Benzoxazole piperidines as selective and potent somatostatin receptor subtype 5 antagonists

Article abstract of DOI:10.1016/j.bmcl.2009.09.024

SAR studies of a recently described SST5R selective benzoxazole piperidine lead series are described with particular focus on the substitution pattern on the benzyl and benzoxazole side-chains. Introduction of a second meta substituent at the benzyl unit

Full text of DOI:10.1016/j.bmcl.2009.09.024

Bioorganic & Medicinal Chemistry Letters 19 (2009) 6106–6113
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Benzoxazole piperidines as selective and potent somatostatin receptor
subtype 5 antagonists
a
a
a
a
Rainer E. Martin a, , Peter Mohr , Hans Peter Maerki , Wolfgang Guba , Christoph Kuratli ,
Olivier Gavelle a, Alfred Binggeli a, Stefanie Bendels b, Rubén Alvarez-Sánchez b, André Alker c,
Liudmila Polonchuk b, Andreas D. Christ d
*
a F. Hoffmann-La Roche Ltd, Pharmaceuticals Division, Medicinal Chemistry, CH-4070 Basel, Switzerland
b F. Hoffmann-La Roche Ltd, Pharmaceuticals Division, Nonclinical Drug Safety, CH-4070 Basel, Switzerland
c F. Hoffmann-La Roche Ltd, Pharmaceuticals Division, Molecular Structure Research, CH-4070 Basel, Switzerland
d F. Hoffmann-La Roche Ltd, Pharmaceuticals Division, Discovery Metabolic and Vascular Diseases, CH-4070 Basel, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 28 July 2009
Revised 3 September 2009
Accepted 5 September 2009
Available online 12 September 2009
SAR studies of a recently described SST5R selective benzoxazole piperidine lead series are described with
particular focus on the substitution pattern on the benzyl and benzoxazole side-chains. Introduction of a
second meta substituent at the benzyl unit significantly lowers residual hH1 activity and insertion of sub-
stituents onto the benzoxazole periphery entirely removes remaining h5-HT2B activity. Compounds with
single digit nM activity, functional antagonism and favorable physicochemical properties endowed with a
good pharmacokinetic profile in rats are described which should become valuable tools for exploring the
pharmacological role of the SST5 receptor in vivo.
Keywords:
Somatostatin receptor 5
G-protein-coupled receptors
Antagonist
Ó 2009 Elsevier Ltd. All rights reserved.
Benzoxazole
Medicinal chemistry
Chemogenomics
Somatostatin (SST) or somatotropin release-inhibiting factor
(SRIF) was first described in 1973. It is a mammalian peptide hor-
mone existing in two isoforms with 14 or 28 amino acids (SST-14
and SST-28, respectively) that is widely distributed throughout the
body.1,2 It exhibits multiple biological functions including antipro-
liferative, hormonal, and neuron-transmitter activity and was
demonstrated to reduce splanchnic perfusion.3 Peripheral SST-28
is predominantly produced in the mucosa of ileum and colon and
in d-cells of the endocrine pancreas, whereas SST-14 can primarily
be found in the foregut and enteric nervous system.4 As a hormone
SST is most commonly inhibitory in nature and impedes for in-
stance the release of growth hormone (GH), pancreatic insulin,
glucagon, and gastrin.2 SST acts via five distinct G-protein-coupled
receptors (GPCR) SST1-5 that have been cloned and characterized.2
Particularly, SST acting via SST5 receptors has been found to acti-
vate and up-regulate NMDA receptor function5 and to control hor-
monal secretions (e.g., insulin, GLP-1, growth hormone).4,6 In the
pancreas, SSTR5 is prominently expressed on 87% of all insulin
secreting cells.7–10 Due to the numerous physiological functions
of SST, selective receptor specific ligands and particularly antago-
nists would be highly interesting tool compounds to study and elu-
cidate the diverse pharmacology of SST in more detail.11
We have recently reported on the identification of receptor sub-
type 5 selective SST5 antagonists containing a benzoxazole head-
group connected to a 4-amino-piperidine unit.12 This lead series
was identified following a chemogenomics approach. Biogenic
amine receptors such as opioid, histamine, dopamine, and seroto-
nine receptors were identified as closest neighbors of the SST5
receptor based on an analysis of the homology of amino acids
delineating the putative consensus drug binding site in the trans-
membrane region of GPCRs.13 After a focused screen of the Sigma
LOPAC14 and the Cerep BioPrint15 compound databases the well-
known hH1R antagonist Astemizole (1) was selected as a seed
structure for lead generation activities and successfully trans-
formed into the first small molecule selective hSST5 receptor
antagonist series (Fig. 1). For example, structure 2 provided high
affinity toward hSST5 (13 nM) and displayed a 200-fold selectivity
versus hH1 (Ki = 2.6
other four SST subtype receptors revealed some weak affinity
toward hSST1 (Ki = 1.75 M), but essentially no interaction with
hSST2, hSST3, and hSST4 (Ki > 10 M). A second and structurally
lM). Assessing cross-selectivity against the
l
l
very closely related nicotinamide series with nanomolar binding
was identified by a similarity analysis of GPCR affinity profiles of
* Corresponding author. Tel.: +41 61 6887350; fax: +41 61 6888714.
