Development of a presynaptic 5-HT1A antagonist.

Article abstract of DOI:10.1016/S0960-894X(02)00879-X

A new 5-HT(1A) silent antagonist 14 (5-HT(1A) IC(50)=2.2 nM) antagonizes the effects of agonists on reciprocal forepaw treading behavior, on neuronal firing in the rat dorsal raphe, and on 5-HT(1A) release in the raphe and hippocampus. While 14 alone was inactive in the social interaction paradigm, it completely reversed the social interaction activity of the serotonergic compounds (buspirone, 1, and 2).

Full text of DOI:10.1016/S0960-894X(02)00879-X

Bioorganic & Medicinal Chemistry Letters 13 (2003) 285–288
Development of a Presynaptic 5-HT1A Antagonist
Ronald J. Mattson,a,* John D. Catt,a Charles P. Sloan,a Qi Gao,d Richard B. Carter,b
Anthony Gentile,a Cathy D. Mahle,c F. Fatima Matos,d Rachel McGovern,a
Cam P. VanderMaelene and Frank D. Yoccaa
aBristol-Myers Squibb Pharmaceutical Research Institute, Wallingford, CT 06492-7660, USA
bCoCensys, Inc., Irvine, CA 92718, USA
cBayer Pharmaceutical Research Center, West Haven, CT, USA
dPfizer Global Research and Development, Groton, CT, USA
eBoehringer Ingelheim Pharmaceuticals, Inc., Ridgefield, CT 06877, USA
Received 16 July 2002; accepted 17 September 2002
Abstract—A new 5-HT1A silent antagonist 14 (5-HT1A IC50=2.2 nM) antagonizes the effects of agonists on reciprocal forepaw
treading behavior, on neuronal firing in the rat dorsal raphe, and on 5-HT1A release in the raphe and hippocampus. While 14 alone
was inactive in the social interaction paradigm, it completely reversed the social interaction activity of the serotonergic compounds
(buspirone, 1, and 2).
# 2002 Elsevier Science Ltd. All rights reserved.
Defining a 5-HT1A antagonist is made difficult by the
complex pharmacology and neuroanatomy of these
serotonergic pathways. Serotonergic 5-HT1A receptors
are expressed both as presynaptic autoreceptors on the
soma and dendrites of serotonergic neurons and as
postsynaptic receptors on a variety of other neurons.
Activation of either the presynaptic autoreceptors or the
postsynaptic receptors causes hyperpolarization and
inhibition of neuronal firing.1 At the postsynaptic
receptors, partial agonists tend to have weak agonist
effects and block the effects of full antagonists. At the
presynaptic autoreceptors, partial agonists tend to
exhibit a full agonist effect, due to the high receptor
reserve2 at these autoreceptors. The term ‘silent
antagonist’3 has been used to distinguish full antago-
nists, which have no intrinsic activity at either pre- or
postsynaptic 5-HT1A receptors, from partial agonists,
which have intrinsic agonist activity at the presynaptic
autoreceptors.
The atypical antipsychotic agent, BMS-1811004 (1,
BMY 14802), binds to 5-HT1A receptors, where it pro-
duces pharmacologic actions consistent with partial
agonist activity. Compound 1 inhibits the firing of
serotonergic dorsal raphe neurons5 (a presynaptic ago-
nist effect), and inhibits forskolin-stimulated adenylyl
cyclase activity in rat hippocampal membranes similar
to buspirone6 (a postsynaptic antagonist effect). In
addition, 1 induces reciprocal forepaw treading in
reserpinized rats, and produces dose-dependent hypo-
thermia in mice similar to agonists and partial agonists.7
Like buspirone, 1 produces anxiolytic-like effects in a
modified rat social interaction paradigm.8
In an investigation of cyclohexyl analogues9 of 1, two
compounds, 2 and 3, were distinguished from the rest
by their affinities for 5-HT1A receptors. These com-
pounds are of particular interest because they contain a
benzyl piperazine, rather than the well-known hetero-
aryl piperazine pharmacophore. These findings sparked
our interest and led to the investigation of the SAR of
these benzyl piperazine 5-HT1A ligands, and subse-
quently produced 14, a new 5-HT1A silent antagonist.
