γ-Carbolines: A novel class of cannabinoid agonists with high aqueous solubility and restricted CNS penetration

Article abstract of DOI:10.1016/j.bmcl.2011.12.124

An oral, peripherally restricted CB1/CB2 agonist could provide an interesting approach to treat chronic pain by harnessing the analgesic properties of cannabinoids but without the well-known central side effects. γ-Carbolines are a novel class of potent mixed CB1/CB2 agonists characterized by attractive physicochemical properties including high aqueous solubility. Optimization of the series has led to the discovery of 29, which has oral activity in a rat inflammatory pain model and limited brain exposure at analgesic doses, consistent with a lower risk of CNS-mediated tolerability issues.

Full text of DOI:10.1016/j.bmcl.2011.12.124

Bioorganic & Medicinal Chemistry Letters 22 (2012) 1619–1624
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
c-Carbolines: A novel class of cannabinoid agonists with high aqueous
solubility and restricted CNS penetration
Yun-Xing Cheng a, Mehrnaz Pourashraf a, Xuehong Luo a, Sanjay Srivastava a, Christopher Walpole a,
Dominic Salois b, Stephane St-Onge b, Kemal Payza b, Etienne Lessard c, Xiao Hong Yu d,
Mirosław J. Tomaszewski a,
⇑
a Department of Medicinal Chemistry, AstraZeneca R&D Montreal, 7171 Frederick-Banting, Saint-Laurent, Quebec, Canada H4S 1Z9
b Department of Molecular Pharmacology, AstraZeneca R&D Montreal, 7171 Frederick-Banting, Saint-Laurent, Quebec, Canada H4S 1Z9
c Department of DMPK, AstraZeneca R&D Montreal, 7171 Frederick-Banting, Saint-Laurent, Quebec, Canada H4S 1Z9
d Department of Bioscience, AstraZeneca R&D Montreal, 7171 Frederick-Banting, Saint-Laurent, Quebec, Canada H4S 1Z9
a r t i c l e i n f o
a b s t r a c t
Article history:
An oral, peripherally restricted CB1/CB2 agonist could provide an interesting approach to treat chronic
pain by harnessing the analgesic properties of cannabinoids but without the well-known central side
effects. c-Carbolines are a novel class of potent mixed CB1/CB2 agonists characterized by attractive phys-
icochemical properties including high aqueous solubility. Optimization of the series has led to the discov-
ery of 29, which has oral activity in a rat inflammatory pain model and limited brain exposure at
analgesic doses, consistent with a lower risk of CNS-mediated tolerability issues.
Ó 2012 Published by Elsevier Ltd.
Received 1 November 2011
Revised 22 December 2011
Accepted 27 December 2011
Available online 4 January 2012
Keywords:
Cannabinoid
Carbolines
Chronic pain
The plant derived cannabinoid
D
9-tetrahydrocannabinol (
D
9-
a THC derivative possessing a carboxylic acid moiety, has been re-
ported active in phase II clinical trial for neuropathic pain with no ma-
jor adverse effect17 and Ajulemic acid has reduced CNS penetration (2,
THC, 1) and its synthetic analogues act as agonists at both cannab-
inoid receptor subtype 1 (CB1) and subtype 2 (CB2) (Fig. 1).1. CB1
receptors are widely expressed in the central nervous system
(CNS), and in the peripheral nervous system, especially in sensory
nerve fibers.2. CB2 receptors, on the other hand, are mainly located
in immune tissues and cells.2,3 Recently, it has been shown that
CB2 is also expressed in the CNS.4 Cannabinoid receptors and li-
gands have been implicated in pain transduction and perception5
as well as in neuroinflammation.6 There are published studies
demonstrating that cannabinoids, CB1/CB2 mixed agonists, are po-
tent analgesics not only in a variety of pre-clinical pain models but
also in the clinic.7–9 Unfortunately, in the clinic these agents also
show undesirable CNS side effects such as euphoria and drowsi-
