Synthesis and antitumor activity of quinonoid derivatives of cannabinoids

Article abstract of DOI:10.1021/jm040042o

Three cannabis constituents, cannabidiol (1), Δ⁸ -tetrahydrocannabinol (3), and cannabinol (5), were oxidized to their respective para-quinones 2, 4, and 6. In the 1960s, the oxidized product 4 had been assigned a para-quinone structure, which was later modified to an ortho-quinone. To distinguish between the two possible quinone structures, a detailed NMR investigation was undertaken. The original para-quinone structure was confirmed. X-ray crystallography elucidated the structures of the crystalline 2 and 6. All three compounds displayed antiproliferative activity in several human cancer cell lines in vitro, and quinone 2 significantly reduced cancer growth of HT-29 cancer in nude mice.

Full text of DOI:10.1021/jm040042o

3
800
J . Med. Chem. 2004, 47, 3800-3806
Syn th esis a n d An titu m or Activity of Qu in on oid Der iva tives of Ca n n a bin oid s
#
‡
‡
#
†
‡
Natalya M. Kogan, Ruth Rabinowitz, Paloma Levi, Dan Gibson, Peter Sandor, Michael Schlesinger, and
,
#
Raphael Mechoulam*
Department of Medicinal Chemistry and Natural Products, School of Pharmacy, The Hebrew University,
J erusalem 91120, Israel, Department of Experimental Medicine and Cancer Research, School of Medicine,
The Hebrew University, J erusalem 91120, Israel and NMR Applications Laboratory, Varian Deutschland GmbH,
D-64289 Darmstadt, Germany
Received February 18, 2004
8
Three cannabis constituents, cannabidiol (1), ∆ -tetrahydrocannabinol (3), and cannabinol (5),
were oxidized to their respective para-quinones 2, 4, and 6. In the 1960s, the oxidized product
4
had been assigned a para-quinone structure, which was later modified to an ortho-quinone.
To distinguish between the two possible quinone structures, a detailed NMR investigation was
undertaken. The original para-quinone structure was confirmed. X-ray crystallography
elucidated the structures of the crystalline 2 and 6. All three compounds displayed antipro-
liferative activity in several human cancer cell lines in vitro, and quinone 2 significantly reduced
cancer growth of HT-29 cancer in nude mice.
In tr od u ction
Sch em e 1. Synthesis of Compounds 2, 4, and 6
Quinones of various chemical families serve as bio-
logical modulators,1-3 and both natural and synthetic
quinones are widely used as drugs.4 Anthracyclines, a
large group of quinonoid compounds produced by dif-
ferent strains of streptomyces, exert antibiotic and
antineoplasic effects and are used to treat some forms
,5
5
of cancer. The best known members of this family are
daunorubicin and doxorubicin, the first identified an-
6
thracyclins. Other quinones are also used as anticancer
drugs. Mitomycin C and streptonigrin produced by
streptomyces, and the synthetic epirubicin and mitox-
7
antron are well-known examples. Although these and
other quinonoid compounds are effective in the treat-
ment of many different forms of cancer, their side effects
(
the most severe being cumulative heart toxicity) limit
8
,9
their use. The development of quinonoid compounds
that display antineoplastic activity, but are less toxic,
is a major therapeutic goal.
A large number of cannabinoids have been synthe-
sized and tested in various in vitro and in vivo assays,
including models of several diseases.10-13 However,
although cannabinoid quinones were prepared by our
group already back in 1968, in connection with an
investigation on the chemical basis of the Beam test (a
color test for cannabinoids),14 they have not been
evaluated as medicinal agents. We report now that such
quinones exhibit potent antineoplastic activity in vitro
and at least one of them is also active in vivo.
now found that a slight change in the reaction condi-
tions (lowering the temperature to 0 °C) raised the yield
to ∼20% and brought the reaction time down to 3 h
(
when no more starting material was detected) (Scheme
1
). The hydroxyquinone crystallized from heptane. We
have previously reported that the quinone 2 cyclized to
14
the para-quinone 4 under acid conditions. Hodjat-
1
5
Kashani et al. have suggested, however, that 4 is an
ortho-quinone. Their structural assignment was based
on nuclear Overhauser effect (NOE) NMR data (see
below). If indeed 4 is an ortho-quinone, then 2 may also
be an ortho-quinone. We have now confirmed by X-ray
crystallography the structure of 2, originally proposed
by our group. The X-ray data are submitted as Sup-
porting Information. Quinone 2 was code-named HU-
Ch em istr y
We have reported that oxidation of cannabidiol (CBD)
(1) by air in an alcohol solution in the presence of 5%
potassium hydroxide over 24 h led to the formation of
the hydroxyquinone 2 in about 5-10% yield.14 We have
*
To whom correspondence should be addressed. Phone: 972-2-
6
758634. Fax: 972-2-6757076. E-mail: mechou@cc.huji.ac.il.
3
31 (HU ) Hebrew University).
#
School of Pharmacy, The Hebrew University.
School of Medicine, The Hebrew University.
NMR Applications Laboratory.
8
‡
Oxidation of ∆ -tetrahydrocannabinol (3) with m-
†
14
chloroperbenzoic acid as originally reported gave the
1
0.1021/jm040042o CCC: $27.50 © 2004 American Chemical Society
Published on Web 06/10/2004

Quinonoid Derivatives of Cannabinoids
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 15 3801
The remaining resonances of the spin system (H7, H6a,
H10a, and H10) were assigned on the basis of standard
analysis of COSY, NOESY, TOCSY, and GHSQC [for a
review, see ref 22]. The methyl group 11 was assigned
on the basis of a COSY cross-peak to H8 and of NOESY
cross-peaks to H8 and H10. No attempt was made to
distinguish between the methyl groups 12 and 13
(resonating at 1.08 and 1.42 ppm).
