Naphthyl piperazines with dual activity as 5-HT1D antagonists and 5-HT reuptake inhibitors

Article abstract of DOI:10.1016/j.bmcl.2007.03.101

SAR around a known molecule with dual 5-HT_1D antagonist and 5-HT_transporter inhibitory activity has led to the discovery of molecules with improved dual activity and reduced cross-reactivity toward other aminergic receptors (5-HT_1B, α₁, and D₂).

Full text of DOI:10.1016/j.bmcl.2007.03.101

Bioorganic & Medicinal Chemistry Letters 17 (2007) 3344–3348
Naphthyl piperazines with dual activity as 5-HT antagonists
1
D
and 5-HT reuptake inhibitors
b
a,
b
b
b
*
Ana B. Buenob, Jeremy Gilmore, John Boot, Richard Broadmore, Jane Cooper,
b
a
a
b
Jeremy Findlay, Lorna Hayhurst, Alicia Marcos, Carlos Montero, Stephen Mitchell,
b
b
b
b
Graham Timms, Rosemarie Tomlinson, Louise Wallace and Leslie Walton
a
Lilly SA, Avda. de la Industria, 30, 28108, Alcobendas, Madrid, Spain
Eli Lilly & Co. Ltd., Lilly Research Centre, Erl Wood Manor, Windlesham, Surrey GU20 6PH, UK
b
Received 22 February 2007; revised 28 March 2007; accepted 30 March 2007
Available online 5 April 2007
Abstract—SAR around a known molecule with dual 5-HT1D antagonist and 5-HTtransporter inhibitory activity has led to the discov-
ery of molecules with improved dual activity and reduced cross-reactivity toward other aminergic receptors (5-HT1B, a
1 2
, and D ).
Ó 2007 Elsevier Ltd. All rights reserved.
6
Major depression is a common illness affecting 5–10% of
1
the western European population. It is well established
that a significant component of the symptomatology of
depression can be attributed to a reduction in serotoner-
gic function. Serotonin (5-HT) reuptake inhibitors
observed after chronic SRI treatment. It has also been
demonstrated that pindolol, a 5-HT1A antagonist,
7
increases the clinical effectiveness of a SRI, fluoxetine.
These results indicated that the effect of a 5-HT reuptake
inhibitor is more pronounced and long-lasting with the
co-administration of a 5-HT1A or 5-HT1B/1D antagonist.
Therefore, the identification of a drug with dual activity
as a 5-HT reuptake inhibitor and a 5-HT1A, 5-HT1B or a
5-HT1D antagonist should produce a similar effect,
potentially resulting in an antidepressant drug with a
superior profile to an SRI alone, both in terms of effi-
cacy and onset of clinically meaningful antidepressant
effects. While the dual activity 5-HT /SRI and
(
with their effect being believed to be due to an enhance-
SRIs) have been widely used as antidepressant drugs,
2
ment of postsynaptic 5-HT levels. Although SRIs are
effective drugs, their onset of clinical action may be
3
,4
delayed by 2–4 weeks. One hypothesis for this delay
in symptomatic improvement is that time is required
for desensitization of the 5-HT / and 5-HT1A autore-
1
B 1D
5
ceptors. The increased release of 5-HT mediated by the
SRI stimulates these receptors, which in turn results in
an inhibition of further 5-HT release, thus limiting the
effect of the SRI. Only on desensitization of these auto-
receptors does the SRI produce its therapeutic effect.
According to this mechanism of action, blockade of
the autoreceptors by an antagonist would prevent the
initial lack of efficacy of the SRI.
1
B
8
–10
5-HT1A/SRI has been studied elsewhere,
interested in the identification of a drug with dual activity
as a 5-HT reuptake inhibitor and a 5-HT1D antagonist in
we were
1
1
order to determine the role of the 5-HT1D autoreceptor
in the augmentation of extracellular serotonin. With this
aim, selectivity toward 5-HT1A and 5-HT1B was also
required.