ness, which are believed to be mediated through the activation
of CB1 receptors in the brain.10 There is a substantial pre-clinical
literature supporting a peripherally-mediated nociceptive action
of cannabinoids, suggesting that cannabinoid-mediated analgesia
without the well known central side effects is achievable.11–16
One strategy to dissociate the analgesic effect of cannabinoid
agonists from CNS side effects would be to develop ligands with
a low potential to cross the blood brain barrier. Ajulemic acid (2),
rat Cbrain/Cplasma ꢀ0.3–0.4) compared to
D9-THC (1, rat Cbrain/Cpl-
asma 1).18 More recently, the group of Dziadulewicz reported naphtha-
len-1-yl-(4-pentyloxynaphthalen-1-yl)methanone (3) to be a CB1/CB2
mixed agonist with limited brain penetration (rat Cbrain/Cplasma
0.07), which displays good oral bioavailability and antihyperalgesic
activity in animal pain models without obvious side-effects.19 This
anti-nociceptive effect was blocked by pretreatment with a selective
CB1 antagonist, but not by a CB2 selective antagonist. Recently, we re-
ported that AZ11713908 (4), a molecule from a different chemical ser-
ies, which is both a potent rat CB1 agonist and rat CB2 inverse agonist,
exhibited analgesic effects in inflammatory and neuropathic pain mod-
els.20 AZ11713908 showed a low potential to cross the blood brain bar-
rier. Systemic administration of AZ11713908, produced robust
analgesia in rodent pain models via peripheral CB1 receptors. A periph-
eral CB1 receptor mechanism of action proved to be sufficient to re-
verse hypersensitivity in rodent inflammatory pain models. These
results support a strategy aimed at developing CB1 or CB1/CB2 mixed
agonists for the treatment of pain via activation of peripheral CB1
receptors while limiting CNS exposure so as to minimize potential
side effects.
Despite the considerable number of cannabinoid ligands re-
ported thus far7,21 there is a scarcity of potent CB1 or CB1/CB2 ago-
nists which combine low CNS penetration together with good
⇑
Corresponding author.
E-mail address: mirek.tomaszewski@astrazeneca.com (M.J. Tomaszewski).
0960-894X/$ - see front matter Ó 2012 Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2011.12.124

1620
Y.-X. Cheng et al. / Bioorg. Med. Chem. Lett. 22 (2012) 1619–1624
physicochemical and pharmacokinetic (PK) properties. Compounds
3 and 4 have a high binding affinity for human CB1/CB2 receptors
and are potent human CB1 receptor agonist with low CNS penetra-
tion. However, they both possess high lipophilicity (ClogP = 7.8 for
3 and ClogP = 6.5 for 4), and poor aqueous solubility (<0.01 mg/
The c-carboline series was identified by scaffold-hopping from
the bicyclic benzimidazole core.22 Compound 5, a representative
early analogue of the carboline series, displayed a good binding
affinity at both human CB1 (Ki 143 nM) and CB2 (Ki 14 nM), as well
as CB1 agonist activity in rat brain tissue (rat CB1 EC50 1440 nM,
Emax 47%). Interestingly, unlike most known cannabinoid ligands,
the carboline series demonstrates good aqueous solubility. For
example, the solubility of 5 at pH 7.4 is >3 mg/mL. Furthermore,
even though compound 5 has a low polar surface area (PSA) of
21, it shows moderate CNS penetration with a brain to plasma ratio
mL).19 In the present study we describe the
c-carbolines, a new
class of potent CB1/CB2 mixed agonists, and the identification of
compound 29, which shows oral bioavailability, high aqueous sol-
ubility, robust analgesic activity in a rat inflammatory pain model,
and limited brain penetration.
CO2H
OH
OH
C4H9
C4H9
O
O
1
O
S
2
O
O
S
O
N
N
N
N
O
C5H11
3
4
O
O
N
O
N
N
N
N
N
S
O
29
5
O
Figure 1. Several cannabinoid ligands.
