A new class of selective, non-basic 5-HT2A receptor antagonists

Article abstract of DOI:10.1016/j.bmcl.2006.03.050

Based on an existing series of 5-HT_2A receptor ligands containing a basic nitrogen, we designed a non-basic lead that had reduced affinity for both the 5-HT_2A receptor and the IKr potassium channel. The present paper describes the development of this lead to a novel series of non-basic piperidine sulfonamides and amides that have high affinity for the 5-HT_2A receptor, whilst maintaining excellent selectivity over off target activities such as the IKr channel. This work has shown that the proposed pharmacaphore model for the 5-HT_2A receptor which suggests that a basic nitrogen is required for the binding of ligands is questionable.

Full text of DOI:10.1016/j.bmcl.2006.03.050

Bioorganic & Medicinal Chemistry Letters 16 (2006) 3201–3204
A new class of selective, non-basic 5-HT2A receptor antagonists
Tammy Ladduwahetty,* Amanda L. Boase, Andrew Mitchinson, Caroline Quin,
Smita Patel, Kerry Chapman and Angus M. MacLeod
The Neuroscience Centre, Merck Sharp & Dohme, Eastwick Road, Harlow, Essex CM20 2QR, UK
Received 1 February 2006; revised 13 March 2006; accepted 16 March 2006
Abstract—Based on an existing series of 5-HT2A receptor ligands containing a basic nitrogen, we designed a non-basic lead that had
reduced affinity for both the 5-HT2A receptor and the IKr potassium channel. The present paper describes the development of this
lead to a novel series of non-basic piperidine sulfonamides and amides that have high affinity for the 5-HT2A receptor, whilst main-
taining excellent selectivity over off target activities such as the IKr channel. This work has shown that the proposed pharmacaphore
model for the 5-HT2A receptor which suggests that a basic nitrogen is required for the binding of ligands is questionable.
Ó 2006 Elsevier Ltd. All rights reserved.
Serotonin (5-HT) is a major neurotransmitter that has
been linked to a number of physiological processes,
including appetite, emotion, changes in mood and the
regulation of the sleep/wake cycle.1 Hence, it is not too
surprising that selective antagonism of the numerous
5-HT receptors has been sought after, and in a number
of cases has led to useful drugs. Antagonism of the 5-
2
HT2 receptor, in particular, has been implicated in
the beneficial effects of some antidepressants as well as
antipsychotics.3 5-HT2 receptors were first identified in
1979 and since the early 1980s, numerous, structurally
diverse antagonists of this receptor have been published
(Fig. 1). These include MDL-100907 which has good
selectivity for the 5-HT2A subtype. The structural diver-
sity of 5-HT2A ligands presented a challenge in terms of
defining a pharmacophore model. However, pharmaco-
phore models that have been proposed agree that two
aryl rings and a basic nitrogen are required for binding
at the receptor.4 Using this model, Rowley et al. have
suggested that 5-HT2 antagonists can be divided into
two classes: (a) those where the aryl rings and the basic
nitrogen are in a triangular arrangement (Eplivanserin
and 1) and (b) those where the aryl rings and the
basic nitrogen are in a linear disposition (MDL-100907
and 2).5
Figure 1. Selected 5-HT2A antagonists.
Site-directed mutagenesis studies have shown that the
basic nitrogen in these ligands binds to a charged
Asp155 in Helix 3 of the 5-HT2A receptor and is impor-
Keywords: 5-HT2A receptor; Antagonist; Non-basic.
*
Corresponding author. Tel.: +44 208 447 9820; e-mail:
Tammy.Ladduwahetty@glpg.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.03.050

