Structure-activity relationships for 1′,1′-dimethylalkyl-Δ8-tetrahydrocannabinols

Article abstract of DOI:10.1016/S0968-0896(02)00649-1

A series of 1′,1′-dimethylalkyl-Δ⁸-tetrahydrocannabinol analogues with C-3 side chains of 2-12 carbon atoms has been synthesized and their in vitro and in vivo pharmacology has been evaluated. The lowest member of the series, 1′,1′-dimethylethyl-Δ⁸-THC (8, n=0) has good affinity for the CB₁ receptor, but is inactive in vivo. The dimethylpropyl (8, n=1) through dimethyldecyl (8, n=8) all have high affinity for the CB₁ receptor and are full agonists in vivo. 1′,1′-Dimethylundecyl-Δ⁸-THC (8, n=9) has significant affinity for the receptor (K_i=25.8±5.8 nM), but has reduced potency in vivo. The dodecyl analogue (8, n=10) has little affinity for the CB₁ receptor and is inactive in vivo. A quantitative structure-activity relationship study of the side chain region of these compounds is consistent with the concept that for optimum affinity and potency the side chain must be of a length which will permit its terminus to loop back in proximity to the phenolic ring of the cannabinoid.

Full text of DOI:10.1016/S0968-0896(02)00649-1

Bioorganic & Medicinal Chemistry 11 (2003) 1397–1410
Structure–Activity Relationships for 10,10-dimethylalkyl-
D8-tetrahydrocannabinols
John W. Huffman,a,* John R. A. Miller,a John Liddle,a Shu Yu,a Brian F. Thomas,b
Jenny L. Wileyc and Billy R. Martinc
aHoward L. Hunter Laboratory, Clemson University, Clemson, SC 29634-0973, USA
bDepartment of Chemistry and Life Sciences, Research Triangle Institute, PO Box 12194,
Research Triangle Park, NC 27709-2194, USA
cDepartment of Pharmacology and Toxicology, Medical College of Virginia Campus of Virginia Commonwealth University,
Richmond, VA23298-0613, USA
Received 2 July 2002; accepted 15 November 2002
Abstract—A series of 10,10-dimethylalkyl-ꢀ8-tetrahydrocannabinol analogues with C-3 side chains of 2–12 carbon atoms has been
synthesized and their in vitro and in vivo pharmacology has been evaluated. The lowest member of the series, 10,10-dimethylethyl-
ꢀ8-THC (8, n=0) has good affinity for the CB1 receptor, but is inactive in vivo. The dimethylpropyl (8, n=1) through dimethyl-
decyl (8, n=8) all have high affinity for the CB1 receptor and are full agonists in vivo. 10,10-Dimethylundecyl-ꢀ8-THC (8, n=9) has
significant affinity for the receptor (Ki=25.8ꢀ5.8 nM), buthas reduced potency in vivo. The dodecyl analogue ( 8, n=10) has little
affinity for the CB1 receptor and is inactive in vivo. A quantitative structure–activity relationship study of the side chain region of
these compounds is consistent with the concept that for optimum affinity and potency the side chain must be of a length which will
permit its terminus to loop back in proximity to the phenolic ring of the cannabinoid.
# 2003 Elsevier Science Ltd. All rights reserved.
Introduction
ꢀ9-THC (1) has good affinity for the CB1 receptor
(Ki=41ꢀ1.7 nM), and its double bond isomer, ꢀ8-THC
(2) has similar receptor affinity.10 It is known that the
nature of the alkyl side chain has a profound effect upon
the pharmacological activity and CB1 receptor affinities
of cannabinoids related to THC.2ꢁ5,10 Reducing the
length of the side chain by one carbon atom to butyl (3)
results in a reduction in affinity (Ki=65ꢀ13 nM) while
incrementally increasing the length of the side chain to
hexyl (4), heptyl (5) and octyl (6), provides a sys-
tematic increase in affinity from 41ꢀ3.8 to 8.5ꢀ1.4
nM. The in vivo potencies of these ꢀ8-THC homo-
logues are consistent with their relative affinities for
the CB1 receptor.10
Nearly 40 years ago, Gaoni and Mechoulam identified
ꢀ9-tetrahydrocannabinol (ꢀ9-THC, 1) as the major
psychoactive constituent of marijuana.1 Subsequently, a
body of empirical structure–activity relationships
(SARs) was developed for those cannabinoids struc-
turally related to ꢀ9-THC.2ꢁ4 These SAR include a
phenolic hydroxyl group atC-1, an alkyl side chain at
C-3, the absolute stereochemistry depicted in 1 and
substituent effects at C-9. These SAR were incorporated
in the three-point receptor model suggested by both
5
Binder etal. and Howlett et al. which also includes a
C-9 or C-11 hydroxyl group.4,6 Although the nature of
the substituent at C-9 has an effect upon the potency of
cannabinoids structurally related to ꢀ9-THC 1, as noted
by Martin et al., these effects are considerably more
complex than simply the presence or absence of a
hydroxyl group in this region of the molecule.7ꢁ9
Many years ago, Roger Adams found that a 10,20-di-
methylheptyl side chain provides greatly increased
potency in the ꢀ6a,10a-THC series, however the sub-
stitution pattern of this side chain introduces two addi-
tional chiral centers.11 Adams etal. carried outhteir
studies using a mixture 0of stereoisomers. Ultimately the
four isomers of 3-(10,2 -dimethylheptyl)-ꢀ8-THC were
prepared and their stereochemistry elucidated.12
*Corresponding author. Tel.: +1-864-656-3133; fax: +1-864-656-
6613; e-mail: huffman@clemson.edu
0968-0896/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0968-0896(02)00649-1

1398
J. W. Huffman et al. / Bioorg. Med. Chem. 11 (2003) 1397–1410
Although all four are very potent cannabinoids, the
(10S,20R) isomer (7) has the greatest affinity for the CB1
(Ki=0.46ꢀ0.04 nM) receptor and is the most potent in
vivo. Subsequently an enantioselective synthesis of the
resorcinol precursor to this exceptionally potent canna-
binoid was developed.13
a chiral center, and is thus considerably more accessible
synthetically.
In connection with a quantitative structure–activity
relationship (QSAR) analysis of the side chain con-
formation of a variety of ꢀ8-THC analogues, Keimo-
witz et al. found that not only is the length of the
cannabinoid side chain important, but that its ability to
fold back to place the terminus close to the polycyclic
nucleus is critical.18 This conclusion agrees with the
conformation of the 10,10-dimethylheptyl side chain of
the bicyclic nonclassical cannabinoid cannabinoid
CP-47,497 which was determined by NMR19 and the
10,10-Dimethylheptyl-ꢀ6a,10a-THC was prepared by
Adams et al., and, although it was found to be some 20
times more potent than the n-pentyl analogue, it was at
least an order of magnitude less potent than a mixture
of 10,20-dimethylheptyl isomers.11,14 Subsequently, it
was found that 10,10-dimethylheptyl-ꢀ8-THC (ꢀ8-THC-
DMH, 8, n=5) is an exceptionally potent cannabinoid
in vivo, with very high affinity for the CB1 receptor
(Ki=0.77ꢀ0.11 nM).15 In connection with the synthesis
of the synthetic cannabinoid nabilone (9),16 the Lilly
group developed a short, efficient synthesis of 2-(3,5-di-
hydroxyphenyl)-2-methyloctane (10), which has pro-
vided a convenient approach to the synthesis of
cannabinoids with a 10,10-dimethylheptyl side chain.17
The 10,10-dimethylheptyl side chain provides a level of
cannabinoid potency only slightly less than that of the
10,20-dimethylheptyl group, however it does not contain
compactconformaiton of
ꢀ8-THC in a model mem-
brane system.20 The study by Keimowitz et al. included
36 compounds with side chains of four to eight carbon
atoms, and all but one, ꢀ8-THC-DMH (8, n=5), con-
tain unsaturation in the side chain.18
Classical empirical cannabinoid SAR state that a seven-
carbon side chain is optimum for cannabinoid potency
and that 10,10- and 10,20-dimethylation greatly increases
potency.2,4 However, recent data show that 3-octyl-
ꢀ8-THC (6) has somewhatgreaetr affiniyt for ht e CB
1

