Discovery of a series of potent arylthiadiazole H3 antagonists

Article abstract of DOI:10.1016/j.bmcl.2010.11.065

A series of 2-piperidinopiperidine-5-arylthiadiazoles was synthesized and subjected to a structure-activity relationship (SAR) investigation. The potency of this series was improved to the single digit nanomolar range. The key analogs were shown to be free of P450 issues, and they also maintained good ex vivo activity and brain penetration.

Full text of DOI:10.1016/j.bmcl.2010.11.065

Bioorganic & Medicinal Chemistry Letters 21 (2011) 861–864
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of a series of potent arylthiadiazole H3 antagonists
Dong Xiao a, , Anandan Palani a, Michael Sofolarides a, Ying Huang a, Robert Aslanian a,
⇑
Henry Vaccaro a, Liwu Hong a, Brian McKittrick a, Robert E. West Jr. b, Shirley M. Williams b,
Ren-Long Wu b, Joyce Hwa b, Christopher Sondey b, Jean Lachowicz b
a Chemical Research Department, Merck Research Laboratories, 2015 Galloping Hill Road, Kenilworth, NJ 07033, United States
b Cardiovascular/Metabolic Disease, Merck Research Laboratories, 2015 Galloping Hill Road, Kenilworth, NJ 07033, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of 2-piperidinopiperidine-5-arylthiadiazoles was synthesized and subjected to a structure–activ-
ity relationship (SAR) investigation. The potency of this series was improved to the single digit nanomolar
range. The key analogs were shown to be free of P450 issues, and they also maintained good ex vivo activ-
ity and brain penetration.
Received 3 September 2010
Revised 7 November 2010
Accepted 16 November 2010
Available online 21 November 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
H3 antagonists
SAR study
The biogenic amine histamine is a critical neurotransmitter
which mediates a wide range of neurological responses. The effect
of histamine is exerted through receptors that belong to the G pro-
tein-coupled receptors (GPCRs). There are four subtypes of the his-
tamine receptor (HR): H1, H2, H3 and H4.1 HR antagonists have
historically provided medical benefits, along with pharmaceutical
blockbusters, through modulation of the H1 and H2 receptors, in
the treatment of allergic conditions and gastric disorders, respec-
tively. Due to the success of H1 and H2 therapeutics, the potential
of H3 and H4 antagonism has attracted considerable research ef-
forts across the industry.2
Following the pharmacological identification of the H3 receptor
in 1983,3 and its cloning in 1999,4 significant strides were made to
identify novel, selective H3 receptor antagonists.5 Although the
exact therapeutic potential is unknown, H3 antagonists lead to in-
creased histamine levels at histaminergic neurons and subsequently
may be useful in treating obesity and diabetes, as well as other CNS
disorders such as cognitive disorders, Alzheimer’s and Parkinson’s
disease, schizophrenia and sleep disorders.6–8
Among the earliest H3 antagonists investigated were imidazole
based compounds;9 however cytochrome P450 (CYP) enzyme inhi-
bition posed significant drug–drug interaction issues. This finding
subsequently led to research in non-imidazole containing com-
pounds.10 In our effort to discover novel H3 antagonists we identi-
fied a high throughput screen (HTS) lead with prototype structure
1 containing a piperidine thiadiazole core. This series showed mod-
est in vitro potency, and it is likely that the pendent urea moiety
would hinder entry into the brain due to its multiple H bond donor
and acceptor characteristics. At the onset of lead optimization,
based on literature precedents11 and a H3 pharmacophore model,12
we envisioned the use of aryl groups to replace the urea, therefore
increasing brain penetration13(Scheme 1).
Introduction of an unsubstituted phenyl ring in place of the
urea moiety afforded compound 2. Compound 2 had comparable
activity to 1 thus confirming our initial structural hypothesis. Fur-
thermore, the caco2 permeability of 2 was improved to 104 nm/sec
from 10.5 nm/sec for 1, and the absorption was also improved to
good from moderate. These results were consistent to the empiri-
cal rules that reduction of H bond donors and acceptors would im-
prove CNS penetration.13 Encouraged by these findings, we
undertook a full optimization of the phenyl ring. Synthesis of this
phenyl series was relatively straightforward. Commercial com-
pound 3 was converted to the dibromo intermediate 4, which
was then displaced with piperidinopiperidine. The reaction can
be controlled to afford 5 exclusively, since the amino substitution
deactivated further addition. The bromothiadiazole 5 was subse-
quently coupled under Pd catalysis conditions with aryl boron or
zinc species to afford the desired products (Scheme 2).
Since H3 activity is centrally mediated, we focused on small
substitutions on the aryl group to minimize molecular weight.
Table 1 shows H3 in vitro potency14 for substitutions at the
ortho-, meta- and para-positions. Polar substituents such as OMe,
CN and acetyl groups afforded moderate improvement of potency
(Table 1).
Considering these data, especially that of meta-substituted
compounds, we speculated that an additional meta-substitution
would provide a pseudo-symmetric interaction with the receptor,
⇑
Corresponding author.
E-mail address: dong.xiao@merck.com (D. Xiao).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.11.065