3202
T. Ladduwahetty et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3201–3204
tant for the binding of both agonists and antagonists.6
Hence, 5-HT, Gramine (both agonists), Ketanserin
and Spiperone (both antagonists) have no affinity for
D155A, D155N and D155Q mutant receptors. More re-
cent work has suggested that the binding interaction of
the basic nitrogen may be more complex than originally
thought, since there may be an additional interaction of
the protonated amine with a Ser159, also on Helix 3.7
Mutating this serine again affected the binding of some
agonists such as 5-HT, whilst others such as lysergic acid
were not affected.
Figure 3.
affinity for 5-HT2A receptors (Ki 5-HT2A >5000 nM).
The amide linkage in 4 restricts rotation of the pheneth-
yl side chain and it is possible that the phenyl group is
prevented from reaching a suitable position for binding.
Although the basic nitrogen in 5-HT2 receptor antago-
nists has been identified as being important for the bind-
ing of agonists and antagonists of the receptor, it has
also been implicated in a number of undesirable side ef-
fects associated with compounds of this class. These in-
clude affinity for dopamine receptors and the IKr
potassium channel, a voltage gated ion channel involved
in control of the heart rhythm. Regulatory agencies are
increasingly concerned about cardiovascular events fol-
lowing the chronic dosing of drugs, and require strin-
gent tests to confirm their safety. The atypical
antipsychotic sertindole, a compound with dual 5-HT2/
D2 affinity, was withdrawn from the market due to the
prolongation of the corrected QT interval, a conse-
quence of its affinity for the IKr channel.8 Designing
5-HT2A ligands with little or no affinity for the IKr
channel would therefore be beneficial. However, potent
IKr channel blockers such as Dofetilide and Terfenadine
(Fig. 2) clearly show that the pharmacophore for bind-
ing at IKr is similar to the one for binding at the 5-
HT2A receptor. Recent work that involved building a
predictive pharmacaphore model for the IKr potassium
channel concluded that the following features were
important for binding: ‘one positive ionisable feature,
two aromatic rings and one hydrophobic group.’9
From unpublished work in our own laboratories, and
known 5-HT2A ligands such as EMD-281014, we knew
that piperazine amides and sulfonamides,11 such as 3,
provide high affinity antagonists analogous to the
piperidine sulfones represented by 2. Consequently,
we chose to synthesise compound 5, again non-basic,
but without the conformational restriction of the amide
4 and where the hydroxyl group can serve either as a
mimic of the protonated amine or in its own right as
a hydrogen bond donor/acceptor that might interact
with Ser159. Gratifyingly, we found that 5 had greater
affinity for the 5-HT2A receptor (Ki 5-HT2A 190 nM)
than the amide 4. The present paper describes the
development of this lead into a novel series of non-
basic piperidine sulfonamides and amides to give high
affinity, selective ligands, thus disproving the generally
accepted view that a basic nitrogen is crucial for bind-
ing at the 5-HT2A receptor.
The 4-hydroxypiperidine 5 was synthesised by addition
of the requisite Grignard reagent to N-BOC piperidone
to give 6, followed by deprotection, and sulfonylation of
the resulting trifluoroacetate salt (Scheme 1). Treating 5
We have previously shown that in the sulfonylpiperidine
series of 5-HT2A antagonists (2), reducing the basicity of
the piperidine nitrogen together with appropriate aro-
matic substitution can lead to significantly reduced bind-
ing to the IKr channel.10 We investigated the interaction
of the basic nitrogen further by synthesising the amide 4
(Fig. 3) which, although lacking a basic nitrogen that
can interact with the charged Asp155, does have an
amide which could perhaps interact as a hydrogen bond
acceptor with Ser159. The amide 4, however, had no
Scheme 1. Reagents: (i) 2,4-difluorophenethyl bromide, Mg, THF; (ii)
TFA, Et3N, 4-F-phenylsulfonyl chloride, CH2Cl2; (iii) DAST, CH2Cl2.
Figure 2. IKr channel blockers.