E-mail address: rainer_e.martin@roche.com (R. E. Martin).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.09.024

R. E. Martin et al. / Bioorg. Med. Chem. Lett. 19 (2009) 6106–6113
6107
F
O
NH
O
N
N
N
S
O
NH
H2N
O
O
N
NH2
N
1
2
Astemizole
H1 Antagonist
O
hSST5 Ki = 4.07 µM
hSST5 Ki = 13 nM
Figure 1. Astemizole (1) serving as a seed structure for lead generation activities resulted in SST5R antagonist 2 displaying a more than 200-fold selectivity against hH1.
Astemizole (1) versus a set of in-house GPCR-biased combinatorial
Table 1
libraries.16
Modification of the central 4-amino-piperidine spacer
In this Letter we would like to present an in depth analysis of
S
N
the SAR against hSST5 including the selectivity against three off-
target receptors hH1, h5-HT2B, and hSST1 and to discuss some
physicochemical and pharmacokinetic properties of this benzoxaz-
ole series.
spacer
O
Having selected Astemizole (1) as a starting point our efforts fo-
cused on generating the minimal critical chemotype essential for
hSST5R activity. These studies revealed that the p-fluorobenzyl
side-chain can be omitted and that the phenethyl group may
advantageously be replaced by appropriately substituted benzyl
moieties such as 3-ethoxy-4-methoxy-benzyl or 3-ethoxy-4-
chloro-benzyl side-chains. Screening for an optimal replacement
of the p-fluorobenzyl-benzimidazole moiety soon resulted in the
identification of benzothiazole as suitable surrogate of the benz-
imidazole group. With these two substituents on either side of
the 4-amino-piperidine core, the benzothiazole and the 3-eth-
oxy-4-methoxy- benzyl moiety, kept constant we set out for the
search of an optimal spacer replacement.
However, screening various diamino spacer units such as piper-
idines (3–6), diaminocyclohexanes (7 and 8; mixture of cis/trans)
or piperazine (9), and phenyl moieties (10) did not result in any
improvement of binding affinity toward hSST5R (Table 1). The ben-
zothiazoles 3–5, 7–8, and 10 were prepared in analogy to com-
pound 3 whose synthesis is exemplified in Scheme 1.
Nucleophilic aromatic substitution of 2-chloro-benzothiazole
(11) with 4-amino-piperidine-1-carboxylic acid ethyl ester under
microwave-assisted heating provided coupling product 12 in 75%
yield. Deprotection of the ethylcarbamate with HBr afforded the
dihydrobromide salt 13 which under reductive amination condi-
tions was coupled with 3-ethoxy-4-methoxy-benzaldehyde fur-
nishing the benzothiazole 3.
Other target compounds were synthesized following a similar
reaction scheme, but varying the diamino spacer unit appropri-
ately. In the case of amide-linked target 6 the first reaction step in-
volved an amide formation between 2-amino-benzothiazole and
the corresponding BOC-protected piperidine acid under HOBT (1-
hydroxy-1H-benzotriazole) activation. Cleavage of the protecting
group under acidic conditions followed by reductive alkylation fur-
nished the desired product 6. Urea compound 9 was prepared by
reaction of mono-BOC-protected piperazine and 2-isocyanato-ben-
zothiazole in THF using diisopropyl-ethylamine as base.
O
Compound
hSST5 Kia
3
330
N
H
N
H
N
4
>1000
>1000
>1000
>1000
>1000
>1000
>1000
N
H
5
N
N
H
N
6
N
O
H
7
N
H
N
H
8
N
N
H
H
N
9
N
N
O
N
H
10
N
H
Ki in nM, radioligand binding assay.17
a
Since none of the central spacer variations improved binding,
we turned our attention back to hit structure 3 focusing on further
optimization of the benzothiazole moiety. However, it soon
became apparent that modifications of the benzothiazole core
did not significantly improve binding affinity toward hSST5R as
demonstrated by 6-chloro-benzothiazole 14 or thiazolo[5,4-b]pyr-
idine 15 (Table 2). In contrast, a profound jump in binding affinity
was observed when going from benzothiazole to benzoxazole.