Assign m en t of th e Ca r bon Sp ectr u m . The proto-
nated carbons were assigned by analysis of the HSQC
spectrum in a straightforward manner. The assign-
ments of the two nonequivalent protons of H10 (2.92
and 1.67 ppm) were confirmed by their cross-peaks to
the same carbon resonance (35.80 ppm). Similarly, the
two nonequivalent protons of H7 (2.06 and 1.76 ppm)
were confirmed by cross-peaks to the same carbon
resonance (27.35 ppm). C3 was assigned on the basis of
cross-peaks to H2, H1′, and H2′ in the HMBC spectrum.
C5 was assigned on the basis of its low-field chemical
shift (152.63 ppm) and its HMBC cross-peaks to H10a
(2.44 ppm) and to the methyl group at 1.42 ppm. C9,
C6, and C5a were assigned on the basis of the analysis
of the HMBC spectrum. The two carbonyls have reso-
nances at 182.62 and 187.63 ppm.
F igu r e 1. Structure of compound 4 with the atom labeling
scheme and the proton and carbon chemical shifts.
desired quinone (4) in a low yield. To improve the yield,
we now oxidized 3 with bis(trifluoroacetoxy)iodobenzene
(BTIB), an oxidizing agent, which was not available
when the original reaction was performed.
BTIB was first used for oxidation of phenols to
1
6
quinones in 1989 and since then has become a widely
1
7-20
used reagent for the oxidation of phenols to quinones.
8
BTIB oxidation of ∆ -THC (3) led to the desired quinone
Deter m in in g th e Cor r ect Con stitu tion . We adopted
a two-pronged approach: (a) to carry out a detailed
high-field NMR study of 4 to sort out the correct
positioning of two carbonyl groups and (b) to chemically
reduce its two carbonyl groups and then form the
corresponding acetates, which would provide us with
two additional methyl groups to help determine the
constitution by NOE studies.
4
in 30-35% yield (Scheme 1). The compound was code-
named HU-336.
An interesting feature of the oxidation of 3 is that by
adding copper chloride in acetonitrile, which commonly
leads to the formation of ortho-quinones,21 the same
para-quinone 4 is obtained as with BTIB and at ap-
proximately the same yield.
As mentioned above, we originally put forward a para-
We had hoped that the HMBC experiments that show
long-range C-H correlations (two to four bonds) will be
sufficient to sort out the question of the correct constitu-
tion. In practice, the proton at 6.30 ppm had strong
correlations with C1′, C5a, and the carbonyl at 182.62
and weaker correlations with C3 and C5. Since the
intensity, and even the observation of HMBC cross-
peaks, is not merely a function of the number of bonds
separating the two interacting nuclei but also depends
on other structural factors (torsional angles and bond
order), the data did not allow us to arrive at an
unambiguous determination of the constitution. Thus,
we decided to carry out a further experiment to deter-
1
4
quinone structure for 4, while Hodjat-Kashani et al.
1
5
proposed an ortho-quinone structure. To establish the
correct structure, a detailed NMR analysis was per-
formed (see below). We have now established that
compound 4 is indeed a para-quinone as originally
suggested.
The quinone of cannabinol (CBN) (6), like the quinone
8
of ∆ -THC (4), was synthesized by oxidation with BTIB,
with a yield of ∼60%. The structure was determined by
X-ray crystallography (data submitted as Supporting
Information). The compound was code-named HU-345.
NMR An a lysis of
mine carbon-carbon connectivity. We used the 1,1-
∆
8-Tetr a h yd r oca n n a bin olqu in on e (4)
23,24
adequate pulse sequence
detecting C13 single quan-
Assign m en t of th e P r oton Sp ectr u m . The struc-
tum coherences in the indirect domain. The resulting
spectrum is similar to that of an HMBC except that only
ture of 4, the atom labeling scheme, and the proton and
carbon chemical shifts are depicted in Figure 1. The 5′-
methyl group was assigned on the basis of its integra-
tion (3H), chemical shift (0.84 ppm), and multiplicity
1
3
1
two-bond C- H correlations are obtained. The cross-
peaks of the resonance at 6.30 ppm are depicted (see
Supporting Information), clearly demonstrating that the
protonated carbon at 133.42 ppm is adjacent to C3
(146.58 ppm) and to a carbonyl at 187.63 ppm. This
result is only consistent with the para structure where
(
triplet). Methylene groups 1′-4′ were assigned from
analysis of the correlation spectroscopy (COSY) and
gradient heteronuclear single quantum coherence total
correlation spectroscopy (GHSQC-TOCSY) data. H2 was
assigned on the basis of strong NOESY cross-peaks to
2
the protonated carbon C2 is between an sp carbon and
a carbonyl carbon. This experiment further confirms the
1
3
1
′ and 2′, its chemical shift (6.30 ppm) and its multiplic-
assignment of the C resonance at 187.63 ppm as C1.
If the constitution were indeed ortho, the protonated
carbon (C4) would have correlations to C3 (146.58 ppm)
and to C5 (152.63 ppm) and there would be no correla-
tion with a carbonyl carbon. The results of the adequate
experiment prove beyond doubt that the positioning of
the carbonyl groups is indeed para. The adequate
ity, and a triplet with a 1.36 Hz coupling constant
indicating long-range coupling to 1′. The assignment of
the H2 resonance does not, however, allow us to
determine whether the two carbonyls are ortho or para
to each other (vide infra). The broad peak at 5.33 ppm
was assigned to H8 on the basis of its chemical shift.

3
802 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 15
Kogan et al.