J. W. Huffman et al. / Bioorg. Med. Chem. 11 (2003) 1397–1410
1399
receptor than 3-heptyl-ꢀ8-THC (5), and both com-
pounds have comparable potency in vivo.10 However,
little is known concerning the SAR of cannabinoids
with 10,10-dimethylalkyl groups other than ꢀ8-THC-
DMH (8, n=5) and the dimethylpentyl analogue
(8, n=3) which was prepared many years ago by Pet-
rzilka et al. and was found to be more potent than ꢀ8-
(2) or ꢀ9-THC (1) in vivo.21,22 In order to define the
relationship between length of the cannabinoid side
chain and biological activity, we now describe the pre-
paration, pharmacology and QSAR for a series of
10,10-dimethylalkyl-ꢀ8-THC homologues (8, n=0–10).
significant affinity for the receptor with Ki=14ꢀ1.8
nM, itis inacitve in vivo. Since ht is compound is inac-
tive in vivo, but has high affinity for the CB1 receptor,
the possibility that it was an antagonist was explored.
Mice were pretreated with 1, 3 or 10 mg/kg of the ligand
and 10 min later were injected with 3 mg/kg of ꢀ9-THC.
The behavioral effects were evaluated in the usual man-
ner (see the Experimental). 10,10-Dimethylethyl-ꢀ8-THC
(8, n=0) had no effecton ꢀ9-THC induced hypo-
thermia or depression of spontaneous activity. How-
ever, in the tail flick protocol to evaluate
antinociception at the 10 mg/kg level this ligand reduced
the per cent maximum possible effect (% MPE) for 3
mg/kg of ꢀ9-THC from 90 to 11%. At the 3 mg/kg level
of dimethylethyl-ꢀ8-THC (8, n=0), the% MPE was
reduced to 67%, however there was no discernible effect
at the 1 mg/kg level. Thus, dimethylethyl-ꢀ8-THC (8,
n=0) appears to be either devoid of activity or a very
weak agonist in the spontaneous activity and rectal
temperature procedures, and to have antagonist prop-
erties in the tail flick protocol.
Results
All of the target cannabinoids were prepared by the acid
catalyzed condensation of an appropriately substituted
resorcinol (11) and trans-para-menthadienol (12,
Scheme 1).21 Several of these compounds (8, n=0–4 and
n=6, 7) had been prepared previously as intermediates
in the synthesis of a series of CB2 selective 1-deoxy-
cannabinoids.23 The substituted resorcinols were pre-
pared from the appropriate tertiary alcohol and 2,6-
dimethoxyphenol using a modification of the procedure
10,10-Dimethylpropyl-ꢀ8-THC (8, n=1, Ki=14ꢀ0.9
nM) is an agonistin all ht ree in vivo procedures, how-
ever itis considerably less poetntht an
ꢀ8-THC (Table
of Dominianni etal.
1). The 10,10-dimethylbutyl analogue (8, n=2) also has
significant affinity for the CB1 receptor (Ki=10.9ꢀ1.7
nM) and is more potent than ꢀ8-THC (2) in the tail
flick measure of antinociception. It is approximately
equipotent to ꢀ8-THC in its ability to decrease rectal
temperature, however this analogue does not show a
dose-dependentresponse in hte depression of spon-
taneous activity.
17,24
The affinities of cannabinoids 8, n=0–4 and 6–10, for
the CB1 receptor were measured by determining their
ability to displace [3H]CP 55,940 from its binding site in
a ratbrain membrane preparaiot n as described by
Compton et al.25 The in vivo pharmacology was eval-
uated using the mouse model of cannabinoid activity
which measures spontaneous activity (SA), antinocicep-
tion (as tail flick, TF) and rectal temperature (RT).26,27
100,100-Dimethylpentyl-ꢀ8-THC (8, n=3) through
1 ,1 -dimethylnonyl-ꢀ8-THC (8, n=7) all have high
affinity for the CB1 receptor with Ki values of 0.77–3.9
nM. The dimethylheptyl (8, n=5, Ki=0.77 nM) and
dimethyloctyl (8, n=6, Ki=0.9ꢀ0.1) analogues exhib-
ited the highest affinities. The affinities of these two
compounds are virtually identical, within experimental
error. The 10,10-dimethylpentyl (8, n=3, Ki=3.9ꢀ0.9
nM), 10,10-dimethylhexyl (8, n=4, Ki=2.7ꢀ1.3 nM)
and 10,10-dimethylnonyl (8, n=7, Ki=1.6ꢀ0.4 nM)
analogues have only slightly lower affinities for the CB1
receptor. The CB1 receptor affinities of these three can-
nabinoids are effectively identical, within experimental
error. All five of these compounds are full agonists in the
mouse and with the exception of the 10,10-dimethylnonyl-
ꢀ8-THC, all are considerably more potent than ꢀ8-THC.
The 10,10-dimethylnonyl analogue is approximately equal
For those cannabinoids with an unsubstituted alkyl
substituent at C-3, the affinity for the CB1 receptor is
maximum for 3-octyl-ꢀ8-THC (6, Ki=8.5ꢀ1.4 nM),
decreases for the heptyl analogue (5, Ki=22ꢀ3.9 nM)
and then decreases further for 3-hexyl-ꢀ8-THC (4,
Ki=41ꢀ3.8 nM) and ꢀ8-THC (2, Ki=44ꢀ12 nM).
3-Butyl-ꢀ8-THC (6) has significantly lower affinity for
the receptor with Ki=65ꢀ13 nM. The in vivo potency
of these compounds is consistent with their respective
affinities for the CB1 receptor, and the details of their
pharmacology have been discussed previously (Table
1).10
In the 3-(10,10-dimethylalkyl)-ꢀ8-THC series (8, n=0–4
and 6–10), although the 10,10-dimethylethyl (8, n=0) has
Scheme 1. (a) HOTs/C6H6, 80 ꢂC.

1400
J. W. Huffman et al. / Bioorg. Med. Chem. 11 (2003) 1397–1410
Table 1. In vitro and in vivo pharmacology of ꢀ8-THC (2) and 3-(10,10-dimethylalkyl)-ꢀ8-THC analogues (8, n=0–10)
Compd
Ki (nM)
ED50 (95% CL)
SA (mmol/kg)
TF (mmol/kg)
RT (mmol/kg)
ꢀ8-THC (2)
44ꢀ12a
65ꢀ13b
41ꢀ3.8b
22ꢀ3.9b
8.5ꢀ1.4b
14ꢀ1.8
14ꢀ0.9
10.9ꢀ1.7
3.9ꢀ0.9
2.7ꢀ1.2
0.77a
0.09ꢀ0.1
1.6ꢀ0.4
6.1ꢀ1.8
25.8ꢀ5.8
126ꢀ18
2.9a
9.0b
1.2b
0.14b
4.8a
10.1b
4.5a
3-Butyl-ꢀ8-THC (3)
6.3b
3-Hexyl-ꢀ8-THC (4)
1.8b
0.10b
0.16b
3-Heptyl-ꢀ8-THC (5)
0.61b
0.34b
>100
3-Octyl-ꢀ8-THC (6)
0.39b
0.24b
3-(10,10-Dimethylethyl-ꢀ8-THC (8 n=0)
3-(10,10-Dimethylpropyl)ꢀ8-THC (8 n=1)
3-(10,10-Dimethylbutyl)ꢀ8-THC (8 n=2)
3-(10,10-Dimethylpentyl)ꢀ8-THC (8 n=3)
3-(10,10-Dimethylhexyl)ꢀ8-THC (8 n=4)
3-(10,10-Dimethylheptyl)ꢀ8-THC (8 n=5)
3-(10,10-Dimethyloctyl)ꢀ8-THC (8 n=6)
3-(10,10-Dimethylnonyl)ꢀ8-THC (8 n=7)
3-(10,10-Dimethyldecyl)ꢀ8-THC (8 n=8)
3-(10,10-Dimethylundecyl)ꢀ8-THC (8 n=9)
3-(10,10-Dimethyldodecyl)ꢀ8-THC (8 n=10
c
d
e
f
24.5 (14.0–43.3)
19. (1.3–2.8)
0.4 (0.4–0.6)
0.21 (0.11–0.28)
0.14a
0.3 (0.19–0.49)
5.1 (3.6–7.0)
37.6 (15.4–92.0)
73k
g
4,2 (3.2–5.6)
1.5 (1.0–2.1)
0.11 (0.06–0.2)
0.15a
1.2 (0.68–2.0)
3.2 (2.4–4.1)
9.4 (6.0–14.8)
73l
1.1 (0.6–2.1)
0.17 (0.01–0.28)
0.27a
0.24 (0.14–0.4)
2.1 (0.4–10.8)h
5.5i
73j
m
Inactive
Inactive
aRef. 15.
bRef. 10.
cActive only at 3.3 mmol/kg, inactive at higher doses.
dMax ꢁ2 ꢂC, notdose responsive.
eMax 84% at95 mmol/kg.
fDecreased from ꢁ2 t oꢁ3.4 ꢂC over a dose range of 0.95–31.8 mmol/kg, no effectat95.5 mmol/kg.
gMax 78% at91 mmol/kg, notdose dependent.
hMax 76%.
iMax 64%.
jMax 56%.
kMax 25%.
lMax ꢁ3 ꢂC.
mStimulation at 68, 22.7 and 68 mmol/kg.
in potency to ꢀ8-THC. None of these cannabinoids
show any selectivity among the three pharmacological
measures of the in vivo protocol.
than four carbon atoms (8, n=0, 1) in which the side
chains are too short to wrap around the molecule are
inactive or weakly active in vivo, although they have
significant affinity for the CB1 receptor. The 10,10-di-
methylbutyl to 10,10-dimethylnonyl-ꢀ8-THC analogues
(8, n=2–7) all have high receptor affinity, and potent in
vivo activity. It would appear that these ligands have
side chains of sufficient length to wrap around the
molecule to give the requisite compact conformation.
As the length of the side chain increases from 10 to 12
carbon atoms (8, n=8–10) the receptor affinity decrea-
ses incrementally and the in vivo potency decreases to
the point that 10,10-dimethyldodecyl-ꢀ8-THC is effec-
tively inactive. It seems probable that as the side chain
extends beyond nine or 10 carbon atoms it can no
longer adopt the compact conformation necessary for
cannabinoid activity. In order to investigate the possible
validity of this hypothesis a QSAR study employing
methodology similar to that used by Keimowitz et al.
was carried out.18
10,10-Dimethyldecyl-ꢀ8-THC (8, n=8, Ki=6.1ꢀ1.8
nM) has only slightly lower affinity for the CB1 receptor
than the lower homologues: however, it is considerably
less potent than the dimethylnonyl analogue in the tail
flick measure of antinociception and somewhat less
potent in its ability to produce hypoactivity or hypo-
thermia. It also failed to produce complete suppression
of spontaneous activity. The 10,10-dimethylundecyl ana-
logue (8, n=9) has moderate affinity for the CB1 recep-
tor (Ki=25.8ꢀ5.8 nM), but greatly reduced potency in
the mouse. This compound is a partial agonist in the
three procedures of the in vivo protocol, with maximum
effects of 56% in the measure of hypoactivity, 25% in
the tail flick measure of antinociception and a maximum
decrease in temperature of 3 ꢂC. The 10,10-dimethyl-
dodecyl analogue (8, n=10) has considerably dimin-
ished affinity for the CB1 receptor (Ki=126ꢀ18 nM),
and is inactive in the tail flick measure of antinocicep-
tion. It is also inactive in causing hypothermia and is a
stimulant rather than a depressant.
Quenched molecular dynamics
The quenched molecular dynamics approach used for
conformational sampling generated 100 low-energy
conformations for each molecule. The conformations
were quite diverse for certain molecules, while other
molecules repeatedly yielded a small number of simi-
lar conformations. These differences in conforma-
tional mobility can be visualized graphically by
overlaying the conformations for a particular mole-
cule (Fig. 1).
Both the in vitro and in vivo pharmacology data are
consistent with the model suggested by Keimowitz et al.
which concludes that for dibenzopyran-based canna-
binoids, receptor affinity and potency are enhanced by a
side chain of sufficientlength thatitcan wrap backward
along either side of the molecule.18 Those 10,10-di-
methylalkyl-ꢀ8-THC analogues with side chains of less

J. W. Huffman et al. / Bioorg. Med. Chem. 11 (2003) 1397–1410
1401
Figure 1. Stereoviews of the conformational ensembles for the compounds in Table 1.
CoMFA
The cross-validated analysis of the relationship between
the CoMFA molecular fields and the pharmacological
affinity and potency measurements generally indicated
that a model derived with five components was optimal.
The strength of the cross-validated and final models can
be demonstrated through comparison with similar rela-
tionships derived with random pharmacological data
that spanned the same range as the real pharmacological
data. In all but one instance, the cross-validated or final
r-squared values obtained with the real pharmacological
data were higher (indicating a better model) than when
derived using random pharmacological data (Table 2).
Table 2. Comparison of the ability of each QSAR model to fit actual
biological data or randomly generated biological data
Actual
data r2
Random
data r2
Actual
data r2
Random
data r2
Cross-validated
Final
CB1 (KI)
0.464
0.408
0.454
0.370
0.449
0.035
0.296
0.430
0.692
0.596
0.638
0.603
0.580
0.331
0.430
0.550
SA (ED50)
TF (ED50)
RT (ED50)

Figure 2. Predicted versus actual plots and stereoviews of the QSARS derived for the steric fields (yellow and green contours) and eletrostatic fields (blue and red) as defined for cannabinoid receptor (CB1)
binding affinity (A) and as defined for cannabinoid-induced changes in spontaneous locomotor activity (B), tail-flick latency (TF) and rectal temperature (RT). For each model, a 75/25 level of contribution
is shown. The steric plots are depicted so that steric bulk should be moved closer to areas contoured in green and farther from regions in yellow in order to increase the target property being contoured (i.e.,
affinity or potency). The electrostatic plots are contoured such that positive charge should be moved closer to regions contoured in blue and farther from regions contoured in red in order to increase the
target property being contoured.

