Discovery of the first nonpeptidic, Small-molecule, highly selective somatostatin receptor subtype 5 antagonists: A chemogenomics approach

Article abstract of DOI:10.1021/jm701143p

We disclose the first selective, nonpeptidic, small-molecule somatostatin receptor subtype 5 (SST5R) antagonists that were identified by a chemogenomics approach based on the analysis of the homology of amino acids defining the putative consensus drug binding site of SST5R. With this strategy, opioid, histamine, dopamine, and serotonine receptors were identified as the closest neighbors of SST5R. The H1 antagonist astemizole was chosen as a seed structure and subsequently transformed into a SST5 receptor antagonist with nanomolar binding affinity devoid of the original target activity.

Full text of DOI:10.1021/jm701143p

Copyright 2007 by the American Chemical Society
Volume 50, Number 25
December 13, 2007
Letters
demonstrated that SST also shows antiproliferative and neu-
rotransmitter activities and reduces splanchnic perfusion.3 SST
acts via five distinct G-protein-coupled receptors (GPCRs),
SST1–5, which all have been cloned and characterized.2 Detailed
structure–activity relationship (SAR) studies have revealed that
the tetrapeptide motif Phe7-Trp8-Lys9-Thr10 is essential for
biological activity and adopts a ꢀ-II-turn conformation according
to NMR and crystallographic investigations. Several research
groups have capitalized on this structural motif, developing
ꢀ-strand mimetics.4 Because of the numerous physiological
functions of SST, selective ligands are of high interest and might
play an important role in the treatment of various human
diseases. For instance, SST acting via SST5 receptors has been
found to activate and up-regulate NMDA receptor function5 or
to control hormonal secretions (e.g., insulin, growth hormone).6
SST has a very short half-life in the circulation, and therefore,
numerous research groups have focused on the design of
nonpeptidic SST ligands exhibiting improved metabolic stability.
Although a number of peptidic and nonpeptidic hSST5 recep-
tor–ligands have been published,6–13 no small-molecule, selec-
tive hSST5 receptor antagonists have been described so far. We
report here the discovery of compounds of the general structure
1, which are receptor subtype 5 selective somatostatin antago-
nists.
Discovery of the First Nonpeptidic,
Small-Molecule, Highly Selective
Somatostatin Receptor Subtype 5
Antagonists: A Chemogenomics Approach
Rainer E. Martin,*,† Luke G. Green,† Wolfgang Guba,‡
Nicole Kratochwil,‡ and Andreas Christ§
DiscoVery Chemistry, Lead Generation, F. Hoffmann-La Roche Ltd.,
4070 Basel, Switzerland, DiscoVery Chemistry, Cheminformatics
and Molecular Modeling, F. Hoffmann-La Roche Ltd., 4070 Basel,
Switzerland, and DiscoVery, Metabolic and Vascular Diseases, F.
Hoffmann-La Roche Ltd., 4070 Basel, Switzerland
ReceiVed September 13, 2007
Abstract: We disclose the first selective, nonpeptidic, small-molecule
somatostatin receptor subtype 5 (SST5R) antagonists that were identified
by a chemogenomics approach based on the analysis of the homology
of amino acids defining the putative consensus drug binding site of
SST5R. With this strategy, opioid, histamine, dopamine, and serotonine
receptors were identified as the closest neighbors of SST5R. The H1
antagonist astemizole was chosen as a seed structure and subsequently
transformed into a SST5 receptor antagonist with nanomolar binding
affinity devoid of the original target activity.