Compounds 2–29 (Tables 1 and 2) were prepared10 by
the method shown in Scheme 1. The monoketal of 1,4-
cyclohexanedione was reacted with an aryl Grignard
*Corresponding author. Tel.: +1-203-677-6355; fax: +1-203-677-
7702; e-mail: mattsonr@bms.com
0960-894X/03/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00879-X

286
R. J. Mattson et al. / Bioorg. Med. Chem. Lett. 13 (2003) 285–288
Table 1. 5-HT1A receptor binding of aryl cyclohexanol derivatives13
Table 2. 5-HT1A receptor binding of substituted benzyl piperazine
derivatives13
Compd
Ar
5-HT1A IC50 (nM)12
Compd
R
5-HT1A IC50 (nM)12
2
3
4
5
6
7
8
3,4-(–OCH2O–)-Ph–
4-OMe–Ph–
3-OMe–Ph–
2-OMe–Ph–
3,4-(OMe)2–Ph–
4-F–Ph–
20
50
50
60
140
260
460
9
2-Br
2-I
2-Cl
2-F
2-OMe
3-OMe
3-Br
1.2
1.2
1.9
10
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
43
Ph–
2.2
2.8
6
6.7
13
27
79
380
2.6
17
21
34
57
110
0.75
19
3-Cl
3-OEt
3-F
4-F
4-Cl
4-OMe
2,5-F2
2,3-F2
3,5-F2
3,4-F2
2,4-F2
2,6-F2
2-F, 5-OMe
2-F, 3-OMe
Scheme 1. Reagents and conditions: (a) ArMgBr/THF, 1 N HCl; (b)
CBZ-piperazine/Ti(OiPr)4/NaBH4; (c) H2, Pd/C; (d) ArCH2X/K2CO3/
MeCN.
reagent to give the 4-hydroxy cyclohexanone intermediate
after mild hydrolysis. This ketone intermediate was reduc-
tively coupled11 with CBz-piperazine to preferentially give
the cis alcohol intermediate, and the CBz protecting group
was then cleaved by hydrogenation. The substituted benzyl
groups were appended by simple alkylation.
with greater affinity than either the ortho-fluoro com-
pound, 12, or the meta-methoxy compound, 14. These
results establish that only the 2,5-substitution pattern
potentiates 5-HT1A binding.
The 5-HT1A binding was measured by published meth-
odology12 and the data expressed as IC50 values with the
radioligand at the Kd concentration. Compounds with
electron withdrawing substituents, such as 7 (Table 1),
bind poorly to the 5-HT1A receptor as does the unsub-
stituted phenyl analogue, 8. The 3,4-dimethoxyphenyl
analogue, 6 (135 nM), binds to the 5-HT1A receptor
with significantly less affinity than the monomethoxy-
compounds, 3–5 (50–60 nM), and the 3,4-methylene-
dioxy compound, 2 (20 nM), binds with higher affinity.
This suggests that oxygen lone electron pairs might be
ideally situated in the latter compound to take part in a
hydrogen bonding interaction with the 5-HT1A receptor
that is not available to the dimethoxy analogue.
The affinity of 2 and 14 for several other receptor sys-
tems was examined to determine the selectivity of these
compounds (Table 3). Compound 2 demonstrated weak
affinity for a1 and a2 receptors, while 14 bound weakly
to 5-HT1D receptors. Both compounds were inactive at
the other receptors examined.
Compound 2 produced partial agonist effects and 14
produced antagonist effects in blocking reciprocal fore-
paw treading behavior14,15 which is mediated by post-
synaptic 5-HT1A receptors in rats. The presynaptic 5-
HT1A antagonist activity of 14 was tested both electro-
physiologically by examining the firing of dorsal raphe
neurons and neurochemically by measuring the extra-
cellular concentrations of 5-HT using in vivo micro-
dialysis.
The effect of substituents on the benzyl group was also
examined (Table 2). With ortho-substituted compounds,
halogens increase 5-HT1A affinity (9–12), while methoxy
attenuates affinity (13). In the meta-substituted benzyl
compounds (14–18) 5-HT1A affinity is potentiated by
halogen and alkoxy substituents. None of the para-sub-
stituents in 19–21 potentiated 5-HT1A binding.
