Discovery of novel steroidal histamine H3 receptor antagonists/inverse agonists. Part 2. Versatile steroidal carboxamide derivatives

Article abstract of DOI:10.1016/j.bmcl.2019.126643

To further proceed with our previous work, novel steroid-based histamine H₃ receptor antagonists were identified and characterized. Using an ‘amine-to-amide’ modification strategy at position 17, in vitro and in vivo potent monoamino steroid derivatives were found during the lead optimization. Usage of the non-basic amide moiety resulted in beneficial effects both in activity and selectivity. The 15α-carboxamido derivative 10 was not only highly active at human and rat H₃ receptors, but also showed negligible activity at rat muscarinic receptors. Furthermore, it proved to be considerably stable in human and rat microsomes and showed significant in vivo potency in the pharmacodynamic rat dipsogenia test and in the water-labyrinth cognitive model. Based on all of these considerations, compound 10 was appointed to be a preclinical candidate.

Full text of DOI:10.1016/j.bmcl.2019.126643

Bioorganic&MedicinalChemistryLetters29(2019)126643
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of novel steroidal histamine H3 receptor antagonists/inverse
agonists. Part 2. Versatile steroidal carboxamide derivatives
István Ledneczki⁎, Zsolt Némethy, Pál Tapolcsányi, János Éles, István Greiner, Eszter Gábor,
Balázs Varga, Ottilia Balázs, Viktor Román, György Lévay, Sándor Mahó
Chemical Works of Gedeon Richter Plc, 30-32 Gyömrői Street, Budapest H-1103, Hungary
A R T I C L E I N F O
A B S T R A C T
Keywords:
To further proceed with our previous work, novel steroid-based histamine H3 receptor antagonists were iden-
tified and characterized. Using an ‘amine-to-amide’ modification strategy at position 17, in vitro and in vivo
potent monoamino steroid derivatives were found during the lead optimization. Usage of the non-basic amide
moiety resulted in beneficial effects both in activity and selectivity. The 15α-carboxamido derivative 10 was not
only highly active at human and rat H3 receptors, but also showed negligible activity at rat muscarinic receptors.
Furthermore, it proved to be considerably stable in human and rat microsomes and showed significant in vivo
potency in the pharmacodynamic rat dipsogenia test and in the water-labyrinth cognitive model. Based on all of
these considerations, compound 10 was appointed to be a preclinical candidate.
Steroid
Carboxamide
Histamine H3 receptor
Antagonist/inverse agonist
Dipsogenia test
Water-labyrinth test
Histamine exerts its biological actions through the modulation of
four G-protein-coupled receptors (H1-H4).1 Among them, H3 receptors
are deeply involved in the regulation of cognitive processes,2 wake-
fulness,3 feeding behavior,4 and body weight.5 Application of com-
pounds acting on H3 receptors has been proposed for the treatment of
Parkinson’s disease,6 schizophrenia,7 addiction,8 sleeping disorders,9 or
attention deficit hyperactivity disorder (ADHD).10 Therefore, sig-
nificant research activities were concentrated upon the discovery of
novel H3 receptor-acting ligands, resulting in a wide variety of pre-
clinical projects and clinical candidates.11 Until now, a single H3 an-
tagonist, pitolisant proved to be effective in narcolepsy and was regis-
tered.12 and launched under the trademark of Wakix®.
significantly reduce the known risk of phospholipidosis – a feature that
may be attributed to molecules with amphiphilic nature.
Starting from the former lead molecule 1 (although this compound
was already presented in our preceeding paper, its pharmacodynamic
result is disclosed here first, and it is used as a reference molecule for
the new derivatives), an ’amine-to-amide’ modification strategy was
attempted either in position 17 or at the end of the linker connected to
carbon 3. Results of the biological characterization of the representative
examples is summarized in Table 1.
In cases when the left-hand side of the molecule contained a cyclic
amide (2), the in vitro binding affinity to both rat (rH3) and human H3
(hH3) receptors remained high. However, the affinity to muscarinic
receptors also remained significant; it proved to be even higher than
that of compound 1. The usage of lactam in position 17 (3) resulted in
much weaker rH3 binding, but the affinity to muscarinic receptors also
decreased. Based on this promising finding, synthesis and character-
ization of another amide type – namely the 17β-carboxamido deriva-
tives – was attempted. In spite of the fact that compound 4 (a potent
binder to both rat and human H3 receptors) showed only a 25-fold
binding selectivity to rH3 over rat muscarinic receptors, it seemed clear
that the carboxamido group could be a viable alternative of the amino
(pyrrolidine) moiety in terms of potency.
Novel active and selective steroid-based H3 antagonists with po-
tency values in the subnanomolar range have already been identified by
Gedeon Richter Plc. (a detailed summary of this activity was already
published in a previous paper).13 During the optimization campaign,
significant affinity of the dibasic compounds towards rat muscarinic
receptors appeared as a major drawback. This disadvantageous feature
was decreased by reducing the basicity of the molecules.
The present paper aims at giving a summary on the continuation of
this approach by complete elimination of one basic center from the
structure. To achieve this goal, amides – instead of amines – were
synthesized. Usage of the non-basic (but slightly acidic) amide moiety
was expected to provide beneficial effects both on activity and se-
lectivity. Furthermore, application of monobasic compounds can
Another amide derivative – namely the 15β-acetamido compound 5
– showed weaker affinity to muscarinic receptors with strong affinity to
H3 receptors. Although the suboptimal metabolic stability in rat
⁎ Corresponding author. Tel.: +36 1 431 4369; fax: +36 1 260 5000.
E-mail address: ledneczki@richter.hu (I. Ledneczki).
https://doi.org/10.1016/j.bmcl.2019.126643
Received 22 May 2019; Received in revised form 23 August 2019; Accepted 27 August 2019
Availableonline28August2019
0960-894X/©2019ElsevierLtd.Allrightsreserved.