Sch em e 2. Reductive Acetylation of 4
spectrum did confirm all of our previous 1H and 13
assignments.
C
To ensure finally that 4 is a para-quinone, reductive
acetylation with zinc and acetic anhydride was per-
formed (Scheme 2). The two acetyl methyl groups
resonate at 2.27 and 2.29 ppm. The methyl group at 2.27
ppm has NOESY cross-peaks to C11 (1.67 ppm) and a
strong cross-peak to H2 (6.30 ppm) and was therefore
assigned as CH3(1). The methyl group at 2.29 ppm has
a strong NOESY cross-peak to the methyl group at H12
(1.31 ppm), suggesting that it is the 2′′-acetate group.
An energy minimization of the structure in the ortho
geometry revealed that the shortest distance between
the acetyl protons in positions 1 and 2 to the methyl
protons in positions 12 and 13 is 4.08 Å, which would
show weak cross-peaks in the NOESY spectrum, while
in the para position the distance is 2.5-3.2 Å (expected
to result in stronger NOE cross-peaks).
Our data, measured on a Varian Inova 600 MHz
spectrometer, suggest that the proton resonances for the
1
2-methyl group and the 2′-methylene group overlap,
and we assume that perhaps the proximity of the
chemical shifts (1.420 and 1.438 ppm, respectively),
which was not properly resolved on the 300 MHz
spectrometer (a 5.4 Hz difference), may have been the
reason for mistaking15 the NOE between H2 and H2′
for the NOE between H4 and the methyl group 12.
F igu r e 2. Results of in vivo action of 2 on HT-29 cancer
growth: (Exp) injection of 2 from day 2 ip; (Exp-D2) injection
of 2 from day 2 subcutaneously; (Exp-D14) injection of 2
intratumorally from day 14 after cancer cell injection.
It was of interest that the HT-29 and MCF-7 lines
were most susceptible to inhibition by 2, whereas the
SNB-19 and DU-345 lines were more susceptible to
inhibition by 4 and 6 than the other lines tested.
Biologica l Section
The ability of the cannabinoid quinones to inhibit
cancer growth in vitro was assayed on some human
cancer cell lines: Raji (Burkitt’s lymphoma), J urkat
Experiments were carried out to determine the ability
of 2 to inhibit the growth of human tumor cells in vivo.
(T-cell lymphoma), SNB-19 (glioblastoma), MCF-7 (breast
2
was chosen for the in vivo assays because it was active
cancer), DU-145 (prostate cancer), NCI-H-226 (lung
cancer), and HT-29 (colon cancer) (see Supporting
Information). Compound 2 exerted an inhibitory effect
on the in vitro growth of all seven lines of human cancer
cells tested (see Supporting Information). The most
striking inhibition by 2 was found in tests with the Raji
and J urkat lymphoma cells. An inhibition of about 50%
of the growth of both lymphomas was obtained at a
concentration of 2 as low as 0.2 µg/mL. A concentration
of 1.56 µg/mL inhibited the growth of the lymphomas
by over 80%. The most sensitive epithelial cancer cells
were HT-29 colon cancer cells and MCF-7 mammary
cancer cells. A concentration of 3.125 µg/mL inhibited
the growth of these cancer cell lines by about 50%.
Compound 6 inhibited the growth of Raji cells more
effectively than that of the other cell lines tested (see
Supporting Information). For the inhibition of the
growth of J urkat cells and of all other cell lines, a
concentration of 12.5-25.0 µg/mL of 6 was required.
Compound 4 had the weakest capacity to inhibit the
growth of human cancer cell lines in vitro. 4 exerted a
similar inhibitory effect on all cell types at concentra-
tions over 12.5 µg/mL.
at least against three cell lines (HT-29, Raji, and J urkat)
at doses considerably lower than those needed by 4 and
6
. Nude mice received a subcutaneous (sc) injection of
HT-29 human colon cancer cells. At various time
intervals after the administration of the tumor cells, the
mice received three times per week intraperitoneal or
subcutaneous injections of 2 at a dose of 5 mg/kg, which
paralleled a concentration of 5 µg/mL in in vitro experi-
ments, a concentration that killed about 50% of HT-29
cancer cells. Treatment of mice with 2 at a dose of 5
mg/kg did not cause either weight loss or any observable
adverse effects in the treated mice. In one experiment
(Figure 2) 10 nude mice, which received a sc injection
of HT-29, were divided into two groups. Starting from
day 2 after tumor injection, half of the mice received ip
injections of 2. The size of tumors was significantly
smaller in mice injected with 2 than in vehicle-treated
control mice, starting at 25 days after cancer cell
injection (p < 0.05). At 35 days after cancer cell
injection, the tumors in the treated group were half the
size of the tumors in controls, a difference that was
highly significant (p < 0.0029) (Figures 2 and 3).

Quinonoid Derivatives of Cannabinoids
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 15 3803
F igu r e 3. Effect of 5 mg/kg dose of 2 (ip) in vivo.
In another in vivo experiment, nude mice received an
injection of HT-29 cancer cells subcutaneously in the
back. 2 at a concentration of 5 mg/kg was likewise
injected subcutaneously (Figure 2). In one group of nude
mice, 2 was administrated intratumorally starting 14
days after the injection of tumor cells. In another group
of nude mice, 2 was injected subcutaneously into the
region where cancer cells were injected starting 2 days
after the injection of tumor cells. In mice that received
Con clu sion
A large number of cannabinoids have been synthe-
sized and tested in in vitro and in vivo models of
diseases (for recent examples, see refs 13 and 25).