Table 3. Receptor binding specificity13 of 2 and 14
Receptor binding assaya
2
14
5-HT1B (5-HT, rat striatum)
>1000
>1000
13,000
11,000
—
>1000
700
410
>100,000
>100,000
700
28,000
>100,000
3000
The synergism between ortho- and meta-benzyl sub-
stituents was also investigated. The six possible difluoro-
benzyl compounds, 22–27, were prepared. Only the 2,5-
difluoro benzyl derivative, 22, possessed higher 5-HT1A
receptor affinity than either the ortho-fluoro compound,
12, or the meta-fluoro compound, 18. Similarly, when
the ortho-fluoro and meta-methoxy groups are com-
bined in a 2-fluoro, 3-methoxy-compound, 29, and a 2-
fluoro, 5-methoxy-compound, 28, only the latter binds
5-HT1C (mesulergine, bovine choroid plexus)
5-HT1D (5-HT, bovine caudate)
5-HT2 (spiperone, rat cortex)
5-HT3 (GR 65630, bovine area postrema)
5-HT uptake (5-HT, rat cortex)
a1 ( WB 4101, rat cortex)
1500
1100
4800
a2 (clonidine, rat cortex)
D2 (spiperone, rat striatum)
1700
aBinding assay ([3H] ligand, tissue; data reported as IC50 values, nM.

R. J. Mattson et al. / Bioorg. Med. Chem. Lett. 13 (2003) 285–288
Table 4. Rat social interaction
287
Electrophysiological recordings16 of the effects of 14 on
the firing of dorsal raphe neurons support the pre-
synaptic 5-HT1A antagonist activity of this compound
(Fig. 1). The inhibition of dorsal raphe activity induced
by 8-OH-DPAT was studied in combination with 14.
The dose-response curve for 8-OH-DPAT is shifted
markedly to the right in the presence of 14 (ED50
15.8 mg/kg versus 1.5 mg/kg, iv) in the manner of an
antagonist.
Compd
Dose
(mg/kg, ip)
Social interaction
time (sÆSEM)
Motor activity
(countsÆSEM)
Saline control
14
Diazepam
Buspirone
1
2
—
1
56Æ7
5800Æ700
5140Æ5
65Æ8
1.7
0.1
0.1
0.1
1.7
1
120Æ13a
125Æ12a
123Æ6a
104Æ10a
113Æ7a
5100Æ500
6700Æ700
6800Æ500
4700Æ400
4900Æ700
Diazepam
+14
In addition to the electrophysiology experiments, 14
was also tested for presynaptic 5-HT1A antagonist
activity using in vivo microdialysis17 (Fig. 2). Partial
agonists, such as 1 and buspirone, effectively antagonize
the post-synaptically mediated forepaw treading beha-
vior induced by agonists, but act as presynaptic agonists
in decreasing 5-HT release in rat hippocampus.18 Rats
were implanted with microdialysis probes in the hippo-
campus and extracellular concentrations of 5-HT were
measured by HPLC-ECD. While compound 14 alone
(1 mg/kg, ip) did not change hippocampal extracellular
5-HT concentrations, 14 (1.0 mg/kg, ip) significantly (p
<0.01) blocked the decrease in hippocampal 5-HT
induced by buspirone (2.5 mg/kg, ip). Thus compound
14 at 1 mg/kg failed to give an agonist effect and effec-
tively antagonized the buspirone-induced agonist effect.
These data demonstrate that 14 is an effective antago-
nist at 5-HT1A presynaptic autoreceptors.
Buspirone
+14
1
+14
2
+14
0.1
1
0.1
1
0.1
1
51Æ9
49Æ9
53Æ10
4600Æ500
5400Æ700
5100Æ500
aTest significantly different from saline control (p>0.05).
digm,19,20 and 5-HT1A agonists are active in this model
after discrete dorsal raphe injections.21 In addition, such
agonists, given systemically, do not increase social
interaction in animals in which the presynaptic 5-HT1A
neurons in the dorsal raphe were lesioned with 5,7-
dihydroxy tryptamine.22 All of this data suggest the
anxiolytic effect is mediated by a blockade of pre-
synaptically controlled 5-HT release.
Diazepam (1.7 mg/kg), buspirone (0.1 mg/kg), 1 (0.1 mg/
kg), and 2 (0.1 mg/kg), all significantly increased the
social interaction times (Table 4). However, the
antagonist, 14, by itself (1.0 mg/kg, ip), had no effect in
the rat social-interaction paradigm. Of these, only 1
significantly increased total motor activity, while the
others had no effect. To further define the antagonist
activity of the molecule, 14 was studied in combination
with diazepam, buspirone, 1, and 2. Compound 14 was
without effect on social interaction activity of diazepam,
but completely reversed the social interaction activity of
the serotonergic compounds, buspirone, 1, and 2, with-
out significant effects on total motor activity. These
results further confirm the full antagonist effects of 14
on presynaptic 5-HT1A receptors.