Interestingly, the unsubstituted benzoxazole derivative 16 was
found with hSST5 Ki = 121 nM to be the most active one, followed

6108
R. E. Martin et al. / Bioorg. Med. Chem. Lett. 19 (2009) 6106–6113
R1
R2
R1
R2
R1
R2
A
N
A
N
A
N
b
a
NH
NH 2 HBr
NH
Cl
N
O
A = S 11
A = S 12
A = S 13
A = O
64
O
A = O
A = O
62
63
R1
R2
A
N
NH
O
A = S, R1 = R2 = H, R3
A = O
=
3
c
O
N
R3
Scheme 1. Reagents and conditions: (a) 12: 4-amino-piperidine-1-carboxylic acid ethyl ester, TEA, MWH 180 °C, 5 min, 75%. 63: 4-amino-piperidine-1-carboxylic acid ethyl
ester, DMF, rt, 18 h, 45%; (b) 13 and 64: HBr (48%) in water, reflux, 2 h, quant; (c) 3: 3-ethoxy-4-methoxy-benzaldehyde, HOAc, DIEA, NaCNBH3, EtOH, 40 °C, 18 h, 62%.18
A = 0: corresponding benzaldehyde, HOAc, DIEA, NaCNBH3, EtOH, 40 °C, 18 h; purification by prep. HPLC.18
by 5-chloro-benzoxazole 17 with 194 nM and 5-ethylsulfone-
benzoxazole 18 with 196 nM, respectively.
bridge structural database (CSD) revealed that alkoxy groups such
as methoxy and ethoxy substituents adopt an in-plane conforma-
tion with respect to the phenyl ring.19 In addition, a para substitu-
ent between the two meta alkoxy groups enforces an anti
orientation of the alkyl chains in order to minimize steric repul-
sion. This result is confirmed by in-house single crystal X-ray anal-
ysis studies of the p-hydroxyethoxy derivative 48.20,21 Although
this compound was found to be inactive on the hSST5 receptor
due to the o-F and the p-hydroxyethoxy motif, it serves as a nice
example for illustrating the coplanar arrangement of the m-ethoxy
group with the aromatic system. The deviation from planarity of
the exocyclic O(3)–C(28) bond and the aromatic system was mea-
sured to be only 2.3° (Fig. 2). The two planes defined by the benz-
oxazole unit and the coplanar three atoms of the piperidine chair
conformation bisect at an angle of 53.2°. Moreover, the structure
also corroborates the staggered conformation typically observed
for a hydroxyethoxy motif due to the gauche effect (u [O(4)–
C(31)–C(30)–O(2)] = 75.8°).19
Other analogs such as 5-nitro- (19), 5-trifluoromethoxy- (20),
and 5-acetamide- (21) substituted benzoxazoles or benzoxazole
sulfonamides (22–24) exhibited Ki values clearly above 200 nM.
Furthermore, the oxazolo-pyridines 25 and 26 showed consider-
ably lower activity compared with benzoxazole 16, and aza-isomer
27 turned out to be devoid of any activity. Remarkably, the unsub-
stituted benzimidazole 28 lacking the p-fluorobenzyl group pres-
ent in seed structure 1 was also found to be entirely inactive.
Earlier exploration of the SAR around the benzyl side-chain had
revealed that compounds containing an ethoxy group in meta po-
sition are of particular interest as exemplified by benzothiazole 3
and benzoxazole 16, respectively.12 Exchange of the meta ethoxy
in 16 with an ethylamino group (29) resulted in a reduction of
affinity by a factor of 3, whereas introduction of an N atom provid-
ing pyrido derivative 30 gave completely inactive compounds
(Table 3).
Interestingly, larger alkyl groups such as isobutyl (31) or cyclo-
pentyl (32) afforded with a Ki of 23 nM and 83 nM, respectively,
higher binding affinity toward hSST5R, but at the expense of in-
creased MW, higher lipophilicity and lower microsomal stability.
Other variations of the meta ethoxy group in 16 such as n-propoxy
(33), methoxy (34), or fluoroethoxy (35) were noticeably less ac-
tive. Screening of the para position revealed that small substituents
like methyl (36), hydroxy (37), chlorine (38), or fluorine (39) are
clearly preferred, whereby methyl derivative 36 with a Ki of
37 nM was the most active one in this series. Less favorable moie-
ties are sterically more demanding groups such as trifluoromethyl
(40), cyclopropoxy (41), or isopropoxy (42). Interestingly, intro-
duction of an additional ethoxy group into the second meta posi-
tion was equally well tolerated providing very active compounds
like 43 with Ki = 33 nM. Also ethoxycarbonyl (44) and fluorine
(45) in para position as well as the sterically rather demanding tet-
rahydropyranyl-oxy moiety in meta position (46) produced quite
active compounds with Ki = 156 nM, 178 nM, and 199 nM, respec-
tively. Interestingly, moving the p-F atom in 45 to the ortho posi-
tion of the phenyl ring, as in compounds 47 or 48, was not
tolerated and resulted in complete loss of affinity. However, the
bis-isopropoxy derivative 49 was again surprisingly active with a
Ki of 31 nM.