Cannabinoids have been shown to inhibit the growth
of tumor cells in culture and animal models.
are usually well tolerated. However, cannabinoid-
derived quinones have not been investigated in bioas-
says, although quinones of various chemical families are
widely used as drugs. Cannabinoid-derived quinones
were synthesized by our group in 1968, and they were
assigned a para-quinone structure on the basis of UV
26,27
They
2
, starting 2 days after tumor implantation, the tumor
size was significantly smaller than in control mice at
days 17-25 after tumor implantation and remained
significantly smaller throughout the period of observa-
tion, till day 59 (p < 0.05). In mice that received 2
intratumorally starting 14 days after tumor implanta-
tion, the antitumor effect of 2 took a longer time to be
manifested. The tumor size in treated mice was smaller
than in controls from day 31 onward, but this difference
was not significant until day 45 after cancer cell
injection. Starting at day 45 after the cell injection and
throughout the period of observation, the size of the
tumors was significantly smaller than in control mice
1
4
spectroscopy. In a later report, on the basis of an NMR
study (NOE), it was concluded that the two carbonyls
1
5
were in an ortho orientation. We have now carried out
a detailed NMR study in order to determine in an
unequivocal way the correct positioning of the two
carbonyl groups and confirmed the initially established
structure of 4 as para-quinone. The structures of 2 and
6
quinones were established by X-ray crystallography.
The three cannabinoid quinones that were synthe-
sized in our laboratory were tested for their antiprolif-
erative activity on human cancer cell lines. All three
compounds inhibited the in vitro growth of human
cancer cell lines, albeit they differed in their potency. 4
exerted the weakest anticancer activity because con-
centrations of 12.5 µg/mL or higher were required to
inhibit by 50% or more the growth of cells tested. The
growth of SNB-19 cells was inhibited by 4 only at a
concentration of 100 µg/mL. 6 was a more potent
anticancer reagent than 4. The growth of Raji lym-
phoma cells was inhibited by over 50% at a concentra-
tion of 6.25 µg/mL, while that of J urkat lymphoma cells
(p < 0.05).
In a further series of experiments, 2 was administered
intraperitoneally at a dose of 2.5 mg/kg to nude mice
that had received a subcutaneous implant of HT-29 cells
2
days earlier. The tumors developing in mice treated
with 2 were smaller than in vehicle-treated control mice,
but the difference was statistically significant only 39
2
days after tumor implantation (the area in cm is 1.006
(
0.1205 in the experimental group vs 1.893 ( 0.3154
in the control group; p ) 0.024)

3
804 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 15
Kogan et al.
and of DU-145 prostate cancer cells was inhibited by a
concentration of 12.5 µg/mL. At concentrations of 25.0
µg/mL of 6, all lines tested were inhibited. By far the
most potent anticancer activity was displayed by 2. An
inhibition of 50% of the growth of the Raji and J urkat
lymphomas was obtained at a concentration of 2 as low
as 0.2 µg/mL, while 50% inhibition of the growth of HT-
maximum 2θ of 110°. The scan width, ∆θ, for each reflection
was 0.80 + 0.15 tan θ. An aperture with a height of 4 mm
and a variable width, calculated as 2.0 + 0.5 tan θ (in mm),
was located 173 mm from the crystal. Reflections were first
measured with a scan of 8.24 deg/min. The rate of the final
scan was calculated from the preliminary scan results so that
the ratio I/σ(I) would be at least 20 but the maximum scan
time would not exceed 60 s. If in the preliminary scan I/σ(I) <
2, this measurement was used as the datum. Scan rates varied
from 1.27 to 8.24 deg/min. Of the 96 steps in the scan, the
first and the last 16 steps were considered to be background.
During data collection, the intensities of three standard
reflections were monitored after every hour of X-ray exposure.
No decay was observed. In addition, three orientation stan-
dards were checked after 100 reflections to check the effects
of crystal movement. If the standard deviation of the h, k, and
l values of any orientation reflection exceeded 0.10, a new
orientation matrix was calculated on the basis of the recen-
tering of the 24 reference reflections. Intensities were corrected
for Lorentz and polarization effects. All non-hydrogen atoms
2
9 colon cancer and of MCF-7 mammary cancer cells
required a concentration of only 3.125 µg/mL. Further
quinonoid derivatives of cannabinoids will have to be
studied in order to determine the structure-activity
relationships of these compounds.
Compound 2 displayed a marked anticancer activity
not only in vitro but also in vivo in experiments with
nude mice that received a subcutaneous inoculation of
HT-29 colon carcinoma cells. The administration of 2
at a concentration that did not have observable adverse
effects on the hosts (5 mg/kg) resulted in significant
inhibition of the growth of the tumor cells when injected
either intraperitoneally (ip) or subcutaneously (sc) into
the region of the tumor graft. Often, compounds with
anticancer activity are difficult to solubilize. The fact
that 2 exerted a striking antitumor effect in vivo,
following either sc or ip administration, indicates that
the vehicle chosen for its solubilization (composed of 1:1:
were found by using the results of the SHELXS-86 direct
method analysis.36 After several cycles of refinements (all
crystallographic computing was done on a VAX9000 computer
at The Hebrew University of J erusalem, using the TEXSAN
structure analysis software), the positions of the hydrogen
atoms were calculated and added to the refinement process.
Ch em ica l Syn th esis. All chemical reagents were pur-
chased from Sigma-Aldrich. Organic solvents were purchased
from Bio-Lab (J erusalem). Cannabinol (5) and cannabidiol (1)
1
8 ethanol/emulphor/PBS (phosphate buffered saline))
31
8
were extracted from Lebanese-grown hashish. ∆ -THC (3)
was prepared from cannabidiol by boiling with p-toluene-
sulfonic acid in toluene.32 For column chromatography, we used
ICN silica TSC, 60 Å.
enabled its bioavailability at the cancer site.