All known classes of clinically effective anxiolytics
demonstrate activity in the social interaction para-
These data demonstrate that 14 is a 5-HT1A pre- and
postsynaptic antagonist that blocks the activity of 5-
HT1A agonists in the rat social-interaction paradigm.
The data also suggest that buspirone, 1, and 2 show
activity in this paradigm through 5-HT1A partial agonist
effects, and support the presynaptic dorsal raphe 5-HT1A
receptors as a key site through which 5-HT1A anxiolytics
express their activity in the rat social-interaction task.
Figure 1. Dose–response curve for 8-OH-DPAT-induced inhibition of
dorsal raphe firing alone and in the presence of 14.
References and Notes
1. Lanfumey, L.; Hamon, M. Nuc. Med. Biol. 2000, 27, 429.
2. Meller, E.; Goldstein, M.; Bohmaker, K. Mol. Pharmacol.
1990, 37, 231.
3. (a) Routledge, C. Behav. Brain Res. 1995, 73, 153. (b)
Fletcher, A.; Forster, E. A.; Bill, D. J.; Brown, G.; Cliffe, I. A.;
Hartley, J. E.; Jones, D. E.; McLenachan, A.; Stanhope, K. J.;
Critchley, D. J.; Childs, K. J.; Middlefell, V. C.; Lanfumey, L.;
Corradetti, R.; Laporte, A. M.; Gozlan, H.; Hamon, M.;
Dourish, C. T. Behav. Brain Res. 1996, 73, 337.
Figure 2. Effects of buspirone and 14, alone and in combination, on in
vivo hippocampal 5-HT levels as measure by microdialysis.

288
R. J. Mattson et al. / Bioorg. Med. Chem. Lett. 13 (2003) 285–288
4. Yevich, J. P.; New, J. S.; Lobeck, W. G.; Dextraze, P.;
Bernstein, E.; Taylor, D. P.; Yocca, F. D.; Eison, M. S.;
Temple, D. L. J. Med. Chem. 1992, 35, 4516.
5. VanderMaelen, C. P.; Braselton, J. P. Eur. J. Pharmacol.
1990, 179, 357.
6. (a) Eison, A. S.; Eison, M. S. Prog. NeuroPsycho-pharmcol.
Biol. Psychiat. 1994, 18, 47. (b) Taylor, D. P. Ann. N. Y. Acad.
Sci. 1990, 600, 545. (c) Gardner, C. R. Gen. Pharm. 1988, 19,
347. (d) Traber, J.; Glaser, T. Trends Pharmacol. Sci. 1987, 8,
432.
7. Brislow, L. J.; Baucutt, L.; Thorn, L.; Hutson, P. H.;
Noble, A.; Beer, M.; Middlemiss, D. N.; Tricklebank, M. D.
Eur. J. Pharmacol. 1991, 204, 21.
baseline firing rate during the 1-min period preceding initia-
tion of 8-OH-DPAT injections. Only one cell per rat was
recorded, but more than one dose of 8-OH-DPAT could be
given to a single rat, in which case ‘dose’ was cumulative dose.
For antagonism experiments, 14 was injected intravenously
(0.32 mg/kg) at least 10 min prior to beginning dosing with 8-
OH-DPAT. Dose–response curves were calculated by an
iterative computer algorithm in which the logistic equation
was fitted to the data using a least-squares criterion, and con-
strained to pass through 0 and 100%. The ED50 was defined as
the estimated dose required to produce a 50% reduction in the
rate of firing of a serotonergic dorsal raphe neuron.