862
D. Xiao et al. / Bioorg. Med. Chem. Lett. 21 (2011) 861–864
N
N
reduce H-bond
donor
N
N
HN
O
N
N
N
N
S
HN
S
reduce rotatable
bond
2: Ki (H3) = 80.1 12 nM
1: Ki (H3) = 48.6 3 nM
Scheme 1.
HN
N
N
N
N
N
N N
tBuNO2
Br
NH2
Br
R
Br
Br
N
N
DIEA
S
3
S
S
CuBr2
80%
THF, 80 ºC
4
5
65%
N
N
Suzuki/Negishi
conditions
N
N
S
6
40-75%
Scheme 2.
through microwave-accelerated Ir-catalyzed, direct borylation15
of the corresponding aromatics, which were then coupled with
bromothiadiazole 5. (Scheme 3).
Table 1
R
N N
Notable in vitro binding data are shown below. Aside from the
dichloro analog 15a, bis-m,m0-substitution did increase potency
compared to the corresponding mono-substituted analog. Further
confirming our speculation, hetero-m,m0-disubstitutions afforded
activity in the low double digit nanomolar range, as in entries Cl/
OMe, CN/OMe, and F/OMe (Table 2).
Additionally, following our initial heteroatom-containing lead,
we sought to replace the phenyl group with heteroaromatic rings.
Among the pyridyl substitutions, 2-pyridine showed the greatest
increase in potency and was pursued further. Results are shown
below (Table 3):
N
N
S
Entry
R
ortho-
Ki (H3) (nM) meta-
para-
80.1 12
2
7
8
H
80.1 12
80.1 12
OMe
CN
F
Cl
Me
40.0
N/A
36.0
69.0
36.0
8.4
8
31.9
42.6
6
1
40.1
36.4
2
7
9
2
1
4
2
62.0 11
39.0 15
161 69
163 23
10
11
12
49.0
60.0
1
8
COCH3
4.0 0.2
65.0
4
The 2-pyridyl analogs provided the most pronounced improve-
ment in potency, compared to the less active pyrimidine and
pyrazole analogs. Interestingly, the 3-methylpyridine and 3-
methoxypyridine afforded the most potent analogs in this series,
results consistent with the substituted phenyl series summarized
above.
which could further increase the binding potency. This series of
analogs was investigated and synthesized using commercially
available boronic acids and esters. Custom synthesis of non-com-
mercial 3,30-disubstituted boronic esters 14 was also developed
(Bpin)2
IrCl(cod)2
R1
R2
R1
R2
N
N
R1
Br
N
N
N
N
O
O
S
N
N
B
S
Pd(PPh3)2Cl2
K2CO3, 60 ºC
15-20%
tBu
N
tBu
R2
MW 130 ºC
13
14
15
Scheme 3.
Table 2
R1
N N
S
N
N
R2
Entry
R1/R2
15a
15b
OMe/OMe
15.0
15c
Cl/OMe
12.0
15d
15e
F/OMe
16.0
15f
15g
Cl/Cl
72.0
CN/OMe
12.0
F/F
50.0
CN/CN
23.0 1
Ki (H3) (nM)
9
1
1
3
2
1