Many compounds containing quinone groups display
potent anticancer activity (for a recent example, see ref
2
8 and references therein). A number of mechanisms
Syn th esis of 2 in th e P r esen ce of KOH. CBD (1.0 g, 3.18
mmol) was dissolved in 90 mL of petroleum ether (40-60 °C
bp), and 5% KOHaq in ethanol (10 mL, 8.77 mmol) was added.
The reaction mixture was stirred at 0 °C in an open beaker
for 3 h; after that 5% HCl (25 mL) was poured into it. The
organic layer was washed with sodium bicarbonate and water
have been suggested by which quinones may exert cell
damage. These include redox cycling, DNA damage, and
inhibition of topoisomerase, protein damage, and lipid
peroxidation.2 Similar mechanisms were shown to
mediate the antitumor effects of adriamycin and dauno-
rubicin, which have been in clinical use for the treat-
ment of solid tumors for over 30 years. The mechanism
of the anticancer activity of cannabinoid quinones is still
unclear. It may differ from the mechanisms proposed
9
and dried (MgSO ). Removal of the solvent under reduced
4
pressure yielded a glassy oil (1.1 g). The brownish oil obtained
was purified by column chromatography using a petroleum
ether-diethyl ether (95:5) solution. After crystallization from
heptane, 211.0 mg (0.64 mmol, 20.2% yield) of large brown
3
0
crystals were obtained. [R]
.1% w/v). Mp 50-51 °C. MS, m/z: 328, 313, 311, 237, 204.
1
D
in EtOH: -110° (in concentrated
2
6,27
for other cannabinoids that bind to the CB1 receptor
0
because compounds 2, 4, and 6 do not bind to this
receptor (unpublished results). The present study indi-
cates that cannabinoid quinones possess a potential for
development into anticancer drugs that may prove to
be effective not only against lymphoma cells but also
against solid tumors.
H NMR (DMSO-d ): δ 10.396 (s, 1H, OH), 6.415 (t, 1H, J )
6
1.43 Hz, H-4′), 5.080 (s, 1H, H-2), 4.501 (br s, 1H, H-9), 4.442
(br s, 1H, H-9), 3.60 (m, 1H, H-3), 2.750 (dt, 1H, J ) 11.61,
2
.85 Hz, H-4), 2.306 (dt, 2H, J ) 7.50, 1.48 Hz, H-1′′), 2.170
(m, 1H, H-5), 1.950 (m, 1H, H-5), 1.770 (m, 1H, H-6), 1.670
m, 1H,H-6), 1.670 (s, 3H, H-7), 1.546 (s, 3H, H-10), 1.425 (q,
(
2
3
H, J ) 7.50 Hz, H-2′′), 1.263 (m, 4H, H-3′′, H-4′′), 0.849 (t,
H, J ) 6.93 Hz, H-5′′). Anal. (C21 ) C. H: calcd, 8.59;
28 3
H O
found, 9.02.
Exp er im en ta l Section
Syn th esis of 4 in th e P r esen ce of Cu Cl. To a solution of
8
NMR Sp ectr om etr y. NMR data were collected on Varian
Unity Inova 500 and 600 MHz spectrometers using the
standard pulse sequences and processed with the VNMR
software.
∆ -THC (104.0 mg, 0.33 mmol) in 0.9 mL of acetonitrile (ACN)
CuCl (5.5 mg, 0.06 mmol) was added. A thin current of air
was bubbled through the mixture for 1.5 h, after which 50 mL
of ether was added. The reaction mixture was washed with
Ma ss Sp ectr om etr y. The samples were analyzed by GC-
MS in a Hewlett-Packard G1800 A GCD system with HP-5971
gas chromatograph with an electron ionization detector. Ultra-
low-bleed 5% phenyl capillary column (28 m × 0.25 mm (i.d.)
2 4
H O, dried (MgSO ), and concentrated. The yellowish oil
obtained was purified by column chromatography using pe-
troleum ether-diethyl ether (95:5) solution. After crystalliza-
tion from heptane, 33.0 mg (0.10 mmol, 30.5% yield) of very
×
0.25 µm film thickness) based on diphenylmethylsiloxane
thin yellow needles were obtained. [R]
D
in EtOH: -110° (in
chemistry (HP-5MS; Agilent Technologies) was used.
concentrated 0.22% w/v). Mp- 53-54 °C. MS, m/z: 328, 313,
1
X-r a y Cr ysta llogr a p h y. Data were measured on an EN-
RAF-NONIUS CAD4 computer-controlled diffractometer. Cu
KR (λ ) 1.541 78 Å) radiation with a graphite crystal mono-
chromator in the incident beam was used. The standard CAD4
centering, indexing, and data collection programs were used.
The unit cell dimensions were obtained by a least-squares fit
of 24 centered reflections in the range of 20° < θ < 28°.
Intensity data were collected using the θ-2θ technique to a
3
285, 272, 229, 204. H NMR (CDCl ): δ 6.300 (t, 1H, J ) 1.51
Hz, H-2), 5.330 (s, 1H, H-8), 2.920 (dd, 1H, J ) 17.04, 4.74
Hz, H-10), 2.440 (td, 1H, J ) 11.32, 5.24 Hz, H-10a), 2.310
(td, 2H, J ) 7.44, 1.51 Hz, H-1′), 2.060 (d, 1H, J ) 17.21 Hz,
H-7), 1.760 (d, 1H, J ) 17.21 Hz, H-7), 1.670 (m, 1H, H-10),
1.620 (s, 3H, H-11), 1.610 (m, 1H, H-6a), 1.438 (m, 2H, H-2′),
1.420 (s, 3H, H-12), 1.270 (m, 4H, H-3′, H-4′), 1.080 (s, 3H,
H-13), 0.840 (t, 3H, J ) 6.99 Hz, H-5′). Anal. (C21
28 3
H O ) C,H.