Combined administration of a 5-HT reuptake inhibitor
and a 5-HT1B/1D antagonist to guinea pigs demonstrated
a significant increase in the extracellular levels of 5-HT
above those evoked by a SRI alone, and to a level only
Initial studies in our laboratories demonstrated that the
incorporation of a SRI pharmacophore into a known 5-
1
2
HT1D agonist ligand, PNU-109291 1, (Fig. 1) led to
1
compounds with dual 5-HT1D antagonist/SRI activity.
3
Microdialysis studies in guinea pigs demonstrated that 2
gave an elevation of the extracellular 5-HT levels signif-
icantly above that obtained after a maximally effective
acute dose of the SRI fluoxetine. However, this indole
derivative was also a potent a and D ligand. Although
Keywords: 5-HT1D antagonist; Serotonin reuptake inhibitor.
*
Corresponding author. Tel.: +34 91 6633402; fax: +34 91 6633411;
e-mail: bueno_ana_belen@lilly.com
1
2
0
960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.03.101

A. B. Bueno et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3344–3348
3345
H NOC
O
2
O
N
N
1
PNU-109291
F
F
H NOC
2
H NOC
2
O
N
O
N
N
N
2
3
Binding affinities Ki (nM)
Binding affinities Ki (nM)
h5-HT1D
8.3
203
0.11
0.68
42
h5-HT1D
56
281
0.78
16
h5-HT1B
r5-HTtransporter
h5-HT1B
r5-HTtransporter
α
1
α
1
D2
D2
4.7
Figure 1. 5-HT1D agonist PNU-109291 (1) and 5-HT1D/SRI ligand lead compounds (2 and 3).
the cross-reactivity at the a receptor could be somewhat
with Zn(CN) and hydrolysis of the resultant nitrile
2
1
diminished by replacing the indole with a naphthylpiper-
azine, as in 3, further improvement of the selectivity
against both receptors would be desirable.
13 with basic H O provided the primary amide 14.
2
2
Deprotection of the primary alcohol provided the race-
mic alcohol 15. Activation of the primary alcohol of
(RS)-15 or (RS)-16 as the mesylate and reaction with
4 provided compounds 3 and 21, respectively, as a mix-
ture of diastereoisomers.
We report herein the effect on this dual activity of mod-
ifying the basic structure of compound 3 in order to
reduce the cross-reactivity on a and D receptors.
1
2
The two diastereomers of 3 and 21 were separated by
chiral HPLC (3a and 3b from 3; and 21a and 21b from
21, Table 1). The affinity for the 5-HT1D receptor and
A series of naphthyl piperazines 4–8 were synthesized
1
4
17
for this study as potential SRI pharmacophores.
Fig. 2) They were combined with the isochroman pres-
ent in 3 or with a dihydroisobenzofuran.
(
the 5-HTtransporter was greater for one isomer in both
cases, necessitating an enantioselective synthesis of
intermediates 15 and 16. Compound (S)-16 was synthe-
1
2
The racemic isochroman (RS)-16 was synthesized as
sized as described. We also applied these conditions to
resolve the two enantiomers of 10 by lipase-catalyzed
hydrolysis of the ethyl ester (Scheme 2). Thus, hydroly-
sis of (±)-10 by Amano lipase at pH 7 to 50% conversion
provided the ester (+)-10 in a 45% yield and the acid
(À)-23 in a 45% yield, both compounds being of 95%
1
2
described. The isobenzofuran (RS)-15 was synthe-
sized following the reaction conditions shown in
Scheme 1. Commercially available 5-chlorophthalide
1
5
9
formed into the ester 10 by Horner–Emmons reaction.
was reduced to its lactol, which in turn was trans-
1
8
This ester was reduced with DIBAH, followed by
1
ee. A single recrystallization of (À)-23 in ether/hexane
provided this compound essentially enantiomerically
pure (>98%). The absolute configuration of (À)-23 was
determined by derivatization to the corresponding (R)-
and (S)-phenylglycine amides according to the method
6
NaBH , to the corresponding alcohol 11. The alcohol
4
was protected as the TBDMS derivative 12. Palla-
dium(0)-catalyzed coupling of the aromatic chloride
1
9
described in the literature. Reduction of the free acid
with borane, followed by TBDMS protection, provided
enantiomerically pure (À)-12. Standard conditions used
for the synthesis of racemic 15 (Scheme 1) provided (S)-
H
R1
N
R2
15 enantiomerically pure. Synthesis of 3a and 21a from
H
(
S)-16 and (S)-15, respectively, demonstrated that the
N
N
diastereomer that possesses the dual activity in both
scaffolds is (S). Enantiopure (S)-15 and (S)-16 were used
for the synthesis of 17–20 and 22.
