Morpholine containing CB2 selective agonists

Article abstract of DOI:10.1016/j.bmcl.2009.02.033

Identification and optimization of two classes of CB2 selective agonists are described. A representative from each class is profiled in a murine model of inflammation and each shows similar efficacy to prednisolone upon oral dosing.

Full text of DOI:10.1016/j.bmcl.2009.02.033

Bioorganic & Medicinal Chemistry Letters 19 (2009) 1604–1609
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Morpholine containing CB2 selective agonists
a
a
a
c
c
Renée Zindell a, , Doris Riether , Todd Bosanac , Angela Berry , Mark J. Gemkow , Andreas Ebneth ,
*
Sabine Löbbe c, Ernest L. Raymond b, Diane Thome b, Daw-Tsun Shih b, David Thomson a
a Department of Medicinal Chemistry, Boehringer Ingelheim Pharmaceuticals, Inc., 900 Ridgebury Road, PO Box 368, Ridgefield, CT 06877-0368, United States
b Department of Inflammation and Immunology, Boehringer Ingelheim Pharmaceuticals, Inc., 900 Ridgebury Road, PO Box 368, Ridgefield, CT 06877-0368, United States
c Evotec AG, Schnackenburgallee 114, 22525 Hamburg, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Identification and optimization of two classes of CB2 selective agonists are described. A representative
from each class is profiled in a murine model of inflammation and each shows similar efficacy to pred-
nisolone upon oral dosing.
Received 24 October 2008
Revised 3 February 2009
Accepted 4 February 2009
Available online 12 February 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Cannabinoid receptor
Anti-inflammatory
CB1
CB2
Agonists
Selectivity
The therapeutic use and psychotropic effects of cannabis has
been known for several centuries. However, it was only in the last
cacious in both in vitro and in vivo models of inflammation11–13 as
well as inflammatory and neuropathic pain.14–17 Some of these
50 years that the active constituent has been identified as
hydrocannabinol.1 Subsequently the receptors identified as medi-
ating its effects were identified as the protein-coupled
D
9-tetra-
compounds, shown in Figure 1, are GW40583318, GW84216619
,
and PRS-211375 (structure not known)20 with GW842166 having
entered clinical trials for the treatment of inflammatory pain. Our
goal is the identification of CB2 selective agonists for the modula-
tion of inflammatory response and treatment of pain (Fig. 2).
Our starting point to identify CB2 selective agonists origi-
nated from patent application WO 2004/060882. Compound 1,
a representative example, has a binding affinity of 50 nM for
CB2 when evaluated in a competitive binding assay with mem-
branes of HEK cells expressing the human CB2 receptor against
G
receptors cannabinoid receptor 1 (CB1) and 2 (CB2).2,3 The CB1
receptor is expressed primarily in the central nervous system
and is believed to be responsible for the undesirable effects such
as ataxia, hypothermia, and euphoria. The highest density of CB1
is found in the basal ganglia, cerebellum, and hippocampus.4 CB1
is also found to a much lesser extent in peripheral tissues such
as the testis, eye, uterus, ovary, lungs and heart.5–7 The CB2 recep-
tor is implicated in the anti-inflammatory activity of cannabis as it
is expressed primarily in peripheral immune cells such as B-cells,
monocytes, macrophages, and in organs such as the spleen, pan-
creas, thymus, lung, and tonsil.6,8 CB2 is also found to a much lesser
extent in central locations.9
CB2 shares 44% homology with CB1 which, coupled with differ-
ences in receptor densities in various tissues, offers the possibility
to separate the immune-related effects of cannabinoids from the
psychoactive effects typically associated with CB1.10 Selective
CB2 agonists should avoid unwanted psychoactive side effects
attributed to CB1 while maintaining immunomodulatory and anal-
gesic activity. CB2 selective agonists have been reported to be effi-
* Corresponding author.
Figure 1. Examples of CB2 selective agonists and modulators reported in the
E-mail address: renee.zindell@boehringer-ingelheim.com (R. Zindell).
literature.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.02.033

R. Zindell et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1604–1609
1605
Figure 2. Starting point and modification points.