Quinonoid Derivatives of Cannabinoids
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 15 3805
8
Syn th esis of 4 w ith BTIB. To a solution of ∆ -THC (50.1
mg, 0.16 mmol) in acetonitrile/H O (6:1, 0.7 mL) a solution of
BTIB (215.0 mg, 0.50 mmol) in 0.7 mL of ACN/H O (6:1) was
added dropwise. The reaction mixture was stirred at room
temperature for 15 min, neutralized with aqueous NaHCO
saturated solution, and extracted with diethyl ether. The
organic layer was washed with H O, dried (MgSO ), and
initiation of the cultures, using the 3-(4,5-dimethylthiazol-2-
yl)-2,5-diphenyltetrazolium bromide (MTT) assay. The prin-
ciple of this assay is that cells that survived following exposure
to various compounds can reduce MTT to a dark-colored
2
2
3
formazan, while dead cells are incapable of doing so. The assay
was performed as described previously.3
3-35
In each MTT
2
4
assay, every concentration of the cytotoxic substance was
tested in five replicates in microplate wells. Assays with every
cell line were carried out in two to three repeated experiments.
The inhibitory effect of various compounds was calculated as
a percentage inhibition in comparison with the values obtained
in untreated wells to which vehicle (ethanol 0.5%) was added.
In Vivo Exp er im en ts. Tumors were grafted into nude mice
concentrated. The yellowish oil obtained was purified by
column chromatography using a petroleum ether-diethyl
ether (95:5) solution. After crystallization from heptane, 16.8
mg (0.05 mmol, 31.9% yield) of very thin yellow needles were
obtained.
Syn th esis of 6 in th e P r esen ce of Cu Cl. To a solution of
CBN (95.0 mg, 0.31 mmol) in 0.9 mL of ACN CuCl (10.8 mg,
6
by sc flank inoculation of 0.2 × 10 HT-29 cells in RPMI 1640
0
.11 mmol) was added. A thin current of air was bubbled
through the mixture for 6 h, after which 50 mL of ether was
added. The reaction mixture was washed with H O, dried
MgSO ), and concentrated. The red oil obtained was purified
by column chromatography using a petroleum ether-diethyl
ether (93:7) solution. After crystallization from heptane, 15.0
mg (0.05 mmol, 15.0% yield) of very large red crystals were
medium without fetal calf serum. The animals were assigned
randomly to various groups and injected ip, intratumorally,
or sc from day 2 or 14 after cell injection with vehicle (1:1:18
ethanol/emulphor/phosphate buffered saline), 5 or 2.5 mg/kg
of 2. Tumors were measured with an external caliper, and their
areas were calculated by multiplying the length of the tumors
by their width.
2
(
4
1
obtained. Mp 81-82 °C. MS, m/z: 324, 309, 281, 225, 128. H
Ack n ow led gm en t. We thank Professor H. Ben-
Bassat for providing us with human cancer cell lines.
We are grateful to Dr. Shmuel Cohen (Department of
Inorganic Chemistry, The Hebrew University) for per-
forming X-ray analyses. This work was supported in
part by grants from the Israel Science Foundation and
the U.S. National Institute of Drug Abuse (Grant DA
NMR (CDCl ): δ 8.190 (s, 1H, H-1), 7.170 (d, 1H, J ) 7.74
3
Hz, H-8), 7.080 (d, 1H, J ) 7.74 Hz, H-7), 6.480 (t, 1H, J )
1
3
.51 Hz, H-2), 2.440 (td, 2H, J ) 7.46, 1.51 Hz, H-1′), 2.180 (s,
H, H-11), 1.700 (s, 6H, H-12, H-13), 1.540 (q, 2H, J ) 7.46
Hz, H-2′), 1.350 (m, 4H, H-3′, H-4′), 0.910 (t, 3H, J ) 6.99 Hz,
H-5′). Anal. (C21
Syn th esis of 6 w ith BTIB. To a solution of CBN (50.0 mg,
.16 mmol) in ACN/H O (6:1, 0.7 mL) a solution of BTIB (215.0
mg, 0.5 mmol) in 0.7 mL of ACN/H O (6:1) was added dropwise.
The reaction mixture was stirred at room temperature for 15
min, neutralized with aqueous NaHCO saturated solution,
and extracted with diethyl ether. The organic layer was
washed with H O, dried (MgSO ), and concentrated. The red
24 3
H O ) C,H.
0
2
9
789) (to R.M) and by a contribution of the Goldhirsch
2
Foundation (to M.S.). R.M. is affiliated with the David
R. Blum Center for Pharmacy at the Hebrew University.
3
Su ppor tin g In for m ation Available: 1H, adequate, HSQC
2
4
1
oil obtained was purified by column chromatography using
petroleum ether-diethyl ether (93:7) solution. After crystal-
lization from heptane, 29.1 mg (0.09 mmol, 56.1% yield) of very
large red crystals were obtained.
and COSY NMR spectra for 4, NOESY and H NMR spectra
for 7, X-ray crystallography data for 2 and 6, elemental
analyses for 2, 4, 6, and 7, and MTT test results for 2, 4, and
6. This material is available free of charge via the Internet at
http://pubs.acs.org.
Red u ctive Acetyla tion of 4 to 7. 4 (16.9 mg, 0.05 mmol)
2
was dissolved in a solution of Ac O (acetic anhydride) (0.7 mL)
Refer en ces
and AcOH (acetic acid) (0.7 mL). Zn (5.4 mg, 0.83 mmol) was
added, and the mixture was boiled under reflux for 30 min.