Somatostatin (SSTa) or somatotropin release-inhibiting factor
(SRIF) is a cyclic tetradecapeptide hormone originally isolated
from sheep hypothalamus and sequenced by Brazeau et al. in
1973.1 It is widely distributed throughout the body and exhibits
multiple biological functions that are most commonly inhibitory
in nature, such as the release of growth hormone (GH),
pancreatic insulin, glucagon, and gastrin.2 In addition, it was
To quickly identify selective and low molecular weight
hSST5R ligands suitable for lead optimization without having
to commit substantial resources for high-throughput screening
(HTS), we opted for a focused screening approach. Because of
a lack of small-molecule hSST5 receptor antagonists, classical
approaches such as 2D/3D similarity searches or pharmacophore
modeling could not be utilized.14–16 Therefore, we implemented
a chemogenomics strategy based on an analysis of the homology
of amino acids delineating the putative consensus drug binding
site in the transmembrane region of GPCRs. Class A GPCRs
with known small-molecule ligands were ranked with respect
to the similarity of the binding pocket of the hSST5 receptor.17
Biogenic amine receptors such as opioid, histamine, dopamine,
* To whom correspondence should be addressed. Phone: (+41) 61 68
87350. Fax: (+41) 61 68 86965. E-mail: rainer_e.martin@roche.com.
†
Discovery Chemistry, Lead Generation.
Discovery Chemistry, Cheminformatics and Molecular Modeling.
Discovery, Metabolic & Vascular Diseases.
‡
§
aAbbreviations: SST, somatostatin; SRIF, somatotropin release-inhibiting
factor; GH, growth hormone; GPCRs, G-protein-coupled receptors; SAR,
structure–activity relationship; NMDA, N-methyl-D-aspartate; SST5R, so-
matostatin subtype 5 receptor; R, receptor; h, human; HTS, high-throughput
screening; H1, histamine subtype 1 receptor; D2, dopaminergic subtype 2
receptor; DMF, N,N-dimethylformamide; MWH, microwave-assisted heat-
ing; DCE, dichloroethane; DIEA, diisopropylethylamine; TEA, triethylamine.
10.1021/jm701143p CCC: $37.00
2007 American Chemical Society
Published on Web 11/19/2007

6292 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 25
Letters
Scheme 1a
Figure 1. Analysis of amino acid homology of the putative consensus
drug binding site in the transmembrane region of biogenic amine
receptors closely related to hSST5R. The amino acids are colored
according to physicochemical properties. The highly conserved aspartic
acid (D) is highlighted in magenta. The selected 35 helix and EC2
loop positions are taken from Table 1 in ref 17.
a Reagents and conditions: (a) NaH, p-fluorobenzyl bromide, DMF, 60
°C, 18 h, 81%; (b) 4-aminopiperidine-1-carboxylic acid ethyl ester, MW,
180 °C, 1 h, 90%; (c) HBr (48%) in water, reflux, 18 h, quantitative. (d)
9a–d: corresponding alcohol, Dess-Martin periodinane, DCE, 40 °C, 2 h,
then addition of 8, HOAc, DIEA, NaCNBH3, DCE, 40 °C, 18 h. (e) 9e–k:
corresponding aldehyde, HOAc, DIEA, NaCNBH3, EtOH, 40 °C, 18 h.
Table 1. hSST5 Receptor Radioligand Binding Data of Compounds 2
and 9a–k
Figure 2. Identification of hSST5 receptor antagonist seed structures
by screening of the Sigma LOPAC and Cerep BioPrint databases.
and serotonin receptors were identified as closest neighbors of
SST5R (Figure 1).
