Design and Synthesis of Enantiomerically Pure Decahydroquinoxalines as Potent and Selective κ-Opioid Receptor Agonists with Anti-Inflammatory Activity in Vivo

Article abstract of DOI:10.1021/acs.jmedchem.6b01868

In order to develop novel κ agonists restricted to the periphery, a diastereo- and enantioselective synthesis of (4aR,5S,8aS)-configured decahydroquinoxalines 5-8 was developed. Physicochemical and pharmacological properties were fine-tuned by structural modifications in the arylacetamide and amine part of the pharmacophore as well as in the amine part outside the pharmacophore. The decahydroquinoxalines 5-8 show single-digit nanomolar to subnanomolar κ-opioid receptor affinity, full κ agonistic activity in the [³⁵S]GTPγS assay, and high selectivity over μ, δ, σ₁, and σ₂ receptors as well as the PCP binding site of the NMDA receptor. Several analogues were selective for the periphery. The anti-inflammatory activity of 5-8 after topical application was investigated in two mouse models of dermatitis. The methanesulfonamide 8a containing the (S)-configured hydroxypyrrolidine ring was identified as a potent (K_i = 0.63 nM) and highly selective κ agonist (EC₅₀ = 1.8 nM) selective for the periphery with dose-dependent anti-inflammatory activity in acute and chronic skin inflammation.

Full text of DOI:10.1021/acs.jmedchem.6b01868

Subscriber access provided by University of Florida | Smathers Libraries
Article
Design and synthesis of enantiomerically pure
decahydroquinoxalines as potent and selective #-opioid
receptor agonists with anti-inflammatory activity in vivo
Michael Soeberdt, Peter Molenveld, Roy P. M. Storcken, Renaud Bouzanne des Mazery, Geert Jan Sterk,
Reshma Autar, Marjon G. Bolster, Clemens Wagner, Sebastianus N. H. Aerts, Frank R. van Holst, Anita
Wegert, Giovanni Tangherlini, Bastian Frehland, Dirk Schepmann, Dieter Metze, Tobias Lotts, Ulrich Knie,
Kun-Yuan Lin, Tai-Yu Huang, Chih-Ching Lai, Sonja Ständer, Bernhard Wünsch, and Christoph Abels
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 20 Feb 2017
Downloaded from http://pubs.acs.org on February 21, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 113
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 113
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Design and synthesis of enantiomerically pure
decahydroquinoxalines as potent and selective κꢀ
opioid receptor agonists with antiꢀinflammatory
activity in vivo
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a,
b
b
b
Michael Soeberdt, * Peter Molenveld, Roy P. M. Storcken, Renaud Bouzanne des Mazery,
b
b
b
b
Geert Jan Sterk, Reshma Autar, Marjon G. Bolster, Clemens Wagner, Sebastianus N. H.
b
b
b
c,d
c
Aerts, Frank R. van Holst, Anita Wegert, Giovanni Tangherlini, Bastian Frehland, Dirk
c,d
e
e,f
a
g
g
Schepmann, Dieter Metze, Tobias Lotts, Ulrich Knie, KunꢀYuan Lin, TaiꢀYu Huang, Chihꢀ
g
e,f
c,d
a
Ching Lai, Sonja Ständer, Bernhard Wünsch, Christoph Abels
a
Dr. August Wolff GmbH & Co. KG Arzneimittel, Sudbrackstraße 56, Dꢀ33611 Bielefeld,
Germany
b
c
Mercachem, Kerkenbos 1013, NLꢀ6546 BB Nijmegen, The Netherlands
Institut für Pharmazeutische und Medizinische Chemie der Universität Münster, Correnstraße
4
8, Dꢀ48149 Münster, Germany
d
e
CellsꢀinꢀMotion Cluster of Excellence (EXC 1003 – CiM), Universität Münster, Dꢀ48149
Münster, Germany
Klinik für Hautkrankheiten, Universitätsklinikium Münster, VonꢀEsmarchꢀStraße 58, Dꢀ
4
8149 Münster, Germany
ACS Paragon Plus Environment

Page 3 of 113
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
f
Kompetenzzentrum chronischer Pruritus (KCP), Universitätsklinikium Münster, Vonꢀ
EsmarchꢀStraße 58, Dꢀ48149 Münster, Germany
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
g
Eurofins Panlabs Taiwan, Ltd., 158 LiꢀTeh Road, Peitou, Taipei, Taiwan 11259
ABSTRACT
In order to develop novel κ agonists restricted to the periphery, a diastereoꢀ and enantioselective
synthesis of (4aR,5S,8aS)ꢀconfigured decahydroquinoxalines 5ꢀ8 was developed.