J. W. Huffman et al. / Bioorg. Med. Chem. 11 (2003) 1397–1410
1403
Although there was not a single instance of a model
derived with random data possessing a higher r2 value
than a model derived with the actual pharmacological
data, the ability of the model to fit the potency of these
compounds in the rectal temperature assay was quite
similar to that determined for random data. The utiliza-
tion of several conformations for each analogue allows
for unbiased analyses to occur. However, it often can also
produce various estimated potencies for one particular
compound (see predicted vs actual plots in Fig. 2), and
therefore can present a greater challenge for the CoMFA
approach in deriving an accurate QSAR model. How-
ever, it has been reported that use of multiple conformers
generated by molecular dynamics to determine and
compare conformationally accessible regions can lead to
better results in QSAR studies than that determined
utilizing single conformations of each analogue.31
for receptor affinity (correlation coefficient for pKd of
0.82). In this study, chain length was directly related to
receptor affinity, yet the angle made by the side chain
from its attachment point to its terminus was found to
be inversely related to affinity. Thus, this study indi-
cated that while increased side-chain length can be
associated with increased affinity, the side-chain’s con-
formation mobility must not be restricted to extending
straight away from the ring system, but must allow its
wrapping back around towards the ring system. One
might further conclude from the studies presented here
that the addition of the 1,1-dimethyl group on alkyl
chains of sufficientlenght induces hte bentconforma-
tions that appear to be most associated with potent
cannabinoid ligands of high CB1 receptor affinity.
It is intriguing that the analogues with long side chains,
10,10-dimethyldecyl-ꢀ8-THC and 10,10-dimethylundecyl-
ꢀ8-THC, retained receptor affinity while losing biologi-
cal activity. It is possible that these analogues were able
to interact with the receptor when in a broken cell pre-
paration but not in intact tissue. However, the analogue
with the shortest side chain, 10,10-dimethylethyl-ꢀ8-
THC also bound with excellent affinity but lacked in
vivo efficacy. Therefore, a more likely explanation is
that the length of the side chain has a greater influence
on receptor activation than on receptor recognition. As
such, these side-chain analogues demonstrate structural
features that distinguish between receptor affinity and
receptor activation.
Visualization of CoMFA fields
Three-dimensional contour plots of the CoMFA model
allow the visualization of regions where changes in
steric or electrostatic properties are correlated with
experimentally determined differences in biological
properties. The contour plots in Figure 2 display the
QSAR model for both receptor affinity and pharma-
cological potency. Inspection of the steric contour plots
reveals a relatively consistent side-chain SAR for both
receptor affinity and behavioral potency. A large con-
tour in yellow surrounds a smaller contour in green in
the side chain region of the analogues. The green con-
tour, in most of the models, actually starts on one side
of the aromatic ring and wraps around the side-chain
end of the molecule at approximately the level of the
C-30 atom until ending on the opposite face of the aro-
matic ring. Thus, the model predicts decreased affinity
and potency for molecules whose side chains prefer an
extended conformation. Conversely, analogues whose
conformational mobility allows their side chain to bend
so that the terminus is alongside the aromatic ring are
associated with increased predicted affinity and potency.
The electrostatic plots reveal blue contours that indicate
that compounds with positive charge density in this
region would be predicted to possess increased
pharmacological activity. This area corresponds to the
increased positive point charges (derived by simple
Gassteiger–Huckle calculations) in this region when
methyl groups are placed at the C-1 position of the alkyl
side chain.
Experimental
General
IR spectra were obtained using Nicolet 5DX or Magna
1
spectrometers; H and 13C NMR spectra were recorded
on a Bruker 300AC spectrometer. Mass spectral ana-
lyses were performed on a Hewlett-Packard 5890A gas
chromatograph with a mass sensitive detector and
HRMS data were obtained in the Mass Spectrometry
Laboratory, School of Chemical Sciences, University of
Illinois. Ether and THF were distilled from Na-benzo-
phenone ketyl immediately before use, and other sol-
vents were purified using standard procedures. Column
chromatography was carried out on Universal silica gel
(32–63 m) using the indicated solvents as eluents. All
compounds were homogeneous to 13C NMR TLC and/
or glc.
The QSAR results are consistent with the concept that
for optimum affinity and potency the side chain’s con-
formational freedom must include conformations where
its terminus loops back and comes in closer proximity to
the phenolic ring. Thus alkyl chains of only certain
lengths can achieve this ‘active’ conformation, and side-
chains which are too short are unable to adopt these
conformations. For side chains that are too long, it is
likely that these conformations are energetically unfa-
vorable. This conclusion is consistent with the multiple
linear regression analysis of Keimowitz et al.18 where two
descriptive variables describing side chain length and ter-
minus position were able to fit the pharmacological data
3-(10,10-Dimethylethyl)-D8-tetrahydrocannabinol (8, n=0).
The cannabinoid was prepared as described pre-
viously:23 1H NMR (300 MHz, CDCl3) d 1.12 (s, 3H),
1.23 (s, 9H), 1.40 (s, 3H), 1.70 (s, 3H), 1.75–1.94 (m,
3H), 2.14 (m, 1H), 2.72 (dt, J=4.6, 10.8 Hz, 1H), 3.22
(dd, J=4.6, 16.3 Hz, 1H), 5.04 (s, 1H), 5.43 (d, J=4.0
Hz, 1H), 6.30 (d, J=1.6 Hz, 1H), 6.46 (d, J=1.6 Hz,
1H); 13C NMR (75.5 MHz, CDCl3) d 18.5, 23.5, 27.6,
27.8, 31.1, 31.4, 34.3, 35.9, 44.8, 76.7, 104.9, 107.3,
110.3, 119.3, 134.7, 151.1, 154.4, 154.5; MS (EI) m/z 301
(30), 300 (90), 217 (100);[a]2D0 ꢁ83ꢂ (c0.10, CHCl3);
HRMS, calcd for C20H28O2: 300.2084, found 300.2089..

