4084
C. J. Creighton et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4083–4085
effective long-acting formulations. For example, an
osmotically-controlled extended-release form of meth-
ylphenidate has been developed, which overcomes
tachyphylaxis and provides effective long acting treat-
ment, without abuse potential.5
O
Me
O
Me
OH
b
a
Cl
Cl
8
9
Tricyclic antidepressants have also been applied as
alternative medications for ADHD, but the results are
inconsistent and sometimes tragic. The use of imipra-
mine (4) and desipramine (5), the most frequently pre-
scribed antidepressants for the treatment of ADHD, is
currently clouded by concerns about their safety.6
Therefore, while efficacious, these compounds are con-
sidered second-line agents for the treatment of ADHD.
Me
O
Me
c
O
7
O
Cl
10
Scheme 2. Reagents and conditions: (a) DEAD, PPh3, 4-hydroxy-3-
methylacetophenone; (b) mCPBA, CHCl3, reflux; (c) EtOH, MeNH2
(40% in water), 150 °C, 10 min.
Atomoxetine (6) has been recently approved for the
treatment of ADHD. Atomoxetine is structurally re-
lated to fluoxetine yet selectively blocks the presynaptic
norepinephrine transporter causing a measurable in-
crease in extracellular levels of norepinephrine and
dopamine in the prefrontal cortex.7–9 This compound
may represent a new paradigm for the treatment of
ADHD. We are currently interested in understanding
the pharmacology of selective and mixed monoamine
transport inhibitors for the treatment of a variety of
CNS disorders. It has been reported that atomoxetine
(6) undergoes oxidative metabolism in humans leading
to the formation of a major phase I metabolite, 4-hy-
droxyatomoxetine (7), and a minor metabolite resulting
from demethylation, N-desmethylatomoxetine (Scheme
1).10 The primary enzyme responsible for the formation
of 7 from 6 has been identified as cytochrome P450
CYP2D6. We identified 7 as the major component upon
treatment of 6 with human liver microsomes, prepared 7
synthetically, and evaluated it against a panel of recep-
tors, ion channels, and enzymes. We found that com-
pound 7 significantly interacted with the l, d, and j-
opioid receptors at a concentration of 10 lM. Therefore
we examined the pharmacological properties of 7 at
opioid receptors in more detail using binding and
functional assays.
3. Biology
HEK-293 cell membranes from cells stably expressing
the j-opioid receptor and other opioid receptors were
prepared as described, with the exception that the
binding buffer used was 50 mM Tris–Cl pH 7.8, 5 mM
MgCl2 and 1 mM EGTA.11
The j-opioid and other opioid receptors are coupled to
the heterotrimeric G protein Gi, and do not normally
elicit a calcium response. Therefore, stable cell lines
expressing the respective opioid receptors and the Gqi5
construct (molecular devices) were generated. The Gqi5
construct expresses a hybrid Gq/Gi protein that redirects
Gi signaling to the Gq pathway, leading to calcium re-
lease. HEK-293 cells expressing receptor and Gqi5 were
plated onto 96-well plates at a density of 50,000 cells/
well in a total volume of 50 lL. Two days later cells were
prepared for assay using the FLIPR Calcium Assay Kit
(molecular devices) according to manufacturer’s direc-
tions, with the exception that the volume of dye mix
added to each well was 50 lL instead of 100 lL. Cells
were treated with compound at the indicated concen-
trations, added in a total volume of 100 lL at twofold
the final concentration. Data points were collected at
one per second for 120 s, then one every three seconds
for 30 s for a total collection time of 150 s. The data were
used to generate EC50 curves using GraphPad Prizm
v3.0, and the data for the j-opioid receptor is shown in
Figure 2.
2. Chemistry
The synthesis of 7 was initiated by Mitsunobu conden-
sation of 4-hydroxy-3-methylacetophenone with (S)-3-
chloro-1-phenyl-1-propanol
8 to provide ether 9
(Scheme 2). Baeyer–Villiger rearrangement of 9 yielded
ester 10, which was subsequently treated with methyl-
amine in a water/ethanol mixture at 150 °C to afford the
desired 7 in approximately 50% yield for the three steps
(Fig. 1).
4. Discussion
Biological screening against the three opioid receptors
shown in Table 1 indicated that 4-hydroxyatomoxetine
(7) binds to both the l and j-opioid receptors to the
extent of 164 and 88 nM IC50’s, respectively. The values
for the l and j-opioid receptors for 7 were 25- and 50-
fold greater than those corresponding to the parent 6.
Subsequent analysis of the pharmacological function of 7
(Fig. 2) on these receptors showed that 4-hydroxyato-
moxetine is a partial agonist for the j-opioid receptor,
HO
Me
O
Me
O
metabolisim
NHMe
NHMe
7
6
Scheme 1.