N
R1
R2
R3
R3
1
2
3
1
2
3
4
5
6
R =Me, R =H, R =F
7 R =Me, R =H, R =F
The undesired enantiomer (+)-10 was racemized with
NaOEt to provide (±)-10 quantitatively, which could
be resubmitted to the enzymatic hydrolysis conditions
(Scheme 2).
1
2
3
1
2
3
R =Me, R =H, R =CN
8 R =H, R =H, R =F
1
2
3
R =H, R =H, R =CN
Figure 2. Naphthyl piperazines 4–8.

3346
A. B. Bueno et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3344–3348
R2
Cl
Cl
a
d
O
O
O
O
R1
R
9
1
0 R = CO Et
13 R = CH OTBDMS; R = CN
1
2
2
2
b
e
f
1
1 R = CH OH
2
1
2
1
4 R = CH OTBDMS; R = CO NH
2 2 2
c
12 R = CH OTBDMS
2
1
2
1
5 R = CH OH; R = CO NH
2
2
2
O
O
(
)n
g
( )n
H N
H N
2
2
O
O
OH
R
(
(
(
(
RS)-15 n = 1
)-15 n = 1
RS)-16 n = 2
)-16 n = 2
S
3, 17-22
S
Scheme 1. Reagents and conditions: (a) i—DIBAH, CH
THF, À78 °C then NaBH , MeOH, 0 °C, 73%; (c) TBDMS-Cl, imidazole, CH
, NaOH (2N), Bu NHSO ; CH Cl
, 0 °C–rt, 90%; (f) TBAF, THF, rt, 83%; (g) i—MsCl, Et
2
Cl
2
, À78 °C, 84%; ii—(EtO)
2
P(O)CH
2
CO
2
Et, Cs
, Pd
N, DMF, 70%; ii—4–8, K
2
CO
3
, THF, 0 °C–rt, 73%; (b) DIBAH,
, t-Bu
P, dioxane, 120 °C, 70%; (e)
CO , CH
CN, 70 °C.
4
2
Cl
2
, 90%; (d) Zn(CN)
2
2
(dba)
3
3
H
O
2 2
4
4
2
2
3
2
3
3
1 2 i
Table 1. Receptor binding affinities at h5-HT1D, h5-HTtransporter, h5-HT1B, h5-HT1A, rat a and D receptors expressed as K (nM) (mean of runs in
triplicate); functional affinity at h5-HT1D expressed as K
B
or % displacement at 30 nM
O
(
)n
O
H N
2
R
a
R
Compound
n
Configuration
5-HT1D
5-HT1D function
5-HTtransporter
5-HT1B
5-HT1A
a
1
D
2
b
2
3
2
2
2
2
2
2
2
2
1
1
1
RS
RS
S
8
56
7
4.3
0.1
0.8
3.0
7.4
2.1
8.1
2.2
5.7
1.3
8.8
1.3
203
281
N.D.
N.D.
0.7
42
4.7
b
4
4
4
7
8
5
6
4
4
5
0.01
0.01
11.1
4.05
0.86
(124%)
(101%)
0.05
1.9
16
80
3
3
7
8
9
0
a
103
91
115
105
307
71
78
b
R
S
88
14
10
74
91
1
1
1
2
2
2
2
>1000
>1000
21
1170
704
287
88
b
S
N.D.
96
S
0.9
1.0
14
67
1.8
26
16
13
13
353
178
64
S
15
114
1a
1b
2
S
26
R
S
>1000
88
43
338
107
346
b
0.18
N.D.
a
R corresponds to the naphthyl piperazines 4–8 without the H attached to the piperazine.