1404
J. W. Huffman et al. / Bioorg. Med. Chem. 11 (2003) 1397–1410
2-Methyl-2-(3,5-dimethoxyphenyl)butane. A mixture of
2.0 g (22.7 mmol) of tertiary amyl alcohol, and 3.5 g
(22.7 mmol) of 2,6-dimethoxyphenol was warmed to
approximately 50 ꢂC with stirring to effect solution,
cooled to 0 ꢂC, and 5.9 mL (81.0 mmol) of methane-
sulfonic acid was added. The mixture was stirred at 0 ꢂC
for 3 h, allowed to warm to ambient temperature and
stirred for an additional 14 h. The reaction mixture was
poured onto ice, extracted with CH2Cl2, and the
extracts were washed with water, saturated aqueous
NaHCO3 and dried (MgSO4). The solventwas removed
in vacuo to give 4.8 g (94%) of 2-methyl-2-(3,5-di-
methoxy-4-hydroxyphenyl)butane as a brown oil which
was used in the subsequent step without further puri-
give 0.15 g (96%) of crude substituted resorcinol as a
brown oil which was used in the next step without fur-
ther purification: 1H NMR (300 MHz, CDCl3) d 0.58 (t,
J=7.3 Hz, 3H), 1.09 (s, 6H), 1.47 (q, J=7.3 Hz, 2H),
6.22 (s, 1H), 6.42 (s, 2H), 6.70 (br s, 2H); 13C NMR
(75.5 MHz, CDCl3) d 9.0, 28.1, 36.5, 37.9, 100.2, 106.3,
153.5, 155.7.
The cannabinoid was prepared using the procedure
described above for the synthesis of 3-(10, 10-dimethyl-
ethyl)-ꢀ8-tetrahydrocannabinol (8, n=0). From 0.15 g
(0.83 mmol) of crude resorcinol, 0.13 g (0.86 mmol) of
trans-p-menthadienol and 0.02 g (10 mol%) of p-tolue-
nesulfonic acid monohydrate, there was obtained 0.18 g
(69%) of cannabinoid as a viscous pale brown oil: Rf
0.15 (petroleum ether/ethyl acetate 25:1), 0.32 (petro-
leum ether/ethyl acetate 10:1); 1H NMR (300 MHz,
CDCl3) d 0.66 (t, J=7.4 Hz, 3H), 1.11 (s, 3H), 1.16 (s,
6H), 1.39 (s, 3H), 1.52 (q, J=7.4 Hz, 2H), 1.68 (s, 3H),
1.74–1.90 (m, 3H), 2.12 (m, 1H), 2.71 (dt, J=4.7, 11.0
Hz, 1H), 3.20 (dd, J=4.7, 16.5 Hz, 1H), 5.09 (s, 1H),
5.42 (d, J=4.0 Hz, 1H), 6.23 (d, J=1.6 Hz, 1H), 6.39
(d, J=1.6 Hz, 1H); 13C NMR (75.5 MHz, CDCl3) d 9.2,
18.5, 23.4, 27.5, 27.8, 28.2, 31.5, 35.9, 36.6, 37.5, 44.8,
76.7, 105.6, 108.0, 110.2, 119.3, 134.7, 149.5, 154.3,
154.5; MS (CI) m/z 315 (15), 57 (100); [a]2D0 ꢁ126ꢂ
(c0.57, CHCl3); HRMS, calcd for C21H30O2: 314.2251,
found 314.2246.
1
fication: H NMR (300 MHz, CDCl3) d 0.67 (t, J=7.3
Hz, 3H), 1.24 (s, 6H), 1.58 (q, J=7.3 Hz, 2H), 3.86 (s,
6H), 6.52 (s, 2H); 13C NMR (75.5 MHz, CDCl3) d 9.0,
28.5, 36.9, 37.8, 56.2, 103.0, 132.4, 140.6, 146.4.
To a stirred solution of 2.1 ꢂg (9.4 mmol) of the above
phenol in 5 mL of CCl4 at0 C was added 1.5 mL (11.6
mmol) of diethyl phosphite, followed by the dropwise
addition of 1.8 mL (12.9 mmol) of Et3N. The reaction
mixture was stirred at 0 ꢂC for 1 h and atambienttem-
perature for 16 h. After dilution with CH2Cl2, the reac-
tion was washed with water, 10% aqueous NaOH,
water, 10% aqueous HCl, brine and dried (MgSO4).
The solventwas removed in vacuo to afford 3.1 g (92%)
of phosphate ester as an orange-cream solid which was
1
used in the next step without further purification: H
2-Methyl-2-(3,5-dimethoxyphenyl)pentane. The dimethyl
ether was prepared using the procedure described for
NMR (300 MHz, CDCl3) d 0.82 (t, J=7.1 Hz, 3H), 1.25
(s, 6H), 1.35–1.43 (m, 6H), 1.52 (q, J=7.1 Hz, 2H), 3.82
(s, 6H), 4.18–4.35 (m, 4H), 6.53 (s, 2H); 13C NMR
(75.5 MHz, CDCl3) d 15.90, 15.94, 16.5, 28.9, 36.4, 45.2,
55.8, 64.0, 102.3, 127.1, 148.2, 151.0.
the
synthesis
of
2-methyl-2-(3,5-dimethoxy-
phenyl)propane. From 3.31 g (32.75 mmol) of 2-methyl-
2-pentanol and 5.0 g (32.4 mmol) of 2,6-dimethoxy-
phenol, there was obtained 7.1 g (92%) of a brown oil
which was used in the subsequent step without further
purification: 1H NMR (300 MHz, CDCl3) d 0.80 (t,
J=7.4 Hz, 3H), 1.00–1.30 (m, 2H), 1.21 (s, 6H), 1.45–
1.57 (m, 2H), 3.86 (s, 6H), 5.40 (s, 1H), 6.52 (s, 2H); 13C
NMR (75.5 MHz, CDCl3) d 14.7, 17.9, 29.0, 37.7, 47.2,
56.2, 102.9, 132.5, 141.1, 146.5.
To 50 mL of liquid NH3 at ꢁ78 ꢂC was added 0.35 g (50
mg atoms) of lithium. The solution was stirred for 10
min, and 3.0 g (8.3 mmol) of the above phosphate in 3
mL of dry THF was added dropwise. The reaction
mixture was stirred at ꢁ78 ꢂC for 2 h followed by the
careful addition of solid NH4Cl to destroy the excess
lithium. The ammonia was allowed to evaporate at
ambient temperature. The solid residue was slurried
with water, extracted with ether and the combined
organic extracts were washed with brine and dried
(MgSO4). The solventwas evaporaetd in vacuo ot give
1.1 g (63%) of 2-methyl-2-(3,5-dimethoxyphenyl)butane
as a pale yellow oil after distillation (110 ꢂC/0.5 mmHg):
1H NMR (300 MHz, CDCl3) d 0.72 (t, J=7.4 Hz, 3H),
1.28 (s, 6H), 1.64 (q, J=7.4 Hz, 2H), 3.81 (s, 6H), 6.33
(t, J=2.2 Hz, 1H), 6.52 (d, J=2.2 Hz, 2H); 13C NMR
(75.5 MHz, CDCl3) d 9.1, 28.4, 36.7, 38.1, 55.0, 96.5,
104.7, 152.1, 160.4; MS (EI) m/z 208 (25), 179 (100).
From 5.0 g (21.0 mmol) of phenol, there was obtained
7.5 g (95%) of phosphate ester as a yellow oil which was
1
used in the next step without further purification: H
NMR (300 MHz, CDCl3) d 0.80 (t, J=7.2 Hz, 3H),
1.03–1.11 (m, 2H), 1.19 (s, 6H), 1.37 (t, J=7.2 Hz, 6H),
1.49–1.54 (m, 2H), 3.84 (s, 6H), 4.25–4.34 (m, 4H), 6.52
(s, 2H); 13C NMR (75.5 MHz, CDCl3) d 14.67, 15.96,
16.1, 17.8, 28.9, 38.0, 47.0, 56.0, 64.0, 64.1, 103.0, 127.1,
147.1, 151.1.
From 6.0 g (16.0 mmol) of phosphate, there was
obtained 3.30 g (93%) of 2-methyl-2-(3,5-dimethoxy-
phenyl)pentane as
a
pale yellow oil: 1H NMR
3-(10,10-Dimethylpropyl)-D8-tetrahydrocannabinol
(8,
(300 MHz, CDCl3) d 0.81 (t, J=7.3 Hz, 3H), 1.04–1.11
(m, 2H), 1.25 (s, 6H), 1.53 (m, 2H), 3.79 (s, 6H), 6.29 (t,
J=2.2 Hz, 1H), 6.49 (d, J=2.2 Hz, 2H); 13C NMR
(75.5 MHz, CDCl3) d 14.8, 18.0, 28.9, 38.0, 47.0, 55.2,
96.5, 104.7, 152.6, 160.4.
n=1). To 0.18 g (0.87 mmol) of 2-methyl-2-(3,5-di-
methoxyphenyl)butane at 0 ꢂC was added 2.2 mL (22
mmol) of 1.0 M BBr3 in CH2Cl2. The reaction mixture
was warmed to ambient temperature, stirred for 18 h
and carefully poured into ice water. After extraction
with CH2Cl2, the organic extracts were washed with
water, and dried (MgSO4). The solventwas removed to
3-(10,10-Dimethylbutyl)-D8-tetrahydrocannabinol (8, n=2).
The cannabinoid was prepared using the procedure

J. W. Huffman et al. / Bioorg. Med. Chem. 11 (2003) 1397–1410
1405
described for the synthesis of 3-(10,10-dimethylethyl)-ꢀ8-
tetrahydrocannabinol. From 3.20 g (14.4 mmol) of
2-methyl-2-(3,5-dimethoxyphenyl)pentane, there was
obtained 2.80 g (100%) of crude substituted resorcinol
as a brown oil which was used in the next step without
further purification: 1H NMR (300 MHz, CDCl3) d 0.75
(t, J=7.3 Hz, 3H), 0.97–1.03 (m, 2H), 1.14 (s, 6H),
1.40–1.46 (m, 2H), 6.22 (br s, 3H), 6.42 (s, 2H); 13C
NMR (75.5 MHz, CDCl3) d 14.6, 17.8, 28.6, 37.7, 46.8,
100.1, 106.1, 153.7, 155.9.
The phenol was converted to the phosphate ester as
described above. From 7.50 g (28.2 mmol) of phenol,
there was obtained 11.9 g of crude ester as a yellow solid
which was used in the next step without further purifi-
1
cation: H NMR (300 MHz, CDCl3) d 0.83 (t, J=7.1
Hz, 3H), 1.00–1.50 (m, 6H), 1.50–1.68 (m, 2H), 1.26 (s,
6H), 1.39 (t, J=6.9 Hz, 6H), 3.86 (s, 6H), 4.19–4.46 (m,
4H), 6.54 (s, 2H); 13C NMR (75.5 MHz, CDCl3) d 13.8,
15.8, 15.9, 22.2, 24.0, 28.8, 32.3, 37.8, 44.3, 55.8, 63.8,
63.9, 102.8, 127.1, 146.9, 151.0.
From 2.70 g (13.9 mmol) of crude resorcinol, 2.20 g
(14.5 mmol) of trans-p-menthadienol and 0.20 g (10
mol%) of p-toluenesulfonic acid monohydrate, there
was obtained 3.50 g (77%) of cannabinoid as a viscous
pale yellow oil: Rf 0.19 (petroleum ether/ethyl acetate
25:1), 0.38 (petroleum ether/ethyl acetate 10:1); 1H
NMR (300 MHz, CDCl3) d 0.82 (t, J=7.3 Hz, 3H),
1.05–1.31 (m, 2H), 1.12 (s, 3H), 1.22 (s, 6H), 1.40 (s,
3H), 1.47–1.52 (m, 2H), 1.71 (s, 3H), 1.75–1.94 (m, 3H),
2.14 (m, 1H), 2.62–2.75 (m, 1H), 3.12–3.28 (m, 1H), 4.65
(s, 1H), 5.44 (br s, 1H), 6.24 (s, 1H), 6.40 (s, 1H); 13C
NMR (75.5 MHz, CDCl3) d 14.7, 17.9, 18.5, 23.4, 27.5,
27.8, 28.6, 28.7, 31.5, 35.9, 37.3, 44.9, 46.9, 76.8, 105.5,
107.8, 110.2, 119.2, 134.7, 149.8, 154.2, 154.5; MS (EI)
m/z 329 (20), 328 (40), 286 (100); [a]2D0 ꢁ258ꢂ (c2.2,
CHCl3); HRMS, calcd for C22H32O2: 238.2402, found
328.2392.
The phosphate ester was reduced as described above.
From 11.88 g (28.4 mmol) of phosphate, there was
obtained 5.81 g (81% from 2-heptanone) of 2-methyl-2-
(3,5-dimethoxyphenyl)heptane as a colorless oil: 1H
NMR (300 MHz, CDCl3) d 0.82 (d, J=7.0 Hz, 3H),
0.98–1.36 (m, 6H), 1.26 (s, 6H), 1.50–1.65 (m, 2H), 3.79
(s, 6H), 6.30 (t, J=2.2 Hz, 1H), 6.49 (d, J=2.3 Hz, 2H);
13C NMR (75.5 MHz, CDCl3) d 14.1, 22.5, 24.3, 28.9,
32.5, 37.9, 44.4, 55.1, 96.5, 104.6, 152.5, 160.4.29
3-(10,10-Dimethylhexyl)-D8-tetrahydrocannabinol
(8,
n=4). The cannabinoid was prepared from 1.16 g (4.6
mmol) of 2-methyl-2-(3,5-dimethoxyphenyl)heptane by
the procedure described above for the synthesis of
3-(10,10-dimethylbutyl)-ꢀ8-tetrahydrocannabinol. There
was obtained after chromatography (petroleum ether/
ethyl acetate 8:1) 1.44 g (86%) of cannabinoid as a vis-
cous pale yellow oil: Rf 0.25 (hexanes/ethyl acetate
25:1), 0.45 (hexanes/ethyl acetate 10:1); 1H NMR
(300 MHz, CDCl3) d 0.81 (t, J=7.1 Hz, 3H), 0.97–1.30
(m, 6H), 1.10 (s, 3H), 1.17 (s, 6H), 1.39 (s, 3H), 1.68 (s,
3H), 1.41–1.54 (m, 2H), 1.70–1.97 (m, 3H), 2.06–2.23
(m, 1H), 2.62–2.79 (m, 1H), 3.18–3.32 (m, 1H), 5.40 (s,
1H), 5.98 (s, 1H), 6.27 (d, J=1.6 Hz, 1H), 6.38 (d,
J=1.7 Hz, 1H); 13C NMR (75.5 MHz, CDCl3) d 14.0,
18.4, 22.5, 23.4, 24.2, 27.5, 27.8, 28.5, 28.6, 31.5, 32.5,
35.9, 37.1, 44.3, 44.9, 79.6, 105.4, 107.5, 110.2, 119.2,
134.7, 149.7, 154.2, 154.8; [a]2D0 ꢁ174ꢂ (c1.38, CHCl3);
HRMS, calcd for C24H36O2: 356.2715, found 356.2713.
3-(10,10-Dimethylpentyl)-D8-tetrahydrocannabinol
(8,
n=3). The cannabinoid was prepared from 1.06 g (5.08
mmol) of 2-(3,5-dihydroxyphenyl)-2-methylhexane,
0.690 g (4.53 mmol) of trans-p-menthadienol and 0.096
g (10 mol%) of p-toluenesulfonic acid monohydrate as
described by Petrzilka and Sikemeier21 to give 1.36 g
(78%) of cannabinoid as a viscous pale yellow oil after
chromatography (Petroleum ether/ethyl acetate 8:1): 1H
NMR (300 MHz, CDCl3) d 0.82 (t, J=7.3 Hz, 3H),
0.98–1.34 (m, 4H), 1.11 (s, 3H), 1.20 (s, 6H), 1.39 (s,
3H), 1.42–1.56 (m, 2H), 1.70 (s, 3H), 1.73–1.96 (m, 3H),
2.06–2.22 (m, 1H), 2.62–2.77 (m, 1H), 3.12–3.27 (m,
1H), 4.90 (s, 1H), 5.40–5.48 (m, 1H), 6.23 (d, J=1.7 Hz,
1H), 6.39 (d, J=1.6 Hz, 1H); 13C NMR (75.5 MHz,
CDCl3) d 14.1, 18.5, 23.4, 23.5, 26.9, 27.6, 28.7, 31.5,
36.0, 37.2, 44.2, 44.9, 76.6, 105.4, 108.0, 110.2, 119.3,
134.7, 150.0, 154.5; HRMS, calcd for C23H34O2:
342.2559, found 342.2558; [a]2D0 ꢁ212ꢂ (c0.857, CHCl3),
2-Methyl-2-(3,5-dimethoxyphenyl)nonane. The reaction
of 4.39 g (27.7 mmol) of 2-methyl-2-nonanol,30 pre-
pared from 2-nonanone and methylmagnesium bro-
mide, with 4.40 g (28.5 mmol) of 2,6-dimethoxyphenol
was carried outas described above ot give 8.83 g of
crude
2-methyl-2-(4-hydroxy-3,5-dimethoxyphenyl)-
lit.9 [a]2D0 ꢁ237ꢂ. The H NMR data are consistent with
nonane as an oil which was used in the subsequent step
without further purification: 1H NMR (300 MHz,
CDCl3) d 0.86 (t, J=7.1 Hz, 3H), 1.26 (s, 6H), 0.96–1.35
(m, 10H), 1.47–1.61 (m 2), 3.88 (s, 6H), 6.54 (s, 2H); 13C
NMR (75.5 MHz, CDCl3) d 14.0, 22.5, 24.6, 29.1, 30.2,
31.8, 37.6, 44.6, 56.2, 102.9, 132.4, 141.0, 146.4.
1
those reported by Petrzilka and Sikemeier.21
2-Methyl-2-(3,5-dimethoxyphenyl)heptane. The reaction
of 3.86 g (28.7 mmol) of 2-methyl-2-heptanol,28 pre-
pared from 2-heptanone and methylmagnesium bro-
mide, with 4.42 g (28.7 mmol) of 2,6-dimethoxyphenol
was carried outas described above ot give 7.51 g of
2-methyl-2-(4-hydroxy-3,5-dimethoxyphenyl)heptane as
an oil which was used in the subsequent step without
further purification: 1H NMR (300 MHz, CDCl3) d
0.83 (t, J=7.0 Hz, 3H), 1.26 (s, 6H), 0.97–1.36 (m,
6H), 1.48–1.60 (m, 2H), 3.87 (s, 6H), 5.50 (br s, 1H),
6.54 (s, 2H); 13C NMR (75.5 MHz, CDCl3) d 13.9,
22.4, 24.2, 29.0, 32.4, 37.5, 44.5, 56.1, 102.9, 132.4,
140.9, 146.4.
The phenol was converted to the phosphate ester as
described above. From 8.83 g (27.7 mmol) of phenol,
there was obtained 11.5 g of ester as a yellow solid
which was used in the next step without further purifi-
1
cation: H NMR (300 MHz, CDCl3) d 0.86 (t, J=7.2
Hz, 3H), 1.26 (s, 6H), 0.97–1.33 (m, 10H), 1.38 (t,
J=7.1 Hz, 3H), 1.39 (t, J=7.1 Hz, 3H), 1.50–1.63 (m,
2H), 3.85 (s, 6H), 4.30 (q, J=7.2 Hz, 2H), 4.32 (q,
J=7.2 Hz, 2H), 6.53 (s, 2H); 13C NMR (75.5 MHz,