O
O
O
HO OH
O
N
HO
+
HO
a, b
N
H
NHNH2
N
H
6
O
N
O
H
N
O
O
N
N
c
N
d, e
N
H
7
8
R
O
N
N
f
N
9-17
Scheme 1. Reagents and conditions: (a) 1 N HCl, dioxane/H2O, reflux; (b) Boc2O, 5 N NaOH; (c) methylpiperidine, HATU, Hunig base; (d) NaH, DMF, allyl bromide; (e) TFA,
DCM; (f) appropriate aldehyde or ketone, NaBH(OAc)3, DCM.

Y.-X. Cheng et al. / Bioorg. Med. Chem. Lett. 22 (2012) 1619–1624
1621
of 0.74.23,24 One strategy undertaken to decrease CNS penetration
in -carbolines was via increasing the PSA (vide infra).
Compounds 9–40 were all derived from common intermediate
6 as shown in Schemes 1 and 2.
Our SAR investigation started with modifications to the right-
hand side, the basic region of the c-carboline scaffold. The human
CB1 agonist activity and in vitro human liver microsome metabolic
stability (HLM) of selected examples are summarized in Table 1.
Substitution on the tetrahydropyridinyl N-atom with cyclic alkyl
and heterocyclic groups led to improved hCB1 activity (9–14).
The introduction of polar fragments such as pyridinylmethyl
groups was detrimental to CB1 functional activity (15–17 vs 9).
Substitution by tetrahydrothiophenyl ring (12) afforded a very po-
tent agonist although the metabolic stability was lost. The best re-
sults came from cyclopentyl and 4-tetrahydropyran (4-THP)
groups, as 11 and 14 achieved a >10 fold gain in agonist activity
and maintained reasonable metabolic stability. Overall, 4-THP
and cyclopentyl moieties were considered as preferred groups for
the right hand side.
We then turned to building SAR on the southern portion, the in-
dole N-atom (Table 2). Compounds 18–37 were prepared accord-
ing to the general method described in Scheme 2. In order to
increase the likelihood of identifying compounds with limited
CNS penetration, we adopted a design strategy which included
increasing polar surface area of c-carboline analogues. Mahar Doan
et al. analyzed the physical chemical properties of CNS and non-
CNS drugs on the market and found that non-CNS drugs have high-
er PSA values (mean PSA value of 56) than CNS drugs (mean PSA
value of 40).25 To this end, a variety of functional groups were
introduced at the indole nitrogen region to increase the overall
polarity. Gratifyingly, with few exceptions, modifications to this
southern region of the carboline were well tolerated. Alkyl, amide,
sulfonamide (18–31), and sulfamide (32,33) groups all exhibited
good CB1 affinity. Insertion of a carbonyl or sulfonyl group be-
tween the nitrogen and propyl group maintained the CB1 potency
(19 vs 23,24) and, in addition, afforded a substantial increase in
polarity (19 PSA = 30 vs 24 PSA = 67). Although amide 23 and sul-
fonamide 24 were both equipotent in CB1 receptor binding, amide
c
23 was a partial CB1 agonist in the [35S]GTP
cS functional assay
with only 17% efficacy, while sulfonamide 24 was a full agonist
(hCB1 EC50 129 nM and Emax 100%). Encouraged by the results of
sulfonamides 24 and 25, additional sulfonamides (26–31) were
investigated. The ethylsulfonyl analogue, compound 29, was found
to be a potent agonist on hCB1 (EC50 of 49 nM, Emax 120%) and in
rat brain tissue (rat CB1 EC50 85 nM, Emax 156%). The in vitro selec-
tivity of 29 was also examined in 52 diverse targets including a
number of pain targets, and no significant activity was observed.26
Further characterization of 29 alongside the starting
c-carbo-
line hit, 5, are shown in Table 3. Compound 29 is a potent dual
CB1/CB2 agonist which combines restricted CNS penetration in ro-
dents (brain to plasma ratio 0.07 in rat and 0.05 in mice)27 together
with good Caco-2 permeability. Although compound 29 and two
reference compounds 3 and 4 display similar low levels of CNS
penetration, their physicochemical properties and plasma protein
binding profiles are very different (Table 3). Compounds 3 and 4
are characterized by high lipophilicity (ClogP >6), high plasma pro-
tein binding (<1% free)19,28 and low solubility. In contrast, com-
pound 29 shows 13% free in rat plasma protein binding, ClogP of
2.4, and aqueous solubility of 0.48 mg/mL. Furthermore, with an
efflux ratio of 1.5 in MDR1/MDCK assay, compound 29 has a low
potential for being a P-gp substrate in man.