Physicochemical and pharmacological properties were fineꢀtuned by structural modifications in
the arylacetamide and amine part of the pharmacophore as well as in the amine part outside the
pharmacophore. The decahydroquinoxalines 5ꢀ8 show singleꢀdigit nanomolar to subnanomolar
3
5
κꢀopioid receptor affinity, full κ agonistic activity in the [ S]GTPγS assay, and high selectivity
over ꢁ, δ, σ , and σ receptors as well as the PCP binding site of the NMDA receptor. Several
1
2
analogs were selective for the periphery. The antiꢀinflammatory activity of 5ꢀ8 after topical
application was investigated in two mouse models of dermatitis. The methanesulfonamide 8a
containing the (S)ꢀconfigured hydroxypyrrolidine ring was identified as potent (K = 0.63 nM)
i
and highly selective, κ agonist (EC = 1.8 nM) selective for the periphery with doseꢀdependent
5
0
antiꢀinflammatory activity in acute and chronic skin inflammation.
INTRODUCTION
Research on κꢀopioid receptor agonists has been focusing on analgesic activity for a long period.
In contrast to µꢀopioid receptor agonists κ agonists do not produce adverse and dangerous side
effects like euphoria, respiratory depression, constipation, tolerance and physical dependence.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 113
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
However, centrally active κꢀopioid receptor agonists are not devoid of side effects, as they cause
1
,2
dysphoria, sedation and diuresis. Therefore, research has been focussing on κ agonists with
3
peripheral restriction. One of the first peripherally restricted κ agonists in clinical development
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
was asimadoline. Although it failed to show antinociceptive effects at doses, which do not induce
central side effects, efficacy in studies with patients suffering from dyspepsia and from irritable
4
bowel syndrome (IBS) was shown. Recently, a phase III clinical study in patients with diarrheaꢀ
5
predominant irritable bowel syndrome (DꢀIBS) was conducted and currently a phase II study for
6
patients with pruritus associated with atopic dermatitis is recruiting patients. Another peripheral
κ agonist in clinical development is CR845. The peptide is currently being developed in several
7
pain indications and in patients with uremic pruritus. A phase III study in patients with post
abdominal surgery pain is currently recruiting.
The only approved κ agonist to date is nalfurafine (1, Figure 1). It is marketed in Japan for the
8
treatment of uremic pruritus. Its antipruritic activity is suggested to be mediated by central κꢀ
9
opioid receptors.
Along with their analgesic and antipruritic activity, κ agonists are also known to have antiꢀ
1
0
inflammatory activity. Uꢀ50,488 (2) is described to inhibit plasma extravasation, to exhibit
1
1,12
antiꢀinflammatory effects in animal models of adjuvant arthritis,
formalinꢀinduced
1
3
14
temporomandibular joint inflammation, intestinal inflammation, skin inflammation caused by
1
5
reactive oxygen species, to reduce neutrophil accumulation and TNFꢀαꢀinduction by ischemia
16
reperfusion as well as TLR4 and NFꢀκB expression and to reduce myeloperoxidase (MPO)
1
7
levels and myocardial TNFꢀα production. In diabetic rats it was shown to improve vascular
function and to reduce levels of sICAMꢀ1, ILꢀ6 and ILꢀ8. Moreover, Uꢀ50,488 also inhibited the
1
8
translocation of NFꢀκB p65. It was shown in several of these studies that the antiꢀinflammatory
ACS Paragon Plus Environment

Page 5 of 113
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
13,16–18
effects can be partially or fully blocked by the selective κ antagonist norꢀBNI
or by the
10
15
nonselective antagonists naloxone or naltrexone. Following treatment with proꢀinflammatory
1
9–21
stimuli κꢀopioid receptors are upꢀregulated in vitro and in vivo.
Treatment of macrophages
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
19
with IFNꢀγ upregulated κꢀopioid receptors. An upregulation of peripheral κ receptors in TNBSꢀ
2
0
induced colonic inflammation was suggested. In a model of croton oilꢀinduced intestinal
inflammation an upꢀregulation of κ receptors on protein level in jejunum and mucosa was
2
1
observed following administration of two doses of croton oil.