Scheme 1. Reagents and conditions: (i) bromoacetyl bromide, ethyl acetate,
saturated aq NaHCO3, rt, quantitative; (ii) KOtBu, tBuOH, rt, 33%; (iii) LAH, THF,
0 °C to rt, 46%; (iv) 4-bromobenzyl bromide, K2CO3, acetonitrile, rt, 40–80% or
4-bromobenzaldehyde, MP-BH(OAc)3, THF, rt, 40–80%; (v) RB(OH)2, Pd(PPh3)4, aq
Na2CO3, MePh, 120 °C, 3–86%.
[3H]-CP55940 probe.21 The ability of CB2 agonists to activate
Gai, which in turn inhibits cyclic adenosine monphosphate
(cAMP) production is determined in a cellular assay employing
CB2 over-expressing CHO cells stimulated with forskolin.22,23
The CB2 cAMP EC50 of 1 is 2 nM with 95% efficacy and shows
120-fold selectivity over CB1 in a similar functional cAMP assay.
Since interest in this program is identifying and optimizing CB2
selective agonists, the CB2 and CB1 cAMP functional assay is
used to develop SAR with binding data generated on selected
compounds to confirm interaction with the target.
Our initial strategy was to rigidify the amino alcohol core of 1 in
an effort to lock the molecule in the desired bioactive conformation
which we hoped would result in improved CB2 potency and selec-
tivity. To this end, variants of the core were investigated. Morpho-
line (example 2) and piperidine (example 3) cores show agonist
activity, however piperizine, thiomorpholine and thiomorpholine
dioxide lacked measurable agonist activity when tested up to
and reduced to form (S)-2-phenylmorpholine. Reaction of this
intermediate under standard alkylation or reductive amination
conditions followed by palladium cross coupling affords analogs
4 and 6–9. The racemic molecules were synthesized starting from
commercially available 2-phenylmorpholine (or 3-phenylpiperi-
dine) to afford analogs 2 and 3 and those described in the text.
In addition to the core replacement effort, exploration of the
biaryl portion of the molecule was undertaken to establish optimal
spatial orientation and substitution pattern of the aromatic rings.
Any racemates showing desired agonist activity were then synthe-
sized in a chiral fashion and data is shown in Table 2.25 Appending
the B ring from the A ring 4-position was found to be optimal since
appending from the 2- or 3-position results in no measurable CB2
20 lM in the CB2 functional assay. Based on the overall profile,
morpholine was chosen as a core for further optimization. Further
agonist activity when tested up to 20 lM. Appending the B ring to
this optimal 4-position of the A ring and exploring various chloro-
substitutions demonstrated that of the 2-, 3-, and 4-chlorophenyl
derivatives, only the 2-chloro-substituted phenyl (example 6) has
agonist activity. Replacing the 2-chloro substituent with electron
neutral or donating moieties such as hydrogen, methyl, and meth-
oxy lead to compounds which have no measurable agonist activity
results of this effort are described in this letter (Table 1).
While the morpholine containing compound 2 shows a loss of
agonist activity as compared to 1, we recognized that the enantio-
mers might differentiate themselves in the functional assay and
therefore resolved the compound by chiral HPLC. The first eluting
enantiomer 4 has a CB2 cAMP EC50 of 40 nM with 50% efficacy
and 250 fold selectivity over CB1 while the second eluting enantio-
mer 5 is an inverse agonist. The 10-fold improvement in potency
upon testing of the eutomer was not surprising since the racemic
compound is a mixture of agonist and inverse agonist.