The rat pharmacokinetic profile for 29 is shown in Table 4. Com-
pound 29 exhibited moderate plasma clearance (28 mL/min/kg
corresponding to 40% liver blood flow in rat) that is consistent with
predicted hepatic clearance from the in vitro metabolic stability in
rat liver microsomes (28 mL/min/kg while taking plasma protein
binding into account). The oral bioavailability of 29 in rats is 16%,
O
R'
O
O
N
O
N
N
N
N
a-c
N
H
R
7
18-40
Scheme 2. Reagents and conditions: (a) NaH, DMF, RX; (b) TFA, DCM; (c)
cyclopentoanone or 4-tetrahydropyranone, NABH(OAC)3, DCM.
Table 1
The hCB1 agonist activity and in vitro metabolic stability data of 5 and 9–17
R
N
O
N
N
Compound
R
hCB1a EC50
(nM)
hCB1a Emax
(%)
HLM (
mg)
ll/min/
Compound
R
hCB1a EC50
(nM)
hCB1a Emax
(%)
HLM (
mg)
ll/min/
5
9
768
103
71
94
66
13
14
60
60
57
57
135
42
O
>300
N
10
115
63
>300
15
>14000
>300
N
11
12
33
5
65
58
93
16
17
5423
1360
66
76
ND
S
N
>300
>300
a
Mean value of at least 2 independent measurements. EC50 were measured by GTP
line transfected with the cloned human CB1 receptor using Win55,212-2 as the reference agonist for Emax calculations.
c
35S binding on membranes of Human Embryonic Kidney (HEK) 293 EBNA stable cell

1622
Y.-X. Cheng et al. / Bioorg. Med. Chem. Lett. 22 (2012) 1619–1624
Table 2
hCB1 binding affinity, functional activity and PSA of compounds 11, 13 and 18–37
R'
N
O
R'=
N
O
B
N
R
A
Compound
R
R0 PSA hCB1a Ki
hCB1a EC50
(nM)
hCB1a Emax
(%)
Compound
R
R0 PSA hCB1a Ki
hCB1a EC50
(nM)
hCB1a Emax
(%)
(nM)
(nM)
11
13
18
19
20
21
22
23
24
25
Allyl
Allyl
A
B
A
B
A
A
A
B
B
A
21
30
21
30
35
21
21
48
67
58
9
56
15
58
264
32
33
60
45
103
4253
77
65
57
85
27
28
29
30
31
32
33
34
35
36
MethylSO2-
EthylSO2-
EthylSO2-
i-PrSO2-
B
A
B
B
B
A
B
B
B
B
67
58
67
67
67
60
69
49
39
83
239
17
120
170
31
473
11
49
150
59
26
71
ND
53
103
122
120
96
n-Propyl
n-Propyl
H
Methyl
Ethyl
PropylCO-
PropylSO2-
PropylSO2-
72
115
133
118
17
100
97
PhSO2-
46
Me2NSO2-
Me2NSO2-
Me2NCO-
p-EtOPhCH2-
EtNHSO2-
22
114
112
ND
113
ND
34
63
110
2380
51
102
115
22
117
129
22
424
ND
26
MethylSO2-
A
58
36
50
109
37
B
78
592
ND
ND
N
SO2-
N
a
Mean value of at least 2 independent measurements. Ki values were measured by displacement of the agonist 3H-CP55,940 binding on membranes of Human Embryonic
Kidney (HEK) 293s stable cell lines transfected with the cloned human CB1 receptor. EC50 were measured by GTP
c
35S binding on membranes of Human Embryonic Kidney
(HEK) 293 EBNA stable cell line transfected with the cloned human CB1 receptor using Win55,212-2 as the reference agonist for Emax calculations.