1406
J. W. Huffman et al. / Bioorg. Med. Chem. 11 (2003) 1397–1410
CDCl3) d 13.9, 15.9, 16.0, 22.5, 24.5, 28.9, 30.1, 31.7,
37.9, 44.4, 55.9, 63.9, 64.0, 103.0, 127.0, 147.0, 151.1.
(petroleum ether/ethyl acetate 7:1) 3.69 g of recovered
phosphate ester and 2.48 g (31% from 2-decanone,
based on starting material consumed) of 2-methyl-2-
(3,5-dimethoxyphenyl)nonane as a colorless oil: 1H
NMR (300 MHz, CDCl3) d 0.86 (t, J=6.7 Hz, 3H),
0.99–1.37 (m, 12H), 1.25 (s, 6H), 1.50–1.63 (m, 2H),
3.78 (s, 6H), 6.29 (t, J=2.1 Hz, 1H), 6.49 (d, J=2.2 Hz,
2H); 13C NMR (75.5 MHz, CDCl3) d 14.0, 22.6, 24.6,
28.9, 29.3, 29.4, 30.3, 31.8, 37.9, 44.4, 55.0, 96.5, 104.5,
152.4, 160.4. Anal. calcd for C19H32O2: C, 78.03; H,
11.03; Found: C, 77.89; H, 10.90.
The phosphate ester was reduced as described above for
the synthesis of 3-pentylmethoxybenzene. From 11.5 g
(26.7 mmol) of phosphate there was obtained 7.3g (93%
from 2-nonanone) of 2-methyl-2-(3,5-dimethoxy-
1
phenyl)nonane as a colorless oil: H NMR (300MHz,
CDCl3) d 0.85 (t, J=7.0 Hz, 3H), 0.98–1.35 (m, 10H), 1.25
(s, 6H), 1.48–1.62 (m, 2H), 3.79 (s, 6H), 6.30 (t, J=2.2 Hz,
1H), 6.48 (d, J=2.2 Hz, 2H); 13C NMR (75.5 MHz,
CDCl3) d 14.1, 22.6, 24.7, 29.0, 29.2, 30.3, 31.9, 38.0, 44.5,
55.2, 96.6, 104.7, 152.8, 160.4. Anal. calcd for C18H30O2:
C, 77.65; H, 10.86; Found: C, 77.78; H, 10.88.
3-(10,10-Dimethylnonyl)-D8-tetrahydrocannabinol
(8,
n=7). The cannabinoid was prepared from 0.829 g
(2.83 mmol) of 2-methyl-2-(3,5-dimethoxyphenyl)-
decane by the method described above for the synthesis
of 3-(10,10-dimethylbutyl)-ꢀ8-THC. There was obtained
after chromatography (petroleum ether/ethyl acetate
8:1) 0.573 g (69%) of cannabinoid as a viscous pale
yellow oil: Rf 0.29 (hexanes/ethyl acetate 25:1), 0.50
(hexanes/ethyl acetate 10:1); 1H NMR (300 MHz,
CDCl3) d 0.86 (t, J=7.0 Hz, 3H), 1.11 (s, 3H), 1.18 (s,
6H), 1.39 (s, 3H), 1.69 (s, 3H), 0.96–1.39 (m, 12H),
1.42–1.57 (m, 2H), 1.75–1.98 (m, 3H, 2.06–2.23 (m, 1H),
2.61–2.78 (m, 1H), 3.13–3.30 (m, 1H), 5.41 (br s, 1H),
5.44 9s, 1H), 6.24 (d, J=1.6 Hz, 1H), 6.38 (d, J=1.6
Hz, 1H); 13C NMR (75.5 MHz, CDCl3) d 14.1, 18.4,
22.6, 23.4, 27.5, 27.8, 28.6, 28.7, 29.3, 29.5, 30.3, 31.5,
31.8, 35.9, 37.2, 44.4, 44.8, 76.6, 105.4, 107.7, 110.2,
119.2, 134.7, 149.9, 154.3, 154.6; MS (EI) m/z (rel
intensity) 399 (20), 316 (15), 286 (100), 217 (17); [a]D20
ꢁ122ꢂ (c0.57, CHCl3); HRMS, calcd for C27H42O2:
398.3185, found 398.3181.
3-(10,10-Dimethyloctyl)-D8-tetrahydrocannabinol (8, n=6).
The cannabinoid was prepared from 0.737 g (2.65
mmol) of 2-methyl-2-(3,5-dimethoxyphenyl)nonane as
described above for the synthesis of 3-(10,10-dimethyl-
butyl)-ꢀ8-THC. There was obtained after chromato-
graphy (petroleum ether/ethyl acetate 8:1) 0.784 g
(77%) of cannabinoid as a viscous pale yellow oil: Rf
0.27 (hexanes/ethyl acetate 25:1), 0.47 (hexanes/ethyl
1
acetate 10:1); H NMR (300 MHz, CDCl3) d 0.85 (t,
J=7.0 Hz, 3H), 1.11 (s, 3H), 0.98–1.35 (m, 10H), 1.18
(s, 6H), 1.39 (s, 3H), 1.44–1.55 (m, 2H), 1.69 (s, 3H),
1.64–1.96 (m, 3H), 2.13–2.24 (m, 1H), 2.62–2.79 (m,
1H), 3.13–3.31 (m, 1H), 5.16 (s, 1H), 5.36–5.47 (m, 1H),
6.23 (d, J=1.8 Hz, 1H), 6.38 (d, J=1.6 Hz, 1H); 13C
NMR (75.5 MHz, CDCl3) d 14.1, 18.5, 22.6, 23.5, 24.6,
27.6, 27.9, 28.6, 29.2, 30.3, 31.5, 31.9, 35.9, 37.2, 44.4,
44.8, 76.7, 105.4, 107.8, 110.2, 119.3, 134.7, 149.9, 154.4,
154.5; MS (EI) m/z (rel intensity) 385 (18), 328 (11), 301
(23), 286 (100), 217 (13); [a]2D0 ꢁ152ꢂ (c1.24, CHCl3);
HRMS, calcd for C26H40O2: 384.3028, found 384.3026.
2-(3,5-Dimethoxyphenyl)-2-methylundecane. The reac-
tion of 5.56 g (24.2 mmol) of 2-methyl-2-undecanol30
with 3.70 g (24.2 mmol) of 2,6-dimethylphenol was car-
ried out as described above for the synthesis of
2-methyl2-(3,5-dimethoxyphenyl)propane to give 4.68 g
(67%) of 2-(3,5-dimethoxy-4-hydroxyphenyl)undecane
as a pale yellow oil after chromatography (petroleum
2-Methyl-2-(3,5-dimethoxyphenyl)decane. The reaction
of 5.30 g (27.5 mmol) of 2-methyl-2-decanol, prepared
from 2-decanone and methylmagnesium bromide with
4.23 g (27.4 mmol) of 2,6-dimethoxyphenol was carried
out as described above to give 8.62 g of crude 2-methyl-
2-(4-hydroxy-3,5-dimethoxyphenyl)decane as an oil
which was used in the subsequent step without further
purification: 1H NMR (300 MHz, CDCl3) d 0.86 (t,
J=7.0 Hz, 3H), 0.96–1.38 (m, 12H), 1.26 (s, 6H), 1.50–
1.63 (m, 2H), 3.87 (s, 6H), 6.54 (s, 2H); 13C NMR
(75.5 MHz, CDCl3) d 14.0, 22.5, 24.6, 29.0, 29.2, 29.4,
30.2, 31.8, 37.5, 44.6, 56.1, 102.9, 132.4, 141.0, 146.4.
1
ether/ether 2:1): H NMR (300 MHz, CDCl3) d 0.85 (t,
J=7.4 Hz, 3H), 0.97–1.37 (m, 14H), 1.26 (s, 6H), 1.50–
1.62 (m, 2H), 3.87 (s, 6H), 5.48 (br s, 1H), 6.54 (s, 2H);
13C NMR (75.5 MHz, CDCl3) d 14.0, 22.5, 24.6, 29.1,
29.2, 29.4, 29.5, 30.2, 31.8, 37.6, 44.6, 56.2, 102.9, 132.4,
140.9, 146.4.
From 4.45 g (15.3 mmol) of the above phenol, there was
obtained 6.49 g (93%) of phosphate ester as a yellow
solid which was used in the next step without further
purification: 1H NMR (300 MHz, CDCl3) d 0.86 (t,
J=6.2 Hz, 3H), 0.96–1.12 (m, 2H), 1.12–1.47 (m, 12H),
1.25 (s, 6H), 1.38 (t, J=7.1 Hz, 3H), 1.39 (t, J=7.1 Hz,
3H), 1.47–1.62 (m, 2H), 3.85 (s, 6H), 4.30 (q, J=7.3 Hz,
2H), 4.32 (q, J=7.3 Hz, 2H), 6.53 (s, 2H); 13C NMR
(75.5 MHz, CDCl3) d 14.0, 16.0, 16.1, 22.6, 24.6, 29.0,
29.4, 29.5, 30.3, 31.8, 38.0, 44.5, 56.0, 64.1, 64.2, 103.1,
127.1, 147.1, 151.2.
The phenol was converted to the phosphate ester as
described above for the preparation of 3-pentylphenol.
From 8.6 g (27.5 mmol) of phenol, there was obtained 7.4 g
of crude ester as a yellow solid which was used in the next
step without further purification: 1H NMR (300 MHz,
CDCl3) d 0.86 (t, J=6.2 Hz, 3H), 0.96–1.34 (m, 12H), 1.25
(s, 6H), 1.38 (t ,J=7.1 Hz, 3H), 1.39 (t, J=7.1 Hz, 3H),
1.47–1.62 (m, 2H), 3.85 (s, 6H), 4.30 (q, J=7.3 Hz, 2H),
4.32 (q, J=7.3 Hz, 2H), 6.53 (s, 2H); 13C NMR (75.5 MHz,
CDCl3) d 14.0, 16.0, 16.1, 22.6, 24.6, 29.0, 29.3, 29.4, 30.3,
31.8, 38.0, 44.5, 56.0, 64.1, 64.2, 103.1, 147.1, 151.2.
Reduction of 6.49 g (14.2 mmol) of the phosphate ester
by the method described above gave 3.65 g (84%) of
2-(3,5-dimethoxyphenyl)-2-methylundecane as a colorless
Reduction of 7.43 g (16.7 mmol) of phosphate ester by the
method described above gave, after chromatography,

