Synthesis and testing of novel phenyl substituted side-chain analogues of classical cannabinoids

Article abstract of DOI:10.1016/S0960-894X(03)00729-7

A series of novel phenyl substituted side-chain analogues of classical cannabinoids were synthesized and their CB1 and CB2 binding affinities were evaluated relative to Δ⁸-THC and compound 2. CB1 and CB2 binding assays indicate that the dimethy

Full text of DOI:10.1016/S0960-894X(03)00729-7

Bioorganic & Medicinal Chemistry Letters 13 (2003) 3487–3490
Synthesis and Testing of Novel Phenyl Substituted Side-Chain
Analogues of Classical Cannabinoids
Mathangi Krishnamurthy,a Antonio M. Ferreirab and Bob M. Moore, IIc,*
aDepartment of Pharmaceutical Sciences, College of Pharmacy, University of Tennessee-Memphis,
Memphis, TN 38163, USA
bComputational Research on Materials Institute, Department of Chemistry, University of Memphis,
Memphis, TN, 38152, USA
cUniversity of Tennessee, College of Pharmacy, Department of Pharmaceutical Sciences, 847 Monroe Ave., Room 327D,
Memphis, TN 38163, USA
Received 30 April 2003; revised 3 July 2003; accepted 12 July 2003
Abstract—A series of novel phenyl substituted side-chain analogues of classical cannabinoids were synthesized and their CB1 and
CB2binding affinities were evaluated relative to Á8-THC and compound 2. CB1 and CB2binding assays indicate that the dimethyl
and ketone analogues (3) and (6) display selectivity for the CB2receptor in comparison to Á8-THC and compound 2. This study
provides newer insights into the geometrical and functional group requirements of the ligand binding pockets of the CB1 and the
CB2receptors.
# 2003 Elsevier Ltd. All rights reserved.
The cannabinoids are a diverse class of compounds that
possess the ability to bind to the cannabinoid receptor-1
the latter, researchers have introduced a variety of sub-
stituents including branched chain alkyls,14,15 unsatu-
rated alkyls16ꢀ18 and alkyls containing 10,10-cyclic
functionality.19 All these efforts have been directed at
determining the structural requirements of the ligand-
binding pocket (LBP).
1ꢀ4
(CB1) and receptor-2(CB)2,
acting as agonist,
antagonist or inverse agonist. Each of these activities
has gained considerable attention due to their potential
indication in the treatment of obesity,5 cancer,6 glau-
coma7 and several other pathological conditions. The
therapeutic potential of CB1 and CB2ligands has
resulted in the synthesis of a diverse set of compounds
that can be broadly grouped into the classical cannabi-
noids, non-classical cannabinoids,8 alkyl amino indoles 9
and pyrazoles.10
An understanding of the structural requirements of the
ligand-binding pocket is essential for guiding the design
of subtype selective ligands; however, to date neither the
X-ray nor NMR solution structures of either the CB1 or
The classical cannabinoids are, by far, the most exten-
sively studied group in terms of structure–activity rela-
tionships (SAR) and pharmacology. The natural
11
product, Á9-THC
(1) (Fig. 1) is the prototypical
example for this class of cannabinoids and it is a partial
agonist of the CB1 receptor.12 Considerable efforts have
been made to elucidate the SAR of these compounds
with respect to modifications of the tricyclic ring and
side chain substituents at the C3 position.13 To assess
*Corresponding author. Tel.: 1-901-448-6085; fax: 1-901-448-6828;
e-mail: bmoore@utmem.edu
Figure 1.
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00729-7

3488
M. Krishnamurthy et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3487–3490
the CB2receptors have been determined. In the absence
of this data, researchers must utilize site directed muta-
genesis, molecular modeling studies and traditional
medicinal chemistry to provide insights into the
geometric and chemical requirements of the LBP.