Table 3
Characterization of selected carboline compounds and reference compounds
Compound Compound
class
hCB1a Ki hCB1a
hCB1a
hCB2a Ki HLM (
l
l/
RLM(
min/mg)
l
l/
Br/Pl
ratio
(rat)
Caco-2 Papp
PSA ClogP Solubility @pH
7.4 (mg/mL)
(nM)
EC50 (nM) Emax (%)
(nM)
min/mg)
(ꢁ10ꢂ6 cm/s)
29
5
1
3
4
Carboline
Hit carboline
120
143
10
35
3
49
768
17
72
6
120
71
20
111
115
9
52
66
202
>300
0.07
0.74
1.0018
0.04
0.07
13
3
67
21
32
27
69
2.4
4.8
7.2
7.8
6.5
0.48
3.3
<0.01
<0.01
<0.01
14
24
19
0.56
D
9-THC
Napthyl
Benzimidazole
>300
>300
0.6
a
Mean value of at least 2 independent measurements. Ki values were measured by displacement of the agonist 3H-CP55,940 binding on membranes of Human Embryonic
Kidney (HEK) 293s stable cell lines transfected with the cloned human CB1 receptor. EC50 were measured by GTP
35S binding on membranes of Human Embryonic Kidney
(HEK) 293 EBNA stable cell line transfected with the cloned human CB1 receptor using Win55,212-2 as the reference agonist for Emax calculations.
c
example, there is only one methyl group difference between 40
(ClogP 2.6) and 41 (ClogP 2.0), and the hERG channel activity
drops sixfold (IC50 of 2.6 M for 40 and 18 M for 41).
Table 4
Rat pharmacokinetic profile of compound 29
l
l
Compound
Administration route
IV
Parameters
Mean
Before investigation of the in vivo efficacy of 29, it was tested in
the field-stimulated mouse vas deferens (MVD), which is a more
physiological CB1 receptor agonism assay than membrane
29
Cl (mL/min/kg)
T1/2 (h)
Vss (l/kg)
F%
28
0.63
1.2
16
Oral
[
35S]GTP
cS binding assay. Thus, the purpose of assessing 29 in
MVD was to confirm qualitatively its agonist activity rather than
to interpret in vivo activity; the MVD is a mouse model, whereas
we used a pain model in rat, and species differences could affect
the absolute potency of compounds. The agonist activity of 29
was confirmed in the MVD assay, in which it displayed an IC50 of
1.9 nM (data not shown). Thus, compound 29 showed CB1 agonist
activity not only in membranes of cloned rat and human CB1
receptor in heterologous system, but also in a physiologically intact
tissue containing a native receptor.
which suggests that absorption may be incomplete. The CYP inhi-
bition potential is low for compound 29 as IC50 values for CYP1A2,
CYP2C9, CYP2C19, CYP2D6 and CYP3A4 are greater than 20
Compound 29 shows moderate hERG activity with an IC50 of
3.1 M using voltage ion flux electrophysiology assay.
Strategies that have been developed to reduce hERG activity in-
lM.