J. W. Huffman et al. / Bioorg. Med. Chem. 11 (2003) 1397–1410
1407
1
oil: H NMR (300 MHz, CDCl3) d 0.86 (t, J=6.7 Hz,
3H), 0.99–1.14 (m, 2H), 1.14–1.37 (m, 12H), 1.25 (s,
6H), 1.50–1.63 (m, 2H), 3.78 (s, 6H), 6.29 (t, J=2.1 Hz,
1H), 6.49 (d, J=2.2 Hz, 2H); 13C NMR (75.5 MHz,
CDCl3) d 14.1, 22.6, 24.6, 28.9, 29.3, 29.5, 29.6, 30.3,
31.8, 37.9, 44.4, 55.0, 96.5, 104.5, 152.4, 160.4; MS m/z
calcd for C20H34O2: 306.2559, found 306.2560.
2-methyldodecane as
a
colorless oil: 1H NMR
(300 MHz, CDCl3) d 0.86 (t, J=6.7 Hz, 3H), 0.99–1.14
(m, 2H), 1.14–1.37 (m, 14H), 1.25 (s, 6H), 1.50–1.63 (m,
2H), 3.78 (s, 6H), 6.29 (t, J=2.1 Hz, 1H), 6.49 (d,
J=2.2 Hz, 2H); 13C NMR (75.5 MHz, CDCl3) d 14.1,
22.6, 24.6, 28.9, 29.3, 29.5, 29.6, 30.3, 31.8, 37.9, 44.4,
55.0, 96.5, 104.5, 152.4, 160.4; HRMS calcd for
C21H36O2: 320.2715, found 320.2715.
3-(10,10-Dimethyldecyl)-D8-tetrahydrocannabinol
(8,
n=8). From 0.462 g (1.51 mmol) of 2-methyl-2-(3,5-
dimethoxyphenyl)undecane there was obtained 0.430 g
of crude resorcinol as a brown oil which used in the next
3-(10, 10-Dimethylundecyl)-D8-tetrahydrocannabinol (8,
n=9). From 0.262 g (0.819 mmol) of 2-(3,5-dimethoxy-
phenyl)-2-methyldodecane there was obtained 0.245 g
(100%) of crude resorcinol which was used in the sub-
sequent step without further purification: 1H NMR
(300 MHz, CDCl3) d 0.86 (t, J=6.9 Hz, 3H), 0.95–1.11
(m, 2H), 1.11–1.35 (m, 14H), 1.19 (s, 6H), 1.41–1.56 (m,
2H), 6.23 (t, J=2.0 Hz, 1H), 6.35 (br s, 2H), 6.40 (d,
J=2.1 Hz, 2H); 13C NMR (75.5 MHz, CDCl3) d 14.1,
22.6, 24.6, 28.8, 29.3, 29.6, 29.7, 30.3, 31.8, 37.6, 44.4,
99.9, 105.7, 153.2, 156.5.
1
step without further purification: H NMR (300MHz,
CDCl3) d 0.86 (t, J=6.9 Hz, 3H), 0.95–1.11 (m, 2H), 1.11–
1.35 (m, 12H), 1.19 (s, 6H), 1.41–1.56 (m, 2H), 6.23 (t,
J=2.0 Hz, 1H), 6.35 (br s. 2), 6.40 (d, J=2.1 Hz, 2H); 13C
NMR (75.5 MHz, CDCl3) d 14.0, 22.6, 24.6, 28.8, 29.3,
29.6., 30.4, 31.8, 37.6, 44.4, 99.9, 105.7, 153.2, 156.5.
From 0.43 g (1.51 mmol) of crude resorcinol, 0.207 g
(1.36 mmol) of trans-p-menthadienol and 0.015 g (10
mol%) of p-toluenesulfonic acid monohydrate, there
was obtained 0.395 g (71%) of cannabinoid as viscous
pale yellow oil: Rf 0.24 (petroleum ether/ether 20:1),
From 0.262 g (0.819 mmol) of resorcinol, there was
obtained, after chromatography (petroleum ether/ether
10:1), 0.195 g (62%) of cannabinoid as a viscous pale
yellow oil: Rf 0.24 (petroleum ether/ether 20:1), 0.39
(petroleum ether/ether 10:1); 1H NMR (300 MHz,
CDCl3) d 0.86 (t, J=7.0 Hz, 3H), 1.11 (s, 3H), 1.18 (s,
6H), 1.39 (s, 3H), 0.96–1.39 (m, 16H), 1.42–1.57 (m,
2H), 1.69 (s, 3H), 1.75–1.98 (m, 3H), 2.06–2.23 (m, 1H),
2.61–2.78 (m, 1H), 3.13–3.30 (m, 1H), 5.41 (br s, 1H),
5.44 (s, 1H), 6.24 (d, J=1.6 Hz, 1H), 6.38 (d, J=1.6 Hz,
1H); 13C NMR (75.5 MHz, CDCl3) d 14.1, 18.4, 22.6,
23.4, 24.6, 27.5, 27.8, 28.6, 28.7, 29.3, 29.6, 29.7, 30.3,
31.5, 31.8, 35.9, 37.2, 44.4, 44.8, 76.6, 105.4, 107.7,
110.2, 119.2, 134.7, 149.9, 154.3, 154.6; MS m/z calcd
for C29H46O2: 426.3498, found 426.3494; [a]2D0 ꢁ110ꢂ
(c0.57, CHCl3).
1
0.39 (petroleum ether/ether 10:1); H NMR (300 MHz,
CDCl3) d 0.86 (t, J=7.0 Hz, 3H), 1.11 (s, 3H), 1.18 (s,
6H), 1.39 (s, 3H), 1.69 (s, 3H), 0.96–1.39 (m, 14H), 1.42–
1.57 (m, 2H), 1.75–1.98 (m, 3H), 2.06–2.23 (m, 1H), 2.61–
2.78 (m, 1H), 3.13–3.30 (m, 1H), 5.41 (br s, 1H), 5.44 (s,
1H), 6.24 (d, J=1.6 Hz, 1H), 6.38 (d, J=1.6 Hz, 1H); 13C
NMR (75.5 MHz, CDCl3) d 14.1, 18.5, 22.7, 23.5, 24.6,
27.6, 27.9, 28.6, 28.7, 29.3, 29.5, 29.6, 30.4, 31.5, 31.8, 36.0,
37.3, 44.4, 44.9, 76.6, 105.4, 107.7, 110.2, 119.2, 134.7,
149.9, 154.3, 154.6; MS m/z calcd for C28H44O2: 412.3341,
Found 412.3340; [a]2D0 ꢁ128ꢂ (c0.56, CHCl3).
2-(3,5-Dimethoxyphenyl)-2-methyldodecane. The reac-
tion of 4.90 g (24.5 mmol) of 2-methyl-2-dodecanol30
with 3.57 g (23.3 mmol) of 2,6-dimethylphenol was car-
ried outas described above ot give 4.45 g (57%) of
2-(3,5-dimethoxy-4-hydroxyphenyl)-2-methyldodecane
as an oil after chromatography (petroleum ether/ether
2-Methyl-2-(3,5-dimethoxyphenyl)tridecane. The reac-
tion of 4.90 g (24.5 mmol) of 2-methyl-2-tridecanol30
with 3.19 g of 2,6-dimethoxyphenol was carried out as
described above to give, after chromatography (petro-
leum ether/ether 2:1), 4.51 g (62%) of 2-methyl-2-(3,5-
dimethoxy-4-hydroxyphenyl)tridecane as an oil: 1H
NMR (300 MHz, CDCl3) d 0.86 (t, J=7.0 Hz, 3H),
0.96–1.13 (m, 2H), 1.13–1.38 (m, 16H), 1.26 (s, 6H),
1.50–1.63 (m, 2H), 3.87 (s, 6H), 6.54 (s, 2H); 13C NMR
(75.5 MHz, CDCl3) d 14.0, 22.5, 24.6, 29.1, 29.2, 29.6,
30.3, 31.8, 37.6, 44.6, 56.1, 102.9, 132.4, 141.0, 146.4.
1
2:1): H NMR (300 MHz, CDCl3) d 0.86 (t, J=7.0 Hz,
3H), 0.96–1.13 (m, 2H), 1.13–1.38 (m, 14H), 1.26 (s, 6H),
1.50–1.63 (m, 2H), 3.87 (s, 6H), 6.54 (s, 2H); 13C NMR
(75.5 MHz, CDCl3) d 14.0, 22.5, 24.6, 29.1, 29.2, 29.5,
29.6, 30.3, 31.8, 37.6, 44.6, 56.1, 102.9, 132.4, 141.0, 146.4.
From 4.35 g (12.9 mmol) of phenol, there was obtained
6.43 g (100%) of crude phosphate ester as a yellow solid
which was used in the next step without further pur-
From 4.41 g (12.9 mmol) of phenol, there was obtained
8.71 g (100%) of crude phosphate ester which was used
1
ification: H NMR (300 MHz, CDCl3) d 0.86 (t, J=6.2
1
Hz, 3H), 0.96 (m, 2H), 1.12–1.47 (m, 14H), 1.25 (s, 6H),
1.38 (t, J=7.1 Hz, 3H), 1.39 (t, J=7.1 Hz, 3H), 1.47–
1.62 (m, 2H), 3.85 (s, 6H), 4.30 (q, J=7.3 Hz, 2H), 4.32
(q, J=7.3 Hz, 2H), 6.53 (s, 2H); 13C NMR (75.5 MHz,
CDCl3) d 13.9, 15.8, 15.9, 22.5, 24.5, 28.8, 29.1, 29.3,
29.4, 29.5, 30.2, 31.7, 37.9, 44.4, 55.9, 63.9, 64.0, 103.0,
127.3, 147.0, 151.1.
in the next step without further purification: H NMR
(300 MHz, CDCl3) d 0.86 (t, J=6.2 Hz, 3H), 0.96–1.12
(m, 2H), 1.12–1.47 (m, 16H), 1.25 (s, 6H), 1.38 (t, J=7.1
Hz, 3H), 1.39 (t, J=7.1 Hz, 3H), 1.47–1.62 (m, 2H),
3.85 (s, 6H), 4.30 (q, J=7.3 Hz, 2H), 4.32 (q, 7.3H),
6.53 (s, 2H); 13C NMR (75.5 MHz, CDCl3) d 13.9, 15.8,
15.9, 22.5, 24.5, 28.8, 29.1, 29.3, 29.4, 29.5, 30.2, 31.7,
37.9, 44.4, 55.9, 63.9, 64.0, 103.0, 127.3, 147.0, 151.1.
From 6.43 g (12.9 mmol) of phosphate ester, there was
obtained after chromatography (petroleum ether/ethyl
acetate 10:1), 3.12 g (75%) of 2-(3,5-dimethoxyphenyl)-
From 8.71 g (12.9 mmol) of phosphate ester, there was
obtained, after chromatography (petroleum ether/ether