2
2
Cutaneous cells like keratinocytes and fibroblasts express κꢀopioid receptors and κꢀopioid
receptors and its ligands, dynorphin A (1ꢀ17 and 1ꢀ8) were downꢀregulated in the skin of patients
2
3
with atopic dermatitis, a skin condition associated with extreme itch. Levels of dynorphin A
were restored in patients treated with UV light therapy and this was paralleled by a decrease in
skin symptoms. These studies suggest that activation of κꢀopioid receptors may result in
decreased itch and skin inflammation. Recently we were able to show that potent and selective κ
2
+
agonists are able to reduce the release of Ca ions into the cytoplasm of HaCaT cells, a human
24
immortalized keratinocyte cell line, following stimulation with ATP. More recently we found
that topical nalfurafine exhibited antiꢀinflammatory effects in addition to its antiꢀpruritic activity
2
5
in a murine model of atopic dermatitis.
During the last years a broad range of novel κ agonists belonging to various different structural
2
6–35
classes have been reported in the literature.
Recently, we have described
decahydroquinoxaline κꢀopioid receptor agonists representing conformationally restricted
3
6–39
analogs of κ agonists Uꢀ50,488 (2) and GRꢀ89,696 (3).
The relative and absolute
configuration of decahydroquinoxalineꢀbased κ ligands has a strong impact on their κ affinity.
3
8
Whereas cisꢀcisꢀconfigured decahydroquinoxaline analogs show very low κ affinity, the transꢀ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 113
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
transꢀconfigured diasteromer 4 represents a very potent κ agonist (K = 9.7 nM). However, the
i
3
9
highest κ affinity was found for the cisꢀtransꢀconfigured diastereomer. The κ affinity resides
almost exclusively in the (4aR,5S,8aS)ꢀconfigured enantiomer 5a (K = 0.25 nM) with an
i
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
eudismic ratio of 400. Moreover, 5a is a full agonist at κ receptors.
Figure 1. Potent κꢀopioid receptor agonists 1ꢀ5a reported in literature and cisꢀtransꢀ
decahydroquinoxalines 5ꢀ8 prepared and tested for κꢀopioid receptor affinity and agonistic
activity described in this study.
ACS Paragon Plus Environment

Page 7 of 113
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Herein we report the diastereoꢀ and enantioselective synthesis of 5a and 23 analogs by late stage
diversification of a versatile chiral building block. The aim was a straightforward optimization of
5a in a hitꢀtoꢀlead program to generate structureꢀactivity data for this novel class of κ agonists as
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
well as to obtain peripherally restricted compounds. This should be accomplished by
introduction of polar substituents, e.g. a hydroxy group on the pyrrolidine ring of 5a (5b-i, 8).
Moreover, the nitrogen atom bearing the methoxycarbonyl moiety in 5a opens the opportunity to
incorporate additional polar substituents (7, 8). In addition to their synthesis and in vitro profiling
at opioid and related receptors, pharmacokinetic data and in vivo pharmacodynamic properties of
this new class of κꢀopioid receptor agonists are discussed.
RESULTS AND DISCUSSION
Synthesis
We have previously reported the synthesis of all four diastereoisomers of the
decahydroquinoxaline core and the separation of the enantiomers of the racemic mixtures racꢀ4
39
and racꢀ5a. This synthetic route towards the enantiomerically pure decahydroquinoxaline 5a
was neither diastereoselective nor enantioselective. Due to time and material consuming
separation of diastereomers and enantiomers, the yield of 5a was rather low. Obviously, the
reported route was not suitable for the production of large amounts of material required for a
broad hitꢀtoꢀlead campaign. Therefore, an alternative approach for the stereoselective synthesis
of 5a and analogs had to be developed. It was anticipated that a versatile cisꢀ
decahydroquinoxaline building block would be best accessed by a diastereoselective synꢀ
hydrogenation of a 5,6,7,8ꢀtetrahydroquinoxaline, similar to our previous synthesis. In order to
direct enantioselectivity we opted for the prochiral starting material 12 (Scheme 1) because
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 113
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
catalytic asymmetric reduction of ketones is a wellꢀrecognized approach to access chiral alcohol
4
0
building blocks. Racemic alcohol 11, the precursor of ketone 12, was obtained in 48% yield by
a trifluoroacetic anhydride mediated Boekelheide rearrangement of Nꢀoxide 10, which had been
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
41
prepared by mꢀchloroperbenzoic acid oxidation of tetrahydroquinoxaline 9. Alcohol 11 was
4
2
oxidized to ketone 12 by DessꢀMartin periodinane in wet dichloromethane (71%).