l
clude modulating pKa, ClogP, and making discreet structural
Compound 29 was then tested in vivo in rat Carrageenan
inflammatory pain model dosed subcutaneously at 0.35, 0.70,
and 1.4 mg/Kg. These doses resulted in total plasma concentrations
of 126–213 nM and between 10 and 13 nM in brain and exhibited
dose dependent reversal of heat hyperalgesia.30,31 No severe side
effect associated with CB1 receptor at central nervous system
was observed at these doses (Fig. 2).32 This is in contrast to known
CNS acting compounds, such as WIN55,212-2, which shows no
margin between analgesic effect and CNS side effect.20 Although
changes to the scaffolds to introduce hydrophilic groups in key
positions.29 In the case of the
c-carbolines, changes in pKa appear
to strongly influence the hERG activity (Table 5). Compound 39
with a THF moiety on N-1 is 1 log unit less basic than compound
29, and 39 is sixfold less active on hERG (IC50 of 3.1
and 25 M for 39). Similarly compound 38, the corresponding
THF analogue of 24, has a lower potential to interact with hERG
channel than 24 (2 M for 24 and 13 M for 38). The hERG activity
is also found to be sensitive to changes in the amide region. For
lM for 29
l
l
l

Y.-X. Cheng et al. / Bioorg. Med. Chem. Lett. 22 (2012) 1619–1624
1623
Table 5
Reducing activity at hERG: 29?39, 24?38 and 40?41
R1
N
O
N
N
R2
R3
a
Compound
R1
R2
R3
hERG IC50
3.1
(lM)
pKa (measured)
ClogP
hCB1 Ki (nM)
29
O
O
EthylSO2-
Methyl
6.9
2.4
120
39
24
38
EthylSO2-
PropylSO2-
PropylSO2-
Methyl
Methyl
Methyl
25
2
5.9
7.3
5.8
2.5
3
197
115
320
O
O
13
3.1
40
41
Methyl
Methyl
Methyl
H
2.6
18
—
—
2.6
2.0
76
O
O
226
a
Mean value of at least 2 independent measurements. Ki values were measured by displacement of the agonist 3H-CP55,940 binding on membranes of Human Embryonic
Kidney (HEK) 293s stable cell lines transfected with the cloned human CB1 receptor.
2. Howlett, A. C.; Barth, F.; Bonner, T. I.; Cabral, G.; Casellas, P.; Devane, W. A.;
Felder, C. C.; Herkenham, M.; Mackie, K.; Martin, B. R.; Mechoulam, R.; Pertwee,
R. G. Pharmacol. Rev. 2002, 54, 161.
3. Buckley, N. E.; McCoy, K. L.; Mezey, E.; Bonner, T.; Zimmer, A.; Felder, C. C.;
Glass, M.; Zimmer, A. Eur. J. Pharmacol. 2000, 396, 141.
4. Cabral, G. A.; Raborn, E. S.; Griffin, L.; Dennis, J.; Marciano-Cabral, F. Br. J.
Pharmacol. 2009, 153, 240.
5. Chapman, V.; Finn, D. P. Rev. Anal. 2003, 7, 25.
Severe
***
150
Moderate
**
100
*
6. Walter, L.; Stella, N. Br. J. Pharmacol. 2004, 141, 775.
Mild
7. Cheng, Y.; Hitchcock, S. A. Expert Opin. Investig. Drugs 2007, 16, 951.
8. Berman, J. S.; Symonds, C.; Birth, R. Pain 2004, 112, 299.
9. Wissel, J.; Haydn, T.; Muller, J.; Brenneis, C.; Berger, T.; Poewe, W.; Schelosky, L.
D. J. Neurol. 2006, 253, 1337.
10. Burns, T. L.; Ineck, J. R. Ann. Pharmacother. 2006, 40, 251.
11. Agarwal, N.; Pacher, P.; Tegeder, I.; Amaya, F.; Constantin, C. E.; Brenner, G. J.;
Rubino, T.; Michalski, C. W.; Marsicano, G.; Monory, K.; Mackie, K.; Marian, C.;
Batkai, S.; Parolaro, D.; Fischer, M. J.; Reeh, P.; Kunos, G.; Kress, M.; Lutz, B.;
Woolf, C. J.; Kuner, R. Nat. Neurosci. 2007, 10, 870.
12. Burstein, S. H.; Karst, M.; Schneider, U.; Zurier, R. B. Life Sci. 2004, 75, 1513.
13. Walker, J. M.; Hohmann, A. G. In Cannabinoids–Handbook of Experimental
Pharmacology; Pertwee, R., Ed.; Springer: Berlin, 2005; p 509.