It subsequently became apparent that the second meta vector of
the benzyl needle was of particular significance as structures with-
out a substituent at this position still carry significant hH1R activ-
ity. For instance, 16 with an hSST5R activity of 121 nM revealed
only a 5.2-fold selectivity over hH1 (Ki = 0.624
observations were made with analogues bearing a p-Cl (38, hH1
Ki = 0.266 M, 2.4-fold) or p-F group (39, hH1 Ki = 0.826 M, 4.3-
lM) and similar
l
l
fold; Table 4). In addition, substituents at the rim of the benzoxaz-
ole moiety did not resolve this issue as shown by examples 22
(hH1 Ki = 0.520
Ki = 0.923 M, 5.1-fold) containing a cyclopropylamide group, 24
(hH1 Ki = 1.013 M, 3.0-fold) and 51 (hH1 Ki = 1.374 M, 5.5-fold)
with chloro/sulfonamide substitution pattern, or 52 (hH1
lM, 1.0-fold) carrying a sulfonamide, 50 (hH1
l
l
l
a
Ki = 0.199 lM, 1.0-fold) bearing an inverse phenyl-sulfonamide
group.
However, this off-target activity was found to be significantly
reduced, in some cases almost eliminated, by introducing a second
vector onto position R4 of the benzyl moiety. Compound 2 (hH1
Ki = 2.6
l
M, 200-fold) and 53 (hH1 Ki = 5.0
lM, 96-fold) containing
an ethoxy moiety or 54 (hH1 Ki = 38.4
l
M, 698-fold) with a more
bulky tetrahydropyranyl-oxy group nicely illustrate this. In gen-
eral, the sterically more demanding the R4 substituent, the more
efficiently hH1 activity was found to be suppressed, whereas a di-
rect comparison between 46 and 54 indicated for selected side-
chain combinations a synergistic effect of substituents on R2 and
R4 with respect to reduction of hH1R activity. Whereas for 46
The observation that the presence of an ortho (cf. 45 with 47) as
well as the increasing size (cf. series 36–42) of a para substituent
exhibits a detrimental influence on the activity of these com-
pounds might be explained by an appropriate preorganization of
the adjacent m-ethoxy side-chain. Indeed, a search in the Cam-
(hH1 Ki = 3.6
lM) only an 18-fold lowering in hH1R binding was
observed due to the reduced activity on the hSST5 receptor, the

R. E. Martin et al. / Bioorg. Med. Chem. Lett. 19 (2009) 6106–6113
6109
Table 2
Table 3
Refinement of the benzothiazole side-chain keeping the 3-ethoxy-4-methoxy benzyl
motif connected to the 4-amino piperidine spacer constant
Variation of the substitution pattern of the benzyl side-chain
O
H
R
N
N
N
N
H
O
N
R
O
Compound
hSST5 Kia
Compound
hSST5 Kia
S
3
330
N
16
29
30
31
32
33
34
35
36
37
38
121
371
>1000
23
O
Cl
S
N
14
15
16
17
18
361
309
121
194
196
O
O
O
O
O
O
O
O
N
S
N
O
N
N
H
O
O
O
O
O
O
O
N
N
Cl
O
O
N
S
O
O
N
N+
O
19
20
21
22
23
O
F
252
354
357
529
226
83
O
F
N
O
F
O
293
472
487
37
O
N
N
H
O
O
N
S
H2N
O
O
O
F
N
S
N
O
O
H2N
S
O
24
25
26
27
28
343
O
N
Cl
N
O
O
O
633
N
O
N
720
66
N
O
OH
>1000
>1000
N
N
H
N
113
O
N
Cl
a
Ki in nM, radioligand binding assay.17
(continued on next page)

6110
R. E. Martin et al. / Bioorg. Med. Chem. Lett. 19 (2009) 6106–6113
Table 3 (continued)
Compound
hSST5 Kia
39
40
41
194
524
537
O
F
O
F
F
F
O
Figure 2. Single crystal X-ray diffraction analysis of benzoxazole 48 illustrating the
O
O
anti in-plane conformation of the m-ethoxy group and the significant twist of the
benzoxazole unit with respect to the piperidine core. The ORTEP drawing depicts
thermal ellipsoids at a 30% probability level.
additional ethylsulfone moiety in 54 notably helped to enhance
this ratio to 698.
In addition to the expected hH1 liability given the origin of the hit
series, a selectivity screen on 8 different serotonine receptors
showed that unsubstituted benzoxazoles such as 16 or 38 are rather
O
42
>1000
potent ligands for h5-HT2B (Ki = 0.041
lM
and 0.074 lM,
respectively), whereas no affinity was found for all other serotonine
receptor subtypes. In contrast to the hH1 liability, attachments onto
the benzoxazole rim were able to efficiently resolve this issue (see
cyclopropylamide 50 or free sulfonamide 51, both compounds h5-
43
44
33
O
O
O
NH2
HT2B Ki >10 lM). Interestingly, this effect seems to be rather inde-
pendent on both the steric and electronic nature of the additional
substituents. In contrast to the SAR observed for hH1R binding, sub-
stituents on R4 seemed to have little influence on h5-HT2B activity as
O
156
O
demonstrated for instance by 46 (h5-HT2B Ki = 0.094 lM).