reduced reactivity of aromatic Grignard reagent as
compared to alkyl Grignard reagent, we used 3,5-dime-
thoxy benzaldehyde as the starting material and reacted
this with phenyl magnesium bromide to obtain the cor-
responding alcohol (7)23 as shown in Scheme 1. Oxida-
tion of alcohol with PCC yielded the key intermediate
ketone (8).23 The dithiolane group was introduced at the
C10 position (10) by reacting the ketone with ethane
dithiol in presence of boron trifluoride.18 The ketone
intermediate was also reacted with dimethyl zinc and
titanium tetrachloride to form the dimethyl substituent
at the C10 position (9).22 Desulfurisation of intermediate
(10) with Raney–Nickel yielded the methylene inter-
mediate (11).24 The 1-substituted 3,5-dimethoxy inter-
mediates (9–11) were deprotected with boron tribromide
to yield the corresponding resorcinols (12–14).22 Á8-
THC analogues (3–5) were then obtained from these
resorcinols (12–14) by reacting them with cis-Á2-p-
menthene-1, 8-diol, prepared from (+)-Á2-carene,25 in
presence of p-toluene sulfonic acid. The Á8-THC ana-
logue with a ketone functionality at the C10 position (7)
was obtained by deprotecting the analogue (4) with
silver nitrate (Scheme 2).26,28
As part of our efforts to characterize the structural
requirements of the CB1 and the CB2receptors, we had
previously synthesized a series of cycloalkyl side-chain
analogues of Á8-THC.20 The series consisted of cyclo-
pentyl, cyclohexyl and cycloheptyl side chains with 10,
10-dimethyl or dithiolane functionalities. The 10,10-
dimethyl series exhibited, on an average, a 100 fold
increase in binding affinity for both the CB1 and CB2
receptors, while the dithiolanes had an average 10-fold
higher affinity. Interestingly, compound (2) had com-
parable affinity to the CB1 receptor as its linear carbon
equivalent congener 10,10-dimethyl heptyl THC
(DMHT).13
The high affinities of this series of compounds prompted
us to question if a phenyl ring could be substituted for
the cyclohexyl ring thus generating a novel series of Á8-
THC analogues. We hypothesized that the introduction
of a phenyl side chain could significantly alter the elec-
tronic properties of both the side chain and A ring while
maintaining steric bulk and conformational restraint.
This substitution was in part designed to study potential
p–p interactions of the ligands with aromatic amino
acids in the LBP.21 To the best of our knowledge, there
have been no previous reports of a 10-phenyl substitu-
tion in the classical cannabinoid series. Therefore a ser-
ies of C10 substituted phenyl derivatives with dimethyl,
dithiolane, methylene and ketone substituents at the C10
position were synthesized (3–6).
The CB1 and CB2binding affinities of these novel Á8-
THC analogues with phenyl side chains were deter-
mined using membrane preparations of the human
receptors transfected into HEK293 EBNA cells. Recep-
tor binding assays were carried out using tritiated
CP55,940 as the competing radioactive ligand and 10
mm WIN 55212-2 was used for determining non-specific
binding.20
The CB2binding affinities of these novel analogues were
in the range of 0.9–86 nm while the CB1 binding affi-
nities ranged from 12to 297 nM ( Table 1). Interestingly,
these compounds exhibited significantly different bind-
ing profile when compared to the lead compound (2).
The dimethyl analogue (3) exhibited good binding affi-
nities for both the CB1 and the CB2receptors with a 13-
fold selectivity for the CB2receptor. This selectivity is in
contrast to the lead compound (2) that binds both the
The side chains of classical cannabinoids have often
been made by reaction of 3,5-dimethoxy benzonitrile
with a suitable Grignard reagent and acid hydrolysis of
the intermediate imine salt to the ketone.22 Due to the
Scheme 1. Reagents and conditions: (a) Phenyl magnesium bromide,
THF, 10% HCl/ꢀ20ꢁC; (b) PCC, CH2Cl2/rt; (c) dimethyl zinc, tita-
nium(IV) chloride, CH2Cl2/ꢀ40ꢁC; (d) ethane dithiol, BF3-diethyl
etherate, CH2Cl2/rt; (e) Raney–Nickel, EtOH/EtOAc, reflux.
Scheme 2. Reagents and conditions: (f) BBr3, CH2Cl2/0ꢁC–rt; (g) cis-
p-menthene-1,8-diol, p-toluene sulfonic acid, benzene/80ꢁC; (h)
AgNO3, MeOH/rt.