The synthesis of chiral analogs is depicted in Scheme 1 and can
be applied to both enantiomerically pure and racemic starting
materials.24 Commercially available (S)-2-amino-1-phenylethanol
is alkylated with bromoacetyl bromide and subsequently cyclized
when tested up to 20 lM. Maintaining an electron withdrawing
group in the 2-position while adding an additional substituent to
the 3-position (example 7) or 5-position (example 8) not only im-
proves potency but enhances the selectivity of the molecule as
compared to the monosubstituted 2-trifluoromethyl and 2-chloro-
phenyl analogs (examples 4 and 6). This data implies substitution
in other positions is tolerated but requires 2-substitution for de-
sired agonist activity. The nitrile substituted phenyl in compound
Table 1
Amino alcohol core variant
Core
CF3
Example
Stereo-chemistry
Core
CB2 cAMP EC50 (nM)
2
Efficacy (%)
95
CB1 EC50/CB2 EC50
120
1
rac
2
3
4
5
rac
rac
S
330
50
57
50
>60
195
250
30
40
R
Inverse agonist

1606
R. Zindell et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1604–1609
Table 2
B ring SAR
Example
R
CB2 cAMP EC50 (nM)
40
Efficacy (%)
53
CB1 EC50/CB2 EC50
250
4
6
7
25
6
51
77
306
960
8
9
3
3
73
56
781
715
9 gave similar potency and selectivity to the disubstituted contain-
ing derivatives suggesting that in addition to the polarizing effects
of the substituents on the B ring, a certain bulk imparts both the
desired potency and selectivity. Thus, rigidifying the central por-
tion of the molecule as well as manipulating the substituents of
the B ring led to novel compounds showing improvement in selec-
tivity whilst maintaining potency as compared to 1.
In addition to developing B ring SAR, we investigated incorpo-
rating a linker between the A and B rings to reduce the potential
for non-specific protein binding in vivo which has been reported
for biaryl containing molecules.26 Given the differences in func-
tional activity seen between enantiomers 4 and 5, we chose to
work with the eutomer directly in this series. Table 3 outlines
our work towards this end. Replacement of the bond between
the A and B rings with a nitrogen (compound 10) or oxygen (com-
pound 11) shows measurable activity despite having no substitu-
ents on the B ring. This is in contrast to phenyl described above
in the biaryl system which lacked agonist activity when the A
and B rings are separated with a bond. Considering the wide range
of activities seen within the biaryl system as described above, it
was plausible to expect the potency could be improved with the
appropriate substitution of the B ring. Encouraged by the activity
of 10 A ring replacements were investigated which led to the iden-
tification of compound 12. Significant improvement in the potency
and selectivity is observed with this change. Unfortunately, incor-
porating the optimized substituted B ring phenyl moiety identified
in Table 2 results in a significant loss in potency for compound 13.
This suggests introduction of a linker leads to significantly different
SAR and prompted a more detailed investigation of the B ring
whilst maintaining the pyridyl A ring.
Table 4 illustrates the SAR of the N-linked compounds.27 The
synthesis was similar to that shown in Scheme 1 with the excep-
tion of the last step which was a SNAr reaction instead of the pal-
ladium cross-coupling conditions described above.28
Unlike the biaryl system, removing substituents from the B ring
(example 12) results in modest functional activity. Mono-substitu-
tion on the B ring (examples 14–16), demonstrates compounds
bearing substitution in the 2- or 4-position are more potent than
Table 3
Incorporation of linker between aryl rings
R
N
O
Example
R
CB2 cAMP EC50 (nM)
Efficacy (%)
60
CB1 EC50/CB2 EC50
>30
10
615
11
12
13
15,000
140
100
70
>150
>10
1600
70

R. Zindell et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1604–1609
1607
Table 4
Exploration of the B ring in the bis-aryl amine series
N
R
O
N
H
N
Example
R
CB2 cAMP EC50 (nM)
140
Efficacy (%)
70
CB1 EC50/CB2 EC50
>150
12
14