O
With respect to the other four SST subtype receptors, com-
pounds with an unsubstituted benzoxazole moiety such as 38 or
39 were devoid of any hSST1 activity (Ki >10
bearing slightly polar R2 groups, some minor or even no affinity for
hSST1 was observed (e.g., 2: Ki = 1.75 M, 134-fold; 50:
Ki = 4.95 M, 27-fold; 59: Ki = 4.50 M, 888-fold; 60: Ki > 10 M;
61: Ki > 10 M), whereas substituents at R4 seem to have little
influence on binding potency. However, essentially no interaction
was observed for all compounds investigated against the other
three subtype SST receptors hSST2, hSST3, and hSST4 (Ki
lM). For compounds
45
46
178
199
O
l
O
F
l
l
l
O
l
O
O
>10 lM). Having successfully addressed the known selectivity is-
O
O
sues we turned our attention to the hERG potassium channel liabil-
ity, a well known issue of compounds containing basic amines
flanked by lipophilic side-chains.22–25 Indeed, several revealed
rather high hERG binding exemplified by compounds 16 (IC50
F
47
>1000
<0.1 lM), 18 (IC50 = 0.18 lM), or 58 (IC50 = 0.43 l
M).26 As ex-
pected, attachment of polar groups such as a free sulfonamide moi-
ety to the periphery of the benzoxazole unit like in compound 22
were able to significantly reduce, albeit not completely eliminate,
the hERG binding potential (IC50 = 3.5 lM). Extensive QSAR studies
showed that more hydrophilic compounds bearing carboxylate
groups are required in order to entirely eliminate hERG channel
affinity (59 and 60 both IC50 >10 lM). Similar observations have
O
F
48
>1000
O
been made by other research groups and were explained by unfa-
vorable electrostatic interactions of the carboxylic acid group with-
in the hERG channel binding cavity.27 Interestingly, the distance
between the positively charged piperidine nitrogen and the car-
boxylic acid group turned out to be crucial. Connection of the
carboxylate moiety via alkyoxy-spacers to the benzoxazole group
(e.g.; 3-carboxy-propoxy side-chain) resulted in a return to single
HO
49
31
O
O
Ki in nM, radioligand binding assay.17
digit lM hERG binding affinity (data not shown).
a

R. E. Martin et al. / Bioorg. Med. Chem. Lett. 19 (2009) 6106–6113
6111
Table 4
Physicochemical properties of selected benzoxazoles 2, 16, 18, 22, 24, 38, 39, 46, and 50–61
R1
R2
O
N
H
N
O
N
R3
R4
Compound
R1
R2
R3
R4
hSST5 Kia
(nM)
Lysa Sb
Pampa Pec
log Dd
KOW
(clog P)
pKae(B = base,
A = acid)
(lg/mL)
(10À6 cm/s)
2
H
H
H
H
SO2NH2
H
SO2C2H5
SO2NH2
Cl
H
H
NH2
OCH3
OCH3
OCH3
OCH3
Cl
OC2H5
13
335
130
>530
>614
338
44
0.00
4.43
1.63
1.64
1.10
0.49
4.33
1.62
2.92
2.13
1.51
2.67
3.44
3.63
2.74
3.92
3.36
2.80
3.31
4.80
4.26
8.28
8.30
8.10
8.05
7.92
7.63
7.79
16
18
22
24
38
39
H
H
H
H
H
H
121
196
529
343
113
194
SO2NH2
H
H
F
>452
O
46
50
H
H
H
H
199
182
212
2.05
1.10
3.02
2.64
3.90
4.47
7.76
—
O
O
OCH3
H
>638
N
H
51
52
SO2NH2
H
Cl
F
H
H
252
205
32
145
2.45
2.20
2.78
1.93
3.65
4.83
7.61
7.72
NHSO2Ph
OCH3
N
53
54
H
H
SO2N(CH3)2
SO2C2H5
OC2H5
52
55
22
1.31
1.75
2.89
3.13
5.22
3.34
—
O
H
183
7.55
O
55
56
SO2NH2
Cl
Cl
H
F
Cl
OC2H5
H
252
239
32
44
2.45
4.81
2.78
precip
3.65
5.61
7.61
—
H
O
N
S
57
58
H
H
OCH3
F
H
175
8
>720
148
0.49
1.61
1.48
2.96
3.51
3.64
7.91
7.72
N
O
N
CONH2
OC2H5
59
60
61
H
COOH
Cl
Cl
F
OC2H5
H
3
14
7
2.09
3.60
2.89
1.82
1.39
1.33
2.94
2.63
2.43
7.59 (B)
3.89 (A)
7.67 (B)
4.03 (A)
7.70 (B)
4.01 (A)
COOH
COOH
H
H
23
18
OC2H5
6
a
b
c
Radioligand binding assay.17
Lyophilization solubility assay. Solubility was measured from lyophilized DMSO stock solutions spectrophotometrically at pH 6.5 in a 50 mM phosphate buffer.