D. Xiao et al. / Bioorg. Med. Chem. Lett. 21 (2011) 861–864
863
Table 3
N N
heteroaryl
N
N
S
Entry
16a
16b
16c
16d
16e
16f
N
N
NH
N
OMe
S
Heteroaryl
Ki (H3) (nM)
N
N
N
N
OMe
30.0 0.1
3.5
1
6.1
2
103.0 22
192.0 27
28
7
Table 4
N N
MeO
N Et
R
S
Entry
17a
17b
17c
17d
17e
N
N
N
N
N
N
N
R
N
N
Ki (H3) (nM)
40.1
2
161.8
2
182.3 15
652.6 79
600.9 96
To complete the SAR study of this thiadiazole series, we also
investigated the modifications at the piperidinopiperidine moiety.
For example, the piperazine analogs 17b–17d were aimed at mod-
ulating the basicity of the nitrogen. Additional analog such as 17e
was directed at reducing molecular weight. However, all these ana-
logs showed drastic reduction of H3 potency compared to the ori-
ginal analog 17a. Indeed, only the piperidinopiperidine as the right
hand side consistently afforded good potency (Table 4).
As mentioned previously, the earlier imidazole based H3 com-
pounds exhibited P450 inhibition issues. Therefore, in the current
series, the most potent compounds 15c and 16b were screened
against three CYP isoforms—3A4, 2D6 and 2C9. Fortunately, neither
a modest AUC in rat per oral dosing, a significant improvement
from the original lead 1 (Table 5).
In conclusion, a series of 2-piperidinopiperidine-5-arylthiadiaz-
oles as centrally active H3 antagonists was designed and was
subjected to a SAR investigation. First, we replaced the urea moiety
in 1 with a phenyl group, which significantly improved the absorp-
tion and permeability. We then systematically investigated the
effects of ortho-, meta- and para-substituted phenyl rings. The
use of polar groups such as OMe, CN, and COCH3 and particularly
with the m,m0-disubstitutions afforded improvement of potency
over phenyl analog. Aditionally, heteroaryl replacement of the
phenyl ring was investigated. 2-Pyridyl substitutions increased
potency significantly compared to the initial lead, while pyrimi-
dine and pyrazole offered less activity. Further substitutions,
3-methylpyridine and 3-methoxypyridine offered potency in the
single digit nanomolar range. The best compounds, 15c and 16b,
compound showed greater than 20 lM inhibition for 3A4, 2D6 and
2C9 under pre- or co-incubation conditions. Compound 15c was
particularly noteworthy due to its ex vivo potency.16 This
compound also showed high brain concentrations in mouse and
Table 5
N N
R
N
N
S
Entry
Ki (H3) (nM)
Rat AUC @10 mpk,
po (ng h/ml 0–6 h)
Mouse brain level@10 mpk,
po (ng/g @ 4 h)
Ex vivo @ 4 h
P450 inhib
3A4, 2D6, 2C9
O
NH
Cl
48.6
3
0
ND
ND
>20
lM
NH
1
12.0
3.5
1
1
379
9027
1527
90%
74%
>20
>20
l
M
M
OMe
15c
N
0
l
OMe
16b