Recently, a related method to classify and characterize GPCRs
according to their ligand recognition properties has been
published by Högberg et al.18
The Sigma LOPAC19 and the Cerep BioPrint20 compound
databases, which contain pharmacologically active reference
compounds and marketed/withdrawn drugs targeting biogenic
amine receptors, were selected for a focused screen. As predicted
by the above sequence comparison of the consensus drug
binding site, three different, albeit structurally closely related,
biogenic amine receptor–ligands with a micromolar Ki against
hSST5R were identified: astemizole (2),21 a second-generation
hH1R antihistamine; fluspirilene (3),22 a hD2R antagonist used
for the treatment of psychotic disorders; and loperamide (4),23
a peripherally acting opioid receptor agonist used for the
management of chronic diarrhea. All had affinities in the low
micromolar range of 4.07, 4.51, and 4.52 µM against hSST5R,
respectively (Figure 2). On the basis of hSST5R potency and
structural simplicity, we selected astemizole (2) as our seed
structure to initiate lead generation activities. Primary SAR
studies on 2 focused on the modification of the p-methoxyphen-
ethyl group as outlined in Scheme 1. The synthetic strategy was
mainly adopted from the published synthesis of astemizole (2),
although several steps were slightly modified, providing in-
creased yields of intermediates.24 Deprotonation of benzimida-
zole 5 with NaH in DMF and subsequent reaction with
p-fluorobenzyl bromide provided alkylated intermediate 6. Neat
reaction of benzimidazole 6 with 4-aminopiperidine-1-carboxylic
acid ethyl ester under microwave-assisted heating to 180 °C
provided piperidine 7 in excellent yield. Removal of the ethyl
carbamate protection group by heating to reflux with hydro-
bromic acid in water furnished dihydrobromide salt 8, which
was used directly in the consecutive reaction step without further
purification. Finally, preparation of target structures 9a–d was
achieved by in situ oxidation of the corresponding alcohols with
Dess-Martin periodinane in dichloroethane followed by reduc-
tive amination with piperidine building block 8 and sodium
cyanoborohydride. Similarily, compounds 9e–k were synthe-
sized by reductive amination from piperidine 8 and the
corresponding aldehydes.
Interestingly, replacement of the p-methoxy group in astemi-
zole (2) with relatively small substituents such as fluorine (9a)
or trifluoromethoxy (9b) leads to hSST5R inactive compounds,
whereas attachment of a simple unsubstituted phenethyl group
(9c) mainly restored activity in the low micromolar range (Table
1).
Reducing chain flexibility by fusion of the ethylene bridge
into a 1,2,3,4-tetrahydronaphthalene system again resulted in a
clear loss of activity (9d). Further elongation of the carbon chain
yielding a phenpropyl side chain (9e) as well as attachment of

Letters
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 25 6293
Scheme 2a
Table 2. hSST5 Receptor Radioligand Binding Data of Compounds
16a–e and 17a–n
a Reagents and conditions. (a) 12: 4-aminopiperidine-1-carboxylic acid
ethyl ester, TEA, MW, 180 °C, 5 min, 75%. 13: 4-aminopiperidine-1-
carboxylic acid ethyl ester, DMF, room temp, 18 h, 45%. (b) 14 and 15:
HBr (48%) in water, reflux, 2 h, quantitative. (c) Corresponding benzal-
dehyde, HOAc, DIEA, NaCNBH3, EtOH, 40 °C, 18 h.
short linear or branched saturated alkyl side chains (not shown)
did not provide active compounds. Quickly, it became apparent
that the phenethyl moiety is not mandatory for binding and
substitution by a simple benzyl side chain (9f) is tolerated well
without major loss of activity when compared with seed
structure 2.
Bringing in para substituents at the benzyl group usually
showed some improvement in binding as evidenced by 9g. A
further gain in affinity was achived by a dihydrobenzofurane
moiety (9h), but also indoles connected via positions 5 (9i) and
3 (9j) provided low micromolar compounds. Finally, the 1,4-
dimethoxy-2-naphthyl derivative 9k was the compound with
the highest activity identified after this first round of optimiza-
tion. With this improved hit structure 9k in hand, we then turned
our attention to the optimization of the benzimidazole part,
keeping the naphthyl side chain fixed (Scheme 2).
Nucleophilic aromatic substitution of benzothiazole 10 and
benzoxazole 11 with 4-aminopiperidine-1-carboxylic acid ethyl
ester under microwave-assisted heating (MWH) or at room
temperature furnished coupling products 12 and 13, respectively.