1408
J. W. Huffman et al. / Bioorg. Med. Chem. 11 (2003) 1397–1410
10:1), 3.27 g (76%) of 2-methyl-2-(3,5-dimethoxy-
phenyl)tridecane as a colorless oil: Rf 0.18 (petroleum
ether); H NMR (300 MHz, CDCl3) d 0.86 (t, J=6.7
Hz, 3H), 0.99–1.14 (m, 2H), 1.14–1.37 (m, 16H), 1.25 (s,
6H), 150–1.63 (m, 2H), 3.78 (s, 6H), 6.29 (t, J=2.1 Hz,
1H), 6.49 (d, J=2.2 Hz, 2H); 13C NMR (75.5 MHz,
CDCl3) d 14.1, 22.6, 24.6, 28.9, 29.3, 29.5, 29.6, 30.4,
31.9, 37.9, 44.4, 55.0, 96.5, 104.5, 152.4, 160.4; MS m/z
calcd for C22H38O2: 334.2872, found 334.2870.
tested for effects on the following procedures: sponta-
neous (locomotor) activity at 5–15 min, tail-flick latency
(antinociception) response at 20 min, core (rectal) tem-
perature at 30 min.
1
Spontaneous activity. Inhibition of locomotor activity
was accomplished by placing mice into individual activ-
ity cages (6.5ꢃ11 in), and recording interruptions of the
photocell beams (16 beams per chamber) for a 10-min
period using a Digiscan Animal Activity Monitor
(Omnitech Electronics Inc., Columbus, OH, USA).
Activity in the chamber was expressed as the total
number of beam interruptions.
3-(10,10-Dimethyldodecyl)-D8-tetrahydrocannabinol
(8,
n=10). From 0.280 g (0.838 mmol) of 2-methyl-2-(3,5-
dimethoxyphenyl)tridecane there was obtained 0.265 g
(100%) of resorcinol which was used in the subsequent
1
step without further purification: H NMR (300 MHz,
Tail-flick latency. Antinociception was assessed using
the tail-flick procedure. The heat lamp of the tail-flick
apparatus was maintained at an intensity sufficient to
produce control latencies of 2–3 s. Control values for
each animal were determined prior to drug administra-
tion. Mice were then re-evaluated following drug
administration and latency (s) to tail-flick response was
recorded. A 10-s maximum was imposed to prevent tis-
sue damage. The degree of antinociception was expres-
sed as the% MPE (maximum possible effect) which was
calculated as:
CDCl3) d 0.86 (t, J=6.9 Hz, 3H), 0.95–1.11 (m, 2H),
1.11–1.35 (m, 16H), 1.19 (s, 6H), 1.41–1.56 (m, 2H),
6.23 (t, J=2.0 Hz, 1H), 6.35 (br s, 2H), 6.40 (d, J=2.1 Hz,
2H); 13C NMR (75.5 MHz, CDCl3) d 14.1, 22.6, 24.6,
28.8, 29.3, 30.3, 31.8, 37.6, 44.4, 99.9, 105.7, 153.2, 156.5.
From 0.28 g (0.84 mmol) of crude resorcinol, 0.13 g
(0.86 mmol) of trans-p-menthadienol and 0.02 g (10
mol%) of p-toluenesulfonic acid monohydrate there was
obtained 0.23 g of cannabinoid as a viscous pale yellow
oil: Rf 0.24 (petroleum ether/ether 20:1), 0.39 (petro-
1
leum ether/ether 10:1); H NMR (300 MHz, CDCl3) d
ꢀ
ꢁ
ðtest latency À control latencyÞ
10 s À test latency
0.86 (t, J=7.0 Hz, 3H), 1.11 (s, 3H), 1.18 (s, 6H), 1.39
(s, 3H), 1.69 (s, 3H), 0.96–1.39 (m, 18H), 1.42–1.57 (m,
2H), 1.75–1.98 (m, 3H), 2.06–2.23 (m, 1H), 2.61–2.78
(m, 1H), 3.13–3.30 (m, 1H), 5.41 (br s, 1H), 5.44 (s, 1H),
6.24 (d, J=1.6 Hz, 1H), 6.38 (d, J=1.6 Hz, 1H); 13C
NMR (75.5 MHz, CDCl3) d 14.1, 18.4, 22.6, 23.4, 24.6,
27.5, 27.8, 28.6, 28.7, 29.3, 29.6, 29.7, 30.3, 31.5, 31.9,
35.9, 37.2, 44.4, 44.8, 76.6, 105.4, 107.7, 110.2, 119.2,
134.7, 149.9, 154.3, 154.6; MS m/z calcd for C30H48O2:
440.3654, found 440.3655; [a]2D0 ꢁ137ꢂ (c1.1, CHCl3).
% MPE ¼
ꢀ 100
Core temperature. Hypothermia was assessed by first
measuring baseline core temperatures prior to drug
treatment with a telethermometer (Yellow Springs
Instrument Co., Yellow Springs, OH, USA) and a rectal
thermistor probe inserted to 25 mm. Rectal tempera-
tures were also measured following drug administration.
The temperature difference (ꢂC) between values was
calculated for each animal.
Receptor binding assays
[3H]CP-55,940 (KD=690 nM) binding to P2 membranes
was conducted as described elsewhere,32 exceptwhole
brain (rather than cortex only) was used. Displacement
curves were generated by incubating drugs with 1 nM of
[3H]CP-55,940. The assays were performed in triplicate,
and the results represent the combined data from three
individual experiments.
Computational methods
Molecular modeling and energy minimization. Molecules
were modeled using SYBYL (Tripos Inc., St. Louis,
MO, USA). Electrostatic charges of each compound
were calculated with the Gasteiger–Huckel method.
Each compound was energy minimized using the
SYBYL force field with a conjugate gradient of 0.001
kcal/mol or a maximum of 100,000 iterations as termi-
nation criteria.
Pharmacology
Male ICR mice (Harlan Laboratories, Dublin, VA,
USA) weighing 18–25 g were maintained on a 14:10 h
light/dark cycle with free access to food and water.
ꢀ9-THC and ꢀ8-THC were obtained from the National
Institute on Drug Abuse. All compounds were dissolved
in 1:1:18 (emulphor–ethanol–saline) for in vivo admin-
istration. Emulphor (EL-620, a polyoxyethylated vege-
table oil, GAF Corporation, Linden, NJ, USA) is
currently available as Alkmulphor. All drug injections
were administered iv (tail vein) at a volume of 0.1 mL/
10 g of body weight. Mice were acclimated in the eval-
uation room overnight without interruption of food and
water. Following drug administration each animal was
Quenched molecular dynamics. In order to characterize
the conformational mobility of each analogue, mole-
cular dynamics were computed on each energy-mini-
mized structure at temperatures from 100 to 1000 K.
During the molecular dynamics simulation, the mole-
cule was heated at 100 K steps, and allowed to remain
at the specified temperature for 1 ps. Upon reaching
1000 K, the molecule was held at this temperature for
100 ps while snapshots were acquired at 1-ps intervals.
Each conformation obtained for a particular molecule
was energy minimized again using a conjugate gradient
of 0.01 kcal/mol or a maximum of 100,000 iterations as