N.D., not determined.
b
Cl
Cl
Cl
a
b
c
d-f
(
±)-10
O
O
S
O
(S)-15
R
+
for 23
O
O
(
+)-10
(-)-23
(-)-12
OEt
OH
OTBDMS
Scheme 2. Reagents and conditions: (a) Amano lipase, buffer, pH 7, 24 h; (b) NaOEt, EtOH, quant.; (c) i—BH
3 2
.SMe ; ii—TBDMS-Cl, imidazole,
CH Cl , 92%;(d)Zn(CN) , Pd (dba) , t-Bu , NaOH(2N), Bu NHSO ;CH Cl , 0 °C–rt, 90%;(f)TBAF, THF, rt, 83%.
2
2
2
2
3
3
P, dioxane, 120 °C, 70%;(e)H
2
O
2
4
4
2
2
In vitro inhibition of the serotonin transporter was mea-
sured by the ability of compounds to displace [ H]-
L-M(tk-) cells expressing the cloned human 5-HT or
1B
3
20
5-HT1D receptors. Human 5-HT
was determined using [ H]-8-OH-DPAT as radioligand
in membrane homogenates prepared from L-M(tk-) cells
expressing the cloned human 5-HT1A receptor. Adren-
ergic a binding affinity was determined in rat cortex
binding affinity
1A
3
citalopram from human 5-HT
2
expressed in
HEK293 cells. Human 5-HT1B and 5-HT1D binding
transporter
0
3
affinities were determined using [ H]-GR125743 as radi-
oligand in membrane homogenates prepared from
21
1

A. B. Bueno et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3344–3348
3347
3
20
membranes using [ H]-prazosin, and D binding activ-
2
3
ity was determined in rat caudate membranes using [ H]-
2
0
raclopride.
Functional 5-HT1D antagonism was measured by the
dextral shift of the 5-HT dose–response curves for
3
5
the binding of GTP-c [ S] to human 5-HT receptor
1
D
stably expressed into LM (tk-) cells, and the results
were expressed as K values or % displacement at
B
2
2
3
0 nM.
The biological results of compounds 3 and 17–22 are
summarized in Table 1.
The two diastereomers of 3 (3a and 3b) show that one
diastereomer (3a, S configuration at the isochroman
center) has higher affinity for the 5-HT1D receptor and
the 5-HTtransporter, while both isomers display similar
cross-reactivity on the other receptors.
Figure 3. Effect of 3a or 19 at 10 mg/kg po and fluoxetine at 20 mg/kg
po on the elevation of 5-HT in the hypothalamus of the freely moving
guinea pig. Drug administered at arrow. Data expressed as a percentage
of a pre-injection control period and represent means + SEM. The
effect of 3a was not statistically different to fluoxetine.
Comparison of the 2-naphthylpiperazine 17 with the 1-
naphthylpiperazine 3a shows that the 2-naphthylpiper-
azine maintains the binding affinity for the 5-HT1D
receptor, although there is a significant reduction in
the functional 5-HT1D antagonism.
the potency of 19 is sevenfold higher toward the 5-
2
6
HT1D receptor than that of 3a. Dose–response studies
performed with 19 show that this compound displays a
considerable elevation of 5-HT at all doses (1–20 mg/
kg) (Fig. 4).
The substitution of the naphthyl fluorine in 3a by nitrile
(
In summary, we have identified a series of isochromans
and isobenzofurans with dual 5-HT1D antagonism and
5-HTtransporter inhibition. Isomer (S) shows a better pro-
file in both series. 1-Napththyl- and 2-naphthyl-pipera-
zines have been shown to be good 5-HT1D/SRI
pharmacophores in this study, with the presence of a
compound 19) retains the 5-HT1D and SRI activity, and
reduces the binding affinity of the compound for the a
1
2
3
and D receptors.
2
Elimination of the chiral center on the piperazine of 17
or 19 by removal of the methyl substituent provided
compounds 18 and 20, respectively. The absence of the
methyl group reduced the SRI activity while increasing
3
-methyl group in the piperazines improving the affinity
for the transporter. We have identified two compounds
9 and 22 with K < 3 nM at 5-HT and 5-HTtransporter
1
,
the affinity for the D receptor.