I. Ledneczki, et al.
Bioorganic&MedicinalChemistryLetters29(2019)126643
Table 1
Derivatives with ‘amine to amide’ modification strategy.
Compound
Steroid structure
hH3 Ki
(nM)
rH3 Ki
(nM)
rmAChR
µS FM
(h/r)
%
hERG inh.
IC50 (µM)
Dipsogenia inh. (p.o)
ED50 (mg/kg)
18
CH3 17
Ki (nM) or
%@3 µM
12
11
13
H
16
1
9
14
10
2
8
15
H
H
5
3
7
4
6
1
2
3
0.6
9
2.4
22
110
700
51%
–
7.5
–
a19%
12 (i.p.)
N
H3
C
H
N
H
H
O
77/85
–
–
N
H3C
H
O
N
H
H
O
–
135
–
–
N
H3
C
O
H
CH3
N
H
H
O
4
5
O
2
18
39
454
–
–
–
–
–
N
H3C
H
H
CH3
N
N
H
H
O
O
10
20%
85/46
H3C
H
CH3
H
H
O
HN
O
CH3
6
9
24
24%
62/66
4,1
2.26
a
Measured at 30 mg/kg (p.o.).
Scheme 1. General synthetic route to 15-carboxamides (6–20): (a) 1. (COCl)2, DMF, CH2Cl2, 2. amine2, CH2Cl2, Et3N; (b) H2, Pd/C, EtOH; (c) Br(CH2)nBr, KOH, H2O,
THF, reflux; (d) amine1, EtOH, reflux.
2

I. Ledneczki, et al.
Bioorganic&MedicinalChemistryLetters29(2019)126643
Table 2
15-Carboxamido derivatives.
X1
X2
H3C
H
O
H
H
m
O
R4
N
R1
R3
N
n
R2
ID
6
n
3
3
3
3
3
3
3
m
0
0
0
0
0
0
0
R1R2N
R3R4N
X1
X2
hH3
rH3 Ki
(nM)
rmAChR
µS FM
(h/r)
%
hERG inh.
IC50 (µM)
Dipsogenia inh.
ED50 (mg/kg)
Ki (nM)
9
Ki (nM) or %@3 µM
H
H
H
H
H
24
13
19
28
9
24%
27%
1708
7%
4700
–
62/66
90/76
48/50
96/96
99/99
–
4.1
4.1
–
2.3
7
H
9
–
8
H
5
–
9
OMe
O
9
11.4
17.5
–
2.5
0.47
–
10
11
12
3
O
21
230
–
O
–
–
–
–
13
14
3
3
0
0
O
O
530
–
–
–
–
–
–
–
–
1371
15*
16
17
18
19
20
3
3
4
2
3
3
0
1
0
0
0
0
O
43
4
–
–
–
–
–
OMe
O
H
20
6%
17%
4%
16%
–
59/80
–
–
234
260
30
34
–
–
–
O
–
–
–
O
9
75/85
–
80
–
0.6
–
O
44%
**Measured at @300 nM.
* In this single case the configuration of C15 was beta.
microsomes averted further development of this molecule, synthesis of
certain 15-carboxamido derivatives seemed to be a rational approach.
Indeed, 15α-carboxamido compound 6 exhibited one-digit nanomolar
affinity to hH3 receptors with good selectivity vs. rat muscarinic re-
ceptors. Furthermore, it showed a robust efficacy in the pharmacody-
namic dipsogenia test in vivo (ED50 = 2.26 mg/kg, p.o.) which is a
significant improvement compared to compound 1 (19% inhibition at
30 mg/kg, p.o. or ED50 = 12 mg/kg, i.p.). Unfortunately, two draw-
backs – namely the relatively low microsomal metabolic stability and
the robust inhibitory potential on the hERG channel activity – were also
identified. Still, usage of the carboxamido moiety in position 15α re-
sulted in a very promising compound that was selected for further
3