Enantioselective reduction of the ketone 12 to (R)ꢀalcohol 13 was initially attempted by the CBS
4
3,44
reduction
using (S)ꢀconfiugred methyl oxazaborolidine and diborane. However, this
reduction led only to moderate yields and low enantiomeric excess of 13. The (R)ꢀconfigured
alcohol 13 was obtained in excellent yield (93%) and ee (>98%) by a variation on the Noyori
4
5,46
asymmetric transfer hydrogenation
cymene)ruthenium(II)dimer
using the catalyst ligand combination dichloro(pꢀ
and
(1R,2R)ꢀ(+)ꢀNꢀ(4ꢀtoluenesulfonyl)ꢀ1,2ꢀ
47,48
diphenylethylenediamine, and triethylammonium formate as hydrogen donor.
ACS Paragon Plus Environment

Page 9 of 113
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme 1. Enantioselective synthesis of cisꢀcis–configured sulfamidate building blocks 18a and
18b and modification of the northern part (amino moiety, phenylacetyl moiety) to afford cisꢀ
a
transꢀdecahydroquinoxalines 5 and 6.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
OH
O
OH
N
N
N
N
N
N
N
N
N
a
b
g
c
d
(R)
e
N
9
10
11
12
13
O
S
OTBS
OH
O
H
N
H
N
OTBS
O
OTBS
H
N
f
h
i
N
N
j
N
N
N
N
N
H
AcOH
RO
O
RO
O
RO
O
14
15
16a R = Me
6b R = tBu
17a R = Me
17b R = tBu
18a R = Me
18b R = tBu
1
R3
l
Cl
R2
R2
R4
Cl
N
H
O
O
N
N
N
H
N
N
k
k
N
N
N
N
N
N
O
O
1
R O
O
1
R O
O
O
O
1
2
1
2
3
4
6
20
1
1
1
1
1
1
9a R = Me, R = H
5a R = Me, R = H, R = R = Cl
1
2
1
2
3
4
9b R = Me, R = (R)-OH
5b R = Me, R = (S)-OH, R = R = Cl
m
1
2
1
2
3
4
9c R = Me, R = (S)-F
5c R = Me, R = (S)-OH, R = F, R = Cl
1
2
1
2
3
4
9d R = Me, R = (S)-OH
5d R = Me, R = (S)-OH, R = Cl, R = F
1
2
1
2
3
4
9e R = tBu, R = H
5e R = Me, R = (S)-OH, R = R = F
1
2
1
2
3
4
9f R = tBu, R = (S)-OH
5f R = Me, R = (S)-OH, R = Cl, R = H
1
2
3
4
5g R = Me, R = (S)-OH, R = H, R = Cl
1
2
3
4
5h R = Me, R = (S)-OH, R = CF , R = Cl
3
1
2
3
4
5
5
5
i R = Me, R = (R)-OH, R = R = Cl
1
2
3
4
j R = Me, R = (S)-F, R = F, R = Cl
1
2
3
4
k R = tBu, R = H, R = R = Cl
a
Reagents and conditions: (a) mꢀCPBA, CH Cl , rt, 22 h, 86%; (b) 1: trifluoroacetic anhydride,
2
2
CH Cl , 22 h; 2: aq. LiOH, CH Cl , rt, 1 h, 48%; (c) DessꢀMartin periodinane, water, CH Cl , rt,
2
2
2
2
2
2
2
0 h, 71%; (d) dichloro(pꢀcymene)ruthenium(II)dimer, (1R,2R)ꢀ(+)ꢀNꢀ(4ꢀtoluenesulfonyl)ꢀ1,2ꢀ
diphenylethylenediamine, triethylammonium formate, DMF, rt, 5 h, 93%; (e) TBSOTf, 2,6ꢀ
o
lutidine, CH Cl , rt, 1 h, 98%; (f) H (5 bar), PtO , MeOH, AcOH, 50 C, 5 h, 99%; (g) methyl
2
2
2
2
chloroformate or Boc O, Et N, CH Cl , rt, 1 h, 91% (16a), 95% (16b); (h) NH F, MeOH, reflux,
2
3
2
2
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 113
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
0 h/2 d, 84% (17a), 89% (17b); (i) SO Cl , Et N, CH Cl , rt, 16 h, 66% (18a), 64% (18b); (j)
2 2 3 2 2
o
pyrrolidine or 3ꢀhydroxypyrrolidine, acetonitrile, 82 C, 16ꢀ22 h, 76% (19a), 93% (19d) 85%
o
(19e); 3ꢀhydroxypyrrolidine, K CO , N,Nꢀdimethylformamide, 70 C, 24 h, 72% (19b), 88%
2
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(19f); (k) substituted phenylacetyl chloride, CH Cl , rt, 0.5ꢀ2 h, 82% (5a), 39 % (5b), 62% (5c),
2
2
36% (5d), 52% (5e), 38% (5f), 96% (5g), 52% (5h), 39% (5i), 67% (5j), 66% (5k), 71% (6); (l)
o
2
M HNMe in THF, 70 C, 3 d (78%); (m) diethylaminosulfur trifluoride, CH Cl , rt, 4 d (42%).