2
i
1D
and >50-fold selectivity over a and D .
1
2
Interesting results were obtained when the isochroman
was replaced by an isobenzofuran. As observed on the
isochroman series, a higher affinity for the 5-HT1D and
the 5-HTtransporter on the isobenzofuran series resides
in the (S) isomer.
These two compounds are viable candidates for further
development: details will be reported in due course.
A single change on 21a, from fluorine to nitrile in the 6-
position of the aromatic ring, generated 22, a compound
with a high affinity for 5-HT1D and the 5-HTtransporter
,
and greater than 50-fold selectivity over the undesired
receptors.
Compounds 19 and 22 have the best profile in terms of
potency on 5-HT1D and the 5-HTtransporter (K < 3 nM)
i
and selectivity (>100-fold over a and D , and >20-fold
1
2
of 5-HT1D affinity over 5-HT1B).
2
4
Microdialysis studies (Fig. 3) demonstrated that the
dual pharmacology of 3a and 19 leads to an elevation
2
5
of extracellular 5-HT levels in the guinea pig hypothal-
amus (at 10 mg/kg po). For 19, the increase obtained
(
for a 3 h period post-drug administration) was signifi-
cantly greater than that obtained after a maximally
effective acute dose of the SSRI fluoxetine. Interestingly,
the effect is higher with 19 than with 3a. These two com-
pounds display similar affinity for the transporter but
Figure 4. Dose–response effect of 19 (1–20 mg/kg po) on the elevation
of 5-HT in the hypothalamus of the freely moving guinea pig. Drug
administered at arrow. Data expressed as a percentage of a pre-
injection control period and represent means + SEM (p < 0.05).

3348
A. B. Bueno et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3344–3348
Acknowledgments
16. DIBAH reduction stopped at the aldehyde. Further
reduction with NaBH after removal of the aluminum
salts was necessary to obtain the alcohol.
4
We thank Dr. Maria Luz de la Puente for the HPLC chi-
ral purifications and e.e. determinations. We thank Dr.
Juan F e´ lix Espinosa and Dr. Paloma Vidal for the deter-
mination of the absolute configurations.
1
1
7. Chiral HPLC conditions: Chiracel OJ column, eluting
with 1:1 hexane (0.2% dimethylethylamine)/ethanol at a
1
ml/min flow rate.
8. The enantiomeric excess was analyzed by chiral HPLC
using a Chiralpak AD column, eluting with hexane/
ethanol in isocratic mode (70% of B) at a 1 ml/min flow
rate. Derivative (+)-10 was hydrolyzed to the correspond-
ing acid to determine its ee.
References and notes
1
. M o¨ ller, H.J.; Henkel, V. Copenhagen, WHO Regional
Office for Europe, 2005.
. Leonard, B. E. CNS Drugs 1995, 4, 1.
19. Yabuuchi, T.; Kusumi, T. J. Org. Chem. 2000, 65, 397; for
the application of this technique to the configurational
determination of a (1,3-dihydro-isobenzofuran-1-yl)-acetic
acid, see: Holler, U.; Gloer, J. B.; Wicklow, D. T. J. Nat.
Prod. 2002, 65, 876.
20. Torrado, A.; Lamas, C.; Agejas, F. J.; Jimenez, A.; Diaz,
N.; Gilmore, J.; Boot, J.; Findlay, J.; Hayhurst, L.;
Wallace, L.; Broadmore, R.; Tomlinson, R. Bioorg. Med.
Chem. 2004, 12, 5277.
2
3
. Montgomery, S. A. In Psychopharmacology; Bloom, F. E.,
Kupfer, D. J., Eds.; The Fourth Generation of Progress;
Raven: New York, 1995; p 451.
4
5
. Goodwin, G. M. J. Clin. Psychiatry 1996, 57, 9.
. Davidson, C.; Stamford, J. A. Br. J. Pharmacol. 1996, 353,
281.
6
7
8
. Rollema, H.; Clarke, T.; Sprouse, J.; Schulz, D. W.
J. Neurochem. 1996, 67, 2204.
. Perez, V.; Gilaberte, I.; Faries, D.; Alvarez, E.; Artigas, F.