15
16
17
18
19
20
21
22
23
24
30
200
80
80
80
70
80
250
>100
30
>670
35
CB1 IA
>200
100
>20,000
65
90
>310
>20,000
>20,000
50
80
>400
10
100
>2000
those which have 3-substitution. In addition to potency, 4-substi-
tution offers improvement in selectivity over CB1. Subsequent
investigation of the 4-position demonstrate that small substituents
with varying electronic properties are tolerated (examples 16–18,
and 20) but steric limitation exist as evidenced by isopropyl-
(example 19), dimethylamine- (example 21) and phenyl-substitu-
ents (example 22). Inspired by the potency and selectivity of ana-
log 16, di-substituted aromatics which contain a small substituent
in the 4-position were investigated. Example 23 demonstrates that
the 3-substituent which was detrimental to potency in example
15, is better tolerated in combination with 4-substitution. Finally,
adding the small substituent to ortho-position, which was well tol-
Table 5
Data of advanced compounds and GW842166
Example
CB2 cAMP EC50 (nM) (%)
CB1 EC50/CB2 EC50
CB2 Ki (nM) CP55940/Win55212
GW842166
7
350 (90)
5 (80)
>60
960
>5000/>10,000
312/104
24
10 (100)
>2000
>10,000/2650

1608
R. Zindell et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1604–1609
Figure 3. In vivo activity of compounds in zymosan induced paw inflammation model. Change in paw swelling of compound or prednisolone dosed mice compared to vehicle
dose mice are represented. Values are mean of six mice per group SEM. *p < 0.001 versus vehicle control.
erated in example 14, leads to compound 24 which has both excel-
lent potency and selectivity.
References and notes
1. Gaoni, Y.; Mechoulam, R. J. Am. Chem. Soc. 1964, 86, 1646.
Based on overall profile, both 7 and 24 were chosen for further
profiling as shown in Table 5. The SAR described has been devel-
oped based on the functional assays, however the binding data is
also collected on selected compounds to confirm the receptor
interaction. Both 3H-CP55940 and 3H-Win55212 are CB1/CB2 dual
agonists which are used in competition assays towards this end.21
Compound 7 shows modest competition against both radiolabeled
ligands whereas 24 competes only against 3H-Win55212. In addi-
tion to competing against only one of the probes, 24 shows a 25
fold loss in binding affinity as compared to 7 suggesting different
binding modes of the two compounds.29 Each of the compounds
is a very potent CB2 agonist based on the functional data with very
good selectivity over CB1. To examine the anti-inflammatory activ-
ity in vivo, an acute rodent inflammation model was used. Swelling
is induced by zymosan and a compounds’ efficacy is assessed by its
ability to inhibit the swelling after an oral dose of compound.30 The
results reported in Figure 3 show compound 7 dosed at 30 mpk
inhibits 54% of paw swelling as compared to vehicle control. Com-
pound 24 dosed at 100 mpk is also efficacious with 85% inhibition
of paw swelling. Both structural classes of compound show very
good potency as CB2 agonists in the functional assay as well as effi-
cacy in a murine model of inflammation.
In summary, by both rigidifying the amino-alcohol 1 to reduce
the entropy of the molecule, and incorporating a linker between
the aryl rings to reduce potential for unproductive protein binding,
we were able to identify active molecules with improved selectiv-
ity over CB1. These compounds were dosed orally in a mouse mod-
el of inflammation and showed efficacy comparable to
prednisolone. Consequently, these CB2 selective agonists should
be devoid of CB1-mediated psychotropic effects with data support-
ing their use for the treatment of inflammatory indications.
2. Devane, W. A.; Dysarz, F. A., III; Johnson, M. R., et al Mol. Pharmacol. 1988, 605.
3. Munro, S.; Thomas, K. L.; Abu-Shaar, M. Nature 1993, 365, 61.
4. Herkenham, M.; Lynn, A. B.; Little, M. D.; Johnson, M. R.; Melvin, L. S.; de Costa,
B. R.; Rice, K. C. Proc. Natl. Acad. Sci. 1990, 87, 1932.
5. Gerard, C. M.; Mollereau, C.; Vassart, G.; Parmentier, M. Biochem. J. 1991, 279,
129.