Pampa (Parallel Artificial Membrane Permeation Assay): low: Pe < 0.1, medium: 0.1 < Pe < 1.0, high: Pe > 1.0.
log D values were measured spectrophotometrically at pH 7.4 in a 1-octanol/50 mM TAPSO buffer system containing 5% (v/v) DMSO.
pKa values were determined spectrophotometrically on a ProfilerSGA instrument in a SGA buffer system containing 10% (v/v) methanol at an ionic strength of 150 mM.
d
e
A substantial subset of compounds with high affinity to hSST5
electron-withdrawing substituents in para-position of the benzyl
group such as Cl (38), or F (39) lowered the pKa value in compari-
son with the methoxy derivative 16 from 8.30 to 7.63 and 7.79,
respectively. Interestingly, a similar reduction in basicity of the
piperidine nitrogen can be achieved by introducing an additional
alkoxy group in meta position of the benzyl moiety and without
presence of a para substituent as a comparison of structures 39
(7.79) with 46 (7.76) or 54 (7.55) demonstrates. Remarkably, these
effects are not additive, and combination of both structural ele-
ments such as a p-F atom and a m-alkoxy group in one single struc-
was further profiled by functional in vitro assays. The overwhelm-
ing majority of compounds are full antagonists devoid of agonistic
or inverse agonistic activity. The functional IC50 was, as expected,
maximally up to one magnitude higher than the corresponding Ki
value (e.g., 60: IC50 = 98 nM vs Ki = 23 nM and 61: IC50 = 72 nM vs
Ki = 18 nM).
The pKa value of the piperidine nitrogen in this series was found
to be typically in the range of 7.6–8.3, rendering the compounds
protoned under physiological conditions (Table 4). As expected,

6112
R. E. Martin et al. / Bioorg. Med. Chem. Lett. 19 (2009) 6106–6113
Table 5
benzoxazole 62 by 4-amino-piperidine-1-carboxylic acid ethyl
ester to give intermediate 63. For some derivatives reaction rates
may be enhanced significantly by conducting the reaction at ele-
vated temperatures using either microwave irradiation or conven-
tional heating. Deprotection of the ethylcarbamate with HBr
afforded the dihydrobromide salt 64 which under reductive amina-
tion conditions was reacted with the corresponding aldehyde to
yield the desired benzoxazole compound, respectively. Alterna-
tively, target structures can also be prepared by alkylation of piper-
idine 64 with appropriately substituted benzyl halides.
Pharmacokinetic parameters in male wistar rat of representative hSST5R antagonists
CLa (mL minÀ1 kgÀ1
)
Vssb (L kgÀ1
)
t1/2 (h)
%Fd
c
Compound
16
18
22
56
57
58
59
60
61
84
126
103
37
41
28
18
27
20
4.1
3.8
8.2
4.6
2.2
3.3
1.8
4.3
1.5
0.7
0.4
1.7
2.6
1.5
2.5
4.9
3.1
2.4
6
6
4
33
2
38
12
57
39
In conclusion, the SAR of the benzoxazole series has been
expanded and a number of potent and selective hSST5 antagonists
were identified. The zwitterionic compounds 60 and 61 combine
a
b
c
Clearance: <13 low, 13–40 medium, >40 high.
Volume of distribution at steady state.
Terminal half-life.
high activity, excellent selectivity against hH1, hSST1, h5-HT2B
,
d
Oral bioavailability.
and hERG, favorable physicochemical properties and good single-
dose PK properties in rat. These further improved compounds
might be useful tools in the in vivo assessment of the biological
role of the SST5 receptor.
ture does not result in a pronounced further reduction of the pKa
value as can be seen for instance from a comparison of 39 (7.79)
and 46 (7.76) with 55 (7.61) or 58 (7.72).
Acknowledgments
The solubilities in phosphate buffer were in a good to excellent
range being strongly influenced by the nature of the side-chains at-
tached (Table 4). The dimethylsulfonamide analogue 53, for
We would like to thank Tim Mamié and Lisa Stoller for their
excellent technical support, Bjoern Wagner for pKa determinations
and Dr. Christoph Ullmer for the hH1 measurements.
instance, showed a solubility at the lower end of 22
whereas compound 50 containing a cyclobutylamide group exhib-
ited excellent solubility with >638 g/mL, respectively. Surpris-
lg/mL,
l
References and notes
ingly, exchange of the p-Cl in compound 38 for a p-F atom (39)
had a dramatic impact on the aqueous solubility increasing it from
44 lg/mL to >452 lg/mL. The permeability in the Pampa assay was
1. Brazeau, P.; Vale, W.; Burgus, R.; Ling, N.; Butcher, M.; Rivier, J.; Guillemin, R.
Science 1973, 179, 77.