T. Ladduwahetty et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3201–3204
3203
Table 1. Binding affinities
with DAST gave a mixture of the fluoro compound 7
and the tetrahydropyridine 8 which were separated by
HPLC.
Compound
h5HT2A h5HT2C Selectivityb hD2
Ki/nMa Ki/nMa Ki/nMa Ki/nMa
hIKr
2
EMD281014
0.42
0.87
4.4
92
557
3785
>4000
806
449
2033
997
239
404
70
219
641
860
—
33
22
39
73
24
66
9
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
710
1268
The unsubstituted piperidine analogues were synthesised
as shown in Scheme 2.
3
4869
5
190
>9000
>9000
>9000
>9000
>9000
>9000
>9000
>9000
>9000
>9000
>9000
>9000
7
24
22
52
14
The substituted vinyl pyridine 9 was synthesised from
commercially available vinyl pyridine via a Heck reac-
tion.12 Subsequent hydrogenation gave 10, which was
converted to the TFA salt of piperidine 11 using stan-
dard methodology. A small library of amides and sul-
fonamides was then synthesized, a selection of which is
shown.
8
12a
12b
12c
12d
12e
12f
12g
13a
13b
9.9
6.1
8.1
1.7
2.1
3.9
2.9
96
58
84
28
10
184
110
28
The SAR for this series of compounds with regard to
binding at 5-HT2A and other GPCRs (5-HT2C, hD2
and hIKr) is summarised in Table 1. The non-basic
piperidine 5 lost two orders of magnitude in binding
compared to the basic compounds 3 and EMD-
281014. Removing the hydrogen bonding hydroxyl
group gave an increase in affinity and to our surprise
both the fluoro analogue 7 and the tetrahydropyridine
8 were equipotent at the 5-HT2A receptor. This suggest-
ed to us that the central ring only functions as a spacer
group and that it might be replaced with an unsubstitut-
ed piperidine. In fact, 12a had a similar affinity to 7 and
8. The 3-fluoro analogue 12b was shown to be antago-
nist in the functional assay.13 A modest increase in affin-
a h5-HT2A, h5HT2C, hD2 (dopamine) and hIKr affinities were deter-
mined as described in Ref. 3 (n = 2).
b h5-HT2C/h5-HT2A
.
ity was obtained with small substituents at the 2-
position of the aryl sulfonamide (see 12c and 12d), The
2-cyano analogue 12e has reduced selectivity for 5-
HT2A over 5-HT2C. Both the naphthalene 12f and the
quinoline 12g are comparable in 5-HT2A affinity to the
basic compound 3. In contrast to the basic compounds,
however, the non-basic compounds such as 12g have im-
proved selectivity over IKr (% inhibition 36% at 10 lM).
The quinoline substituent gave the highest selectivity
over 5-HT2c. Replacing the sulfonamide linkage with
an amide as in 13a retains affinity for 5-HT2A receptors.
However, the quinoline substituent, which gave the best
selectivity in the case of the sulfonamide linkage, gave
reduced selectivity over 5-HT2C with an amide linkage.
All the compounds in this series had excellent selectivity
over both the hD2 receptor and the hIKr potassium
channel.
In conclusion, this work has shown that the proposed
pharmacophore model for the 5-HT2A receptor which
suggests that a basic nitrogen is required for the binding
of ligands is questionable. At least in the piperidine sul-
fone series exemplified by 2, it is possible to replace the
basic piperidine with a non-basic core and still retain
high affinity for the 5-HT2A receptor. It could be that
in this series of molecules the sulfone or amide moieties
present an additional key point of interaction that
makes the binding of the basic amine less important.
This work also shows that our initial assumption that
the amide 4 was not the best compound to test whether
a basic nitrogen was necessary for binding at the 5-HT2A
receptor was correct.
In fact, replacing the basic piperidine core with a
non-basic one has enabled us to retain 5-HT2A affinity
whilst significantly reducing binding to the IKr
channel. The fact that the basic nitrogen is not needed
for 5-HT2A binding in this series of compounds has
opened up several possibilities that were hitherto
unavailable. This work will be reported in future
communications.
Scheme 2. Reagents and conditions: (i) 2,4-difluoro-iodobenzene,
Pd(OAc)2, P(o-tolyl)3, Et3N, toluene, 150 °C, 5 h; (ii) Pd/C, EtOAc,
H2 (45 psi); (iii) BnBr, acetone, reflux, 16 h; (iv) NaBH4, MeOH,
reflux, 16 h; (v) H2, HCl (conc.), EtOH, Pd(OH)2, then NaOH,
(BOC)2O, dioxane; (vi) TFA; (vii) Et3N, ArSO2Cl, CH2Cl2 or Et3N,
ArCOCl, CH2Cl2.