The residue was filtered off, pyridine (2.2 mL) was added to
the filtrate, and the solution was left at room temperature
(
1) Thomson, R. H. Naturally Occurring Quinones; Routledge,
Chapman & Hall: London, 1987.
(2) Meganathan, R. Biosynthesis of Menaquinone (Vitamin K2) and
Ubiquinone (Coenzyme Q): A Perspective on Enzymatic Mech-
anisms. Vitam. Horm. 2001, 61, 173-218.
overnight under an N
was poured into ice-cold 5% HCl, the organic layer was washed
with NaHCO and water, dried (MgSO ), and concentrated.
2
atmosphere. After that, the solution
(
3) McIntire, W. S. Newly Discovered Redox Cofactors: Possible
Nutritional, Medical and Pharmacological Relevance to Higher
Animals. Annu. Rev. Nutr. 1998, 18, 145-177.
3
4
The red oil obtained was purified by column chromatography
using a petroleum ether-diethyl ether (90:10) solution. The
diacetate (10.0 mg, 0.02 mmol, 46.5% yield) was obtained as
(4) Lee, K. H. Novel Antitumor Agents from Higher Plants. Med.
Res. Rev. 1999, 19, 569-596.
1
(5) Begleiter, A. Clinical Applications of Quinone-Containing Alky-
lating Agents. Front. Biosci. 2000, 5, E153-E171.
a yellow oil. MS, m/z: 414, 372, 330, 287, 262, 247, 209. H
NMR (CDCl
2
7
3
): δ 6.440 (br s, 1H, H-2), 5.330 (br s, 1H, H-8),
.700 (m, 1H, H-10), 2.650 (m, 1H, H-10a), 2.440 (n, 2H, J )
.75 Hz, H-1′), 2.290 (s,3H, H-2′′), 2.270 (s,3H, H-1′′), 2.100
(
6) Di Marco, A.; Cassinelli, G.; Arcamone, F. The Discovery of
Daunorubicin. Cancer Treat. Rep. 1981, 65 (Suppl. 4), 3-8.
(7) Arcamone, F.; Cassinelli, G. Biosynthetic Anthracyclines. Curr.
Med. Chem. 1998, 5, 391-419
(m, 1H, H-7), 1.960 (m, 1H, H-10), 1.770 (m, 1H, H-7), 1.750
(
8) Zucchi, R.; Danesi, R. Cardiac Toxicity of Antineoplastic An-
(
m, 1H, H-6a), 1.680 (s, 3H, H-11), 1.518 (q, 2H, J ) 7.75 Hz,
thracyclines. Curr. Med. Chem.: Anti-Cancer Agents 2003, 3,
H-2′), 1.360 (s, 3H, H-12), 1.295 (m, 4H, H-3′, H-4′), 1.080 (s,
1
51-171.
3
H, H-13), 0.880 (t, 3H, J ) 6.87 Hz, H-5′). Anal. (C25
34 5
H O )
(
9) Thomas, X.; Le, Q. H.; Fiere, D. Anthracycline-Related Toxicity
Requiring Cardiac Transplantation in Long-Term Disease-Free
Survivors with Acute Promyelocytic Leukemia. Ann. Hematol.
C. H: calcd, 8.27; found, 8.19.
Biologica l Eva lu a tion . Raji and J urkat cells were sus-
pended in RPMI 1640 medium, supplemented with 20% heat-
inactivated fetal calf serum, 2 mM L-glutamine, 100 U/mL
2
002, 81, 504-507.
(
10) Razdan, R. K. Structure-Activity Relationships in Cannab-
inoids. Pharmacol. Rev. 1986, 38, 75-149.
penicillin, and 0.01 mg/mL streptomycin at 37 °C in a 5% CO
2
(11) Mechoulam, R.; Hanus, L.; Fride, E. Towards Cannabinoid
humidified atmosphere. Other cell lines were suspended in
RPMI 1640 medium, supplemented with 10% heat-inactivated
fetal calf serum, 2 mM L-glutamine, 100 U/mL penicillin, and
DrugssRevisited. Prog. Med. Chem. 1998, 35, 199-243.
(
12) Barth, F.; Rinaldi-Carmona, M. The Development of Cannab-
inoid Antagonists. Curr. Med. Chem. 1999, 6, 745-55.
13) Bagshaw, S. M.; Hagen, N. A. Medical Efficacy of Cannabinoids
and Marijuana: A Comprehensive Review of the Literature. J .
Palliat. Care 2002, 18, 111-122.
(
0
.01 mg/mL streptomycin at 37 °C in a 5% CO
2
humidified
atmosphere.
(
14) Mechoulam, R.; Ben-Zvi, Z.,; Gaoni, Y. Hashish-13. On the
Nature of the Beam Test. Tetrahedron 1968, 24, 5615-5624.
15) Hodjat-Kashani, H.; Lambert, G.; Duffley, R.; Razdan, R. Hash-
Cell P r olifer a tion Test. Aliquots of suspensions of cancer
cells were dispensed at 200 µL volumes into wells of 96-well
(
6
tissue culture plates at densities of 0.02 × 10 cells/well.
8
ish: Oxidation of ∆ -Tetrahydrocannabinol (THC); Synthesis of
Various concentrations of cannabinoid quinones were intro-
duced into the wells, and their efficacy was tested 3 days after
∆
8
-THC-1,2-dione and 2-Hydroxy-∆ -THC. Heterocycles 1986, 24,
8
1973-1976.

3
806 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 15
Kogan et al.