17. In vivo microdialysis. Method. Male Sprague–Dawley rats
were anesthetized (im) with a mixture of ketamine HCl (80 mg/
kg) and xylazine (6 mg/kg) and stereotaxically implanted with
I-shaped fused silica microdialysis probes in the ventral hip-
pocampus. Microdialysis experiments were carried out 24 h
after surgery when the hippocampus was perfused with artifi-
cial cerebrospinal fluid (aCSF, NaCl 140.3 mM, KCl 2.7 mM,
CaCl2 2.2 mM, pH 6.5) at 1.0 mL/min using a microinfusion
pump. The dialysate was collected directly into a 25 mL loop of
the injection valve and automatically injected (Digital Valve
Sequence Programmer, Valco Instruments Inc., Houston, TX,
USA) every 25 min onto the HPLC column. After stabilization
of baseline dialysate levels, 14 (1–20 mg/kg, ip) was injected
and the dialysates continuously collected. In the antagonism
experiment, 14 (1.0 mg/kg, ip) was administered 50 min before
buspirone (2.5 mg/kg, ip). Chromatographic conditions. Dialy-
sate concentrations of 5-HT was measured by HPLC (ODS-
Hypersil column; 3 m, 150 Â 3.0 mm; Keystone Scientific, Bel-
lefonte, PA, USA), maintained at 30 ꢀC with a column heater
and detected coulometrically (ESA Model 5200 Coulochem II,
Bedford, MA, USA) using a high sensitivity analytical cell
(ESA Model 5011). The first electrode was set at +0.02 Vand
the second at +0.35 V; a guard cell placed before the injector
was set at +0.45 V. The mobile phase consisted of 59 mM
potassium phosphate (pH 4.8), 2.3 mM 1-heptanesulfonic acid
sodium salt, 1 mL/L triethylamine, 17% methanol and 2%
acetonitrile and was delivered at a flow rate of 0.5 mL/min by
a Waters 610 pump. Retention time, peak area and peak
height were measured with a data management software sys-
tem (Baseline 810, Waters).
8. Smith, H. L.; Matos, F.; Yocca, F. D.; Carter, R. B. Soc.
Neurosci. Abs. 1992, 18, 1382.
9. Catt, J. D.; Mattson, R. J.; Sloan, C. P.; Taylor, D. P.;
Moon, S. L. Abstracts of Papers, 24th National Med. Chem.
Symposium, Salt Lake City, 1994; Abstract 16.
10. (a) Mattson, R. J.; Catt, J. D. US Patent 5,387,593, 1995;
Chem Abstr. 123, 83213. (b) Compounds were characterized
by IR, NMR, MS, and CHN.
11. Mattson, R. J.; Pham, K. M.; Leuck, D. J.; Cowen, K. A.
J. Org. Chem. 1990, 55, 2552.
12. Mahle, C. D.; Wiener, H. L.; Yocca, F. D.; Maayani, S. J.
Pharmacol. Exp. Ther. 1992, 263, 1275.
13. IC50 values are the mean of 2 determinations run at five
different concentrations with the radioligand at the Kd con-
centration. Each experiment was carried out in triplicate.
Standard errors were typically Æ20% of the mean value.
14. Tricklebank, M. D.; Forler, C.; Middlemiss, D. N.;
Fozard, J. R. Eur. J. Pharmacol. 1985, 117, 15.
15. Tricklebank, M. D.; Forler, C.; Fozard, J. R. Eur. J.
Pharmacol. 1985, 106, 271.
16. In vivo electrophysiology. Adult male albino Sprague–
Dawley rats were anesthetized with chloral hydrate and placed
in a stereotaxic device (David Kopf, Model 1700). Extra-
cellular single-unit recordings were obtained from sponta-
neously active serotonergic neurons in the dorsal raphe
nucleus using glass micropipettes containing 1.0 M NaCl and
0.5% Pontamine Sky Blue dye. A dye spot was deposited at
the tip of the electrode by iontophoresis at the conclusion of
each recording session, and data was used only for recording
locations confirmed histologically to be in the dorsal raphe
nucleus. Changes in neuronal firing rates caused by intra-
venous injections of 8-OH-DPAT were determined by first
recording stable baseline activity for 3–15 min from a single
cell, and then injecting 8-OH-DPAT through a tail vein via a
syringe pump. Percent change in firing rate was calculated by
comparing the discharge rate taken over a 1-min period at
peak drug effect following injection of 8-OH-DPAT to the
18. Sharp, T.; Backus, L. I.; Hjorth, S.; Bramwell, S. R.;
Grahame-Smith, D. G. Eur. J. Pharmacol. 1990, 176, 331.
19. Gardnier, C. R.; Guy, A. P. Drug Dev. Res. 1984, 4, 207.
20. Guy, A. P.; Gardner, C. R. Neuropsychobiology 1985, 13, 194.
21. Higgins, G. A.; Jones, B. J.; Oakley, N. R. Psycho-
pharmacology 1992, 106, 261.
22. Picazo, O.; Lopez-Rubalcava, C.; Fernandez-Guasti, A.
Brain Res. Bull. 1995, 37, 169.