Lancet 1997, 349, 1594.
. Matzen, L.; van Amsterdam, C.; Rautenberg, W.; Greiner,
H. E.; Harting, J.; Seyfried, C. A.; B o¨ ttcher, H. J. Med.
Chem. 2000, 43, 1149.
21. Taylor, E. W.; Duckles, S. P.; Nelson, D. L. J. Pharmacol.
Exp. Ther. 1986, 236, 118.
22. Pullar, I. A.; Boot, J. R.; Carney, S. L.; Cohen, M. L.;
Colvin, E. M.; Hardy, C. H. L.; Lucaites, V. L.; Nelson,
D. L.; Schenck, K. W.; Tomlinson, R.; Wedley, S. Eur. J.
Pharmacol. 2001, 432, 9.
23. Substituents bigger than H or F in this position produce a
decrease in 5-HT1D, a , and D affinity (unpublished
9
. Malagi e´ , I.; Trillat, A-C.; Bourin, M.; Jacquot, C.; Hen,
R.; Gardier, A. M. J. Neurochem. 2001, 76, 865.
1
2
results). Cyanide seems to be too big for a and D , but it
1
2
1
0. Lovell, P. J.; Blaney, F. E.; Goodacre, C. J.; Scott, C. M.;
Smith, P. W.; Starr, K. R.; Thewlis, K. M.; Vong, A. K.
K.; Ward, S. E.; Watson, J. M. Bioorg. Med. Chem. Lett.
has the right size and shape for the 5-HT1D receptor,
decreasing the cross-reactivity.
24. Mitchell, S. N.; Greenslade, R. G.; Cooper, J. Eur. J.
Pharmacol. 2001, 432, 19.
2007, 17, 1033.
1
1
1. Pullar, I.; Boot, J. R.; Broadmore, R. J.; Eyre, T. A.;
Cooper, J.; Sanger, G. J.; Wedley, S.; Mitchell, S. N. Eur.
J. Pharmacol. 2004, 493, 85.
2. Ennis, M. D.; Ghazal, N. B.; Hoffman, R. L.; Smith, M.
W.; Schalchter, S. F.; Lawson, C. F.; Im, W. B.;
Pregenzer, J. F.; Svensson, K. A.; Lewis, R. A.; Hall, E.
D.; Sutter, D. M.; Harris, L. T.; McCall, R. B. J. Med.
Chem. 1998, 41, 2180.
25. For homology and pharmacology similarities between
human and guinea pig 5-HT1B and 5-HT1D receptors, see
B u¨ hlen, D.; Fink, K.; B o¨ ing, C.; G o¨ thert, M. Pharmacol-
ogy 1996, 353, 281; Wurch, T.; Palmier, C.; Colpaert, F.
C.; Pauwels, P. J. J. Neurochem. 1997, 68, 410; Barnes, N.
M.; Sharp, T. Neuropharmacology 1999, 38, 1083; for
correlation between in vitro binding affinity to human and
guinea pig 5-HT1D receptors, see: Pregenzer, F. J.; Alberts,
G. L.; Bock, J. H.; Slightom, J. L.; Im, W. B. Neurosci.
Lett. 1997, 235, 117.
1
3. Timms, G. H.; Boot, J. R.; Bradmore, R. J.; Carney, S. L.;
Cooper, J.; Findlay, J. D.; Gilmore, J.; Mitchell, S.; Moore,
M. A.; Pullar, I.; Sanger, G. J.; Tomlinson, R.; Tree, B. B.;
Wedley, S. Bioorg. Med. Chem. Lett. 2004, 14, 2469.
4. Bueno, A. B.; Flynn, C. J.; Gilmore, J.; Marcos, A.;
Montero, C.; Porter, W.; Williams, A. C. Tetrahedron
Lett. 2005, 46, 7769.
26. Although the affinity of 19 for 5-HT1D is 20-fold
higher than for 5-HT1B or 5-HT1A, we cannot rule out
that this affinity for these receptors (K = 21 and
1
1
i
26 nM, respectively) could be partially responsible for
these differences between 3a and 19 in the serotonin
elevation.
5. Provided by Lancaster.