Deprotection of the ethyl carbamate with HBr afforded the
dihydrobromide salts 14 and 15, which under reductive ami-
nation conditions were coupled with the corresponding alde-
hydes, providing access to target compounds 16a and 17a,
respectively.
of 8.3, and a second basic transition at the benzoxazole moiety
was observed at 3.3. Microsomal stability in human and mouse
microsomes was in the medium range, showing clearances of
CL ) 9.11 and 38.28 mL min-1 kg-1, respectively. However,
17c inherited considerable hH1R affinity (Ki ) 0.624 µM) from
seed structure 2 and also displayed substantial potency toward
the hERG potassium channel (IC50 < 100 nM), a well-known
liability of astemizole (2).21,26 Exploration of the chemical space
of the meta position showed that larger alkyl groups such as
cyclopentyl (17d) or isobutyl (17e) result in a slight further
improvement in binding, whereas n-propyl (17f) was notably
less active. Surprisingly, a fluoroethyl side chain (17g) was not
well tolerated, resulting in a 4-fold decrease in binding compared
with 17c. At the para position small lipophilic groups such as
fluorine (17h) and chlorine (17i) worked best, with methyl (17j)
being the most active in the series (Ki ) 37 nM). Interestingly,
the larger lipophilic OCF3 moiety (17k) reduced affinity
markedly, probably because of its large electron-withdrawing
potential lowering the basicity of the piperidine moiety below
a certain threshold critical for receptor interaction.
With the potency of the series having been considerably
improved, there only remained the outstanding issue of whether,
having started with an hH1R antagonist with a hERG liability,
it would be possible to move away from this undesirable
polypharmacology. In keeping with current opinion,27 attach-
ment of the polar sulfonamide group to the oxazole in 17l did
much to improve the hERG liability, raising the IC50 to 3.5 µM,
although at the expense of a slight decrease of hSST5R binding
compared with 17c. However, exchange of the p-methoxy
moiety in 17l to a chlorine group (17m) resulted in a significant
boost in hSST5R affinity to 17 nM. Removing hH1R activity
was equally facile; introduction of a second ethoxy group to
the benzylamine (17n) not only resulted in high hSST5R affinity
Exchange of the entire p-fluorobenzylbenzimidazole moiety
in 9k with benzothiazole (16a) resulted in an instant gain in
affinity by a factor of 1.8 (Table 2). A similar reduction,
although with a factor of 1.5 less pronounced, was obtained
when switching to a benzoxazole group (17a). As already
observed in the benzimidazole series 9, the physicochemically
rather unfavorable 1,4-dimethoxy-2-naphthyl group can also be
replaced in series 16 by benzyl substituents, as demonstrated
with 16b,c. However, affinities tended to be in the low
micromolar rather than submicromolar range. After several
benzyl side chains with various substitution patterns were
screened, a breakthrough was observed for compounds contain-
ing an ethoxy side chain at the meta position, which provided
two promising hits with 0.149 µM (16d) and 0.326 µM (16e).
A further gain in affinity of about a factor 2.5 was noted when
going to the corresponding benzoxazole derivatives (17b,c).
The functional response of benzoxazole 17c was determined
in a cAMP assay, revealing IC50 ) 0.74 µM. Remarkably, the
overwhelming number of compounds measured were full
antagonists; agonism or partial agonism was very rarely
observed. The physicochemical profile of optimized hit structure
17c turned out to be already quite promising, displaying
solubility (Lysa) S ) 130 µg/mL, a distribution coefficient log D
) 2.92, and a passive membrane permeability (Pampa) of Pe
) 4.44 × 10-6 cm/s.25 The tertiary piperidine N showed a pKa

6294 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 25
Letters
(10) Scicin´ski, J. J.; Barker, M. D.; Murray, P. J.; Jarvie, E. M. The solid
phase synthesis of a series of tri-substituted hydantoin ligands for the
somatostatin SST5 receptor. Bioorg. Med. Chem. Lett. 1998, 8, 3609–
3614.