2. Reisine, T.; Bell, G. I. Endocrinol. Rev. 1995, 16, 427.
found to be high for all compounds with the only exemption of sul-
fonamide 2 (Table 4).28 Lipophilicity values range from 1.33 (61) to
3.63 (39) and again vary depending on the substitution pattern.
In general, the correlation of rat in vitro microsomal and hepa-
tocyte data to in vivo clearance values was very good (data not
shown). Based on in vitro clearance data, a few compounds were
selected for single dose pharmacokinetic studies in rat. The results
are summarized in Table 5 and show medium to high clearance
indicating in some cases extrahepatic elimination pathways (16,
18, and 22, respectively). Incorporation of a chlorine substituent
on the benzoxazole moiety as well as on the para position of the
benzyl group (56) led to a significant reduction of clearance and
improvement of oral bioavailability. Parent benzoxazole 16, ethyl-
sulfone 18 as well as sulfonamides 22 and 57 showed very poor
bioavailability, which might be explained by a combination of
medium to high clearance and substantially increased polarity
moving them into a critical range for absorption. Interestingly, in
the Pampa permeation assay compounds 16, 18, and 22 exhibited
high and compound 57 medium to high permeability (Table 4)
which indicates that absorption should not be the primary limiting
factor. In contrast, the less polar amide compound 58 (log D = 2.96)
showed higher bioavailability and medium plasma clearance, but
3. Gillies, G. Trends Pharmacol. Sci. 1997, 18, 87.
4. Chisholm, C.; Greenberg, G. R. Am. J. Physiol. Endocrinol. Metab. 2002, 283, E311.
5. Pittaluga, A.; Feligioni, M.; Longordo, F.; Arvigo, M.; Raiteri, M. J. Pharmacol. Exp.
Ther. 2005, 313, 242.
6. Ösapay, G.; Ösapay, K. Exp. Opin. Ther. Patents 1998, 8, 855.
7. Kumar, U.; Sasi, R.; Suresh, S.; Patel, A.; Thangaraju, M.; Metrakos, P.; Patel, S.
C.; Patel, Y. C. Diabetes 1999, 48, 77.
8. Mitra, S. W.; Mezey, É.; Hunyady, B.; Chamberlain, L.; Hayes, E.; Foor, F.; Wang,
Y.; Schonbrunn, A.; Schaeffer, J. M. Endocrinology 1999, 140, 3790.
9. Portela-Gomes, G. M.; Stridsberg, M.; Grimelius, L.; Oberg, K.; Janson, E. T. Appl.
Immunohistochem. Mol. Morph. 2000, 8, 126.
10. Strowski, M. Z.; Blake, A. D. Mol. Cell Endocrinol. 2008, 286, 169.
11. Wolkenberg, S. E.; Thut, C. J. Curr. Opin. Drug Discovery Dev. 2008, 11, 446.
12. Martin, R. E.; Green, L. G.; Guba, W.; Kratochwil, N.; Christ, A. J. Med. Chem.
2007, 50, 6291.
13. Kratochwil, N. A.; Malherbe, P.; Lindemann, L.; Ebeling, M.; Hoener, M. C.;
Mühlemann, A.; Porter, R. H. P.; Stahl, M.; Gerber, P. R. J. Chem. Inf. Model. 2005,
45, 1324.
14. LOPAC = Library of Pharmacologically Active Compounds. See website:
www.sigmaaldrich.com.
15. BioPrintÒ is a registered trademark of Cerep SA. See website: www.cerep.com.
16. Guba, W.; Green, L. G.; Martin, R. E.; Roche, O.; Kratochwil, N.; Mauser, H.;
Bissantz, C.; Christ, A.; Stahl, M. J. Med. Chem. 2007, 50, 6295.
17.
A CHO cells stably expressing human SSTR5 (GenBank accession number
D16827, Euroscreen, Brussels, Belgium) was used for binding and functional
assays; cells expressing human SSTR1, 2, and 3, and rat and mouse SSTR5 were
established in-house. SST-14 was purchased from Bachem (Bubendorf,
Switzerland). Membranes from cells expressing SST receptors were prepared
by sonication and incubated with radiolabeled tracer (11-Tyr SST-14; Perkin–
Elmer, Schwerzenbach, Switzerland, or Amersham, Dübendorf, Switzerland)
and either test compound in varying concentration or, for the quantification of
non-specific binding, non-labeled SST-14. The incubation was stopped by
filtration through glass-fiber filters and the bound radioactivity measured to
estimate the concentration of test compound required for half maximal
unfortunately also very high hERG receptor activity (IC50
0.43 M). The attachment of a carboxylic acid moiety on the benz-
oxazole periphery resulted in 59 which is devoid of any hERG
activity (IC50 >10 M) and, due to the second m-ethoxy group, very
=
l
l
inhibition of binding (IC50) and the binding affinity (Ki). For functional
selective against hH1 (>200-fold). This modification also improved
clearance, but again at the expense of lower oral bioavailability. In
contrast, the carboxylic acid in regioisomers 60 and 61 did not sig-
nificantly affect the volume of distribution of the resulting zwitter-
ions and allowed for a significant reduction of the clearance values.