6. Galiegue, S.; Mary, S.; Marchand, J.; Dussossoy, D.; Carriere, D.; Carayon, P.;
Bouaboula, M.; Shire, D.; Fur, G.; Casellas, P. Eur. J. Biochem. 1995, 232, 54.
7. Straiker, A.; Stella, N.; Piomelli, D.; Mackie, K.; Karten, H. J.; Maguire, G. Proc.
Nat. Acad. Sci. 1999, 96, 14565.
8. Bouaboula, M.; Rinaldi, M.; Carayon, P.; Carillon, C.; Delpech, B.; Shire, D.; Fur,
G.; Casellas, P. Eur. J. Biochem. 1993, 214, 173.
9. Van Sickle, M. D.; Duncan, M.; Kingsley, P. J.; Mouihate, A.; Urbani, P.; Mackie,
K.; Stella, N.; Makriyannis, A.; Piomelli, D.; Davison, Joseph S.; Marnett, L. J.; Di
Marzo, V.; Pittman, Q. J.; Patel, K. D.; Sharkey, K. A. Science 2005, 310, 329.
10. Croxford, J. L.; Yamamura, T. J. Neuroimmunol. 2005, 166, 3.
11. Ashton, J. C. Curr. Opin. Investig. Drugs 2007, 8, 373.
12. Croxford, J. L.; Miller, S. D. J. Clin. Invest. 2008, 111, 1231.
13. Fernandez-Ruiz, J.; Romero, J.; Velasco, G.; Tolon, R. M.; Ramos, J. A.; Guzman,
M. Trends Pharmacol. Sci. 2007, 28, 39.
14. Whiteside, G. T.; Lee, G. P.; Valenzano, K. J. Curr. Med. Chem. 2007, 14, 917.
15. Cheng, Y.; Hitchcock, S. A. Expert Opin. Investig. Drugs 2007, 16, 951.
16. Guindon, J.; Hohmann, A. G. Br. J. Pharmacol. 2007, 153, 319.
17. Yao, B. B.; Hsieh, G. C.; Frost, J. M.; Fan, Y.; Garrison, T. R.; Daza, A. V.; Grayson,
G. K.; Zhu, C. Z.; Pai, M.; Chandran, P.; Salyers, A. K.; Wensink, E. J.; Honore, P.;
Sullivan, J. P.; Dart, M. J.; Meyer, M. D. Br. J. Pharmacol. 2007, 153, 390.
18. Valenzano, K. J.; Tafesse, L.; Lee, G.; Harrison, J. E.; Boulet, J. M.; Gottshall, S. L.;
Mark, L.; Pearson, M. S.; Miller, W.; Shan, S.; Rabadi, L.; Rotshteyn, Y.; Chaffer, S.
M.; Turchin, P. I.; Elsemore, D. A.; Toth, M.; Koetzner, L.; Whiteside, G. T.
Neuropharmacology 2005, 48, 658.
19. Giblin, G. M. P.; O’Shaughnessy, C. T.; Naylor, A.; Mitchell, W. L.; Eatherton, A. J.;
Slingsby, B. P.; Rawlings, D. A.; Goldsmith, P.; Brown, A. J.; Haslam, C. P.;
Clayton, N. M.; Wilson, A. W.; Chessell, I. P.; Wittington, A. R.; Green, R. J. Med.
Chem. 2007, 50, 2597.
20. Pharmos Corp. WO2006/043260.
21. Human CB2 receptor binding assay: CB2 membranes were isolated from HEK
cells stably transfected with the CB2 receptor. The membrane preparation was
bound to scintillation beads (Ysi-Poly-L-lysince SPA beads, GE Healthcare) in
assay buffer containing 50 mM Tris, pH 7.5, 2.5 mM EDTA, 5 mM MgCl2, 0.8%
fatty acid in free Bovine Serum Albumin. Unbound membrane was removed by
washing in assay buffer. Membrane-bead mixture was added to the 96-well
assay plates in the amounts of 2 mg membrane per well (CB2) and 1 mg SPA
bead per well. Compounds were added to the membrane-bead mixture in
dose-response concentrations ranging from 1 Â 10À5 M to 1 Â 10À10 M. The
competition reaction was initiated with the addition of [3H]-CP55940 or [3H]-
Win55212 as noted (Perkin–Elmer Life and Analytical Sciences) at a final
concentration of 4 nM. IC50 values for each compound were calculated as the
Acknowledgements
The authors would like to thank Drs. John Regan and Matthew
Netherton for their time and effort in proof reading the manuscript.