(11) Rohrer, S. P.; Birzin, E. T.; Mosley, R. T.; Berk, S. C.; Hutchins,
S. M.; Shen, D.-M.; Xiong, Y.; Hayes, E. C.; Parmar, R. M.; Foor,
F.; Mitra, S. W.; Degrado, S. J.; Shu, M.; Klopp, J. M.; Cai, S.-J.;
Blake, A.; Chan, W. W. S.; Pasternak, A.; Yang, L.; Patchett, A. A.;
Smith, R. G.; Chapman, K. T.; Schaeffer, J. M. Rapid identification
of subtype-selective agonists of the somatostatin receptor through
combinatorial chemistry. Science 1998, 282, 737–740.
(12) Bruns, C.; Lewis, I.; Briner, U.; Meno-Tetang, G.; Weckbecker, G.
SOM230: a novel somatostatin peptidomimetic with broad somatotro-
pin release inhibiting factor (SRIF) receptor binding and a unique
antisecretory profile. Eur. J. Endocrinol. 2002, 146, 707–716.
(13) Contour-Galcéra, M.-O.; Sidhu, A.; Plas, P.; Roubert, P. 3-Thio-1,2,4-
triazoles, novel somatostatin SST2/SST5 agonists. Bioorg. Med. Chem.
Lett. 2005, 15, 3555–3559.
(14) Klabunde, T.; Hessler, G. Drug design strategies for targeting
G-protein-coupled receptors. ChemBioChem 2002, 3, 928–944.
(15) Stahl, M.; Guba, W.; Kansy, M. Integrating molecular design resoures
within modern drug discovery research: the Roche experience. Drug
DiscoVery Today 2006, 11, 326–333.
(16) Bleicher, K. H.; Green, L. G.; Martin, R. E.; Rogers-Evans, M. Ligand
identification for G-protein-coupled receptors: a lead generation
perspective. Curr. Opin. Chem. Biol. 2004, 8, 287–296.
(17) Kratochwil, N. A.; Malherbe, P.; Lindemann, L.; Ebeling, M.; Hoener,
M. C.; Mühlemann, A.; Porter, R. H. P.; Stahl, M.; Gerber, P. R. An
automated system for the analysis of G protein-coupled receptor
transmembrane binding pockets: alignment, receptor-based pharma-
cophores, and their application. J. Chem. Inf. Model. 2005, 45, 1324–
1336.
(18) Frimurer, T. M.; Ulven, T.; Elling, C. E.; Gerlach, L.-O.; Kostenis,
E.; Högberg, T. A physicogenetic method to assign ligand-binding
relationships between 7TM receptors. Bioorg. Med. Chem. Lett. 2005,
15, 3707–3712.
(13 nM) but also provided a 200-fold selectivity versus hH1R
(Ki ) 2.6 µM). Assessing cross-selectivity against the other four
SSTRs revealed some minor affinity toward hSST1R Ki ) 1.75
µM but essentially no interaction with hSST2R, hSST3R, and
hSST4R (Ki > 10 µM).
In conclusion, it has been demonstrated that by capitalizing
on the similarity of GPCR transmembrane binding pockets with
known small-molecule ligands, useful starting points for GPCR
lead identification programs can be found without recourse to
resource-demanding HTS campaigns. By careful optimization,
it is possible to readily move away from the original receptor
activity, as in this case where a hH1R ligand was transformed
into the first known, small-molecule, selective hSST5 receptor
antagonist. These ligands are endowed with promising physi-
cochemical and metabolic properties and should serve as useful
tools for the exploration of the biological role of the SST5
receptor.
Acknowledgment. We are grateful to Olivier Gavelle,
Christoph Kuratli, and Tim Mamié for their excellent technical
support and to Drs. Konrad Bleicher and Mark Rogers-Evans
for helpful discussions.
Supporting Information Available: Experimental details and
analytical data for selected compounds and details of physicochem-
ical and biological in vitro assays. This material is available free
of charge via the Internet at http://pubs.acs.org.
References
(1) Brazeau, P.; Vale, W.; Burgus, R.; Ling, N.; Butcher, M.; Rivier, J.;
Guillemin, R. Hypothalamic polypeptide that inhibits the secretion of
immunoreactive pituitary growth hormone. Science 1973, 179, 77–
79.