M. Krishnamurthy et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3487–3490
3489
Table 1. Affinities (Ki) of compounds 3, 4, 5, 6 for CB1 and CB2
receptors
studies with functional assays should contribute to a
better understanding about the differences in the struc-
tural requirements of the LBP of the CB1 and CB2
receptors and may aid in developing more selective
compounds.
Compd
CB1 Ki (nM)a
CB2 Ki (nM)a
Ratio CB1/CB2
Á8-THC
28.5 (ꢂ3.30)
0.57 (ꢂ0.05)
12.3 (ꢂ0.61)
17.3 (ꢂ0.33)
67.6 (ꢂ2.90)
297 (ꢂ10.6)
25.0 (ꢂ4.80)
0.65 (ꢂ0.04)
0.91 (ꢂ0.08)
17.6 (ꢂ1.03)
85.9 (ꢂ0.31)
23.6 (ꢂ1.76)
1.14
0.87
13.5
0.98
0.78
12.6
2
3
4
5
6
Acknowledgements
We would like to thank Dr. Albert Tuinman, Chemistry
Department, University of Tennessee, Knoxville for
HRMS spectra. This research was supported by the
College of Pharmacy, University of Tennessee Health
Sciences Center.
aThe Ki values for Á8-THC and the analogues were obtained from
three independent experiments each of which was run in duplicate and
are expressed as the mean of three values, with the standard error of
mean shown in parentheses.
References and Notes
subtypes with almost equal affinity. The ketone analogue
(6) exhibited similar binding affinity for the CB2receptor
when compared to Á8-THC but almost a 10-fold
decrease in the binding affinity for the CB1 receptor. The
dithiolane analogue (4) exhibited no subtype selectivity,
however there was a 10-fold decrease in affinity relative
to the 10-cyclohexyl congener (CB1 Ki=1.86 nM and
CB2 Ki=1.05 nM).20 The methylene analogue (5) dis-
played significantly reduced binding affinities for both
the subtypes in comparison to Á8-THC.
1. Gernard, C.; Mollereau, C.; Vassart, G.; Parmentier, M.
Biochem. J. 1991, 279, 129.
2. Skaper, S. D.; Buriani, A.; Dal Toso, R.; Petrelli, L.;
Romamello, S.; Facci, L.; Leon, A. Proc. Natl. Acad. Sci.
U.S.A. 1996, 93, 3984.
3. Matsuda, L. A.; Lolait, S. J.; Brownstein, M. J.; Young,
A. C.; Bonner, T. I. Nature 1990, 346, 61.
4. Munro, S.; Thomas, K. L.; Abu-Shaar, M. Nature 1993,
365, 61.
5. Di Marzo, V.; Goparaju, S. K.; Wang, L.; Liu, J.; Batkai,
S.; Jarai, Z.; Fezza, F.; Miura, G. I.; Palmiter, R. D.; Sugiura,
T.; Kunos, G. Nature 2001, 410, 822.
6. Parolaro, D.; Massi, P.; Rubino, T.; Monti, E. Prosta-
glandins Leukot. Essent. Fatty Acids 2002, 66, 319.
7. Porcella, A.; Maxia, C.; Gessa, G. L.; Pani, L. Eur. J.
Neurosci. 2001, 13, 409.
8. Mevin, L. S.; Johnson, M. R.; Herbert, C. A.; Miline,
G. M.; Weissmen, A. A. J. Med. Chem. 1984, 27, 67.
9. D’Ambra, T. E.; Estep, K. G.; Bell, M. R.; Eissentat, M. A.;
Josef, K. A.; Ward, S. J.; Haycock, D. A.; Baizman, E. R.;
Casiano, F. M.; Beglin, N. C.; Chippari, S. M.; Grego, J. D.;
Kullnig, R. K.; Daley, G. T. J. Med. Chem. 1992, 35, 124.
10. Rinaldi-Carmona, M.; Barth, F.; Heaulme, M.; Shire, D.;
Calandra, B.; Congy, C.; Martinez, S.; Maruani, J.; Neliat, G.;
Caput, D.; Farrara, P.; Soubrie, P.; Breliere, J. C.; Le Fur, G.
FEBS Lett. 1994, 350, 240.
11. Gaoni, Y.; Mechoulam, R. J. Am. Chem. Soc. 1964, 86,
1646.