864
D. Xiao et al. / Bioorg. Med. Chem. Lett. 21 (2011) 861–864
Gonzalez, M. I.; Heslop, T.; Hirst, W. D.; Jennings, C.; Jones, D. N. C.; Lacroix, L.
P.; Martyn, A.; Ociepka, S.; Ray, A.; Regan, C. M.; Roberts, J. C.; Schogger, J.;
Southam, E.; Stean, T. O.; Trail, B. K.; Upton, N.; Wadsworth, G.; Wald, J. A.;
White, T.; Witherington, J.; Woolley, M. L.; Worby, A.; Wilson, D. M. J.
Pharmacol. Exp. Ther. 2007, 321, 1032; (b) Cowart, M.; Faghih, R.; Curtis, M. P.;
Gfesser, G. A.; Bennani, Y. L.; Black, L. A.; Pan, L.; Marsh, K. C.; Sullivan, J. P.;
Esbenshade, T. A.; Fox, G. B.; Hancock, A. A. J. Med. Chem. 2005, 48, 38; (c)
Apodaca, R.; Dvorak, C. A.; Xiao, W.; Barbier, A. J.; Boggs, J. D.; Wilson, S. J.;
Lovenberg, T. W.; Carruthers, N. I. J. Med. Chem. 2003, 46, 3938.
were devoid of P450 inhibition issues, and showed good brain
concentrations and ex vivo potency. Further optimization of the
pharmacokinetic profile, as well as other biological parameters,
led to an orally active H3 antagonist in our mouse obesity model.
These results will be disclosed in due course.
Acknowledgments
12. (a) Dastmalchi, S.; Hamzeh-Mivehroud, M.; Ghafourian, T.; Hamzeiy, H. J. Mol.
Graphics Modell. 2008, 26, 834; (b) Narkhede, S. S.; Degani, M. S. QSAR Comb. Sci.
2007, 26, 744; (c) Axe, F. U.; Bembenek, S. D.; Szalma, S. J. Mol. Graphics Modell.
2006, 24, 456.
13. For recent reviews on CNS penetration, see: (a) Klon, A. E. Curr. Comput.-Aided
Drug Des. 2009, 5, 71; (b) Di, L.; Kerns, E. H.; Carter, G. T. Expert Opin. Drug
Discovery 2008, 3, 677; (c) Obrezanova, O.; Gola, J. M. R.; Champness, E. J.;
Segall, M. D. J. Comput. Aided Mol. Des. 2008, 22, 431; (d) Hitchcock, S. A.;
Pennington, L. D. J. Med. Chem. 2006, 49, 7559; (e) Clark, D. E. Annu. Rep. Med.
Chem. 2005, 40, 403.
The authors thank Dr. John J. Piwinski for supporting this re-
search program and Dr. Pauline Ting for helpful comments on this
letter.
References and notes
1. Parsons, M. E.; Ganellin, C. R. Br. J. Pharmacol. 2006, 147, S127.
2. Leurs, R.; Bakker, R. A.; Timmerman, H.; de Esch, I. J. P. Nat. Rev. Drug Disc. 2005,
4, 107.
14.
H
3 binding assay: memberanes (P2 pellet) from rHu H3–HEK cells (3
lg protein)
were incubated in 200 a-
ll
50 mM Tris–HCl, Ph 7.4, with 1 nM [3H] N-
methylhistamine (82 Ci/mmol) and compounds at concetrations equivalent to
half orders of magnitude over a five-order of magnitude range. Nonspecific
binding was determined in the presence of 10ꢀ5 M thioperamide. After 30 min
at 30 °C, assay mixture were filtered through 0.3% polyethylenimine soaked
GF/B glass fiber filters, which were rinsed thrice with buffer, dried,
impregnated with Meltilex wax scintillant and counted. Ki values were
determined from curves fit to the data using GraphPad Prism nonlinear,
least-squares, curve-fitting program. H3 binding Ki values were the average of
at least two independent determinations. The assay-to-assay variation was
generally 2-fold.
3. Arrang, J. M.; Garbarg, M.; Schwartz, J. C. Nature 1983, 302, 832.
4. Lovenberg, T. W.; Toland, B. L.; Wilson, S. J.; Jiang, X.; Pyati, J.; Hurvar, A.;
Jackson, M. R.; Erlander, M. G. Mol. Pharmacol. 1999, 55, 1101.
5. For recent reviews: (a) Raddatz, R.; Tao, M.; Hudkins, R. L. Curr. Top. Med. Chem.
2010, 10, 153; (b) Gemkow, M. J.; Davenport, A. J.; Harich, S.; Ellenbroek, B. A.;
Cesura, A.; Hallett, D. Drug Discovery Today 2009, 14, 509; (c) Stocking, E. M.;
Letavic, M. A. Curr. Top. Med. Chem. 2008, 8, 988; (d) Esbenshade, T. A.;
Browman, K. E.; Bitner, R. S.; Strakhova, M.; Cowart, M. D.; Brioni, J. D. Br. J.
Pharmacol. 2008, 154, 1166; (e) Hudkins, R. L.; Raddatz, R. Annu. Rep. Med.
Chem. 2007, 42, 49; (f) Berlin, M.; Boyce, C. W. Expert Opin. Ther. Pat. 2007, 17,
675; (g) Hancock, A. A. Biochem. Pharmacol. 2006, 71, 1103.
6. Stark, H.; Schlicker, E.; Schunack, W. Drugs Future 1996, 21, 507.
7. Kurose, Y.; Terashima, Y. Brain Res. 1999, 828, 115.
8. Masaki, T.; Chiba, S.; Yoshimichi, G.; Yasuda, T.; Noguchi, H.; Kakuma, T.;
Sadata, T.; Yoshimatsu, H. Endocrinology 2003, 144, 2741.
9. Stark, J.; Kathmann, M.; Schlicker, E.; Schunack, W.; Schlegel, B.; Wolfgang, S.
Mini-Rev. Med. Chem. 2004, 9, 965.
10. Zhang, M.; Ballard, M. E.; Pan, P.; Roberts, S.; Faghih, R.; Cowart, M. D.;
Esbenshade, T. A.; Fox, G. B.; Decker, M. W.; Hancock, A. A.; Rueter, L. E. Brain
Res. 2005, 1045, 142.
15. Harrisson, P.; Morris, J.; Marder, T. B.; Steel, P. G. Org. Lett. 2009, 11, 3586.
16. H3 ex vivo assay: Individual Imprinting Control Region mice (Taconics,
nomenclature IcrTacICR) cortexes were dissected and homogenized in ice
cold assay buffer, 50 mM Na2HPO4–KH2PO4 buffer (pH 6.8). Samples were then
frozen at ꢀ80 °C for at least 12 h. Protein concentration was determined by
BCA Protein Assay. A 30 min room temperature incubation was performed,
containing 140
methylhistamine and 10
l
g/assay homogenized cortex sample, 0.1% BSA, 1 nM 3H N-
M thioperamide for non-specific binding or assay
a-
l
buffer for total binding. Incubations were performed in quadruplet and
stopped by rapid filtration on a Brandel Harvester using Unifilter-96, GF/B
plates presoaked in 0.3% PEI for 30 min. The remining radioactivity was
measured on a Packard TopCount-NTX. Specific binding was calculated by
subtracting the non-specific binding from the total.
11. For examples: (a) Medhurst, A. D.; Atkins, A. R.; Beresford, I. J.;
Brackenborough, K.; Briggs, M. A.; Calver, A. R.; Cilia, J.; Cluderay, J. E.; Crook,
B.; Davis, J. B.; Davis, R. K.; Davis, R. P.; Dawson, L. A.; Foley, A. G.; Gartlon, J.;