R. Zindell et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1604–1609
1609
concentration of compound that inhibits the specific binding of the
radioactively labeled ligand to the receptor by 50% using the XLFit 4.1 four-
parameter logistic model. IC50 values were converted to inhibition constant
(Ki) values using the Cheng–Prusoff equation.
25. Boehringer Ingelheim Corp. WO 2007/070760.
26. Hajduk, P. J.; Bures, M.; Praestgaard, J.; Fesik, S. W. J. Med. Chem. 2000, 43, 3443.
27. Boehringer Ingelheim Corp. WO2008/048914.
28. Synthetic scheme varies from Scheme 1 in that (iv) employs 6-chloro-pyridine-
3-carbaldehyde as the aldehyde to afford (S)-4-(6-chloro-pyridin-3-yl)-2-
phenyl-morpholine. (v) RNH2, NaHMDS, DMSO, 60 °C, 23–60%.
29. Literature suggests there is more than one plausible binding site and
orientation for CB ligands. Xie, X.; Chen, J. Z.; Billings, E. M. Proteins:
Structure, Function, and Genetics 2003, 53, 307; Raitio, K. H.; Salo, O. M.
H.; Nevalainen, T.; Poso, A.; Jarvinen, T. Curr. Med. Chem. 2005, 12,
1217.
30. In vivo zymosan-induced paw edema assay: Male Balb/c mice (Jackson
Laboratories, Bar Harbor, ME), between 6–10 weeks of age were used. Mice
were fed and watered ad libitum, and housed in compliance with IACUC
guidelines. On the day of study, mice were administered the test compounds,
and appropriate vehicles (95%PEG400/5%Tween 80 for po or 80% Cremophor EL
22. CB2 and CB1 cAMP assays: CHO cells expressing human CB2 or CB1
(Euroscreen) were plated at a density of 5000 cells per well in 384-well
plates and incubated overnight at 37 °C. After removing the media, the cells
were treated with test compounds diluted in stimulation buffer containing
1 mM IBMX, 0.25% BSA and 10 mM forskolin. The assay was incubated for
30 min at 37 °C. Cells were lysed and the cAMP concentration was measured
using DiscoveRx XS+ cAMP kit, following the manufacturer’s protocol. The
maximal amount of cAMP produced by forskolin compared to the level of
cAMP inhibited by 200 nM CP-55940 is defined as 100%. The EC50 value of
each test compound was determined as the concentration at which 50% of the
forskolin-stimulated cAMP synthesis was inhibited using a four-parameter
logistic model.
23. Reproducibility of the cAMP assays is assessed using the control compound CP-
55940 which is tested twice on every assay plate to rule out any plate artifacts.
Efficacy at CB1 and CB2 is expressed as a percentage relative to the efficacy of
CP-55940. Each compound is tested in triplicate at least three times with
individual dilutions from the stock. The results reported in the table are the
mean values of the measurements and individual values for the compounds do
not differ by more than a factor of three from the mean.
for ip) approximately 60 min prior to a challenge of a dorsal metatarsal
injection of 0.5% zymosan (Sigma Z-4250) in 0.9% saline in each hind paw.
Volume injected was 0.025 microliters. Paw volumes were ascertained before
challenge and 3 h after challenge by a water-based plethysmometer (Ugo
Basile) by the following method: The hind paws of mice are extended and
placed in the water chamber to the level of the lateral malleous, and the
volume displacement is electronically recorded. Six animals (12 paws) were
used in each group.
24. All compounds described gave satisfactory 1H NMR and LC–MS analytical data.