(2) Reisine, T.; Bell, G. I. Molecular biology of somatostatin receptors.
Endocrinol. ReV. 1995, 16, 427–442.
(3) Gillies, G. Somatostatin: the neuroendocrine story. Trends Pharmacol.
Sci. 1997, 18, 87–95.
(19) LOPAC ) Library of Pharmacologically Active Compounds. See Web
site: www.sigmaaldrich.com.
(20) BioPrint is a registered trademark of Cerep SA. See Web site:
www.cerep.com.
(21) Oppenheimer, J. J.; Casale, T. B. Next generation antihistamines:
therapeutic rationale, accomplishments and advances. Expert Opin.
InVest. Drugs 2002, 11, 807–817.
(22) Gao, K.; Muzina, D.; Gajwani, P.; Calabreses, J. R. Efficacy of typical
and atypical antipsychotics for primary and comorbid anxiety symp-
toms or disorders: a review. J. Clin. Psychiatry 2006, 67, 1327–1340.
(23) Heel, R. C.; Brogden, R. N.; Speightz, T. M.; Avery, G. S. Loperamide:
a review of its pharmacological properties and therapeutic efficacy in
diarrhoea. Drugs 1978, 15, 33–52.
(24) Anaya de Parrodi, C.; Quintero-Cortés, L.; Sandoval-Ramírez, J. A
short synthesis of astemizole. Synth. Commun. 1996, 26, 3323–3329.
(25) Kansy, M.; Senner, F.; Gubernator, K. Physicochemical high through-
put screening: parallel artifical membrane permeation assay in the
description of passive absorption processes. J. Med. Chem. 1998, 41,
1007–1010.
(26) Astemizole (2): IC50 ) 0.9 nM. Cavalli, A.; Poluzzi, E.; De Ponti, F.;
Recanatini, M. Toward a pharmacophore for drugs inducing the long
QT syndrome: insights from a CoMFA study of HERG K+ channel
blockers. J. Med. Chem. 2002, 45, 3844–3853.
(27) Jamieson, C.; Moir, E. M.; Rankovic, Z.; Wishart, G. Medicinal
chemistry of hERG optimizations: highlights and hangups. J. Med.
Chem. 2006, 49, 5029–5046.
(4) Loughlin, W. A.; Tyndall, J. D. A.; Glenn, M. P.; Failie, D. P. Beta-
strand mimetics. Chem. ReV. 2004, 104, 6085–6117.
(5) Pittaluga, A.; Feligioni, M.; Longordo, F.; Arvigo, M.; Raiteri, M.
Somatostatin-induced activation and up-regulation of N-methyl-D-
aspartate receptor function: Mediation through calmodulin-dependent
protein kinase II, phospholipase C, protein kinase C, and tyrosine
kinase in hippocampal noradrenergic nerve endings. J. Pharmacol.
Exp. Ther. 2005, 313, 242–249.
(6) Ösapay, G.; Ösapay, K. Therapeutic applications of somatostatin
analogues. Expert Opin. Ther. Pat. 1998, 8, 855–870.
(7) Rajeswaran, W. G.; Hocart, S. J.; Murphy, W. A.; Taylor, J. E.; Coy,
D. H. Highly potent and subtype selective ligands derived by N-methyl
scan of a somatostatin antagonist. J. Med. Chem. 2001, 44, 1305–
1311.
(8) Hannon, J. P.; Nunn, C.; Stolz, B.; Bruns, C.; Weckbecker, G.; Lewis,
I.; Troxler, T.; Hurth, K.; Hoyer, D. Drug design at peptide receptors.
J. Mol. Neurosci. 2002, 18, 15–27.
(9) Souers, A. J.; Virgilio, A. A.; Rosenquist, Å.; Fenuik, W.; Ellman,
J. A. Identification of a potent heterocyclic ligand to somatostatin
receptor subtype 5 by the synthesis and screening of ꢀ-turn mimetic
libraries. J. Am. Chem. Soc. 1999, 121, 1817–1825.
JM701143P