12. Howlett, A. C.; Barth, T. I.; Bonner, G.; 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.
13. Seltzman, H. H. Curr. Med. Chem. 1999, 6, 685.
14. Huffman, J. W.; Lainton, J. A. H.; Banner, W. K.; Dun-
can, S. G.; Jordan, R. D.; Yu, S.; Dai, D.; Martin, B. R.;
Wiley, J. L.; Compton, D. R. Tetrahedron 1997, 53, 1557.
15. Huffman, J. W.; Liddle, J.; Duncan, S. G., Jr.; Yu, S.;
Martin, B. R.; Wiley, J. L. Bioorg. Med. Chem. 1998, 6, 2383.
16. Busch-Petersen, J.; Hill, W. A.; Fan, P.; Khanolkar, A.;
Xie, X. Q.; Tius, M. A.; Makriyannis, A. J. Med. Chem. 1996,
39, 3790.
The binding affinities of our novel 10-phenyl substituted
Á8-THC analogues provide some new insights into the
functional group requirements of the binding pockets of
the CB1 and CB2receptors. Valuable structural infor-
mation can be gleaned from our dimethyl and ketone
analogues that exhibited modest selectivity for the CB2
receptor. This in combination with our previous data
for the cyclohexyl Á8-THC analogues20 suggests that a
subsite binding pocket of the CB2receptor can tolerate
both cycloalkyl side chains and rigid aromatic side
chains when compared to the CB1 receptor. The selec-
tivity of 3 and 6 is interesting when considering that
several of the short chain C3 analogues reported by
Huffman et al. also exhibited significant CB2selectiv-
ity.27 However, a comparison between the two struc-
tural types is difficult when considering that the short
chain analogues were primarily 1-deoxy and 1-methoxy
compounds as compared to our 1-hydroxy analogues.
Notwithstanding, the modest selectivity of our 1-
hydroxy compounds might suggest the presence of
favorable interactions between the phenyl side chain
and aromatic amino acids that may be present in the
binding pocket of the CB2receptor. Although it is dif-
ficult to draw any direct comparisons aromatic residues
have been proposed to reside in the LBP of the CB1
receptor.21 In contrast, reduced CB1 binding affinities
exhibited by these compounds relative to the cyclo-
hexyl20 and linear chain derivatives18 may suggest that
the compounds cannot adopt a conformation to max-
imize ligand-receptor interactions. The presence of a
polar C10 keto group may also diminish CB1 affinity as
has been previously proposed by Papahatjis and cow-
orkers.18 Further studies utilizing substituted C10 phenyl
groups should provide additional insights into the SAR
of this class of Á8-THC analogues. Combining these
17. Martin, B. R.; Jefferson, R.; Winckler, R.; Wiley, J. L.;
Huffman, J. W.; Crocker, P. J.; Saha, B.; Razdan, R. K. J.
Pharmacol. Exp. Ther. 1999, 290, 1065.
18. Papahatjis, D. P.; Kpurouli, T.; Abadji, V.; Goutopoulos,
A.; Makriyannis, A. J. Med. Chem. 1998, 41, 1195.
19. Papahatjis, D. P.; Nikas, S. P.; Andreou, T.; Makriyannis,
A. Bioorg. Med. Chem. Lett. 2002, 12, 3583.

3490
M. Krishnamurthy et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3487–3490
20. Nadipuram, A.; Krishnamurthy, M.; Ferreira, A. M.; Li,
W.; Moore, B. M. Bioorg. Med. Chem. 2003, 11, 3121.
21. Huffman, J. W.; Mabon, R.; Wu, M.; Lu, J.; Hart, R.;
Hurst, D. P.; Reggio, P. H.; Wiley, J. L.; Martin, B. R. Bioorg.
Med. Chem. 2003, 11, 539.