14. Pertwee, R. G. Br. J. Pharmacol. 2009, 156, 397.
15. Karst, M.; Wippermann, S. Expert Opin. Investig. Drugs 2009, 18, 125.
16. Kunos, G.; Osei-Hyiaman, D.; Batkai, S.; Sharkey, K. A.; Makriyannis, A. Trends
Pharmacol. Sci. 2008, 10, 1.
50
0
None
10
0.1
1
Dose (µmol/kg, s.c.)
Efficacy in carr.
Hypo-activity score
Figure 2. Anti-hyperalgesic effects and hypoactivity score of 29 (sc) in rat
carrageenan model (method see References 20,32). Data were shown as mean
-
SEM (n = 7). ⁄p <0.05, ⁄⁄p <0.01 and ⁄⁄⁄p <0.001 versus vehicle group by one way
ANOVA followed by Holm-Sidak test. Slightly hypo-activity was observed at rats
17. Salim, K.; Schneider, U.; Burstein, S.; Hoy, L.; Karst, M. Neuropharmacology
2005, 48, 1164.
treated with 2.4 lmol/kg dose.
18. Dyson, A.; Peacock, M.; Chen, A.; Courade, J. P.; Yaqoob, M.; Groarke, A.; Brain,
C.; Loong, Y.; Fox, A. Pain 2005, 116, 129.
19. Dziadulewicz, E. K.; Bevan, S. J.; Brain, C. T.; Coote, P. R.; Culshaw, A. J.; Davis, A.
J.; Edwards, L. J.; Fisher, A. J.; Fox, A. J.; Gentry, C.; Groarke, A.; Hart, T. W.;
Huber, W.; James, I. F.; Kesingland, A.; La Vecchia, L.; Loong, Y.; Lyothier, I.;
McNair, K.; O’Farrell, C.; Peacock, M.; Portmann, R.; Schopfer, U.; Yaqoob, M.;
Zadrobilek, J. J. Med. Chem. 2007, 50, 3851.
we can not rule out contribution of central CB1 activation to the
analgesic activity, the fact remains that the more peripherally re-
stricted compound 29 shows considerably improved safety margin
compared to a CNS active CB agonist, WIN55,212-2.20
20. Yu, X. H.; Cao, C. Q.; Martino, G.; Puma, C.; Morinville, A.; St-Onge, S.; Lessard,
E.; Perkins, M. N.; Laird, J. M. A. Pain 2010, 151, 337.
In summary, we have identified c-carbolines as a new structural
class of potent mixed CB1/CB2 agonists, and we have demon-
strated that solubility and potency can be achieved together for a
cannabinoid ligand. Carbolines are a class of cannabinoid agonists
with good physicochemical properties and low CNS penetration,
and compound 29 demonstrates significant anti-hyperalgesia in a
rat inflammatory pain model with significant margin versus CNS
side effects. Further investigation of this promising chemical series
is under way and more detail accounts will be reported in due
course.
21. Thakur, G. A.; Tichkule, R.; Bajaj, S.; Makriyannis, A. Expert Opin. Ther. Patents
2009, 19, 1647.
22. Page, D.; Balaux, E.; Boisvert, L.; Liu, Z.; Milburn, C.; Tremblay, M.; Wei, Z.; Woo,
S.; Luo, X.; Cheng, Y.-X.; Yang, H.; Srivastava, S.; Zhou, F.; Brown, W.;
Tomaszewski, M.; Walpole, C.; Hodzic, L.; St-Onge, S.; Godbout, C.; Salois, D.;
Payza, K. Bioorg. Med. Chem. Lett. 2008, 18, 3695.
23. All the animal uses were approved by AstraZeneca R&D Montreal Animal Care
Committee in accordance with Canadian Council on Animal Care guidelines.
Unless specified otherwise, the in vivo PK study and CNS experiments were
conducted in Sprague–Dawley male rat.