This modification had furthermore a very beneficial influence on
bioavailability and half-life which for these compounds is now
within a satisfactory range.
experiments, transfected cells were incubated with forskolin and test
compound in varying concentration. Subsequently, cellular cAMP
concentration was measured using a FRET (fluorescence resonance energy
transfer) based assay as previously published Roth, D.; Matile, H.; Josel, H.-P.;
Enderle, T. Fast-TRF: Novel Time-Resolved Assays for Drug Discovery. In Society
for Biomolecular Screening, 11th Annual Conference and Exhibition, Geneva, 2005,
p 265. The concentration of the test compound necessary to induce a half
maximal effect (EC50) and the efficacy compared to 0.15 nM SST-14 were
determined from concentration-versus-cAMP graphs. For the determination of
potential antagonism, 0.15 nM SST-14 was applied together with the test
compound, and the concentration of the test compound to half maximally
reverse the effect of SST-14 (i.e., IC50) was deduced from concentration-
versus-cAMP graphs.
The synthesis of benzoxazole analogs followed a preparation
similar to benzothiazole 3 (Scheme 1). The synthesis commences
with a nucleophilic aromatic substitution of the chlorine atom in

R. E. Martin et al. / Bioorg. Med. Chem. Lett. 19 (2009) 6106–6113
6113
18. Compounds were purified by preparative HPLC on a Phenomenex Aqua 5
l
m
24. Lagrutta, A. A.; Trepakova, E. S.; Salata, J. J. Curr. Top. Med. Chem. 2008, 8, 1102.
25. Aronov, A. M. Curr. Top. Med. Chem. 2008, 8, 1113.
C18 125A 60 Â 21.20 mm column equipped with a Gilson Liquid Handler 215
autosampler, two Varian Prep Star Model SD-1 pumps, a Sedex ELSD 75
lightscatter and a Dionex UVD 340S UV detector.
26. ThehERGcurrentmeasurementwasperformedat automated patch clamp system
PatchXpress 7000A (MDS Analytical Technologies, USA). Electrophysiologcal
recordings ofK+ currents (IKhERG)wereconducted atroomtemperature(22–25 °C)
using Aviva Bioscience SealChip16TM (USA). CHO cells stably expressing hERG K+
channels (Roche, USA) were stimulated by a voltage pattern to activate hERG
channels. The IKhERG currents were recorded under control conditions and after
compound application in the extracellular buffer (NaCl 150 mM, KCl 4 mM, CaCl2
1.2 mM, MgCl2 1 mM, HEPES 10 mM, pH 7.4 with NaOH, 300–310 mOsm). Offline
analysis of the peak tail current was performed using DataXpress2 software (MDS
Analytical Technologies, USA). Concentration-response curves were fitted by non-
linear regression analysis and the IC50 values were reported.
19. Brameld, K. A.; Kuhn, B.; Reuter, D. C.; Stahl, M. J. Chem. Inf. Model. 2008, 48, 1.
20. A single crystal was mounted in a loop and cooled to 100 K in a nitrogen
stream. Data were collected on a STOE Imaging Plate Diffraction System (STOE,
Darmstadt) with Mo-radiation (0.71 Å) and data processed with STOE IPDS-
software. The crystal structure was solved and refined with the program SHELXTL
(Bruker AXS, Karlsruhe).
21. Crystallographic data for the structure in this Letter have been deposited with
the Cambridge Crystallographic Data Centre as supplementary publication
numbers CCDC 721103. These data can be obtained, free of charge, on
application to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK [fax: +44(0)-
1223-336033 or e-mail: deposit@ccdc.cam.ac.uk].
27. Zhu, B.-Y.; Jia, Z. J.; Zhang, P.; Su, T.; Huang, W.; Goldman, E.; Tumas, D.;
Kadambi, V.; Eddy, P.; Sinha, U.; Scarborough, R. M.; Song, Y. Bioorg. Med. Chem.
Lett. 2006, 16, 5507.
22. Jamieson, C.; Moir, E. M.; Rankovic, Z.; Wishart, G. J. Med. Chem. 2006, 49, 5029.
23. Waring, M. J.; Johnstone, C. Bioorg. Med. Chem. Lett. 2007, 17, 1759.
28. Kansy, M.; Senner, F.; Gubernator, K. J. Med. Chem. 1998, 41, 1007.