2
2
2
In order to suppress cleavage of the benzylic CꢀO bond during hydrogenation of the pyrazine
ring and to achieve regioselective functionalization of the southern secondary amine in 1ꢀ
position, alcohol 13 was protected as TBS ether 14. Catalytic hydrogenation of
tetrahydroquinoxaline 14 with platinum(II)oxide proceeded via the objected synꢀhydrogen
addition and provided the cisꢀcisꢀdecahydroquinoxaline 15 in almost quantitative yield after
optimization of the reaction conditions. Regioselective carbamate formation (16a,b) and
deprotection of the silyl ether afforded amino alcohols 17a and 17b in high yields. Attempts to
protect the amino group in 17a with Boc or Cbz groups were not successful, but the amino group
could be benzylated or acetylated. However, subsequent transformation of the hydroxy group
into an amine or amine equivalent by, respectively, nucleophilic substitution of a leaving group,
Mitsunobu reaction, or reductive amination after oxidation to the ketone failed. Finally, an
elegant strategy to simultaneously protect the amino group and activate the hydroxy group by
means of a cyclic sulfamidate was developed. The chiral 1,2,3ꢀoxathiazolidine 2,2ꢀdioxide 18a,
was obtained by reaction of methyl carbamate 17a with sulfuryl chloride. The sulfamidate 18a
and in particular the Boc protected derivative 18b represent multifunctional chiral building
blocks allowing decoration at three orthogonal positions. Nucleophilic substitution by small
ACS Paragon Plus Environment

Page 11 of 113
Journal of Medicinal Chemistry
1
2
3
secondary amines, such as pyrrolidine, hydroxypyrrolidine and dimethylamine, proceeded with
inversion of configuration and afforded in one step the deprotected amino compounds 19 and 20
in the objected cisꢀtransꢀconfiguration. Subsequently, the remaining secondary amine in 4ꢀ
position was acylated with phenylacetyl chlorides to provide the final compounds 5a-j and 6
(Scheme 1) and Boc building blocks 5k and 21 (Scheme 2). In case of the hydroxypyrrolidine
derivatives some Oꢀacylation was observed, but the side products could be separated by flash
column chromatography. The fluoropyrrolidine derivative 19c was prepared by treatment of the
secondary alcohol 19b with DAST (diethylaminosulfur trifluoride), which occurred under
inversion of configuration.
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
9
In contrast to the previously reported route, the enantioselective synthesis toward the cisꢀtransꢀ
decahydroquinoxaline 5a was robust and scalable, which is illustrated by the experimental
procedures where the multigram preparation of the key intermediates 18a (100 g), 19a, 19d (20
g) and final compound 5a (30 g) is described.
The Boc building block 5k was deprotected with trifluoroacetic acid (Scheme 2) and the
obtained amine 7a was functionalized by alkylation, acylation or sulfonylation following
standard methods, to provide the final compounds 7b-h. Boc deprotection of 21 followed by Nꢀ
sulfonylation was not chemoselective and provided mixtures of Nꢀ and Oꢀfunctionalized
compounds that were difficult to separate. Therefore, the hydroxy group in Boc building blocks
21 was protected as acetate (22) before the sulfonylation of the secondary amino moiety in 23
was performed, to yield after deprotection of 24 the final sulfonamides 8.