(
16) Tamura, Y.; Yakura, T.; Tohma, H.; Kikuchi, K.; Kita, Y.
Hypervalent Iodine Oxidation of p-Alcoxy- and Related Phe-
nols: A Facile and Efficient Synthesis of p-Quinones. Synthesis
(27) Portella, G.; Laezza, C.; Laccetti, P.; De Petrocellis, L.; Di Marzo,
V.; Bifulco, M. Inhibitory Effects of Cannabinoid CB1 Receptor
Stimulation on Tumor Growth and Metastatic Spreading: Ac-
tions on Signals Involved in Angiogenesis and Metastasis.
FASEB J . 2003, 17, 1771-1773.
(28) Kissau, L.; Stahl, P.; Mazitschek, R.; Giannis, A.; Waldmann,
H. Development of Natural Product-Derived Receptor Tyrosine
Kinase Inhibitors Based on Conservation of Protein Domain
Fold. J . Med. Chem. 2003, 46 2917-2931.
1
989, 126-127.
(
(
(
17) Akai, S.; Kita, Y. Recent Progress in the Synthesis of p-Quinones
and p-Dihydroquinones through Oxidation of Phenol Derivatives.
A Review. Org. Prep. Proced. Int. 1998, 30, 603-629.
18) Barret, R.; Daudon, M. Oxidation of Phenols to Quinones by Bis-
(
4
trifluoroacetoxy)iodobenzene. Tetrahedron Lett. 1990, 31, 4871-
(
29) Ollinger, K.; Kagedal, K. Induction of Apoptosis by Redox-Cycling
Quinones. Subcell. Biochem. 2002, 36, 151-70.
872.
19) Kato, N.; Sugaya, T.; Mimura, T.; Ikuta, M.; Kato, S.; Kuge, Y.;
Tomioka, S.; Kasai, M. Facile and Efficient Synthesis of 7,10-
Dihydroxy-6H-pyrazolo[4,5,1-de] Acridin-6-one via Hypervalent
Iodine Oxidation. Synthesis 1997, 625-627.
(
30) Gewirtz, D. A. A Critical Evaluation of the Mechanisms of Action
Proposed for the Antitumor Effects of the Anthracycline Anti-
biotics Adriamycin and Daunorubicin. Biochem. Pharmacol.
1
999, 57, 727-741.
(
20) Barret, R.; Daudon, M. An Efficient Synthesis of J uglone. Synth.
Commun. 1990, 20, 2907-2912.
1
(
31) Gaoni, Y.; Mechoulam, R. The Isolation and Structure of ∆ -
THC and Other Neural Cannabinoids from Hashish. J . Am.
Chem. Soc. 1971, 93, 217-224.
32) Gaoni, Y.; Mechoulam, R. The Isomerisation of Cannabidiol to
Tetrahydrocannabinoids. Tetrahedron 1966, 22, 1481-1488.
33) Carmichael, J .; DeGraff, W. G.; Gazdar, A. F.; Minna, J . D.;
Mitchell, J . B. Evaluation of a Tetrazolium-Based Semiauto-
mated Colorimetric Assay: Assessment of Chemosensitivity
Testing. Cancer Res. 1987, 47, 936-942.
34) Rubinstein, L. V.; Shoemaker, R. H.; Paull, K. D.; Simon, R. M.;
Tosini, S.; Skehan, P.; Scudiero, D. A.; Monks, A.; Boyd, M. R.
Comparison of in Vitro Anticancer-Drug-Screening Data Gener-
ated with a Tetrazolium Assay versus a Protein Assay against
a Diverse Panel of Human Tumor Cell Lines. J . Natl. Cancer
Inst. 1990, 82, 1113-1118.
35) Rubnov, S.; Kashman, Y.; Rabinowitz, R.; Schlesinger, M.;
Mechoulam, R. Suppressors of Cancer Cell Proliferation from
Fig (Ficus carica) Resin: Isolation and Structure Elucidation.
J . Nat. Prod. 2001, 64, 993-996.
36) Sheldrick, G. M. Crystallographic Computing 3; Oxford Univer-
sity Press: Oxford, U.K., 1985; pp 175-189.
(
21) Capdevielle, P.; Maumy, M. Ortho-Hydroxylation Selective des
Phenols: 2. Un Noveau Systeme Catalytique a Charactere
Preparative (Selective Ortho Hydroxylation of Phenols. 2. A New
Catalytic System for Preparation). Tetrahedron Lett 1982, 23,
(
(
1
577-1580.
(
(
(
22) Reynolds, W. F.; Enriquez, R. G. Choosing the Best Pulse
Sequences, Acquisition Parameters, Postacquisition Processing
Strategies, and Probes for Natural Product Structure Elucidation
by NMR Spectroscopy. J . Nat. Prod. 2002, 65, 221-244.
23) K o¨ ck, M.; Reif, B.; Griesinger, C.; Fenical, W. Differentiation of
HMBC Two- and Three-Bond Correlations: A Method To
Simplify the Structure Determination of Natural Products.
Tetrahedron Lett. 1996, 37, 363-366.
(
24) Reif, B.; K o¨ ck, M.; Kerssebaum, R.; Kang, H.; Fenical, W.;
Griesinger, C. ADEQUATE, a New Set of Experiments To
Determine the Constitution of Small Molecules at Natural
Abundance. J . Magn. Reson., Ser. A 1996, 118, 282-285.
25) Croxford, J . L. Therapeutic Potential of Cannabinoids in CNS
Disease. CNS Drugs 2003, 17, 179-202.
(
(
(
(
26) Guzman, M. Cannabinoids: Potential Anticancer Agents. Nat.
Rev. Cancer 2003, 3, 745-755.
J M040042O