24. For the measurement of brain to plasma ratio of compound 5, the compound
was administered at 1 mg/kg via subcutaneous route (sc) in rats. The blood and
brain samples were collected at 30 min after compound administration.
25. Mahar Doan, K. M.; Humpreys, J. E.; Webster, L. O.; Wring, S. A.; Shamprine, L.
J.; Serabit-Singh, C. J.; Adkinson, K. K.; Polli, J. W. J. Pharmacol. Exp. Ther. 2002,
303, 1029.
References and notes
1. Pertwee, R. G. Pharmacol. Ther. 1997, 74, 129.

1624
Y.-X. Cheng et al. / Bioorg. Med. Chem. Lett. 22 (2012) 1619–1624
26. Compound 29 shows less than 50% activity (10
l
a
M) against the following
1, h 2a, hb1, hD2L, hERa,
30. The determination of the total plasma or brain concentration of 29 was
performed by protein precipitation (after homogenization of brain samples),
followed by reversed-phase liquid chromatography and electrospray mass
spectrometry. Twenty microliter of biological samples from CNS experiments
(plasma and brain) and blank matix (ces) are added manually into a 96 well
targets: hACE, hCOX-1, hMAO-A, mERK2, hPKC
a, r
a
rGABAa, hH1, hM2, hM3, n-AChR, 5-HT2a, hNET, LTCC, NaV site 2, h5HT6,
hA2a, hAT1, hBRS-3, hCCR1, hCCR2b, hCCR5, hCGRP, hCRF1, hCX3CR1, hCXCR2,
hCXCR6, hETA, hFPR1, hGHS-R, hMC4, hMCH1, hM4, hNK3, hGLP-1, hGLP-2,
hMtl, hSST4, hV1B, rA3, rGPR103, rPK2r, mGPR73, hFPRL-1, and mUT II.
plate. All the in vivo samples are precipitated by adding 60 ll of ACN+. The
Compound 29 has an IC50 >10
hKOR receptors.
l
M against 3 opioid receptors hDOR, hMOR and
plate is carefully capped, vortexed and centrifuged at 9000g for 30 min at 4 °C.
The samples are then ready to be analyzed by LC–MS/MS. After preparing an
analytical standard in ACN+, serial dilutions are carried out in order to prepare
a calibration curve. The range of the calibration curve is 1.22 to 1000.
31. Brain samples were corrected for 2% blood contamination according to Heisey
SR. Brain and choroid plexus blood volumes in vertebrates. Comp. Biochem.
Physiol. 1968, Aug; 26(2):489.
32. Method to score hypo-activity: hypo-activity was visually observed and scored
by the experimenters. Rats were observed for the following behaviors: sniffing,
grooming, scratching, vocalization, stereotypy (excessive repetition or lack of
variation in movements), rearing, and body posture. Animals were also
examined for signs of ptosis (drooping of the upper eyelid), exophthalmos
(abnormal protrusion of the eyeball). Hypoactivity was rated 0, absent; 1, mild;
2, moderate; and 3, severe.
27. Animals were treated with 29 (sc and po) and samples were collected at 30 or
60 min after drug administration. The animals are sacrificed by decapitation,
the brain collected immediately and after removing the pons, cut in two to
separate the right and left hemispheres. Visible blood clots are removed. The
right hemisphere is snap-frozen by immersion in liquid nitrogen and stored
frozen at ꢂ80 °C. They are then pulverized with the cryogenic system.
Pulverized samples are then transferred into glass tubes and weighing the
samples, the appropriate amount of buffer is added and samples are
homogenized with the Covaris proce.
28. Unpublished results from this laboratory.
29. For recent reviews on strategies to mitigate hERG activity, see (a) Jamieson, C.;
Moir, E. M.; Rankovic, Z.; Wishart, G. J. Med. Chem. 2006, 49, 5029; (b) Durdagi,
S.; Subbotina, J.; Lees-Miller, J.; Guo, J.; Duff, H. J.; Noskov, S. Y. Curr. Med. Chem.
2010, 17, 3514.