I. Ledneczki, et al.
Bioorganic&MedicinalChemistryLetters29(2019)126643
Table 3
Physico-chemical properties of compound 10.
20
Molecular weight
Melting
point
(°C)
Solubility
in DMSO
(mM)
Kinetic
solubility
(μM)
Measured
Log P
Measured
Log D
[
]
D
(c = 0.1%, CH2Cl2)
(pH 7.4)
466.7
122–124
> 50
250
4.2
1.7
+75,1°
one order of magnitude compared to the 15α-carboxamide compound
10. Thus, the best in vitro potency values were found among the 15α-
carboxamido-R-2-methylpyrrolidine derivatives.
5
4
3
2
1
0
+++
Having attempted to provide the structure with more flexibility by
lengthening the chain with one carbon at position 15 (16) resulted in
high H3 and low muscarinic affinity. However, metabolic stability was
lower than what was seen with the corresponding methoxy derivative
9. Furthermore, lengthening and shortening (n = 2, 4) the original
propoxy linker with one carbon in compounds 17 and 18 seemed to be
not beneficial for rH3 binding.
**
Various R3R4N groups may be tolerated regarding the binding ac-
tivity at hH3 receptors. The overall characteristics of pyrrolidino (19)
derivative was almost as good as that of compound 10. The cyclopro-
pylmethylamino (7) and cyclobutylmethylamino (8) derivatives were
also very potent in vitro. Only the N,N-diethylamino derivative (20)
showed weak affinity towards rH3 receptors.
Ctrl
Scop. (3 mg/kg)
10 (3 mg/kg) + Scop.
As we demonstrated above, the best derivatives were among the 17-
oxo and 17-metoxy analogues. They possessed more advantageous
properties than the 17-deoxo derivatives, however the presence of the
oxo and the hydroxy functions might raise the issue of possible binding
to various hormone receptors. As compound 10 showed ten times
higher affinity to hH3 receptors than that of compound 19, in-depth
characterization of 10 was carried out with special emphasis on its
physico-chemical properties and binding affinity to various steroid re-
ceptors. Physico-chemical properties of compound 10 showed pro-
mising data (molecular weight, solubility, lipophilicity) concerning its
drug-likeness (Table 3). Furthermore, detailed receptor profiling of 10
did not reveal any affinity to other steroid receptors (data not shown, as
a more detailed summary on the pharmacological properties of 10 is
planned to be published in a separate paper). As it is evident from the
data summarized in Table 2, compound 10 was not only highly active at
hH3 and rH3 receptors, but also showed negligible activity at rat mus-
carinic receptors (Ki = 4700 nM). Furthermore, 10 proved to be con-
siderably stable in human and rat microsomes (FM% = 99 in both
species) and acted in a very potent manner in the rat dipsogenia test
following p.o. administration with an ED50 value of 0.47 mg/kg.
Oral bioavailability following 3 or 10 mg/kg (iv. and po.) admin-
istration of compound 10 was investigated in fasted male Wistar rats.
Oral bioavailability of compound 10 was 62.3%, calculated from the
AUC0–24h values of the 3 mg/kg iv. and 3 mg/kg po. treatments. Drug
exposure (AUC0–24h) increased more than dose proportionally between
3 and 10 mg/kg both after iv. and po. administrations; the increase was
64% (iv.) and 39% (po.) higher than it would have been in case of
linearity. Clearance (Cl) and volume of distribution (Vd) decreased,
elimination half-life (Thalf) of the compound slightly increased with
dose. Brain penetration of compound 10 was investigated at 10 mg/kg
oral dose. Compound 10 showed rapid brain penetration with maximal
brain concentrations at 1 hr. Brain to plasma AUC0–inf ratio was 0.24,
thalf was 3.02–3.58.
1
2
Days
Fig. 1. Effect of compound 10 (3 mg/kg) on the scopolamine-induced amnesia
in the rat water labyrinth test (n = 9–10). +++p < 0.001 vs. control,
**p < 0.01 vs. scopolamine-treated group (ANOVA, followed by Duncan-test).
exploration.
Based on our previous findings, the central (estrane) core was left
unchanged, since this structural element had already proved to be the
best choice.9 In order to further explore the chemical space around the
steroidal core, modifications at position 17 (X1 and X2), usage of several
different amines at the end of the linkers (R1R2N and R3R4N) and dif-
ferent linkers’ lengths (m, n) were tested.
The general synthetic route is summarized in Scheme 1. The benzyl-
protected carboxylic acid starting materials were transformed to the
desired amides through the corresponding acyl chlorides by reacting
with various amines. After debenzylation and subsequent formation of
the aminopropoxy moiety, the desired carboxamido derivatives were
easily obtained (detailed chemical syntheses of the different starting
materials (21a–d) and a representative compound (10) are described in
the Supplementary Material).
Main characteristics of the synthesized derivatives are summarized
in Table 2. Compounds 6, 7 and 8 did not have any functional group on
carbon 17 (both X1 and X2 are hydrogens). However, this feature
seemed to be the reason of suboptimal metabolic stability and hERG
liability, as improved characteristics were found in the case of the 17-
methoxy (9) and 17-oxo (10) derivatives. Most probably, increasing the
molecules’ polarity resulted in beneficial effects not just on metabolic
stability and hERG liability (the hERG IC50 values reached [9] or even
exceeded [10] the 10 µM limit), but also on in vitro potency. In addition,
both compounds showed activity in the dipsogenia test in vivo. It was
particularly true for compound 10, which showed extraordinary in vivo
potency with an ED50 value of 0.47 mg/kg. To further proceed with the
optimization, effects of other R1R2N moieties were tested. Usage of the
S-2-methylpyrrolidne enantiomer resulted in a somewhat decreased
(however, still significant) affinity to hH3 receptors in case of com-
pound 11. However, the replacement with several fluorinated pyrroli-
dine and piperidine moieties (compounds 12, 13 and 14) resulted in
significant loss of binding affinity to rH3 receptors. In case of 15 (which
is a 15β-carboxamide compound), the in vitro hH3 Ki was increased with
To demonstrate beneficial effects on cognitive performance, com-
pound 10 was also tested in the water-labyrinth test.14 in rats. At the
dose of 3 mg/kg, compound 10 significantly improved the scopolamine-
induced impairment on Day 2 with significant (52.3%) restoring effect
(Fig. 1). Further models (including place and novel object recognition
paradigms as well as inhibitory avoidance test) will be discussed in
detail in a separate paper.
In summary, as a continuation of our previous work, novel, estrane
4