J. W. Huffman et al. / Bioorg. Med. Chem. 11 (2003) 1397–1410
1409
termination criteria, yielding a group of 100 energy-
minimized conformers per compound.
Acknowledgements
The work at Clemson University was supported by
grant DA03590, that at the Medical College of Virginia
Campus of Virginia Commonwealth University by
grantDA03672, and htatathte Research Triangle
Institue by grant DA11638, all from the National Insti-
tute on Drug Abuse. High resolution mass spectral data
were determined by the Mass Spectrometry Laboratory,
School of Chemical Sciences, University of Illinois.
Molecular and conformational alignment. All confor-
mations from each of the analogues were overlaid using
a single template molecule. Because the phenyl ring
systems of all analogues were identical, it was unim-
portant which molecule or conformation was used for
this template, as long as it was applied consistently
using the six aromatic carbons of the phenol ring
system. The alignment was performed using atom-
by-atom root mean square distance minimization.
This alignmentposiiotned all of hte molecules in
the same three-dimensional space and superposed the
ring systems to as great an extent as possible. Since the
side chains were not part of the alignment rule, this
feature of the molecule could be compared between
conformers of the same compound as well as between
different compounds using the QSAR technique
described below.
References and Notes
1. Gaoni, Y.; Mechoulam, R. J. Am. Chem. Soc. 1964, 86,
1646.
2. Razdan, R. K. Pharmacol. Rev. 1986, 38, 75.
3. Seltzman, H. H. Curr. Med. Chem. 1999, 6, 685.
4. Khanolkar, A. D.; Palmer, S. L.; Makriyannis, A. Chem.
Phys. Lipids 2000, 108, 37.
5. Binder, M.; Witteler, F.-J.; Schmidt, B.; Franke, I.; Bohe-
nenberger, E.; Sandermann, H. In The Cannabinoids: Clinical,
Pharmacologic and Therapeutic Aspects; Agurell, S., Dewey,
W. L., Willette, R. E., Eds.; Academic: New York, 1984; p
709.
6. Howlett, A. C.; Johnson, M. R.; Melvin, L. S.; Milne,
G. M. Mol. Pharmacol. 1988, 33, 297.
7. Martin, B. R.; Compton, D. R.; Prescott, W. R.; Barrett,
R. L.; Razdan, R. K. Drug Alcohol Depend. 1995, 37, 231.
8. Reggio, P. H.; Greer, K. V.; Cox, S. M. J. Med. Chem.
1989, 32, 1630.
9. Huffman, J. W.; Banner, W. K.; Zoorob, G. K.; Joyner,
H. H.; Reggio, P. H.; Martin, B. R.; Compton, D. R. Tetra-
hedron 1995, 51, 1017.
10. 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.
11. Adams, R.; Harfenist, M.; Loewe, S. J. Am. Chem. Soc.
1949, 71, 1624.
12. Huffman, J. W.; Duncan, S. G.; Wiley, J. L.; Martin, B. R.
Bioorg. Med. Chem. Lett. 1997, 7, 1557.
13. Liddle, J.; Huffman, J. W. Bioorg. Med. Chem. Lett. 1999,
8, 2223.
14. Adams, R.; MacKenzie, S.; Loewe, S. J. Am. Chem. Soc.
1948, 70, 664.
15. Martin, B. R.; Compton, D. R.; Semus, S.; Lin, S.; Mar-
ciniak, C.; Grzybowska, J.; Charlambous, A.; Makriyannis, A.
Pharmacol. Biochem. Behav. 1993, 46, 295.
16. Archer, R. A.; Blanchard, W. B.; Day, W. A.; Johnson,
D. W.; Lavagnino, E. R.; Ryan, C. W.; Baldwin, J. E. J. Org.
Chem. 1977, 42, 2277.
17. Dominianni, S. J.; Ryan, C. W.; De Armitt, C. W. J. Org.
Chem. 1977, 42, 344.
18. Keimowitz, A. R.; Martin, B. R.; Razdan, R. K.; Crocker,
P. J.; Mascarella, S. W.; Thomas, B. F. J. Med. Chem. 2000,
43, 59.
Quantitative structure–activity relationship analysis.
Comparative molecular field analysis was used to
characterize the quantitative structure–activity relation-
ships of the compounds shown in Table 1. This tech-
nique has previously been used successfully in QSAR
studies of cannabinoids.18,33ꢁ37 For the QSAR studies, a
total of 50 random conformations from the 100
obtained were selected for each analogue. The descrip-
tive variables were steric and electrostatic descriptions
of the three-dimensional structure of the entire set of
compounds. The dependentvariables for hte models
were receptor affinity (Ki) and potency (ED50) in t he
spontaneous activity (SA), tail flick (TF) and rectal
temperature (RT) assays. The CoMFA analysis was
performed using a proton (H+1) probe atom posi-
tioned at lattice points spaced around the molecules at
1.5A increments. Cross-validation was performed by
randomly selecting 80% of the compounds to form a
training set, developing a QSAR model based on their
three-dimensional steric and electrostatic properties,
and using this model to predict the dependent variables
of the remaining 20% of the compounds that were not
included in the training set. The predicted dependent
variable of the compounds that were omitted from the
training set was then compared against the actual
dependentvariable and
a
correliaotn coefficient
obtained. This process was repeated randomly until
every compound had been omitted from the training set
and had its dependent variables predicted at least once.
The correlation coefficients of the entire process were
tracked throughout this process and the average
r-squared value was calculated as a measure of the
press, or the goodness/robustness of the model.
19. Xie, X. Q.; Yang, D. P.; Melvin, L. S.; Makriyannis, A.
J. Med. Chem. 1994, 37, 1418.
20. Martel, P.; Makriyannis, A.; Mavromoustakos, T.; Kelly,
K.; Jeffrey, K. R. Biochim. Biophys. Acta 1993, 1151, 51.
21. Petrzilka, T.; Sikemeier, C. Helv. Chim. Acta 1967, 50,
1416.
22. Ganz, A. J.; Waser, P. G. Arzneim.-Forsch. Drug Res.
1980, 30, 471.
23. Huffman, J. W.; Liddle, J.; Yu, S.; Aung, M. M.; Abood,
M. E.; Wiley, J. L.; Martin, B. R. Bioorg. Med. Chem. 1999, 7,
2905.
Control studies were also run where randomized
pharmacological data was used in place of real
pharmacological data. These artificial points were ran-
dom numbers generated to fall within the range of the
real data. The same cross-validated and final analyses
were performed in the control studies to check that the
r2 values were higher when pharmacological data were
used than when artificial data were used.

1410
J. W. Huffman et al. / Bioorg. Med. Chem. 11 (2003) 1397–1410
24. Huffman, J. W.; Joyner, H. H.; Lee, M. D.; Jordan, R. D.;
Pennington, W. T. J. Org. Chem. 1991, 56, 2081.
25. Compton, D. R.; Rice, K. C.; De Costa, B. R.; Razdan,
R. K.; Melvin, L. S.; Johnson, M. R.; Martin, B. R. J. Phar-
macol. Exp. Ther. 1993, 265, 218.
26. Martin, B. R.; Compton, D. R.; Little, P. J.; Martin, T. J.;
Beardsley, P. M. In Structure Activity Relationships in Canna-
binoids; Rapaka, R. S., Makriyannis, A. (Eds.); NIDA
Research Monograph 79; National Institute on Drug Abuse:
Rockville, MD, l987; p 108.
27. Little, P. J.; Compton, D. R.; Johnson, M. R.; Melvin,
L. S.; Martin, B. R. J. Pharmacol. Exp. Ther. 1988, 247, 1046.
28. Kell, D. R.; McQuillin, F. J. J. Chem. Soc., Perkin Trans.
1 1972, 2096.
29. Singh, V.; Kane, V. V.; Martin, A. R. Synth. Commun.
1981, 11, 429.
31. Nicklaus, M. C.; Milne, G. W.; Burke, T. R., Jr. J. Com-
put.-Aided Mol. Des. 1992, 6, 487.
32. Martin, B. R.; Compton, D. R.; Thomas, B. F.; Prescott,
W. R.; Little, P. J.; Razdan, R. K.; Johnson, M. R.; Melvin,
L. S.; Mechoulam, R.; Ward, S. J. Pharmacol. Biochem.
Behav. 1991, 40, 471.
33. Schmetzer, S.; Greenidge, P.; Kovar, K. A.; Schulze-
Alexandru, M.; Folkers, G. J. Comput.-Aided Mol. Des. 1997,
11, 278.
34. Shim, J. Y.; Collantes, E. R.; Welsh, W. J.; Subramaniam,
B.; Howlett, A. C.; Eissenstat, M. A.; Ward, S. J. J. Med.
Chem. 1998, 41, 4521.
35. Thomas, B. F.; Adams, I. B.; Mascarella, S. W.; Martin,
B. R.; Razdan, R. K. J. Med. Chem. 1996, 39, 471.
36. Thomas, B. F.; Compton, D. R.; Martin, B. R.; Semus,
S. F. Mol. Pharmacol. 1991, 40, 656.
37. Tong, W.; Collantes, E. R.; Welsh, W. J.; Berglund, B. A.;
Howlett, A. C. J. Med. Chem. 1998, 41, 4207.
´
´
30. Prevost, C.; Singer, E. Bull. Soc. Chim. France 1950,
1068.