22. Singer, M.; Ryan, W. J.; Saha, B.; Martin, B. R.; Razdan,
R. K. J. Med. Chem. 1998, 41, 4400.
23. Frenette, R.; Monette, M.; Bernstein, M. A.; Young,
R. N.; Verhoeven, T. R. J. Org. Chem. 1991, 56, 3083.
24. Sondheimer, F.; Rosenthal, D. Tetrahedron. Lett. 1958,
80, 3995.
acetate: petroleum ether 20:80), 1H NMR (CDCl3) d 7.54 ppm
(m, 2H), 7.18 ppm (m, 3H), 6.65 ppm (d, J=2.1 Hz, 1H), 6.39
ppm (d, J=2.1 Hz, 1H), 5.35 ppm (d, J=4.2Hz, 1H), 4.61
ppm (s, 1H), 3.32ppm (m, 4H), 3.1 ppm (m, 1H), 2.62ppm
(m, 1H), 2.06 ppm (m, 1H), 1.76 ppm (m, 3H), 1.62 ppm (s,
3H), 1.29 ppm (s, 3H), 1.03 ppm (s, 3H); MS: (ESI, Neg), m/z
423 ([M-1]ꢀ). HRMS (FAB), m/z calcd for C25H28O2S2,
424.1531, experimental 424.1533.
5. Rf=0.34 (methylene chloride/hexane 50:50), Rf=0.42(ethyl
1
acetate/petroleum ether 10:90), H NMR (CDCl3) d 7.22 ppm
(m, 3H), 7.13 ppm (m, 2H), 6.22 ppm (m, 1H), 5.98 ppm (m,
1H), 5.35 ppm (d, J=6 Hz, 1H), 4.52ppm (s, 1H), 3.74 ppm
(s, 2H), 3.1 ppm (m, 1H), 2.62 ppm (m, 1H), 2.07 ppm (m,
1H), 1.75 ppm (m, 3H), 1.62ppm (s, 3H), 1.29 ppm (s, 3H),
1.03 ppm (s, 3H); MS: (ESI, Neg), m/z 333 ([Mꢀ1]ꢀ). HRMS
(FAB), m/z calcd for C23H26O2, 334.1933, experimental
334.1928.
25. Prasad, R. S.; Dev, S. Tetrahedron 1976, 32, 1437.
26. Reece, C. A.; Rodin, J. O.; Brownlee, R. G.; Duncan,
W. G.; Silverstein, R. M. Tetrahedron 1968, 24, 4249.
27. Huffman, J. W.; Bushell, S. M.; Miller, J. R. A.; Wiley,
J. L.; Martin, B. R. Bioorg. Med. Chem. 2002, 10, 4119.
28. Selected data of final compounds: 3. Rf=0.42(methylene
chloride/hexane 50:50), Rf=0.6 (ethyl acetate: petroleum ether
6. Rf=0.2(methylene chloride/hexane 60:40), Rf=0.47 (ethyl
1
1
10:90), H NMR (CDCl3) d 7.14 ppm (m, 5H), 6.36 ppm (d,
acetate/petroleum ether 20:80), H NMR (CDCl3) d 7.82ppm
J=1.8 Hz, 1H), 5.91 ppm (d, J=2.1 Hz, 1H), 5.35 ppm (d, J=6
Hz, 1H), 4.44 ppm (s, 1H), 3.1 ppm (m, 1H), 2.61 ppm (m, 1H),
2.05 ppm (m, 1H), 1.75 ppm (m, 3H), 1.62 ppm (s, 3H), 1.54
ppm (m, 6H), 1.31 ppm (s, 3H), 1.04 ppm (s, 3H); MS: (ESI,
Neg), m/z 361 ([M-1]ꢀ). HRMS (FAB), m/z calcd for
C25H30O2, 362.2246, experimental 362.2239.
(m, 2H), 7.58 ppm (m, 1H), 7.48 ppm (m, 2H), 6.92 ppm (d,
J=1.8 Hz, 1H), 6.83 ppm (d, J=1.5 Hz, 1H), 5.48 ppm (m,
2H), 3.34 ppm (m, 1H), 2.81 ppm (m, 1H), 2.18 ppm (m, 1H),
1.88 ppm (m, 3H), 1.74 ppm (s, 3H), 1.41 ppm (s, 3H), 1.12
ppm (s, 3H); MS: (ESI, Neg), m/z 347 ([Mꢀ1]ꢀ). HRMS
(FAB), m/z calcd for C23H24O3, 348.1725, experimental
348.1717.
4. Rf=0.22 (methylene chloride/hexane 50:50), Rf=0.58 (ethyl