I. Ledneczki, et al.
Bioorganic&MedicinalChemistryLetters29(2019)126643
based steroidal H3 receptor antagonists with in vivo activity in the
dipsogenia and water labyrinth tests were identified. The best com-
pound 10 was appointed to be preclinical candidate and detailed
pharmacological characterization was initiated which is also about to
be published in a separate paper.
References
1. Haas H, Panula P. Nat Rev Neurosci. 2003;4:121.
2. Esbenshade TA, Fox GB, Cowart MD. Mol Interv. 2006;6:77.
3. Celanire S, Wijtmans M, Talaga P, Leurs R, de Esch IJ. Drug Discovery Today.
2005;10:1613.
4. Hancock AA, Bennani YL, Bush EN, et al. Eur J Pharmacol. 2004;487:183.
5. Schlicker E, Kathmann M. Handb Exp Pharmacol. 2017;241:277.
6. Arnulf I. Eur Neuropsychopharmacol. 2009;19:S204.
Acknowledgments
7. Fox GB, Esbenshade TA, Pan JB, et al. J Pharmacol Exp Ther. 2005;313:176.
8. Ellenbroek BA, Ghiabi B. Trends Neurosci. 2014;37:191.
The authors acknowledge Éva Schmidt, Mónika Vastag, László
Fodor, Gábor Wágner, Zoltán Béni, György Domány and Béla Kiss for
their contributions, assistance and support on this project.
9. Panula P, Chazot PL, Cowart M, et al. Pharmacol Rev. 2015;67:601.
10. Sander K, Kottke T, Stark H. Biol Pharm Bull. 2008;31:2163.
11. Lazewska D, Kiec-Kononowicz K. Expert Opin Ther Pat. 2014;24:89.
12. Dauvilliers Y, Bassetti C, Lammers GJ, et al. Lancet Neurol. 2013;12:1068.
13. Ledneczki I, Tapolcsányi P, Gábor E, et al. Biorg Med Chem Lett. 2017;27:4525.
14. Paróczai M, Kiss B, Kárpáti E. Brain Res Bull. 1998;45:475.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bmcl.2019.126643.
5