over
20 min thiophosgene (25.0 g, 217.5 mmol). The mixture
was stirred at room temperature for 1 h. The mixture was
poured into ice-cold water (500 ml) and then extracted
with ethyl ether (500 ml). The organic layer was washed
with brine (500 ml), dried over anhydrous magnesium
sulfate, and concentrated under reduced pressure to give
4. (a) Hanus, L.; Breuer, A.; Tchilibon, S.; Shiloah, S.;
Goldenberg, D.; Horowitz, M.; Pertwee, R. G.; Ross, R.
A.; Mechoulam, R.; Fride, E. Proc. Nat. Acad. Sci. U.S.A
1990, 96, 14228; (b) Whiteside, G. T.; Gottshall, S. L.;
Boulet, J. M.; Chaffer, S. M.; Harrison, J. E.; Pearson, M.
S.; Turchin, P. I.; Mark, L.; Garrison, A. E.; Valenzano,
K. J. Eur. J. Pharmacol. 2005, 528, 65.
5. Pertwee, R. G. Prog. Neurobiology 2001, 63, 569.
6. (a) Ichinari, M.; Sato, T.; Hayase, Y. Heterocycles 1988,
27, 227; (b) Tisler, M. Tetrahedron Lett. 1959, 1, 12.
7. Receptor binding assay method: Membrane; recombinant
human CB1 (h-CB1), CB2 (h-CB2), mouse spleen (m-CB2),
brain (m-CB1); radioligand [3H]-CP55940. Membrane
fractions, used for the measurement of binding activity,
were prepared as in the other study and stored in a deep
freezer (ꢀ80 ꢁC). In brief, confluent cultures of the h-CB1,
h-CB2, m-CB1, and m-CB2 cells were harvested. The
harvested cells were sonicated in a buffer for membrane
suspensions (membrane buffer: 20 mM Tris–HCl, pH 7.4,
2 mM EDTA, and 0.25 M sucrose containing protease
inhibitor) on ice, and centrifuged at 3000 rpm for 10 min

4034
H. Kai et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4030–4034
(2-isopropylphenyl)isothiocyanate (38.2 g, yield: 99%) as a
brown oil.
To solution of (2-isopropylphenyl)isothiocyanate
thiazine (22.5 g, yield: 76%) as a white crystals (mp 155–
157 ꢁC). 1H NMR (d ppm TMS/CDCl3 270 MHz) 1.15
(6H, s), 1.20 (6H, d, J = 6.7 Hz), 2.67 (2H, s), 3.09 (2H, s),
3.15 (1H, sept, J = 6.7 Hz), 6.88 (1H, m), 7.05–7.11 (2H,
m), 7.20 (1H, m).
a
(20.0 g, 112.8 mmol) in diethylether (30 ml) was added 3-
amino-2,2-dimethylpropanol (11.63 g, 112.8 mmol). The
mixture was stirred at room temperature for 1 hour and
concentrated under reduced pressure to give N-(2-isopro-
pylphenyl)-N0-(1-hydroxy-2,2-dimethyl)propylthiourea
(31.6 g, yield: 99%) as a yellow oil. 1H NMR (d ppm
TMS/CDCl3 270 MHz) 0.82(6 H, s), 1.25(6H, d,
J = 6.7 Hz), 3.11(1H, sept, J = 6.7 Hz), 3.25(2H, s),
3.55(2H, d, J = 6.3 Hz), 6.05(1H, m), 7.17–7.40 (4H, m).
A mixture of N-(2-isopropylphenyl)-N0-(1-hydroxy-2,2-
dimethyl)propylthiourea (31.6 g, 112.7 mmol) and concen-
trated hydrochloric acid (30 ml) was heated under reflux
with stirring for 3 hours. The reaction solution was cooled
to room temperature and poured into an aqueous solution
of 10% sodium hydroxide (100 ml). The mixture was
poured into water (400 ml) and extracted with chloroform
(100 ml · 2). The organic layer was washed with brine
(100 ml), dried over anhydrous magnesium sulfate, and
concentrated under reduced pressure. The crude product
was recrystallized from ethyl acetate and hexane to give 2-
(2-isopropylphenyl)imino-5,5-dimethyl-5,6-dihydro-4H-1,3-
To a mixture of 2-(2-isopropylphenyl)imino-5,5-dimethyl-
5,6-dihydro-4H-1,3-thiazine (3.15 g, 12.0 mmol), carbon
dioxide (1.08 ml, 18.0 mmol), and N,N-dimethylformam-
ide (36 ml), 60% sodium hydride (0.72 g, 18 mmol) was
added under ice-cooling. The mixture was stirred for 1 h.
Methyliodide (0.17 g) was added to it. The mixture was
stirred at 0 ꢁC for 1 h. To the solution was added water
(150 ml), extracted with diethyl ether (150 ml), dried over
anhydrous magnesium sulfate, and concentrated under
reduced pressure. The residue was purified by silica gel
column chromatography (ethyl acetate/hexane) to give
compound 13 (2.53 g, yield: 72%). The product was
recrystallized from ethyl acetate and hexane to give yellow
crystals, mp 86–87 ꢁC. Anal. Found: C, 57.85; H, 6.80; N,
8.13; S, 27.28, Calcd for C17H24N2S3: C, 57.91; H, 6.86; N,
1
7.95; S, 27.28%. H NMR (d ppm TMS/CDCl3 270 Mz)
1.20 (6H, d, J = 6.9 Hz), 1.23 (6H, s), 2.65 (3H, s), 2.68
(2H, s), 3.11 (1H, sept, J = 6.9 Hz), 4.51 (2H, s), 6.87 (1H,
m), 7.11–7.18 (2H, m), 7.32 (1H, m).