3204
T. Ladduwahetty et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3201–3204
6. Kristiansen, K.; Kroeze, W. K.; Willins, D. L.; Gelber, E.
Acknowledgments
I.; Savage, J. E.; Glennon, R. A.; Roth, B. L.
J. Pharmacol. Exp. Ther. 2000, 293, 735.
7. Wang, C.-D.; Gallaher, T. K.; Shih, J. C. Mol. Pharmacol.
1993, 43, 931.
8. Fritze, J.; Bandelow, B. Int. J. Psychiatry Clin. Pract.
1998, 2, 265; (b) Rampe, D.; Murawsky, M. K.; Grau, J.;
Lewis, E. W. J. Pharmacol. Exp. Ther. 1998, 286, 788.
9. Du, L.-P.; Tsai, K.-C.; Li, M.-Y.; You, Q.-D.; Xia, L.
Bioorg. Med. Chem. Lett. 2004, 14, 4771.
The authors thank J. Kingston for the separation of
7 and 8, K. Wilson for the synthesis of 4 and the
CIBE Lead Evaluation group for testing our
compounds.
References and notes
10. Fish, L. R.; Gilligan, M. T.; Humphries, A. C.; Ivarsson,
M.; Ladduwahetty, T.; Merchant, K.; O’Connor, D.;
Patel, S.; Philipps, E.; Vargas, H. M.; Hutson, P. H.;
MacLeod, A. M. Bioorg. Med. Chem. Lett. 2005, 15, 3665.
11. S.R. Fletcher, unpublished results. Burkamp, F.; Cheng,
S. K.-F.; Fletcher, S. R. WO2001074797.
1. Schloss, P.; Williams, D. C. J. Psychopharmacol. 1998,
12(2), 115.
2. Denotes 5-HT2A and 5-HT2C
.
3. Fletcher, S. R.; Burkamp, F.; Blurton, P.; Cheng, S. K.
F.; Clarkson, R.; O’Connor, D.; Spinks, D.; Tudge,
M.; van Niel, M. B.; Patel, S.; Chapman, K.;
Marwood, R.; Shepheard, S.; Bentley, G.; Cook, G.
P.; Bristow, L. J.; Castro, J. L.; Hutson, P. H.;
MacLeod, A. M. J. Med. Chem. 2002, 45, 492, and
references cited therein.
4. Westkaemper, R. B.; Glennon, R. A. Curr. Top. Med.
Chem. 2002, 2, 575.
5. Rowley, M.; Bristow, L. J.; Hutson, P. H. J. Med. Chem.
2001, 44, 477.
12. Frantz, R.; Granier, M.; Durand, J.-O.; Lanneau, G. F.
Terahedron. Lett. 2002, 43, 9115.
13. Determined using an Aurora b-lactamase reporter gene
assay with CHO cells stably transfected with the h5-HT2A
receptor. Methodology as described in Thomson, C. G.;
Beer, M. S.; Curtis, N. R.; Diggle, H. J.; Handford, E.;
Kulagowski, J. J. Bioorg. Med. Chem. Lett. 2004, 677
(antagonist affinity: IC50 319 nM, agonist affinity: no effect
up to 30 lM, n = 2).