Scheme 2. Modification of the southern part of (hydroxy)pyrrolidine substituted cisꢀtransꢀ
a
decahydroquinoxalines to afford the final compounds 7 and 8.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 113
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Reagents and conditions: (a) CF CO H, CH Cl , 2 h, rt, 83%; (b) methyl 2ꢀbromoacetate, ethyl
3
2
2
2
2ꢀbromoacetate, or isopropyl 2ꢀbromoacetate, K CO , MeCN, rt, 16 h, 90% (7b), 58% (7c), 51%
2
3
(7d); 2ꢀ(methylamino)ꢀ2ꢀoxoacetic acid, EDCI, HOAt, CH Cl , rt, 16 h, 85% (7e); isopropyl
2 2
isocyanate, CH Cl , 50% (7f); methanesulfonyl chloride, Et N, CH Cl , rt, 2 h, 64% (7g);
2
2
3
2
2
o
dimethylsulfamoyl chloride, Et N, MeCN, 50 C, 2 d, 57% (7h); (c) 2ꢀ(3,4ꢀdichlorophenyl)acetyl
3
ACS Paragon Plus Environment

Page 13 of 113
Journal of Medicinal Chemistry
1
2
3
chloride, 2ꢀ(3ꢀchloroꢀ4ꢀfluorophenyl)acetyl chloride, 2ꢀ(4ꢀchloroꢀ3ꢀfluorophenyl)acetyl chloride,
or 2ꢀ(4ꢀchlorophenyl)acetyl chloride, CH Cl , rt, 2 h, 88% (21a), 70% (21b), 80% (21c), 62%
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
2
(21d); (d) acetic anhydride, pyridine, rt, 16 h, 92% (22a), 92% (22b), 93% (22c), 95% (22d); (e)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CF CO H, CH Cl , 2 h, rt, 92% (23a), 97% (23b), 92% (23c), 95% (23d); (f) methanesulfonyl
3
2
2
2
chloride, Et N, DMAP, CH Cl , rt, 2 h, 88% (24a), 77% (24b), 63% (24c); or 1,2ꢀoxazolꢀ3ꢀ
3
2
2
o
ylmethanesulfonyl chloride, pyridine, 60 C, 3 d, 49% (24d), 57% (24e); (g) NaOMe, MeOH, rt,
.5 h, 58% (8a), 69% (8b), 81% (8c), 55% (8d), 55% (8e).
1
Pharmacological Evaluation
ꢀOpioid receptor affinity
κ
Competitive radioligand receptor binding studies were used to determine the κꢀopioid receptor
affinity of decahydroquinoxalines 5ꢀ8. Guinea pig brain membrane preparations were employed
as receptor material and tritiumꢀlabeled Uꢀ69,593 as radioligand. A large excess of nonꢀlabeled
Uꢀ69,593 (10 µM) was used to determine the nonꢀspecific binding of the radioligand. The κ
affinity of the test compounds 5ꢀ8 together with the κ affinity of the reference compounds Uꢀ
69,593 and naloxone are summarized in Table 1.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 113
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 1: Affinities of decahydroquinoxalines towards κꢀopioid and related receptors.
affinity K ± SEM (nM)
κ agonistic activity
i
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
NMDA
EC50
(nM)
% at 1
compd.
κ
µ
δ
σ1
σ2
a
(PCP)
µM
a
a
a
a
b
5a
0.25 ± 0.08 43 ± 9.2 58 ± 8.4
0%
0%
0%
0%
252
12%
35%
2%
2.0
0.9
2.2
3.9
11
107
113
110
93
a
a
b
5b
0.45 ± 0.23
0.82 ± 0.35
1.8 ± 0.67
1.2 ± 0.56
0.65 ± 0.13
1.7 ± 0.40
0.27 ± 0.09
1.3 ± 0.74
0.56 ± 0.52
0.16 ± 0.07
0.35 ± 0.24
0.70 ± 0.23
0.43 ± 0.19
1.4 ± 0.39
1.5 ± 0.20
2.5 ± 1.4
184
1300
551
101
61 ± 16
250
21%
13%
11%
11%
b
b
b
5c
a
5d
0%
a
a
a
a
5e
4%
623
34%
21%
10%
11%
108
103
102
100
109
92
a
a
5f
365
512
612
451
7%
14
a
a
b
5g
1120
394
0%
0%
5%
4%
0%
15
a
a
b
5h
30%
0.3
4.7
0.4
1.0
3.7
1.7
3.2
5.0
8.1
4.5
a
a
a
b
5i
5%
198
0%
13%
55%
a
a
a
b
5j
0%
74 ± 26
7.9 ± 2.7
915
0%
0%
0%
0%
0%
a
a
a
b
6
0%
4%
0%
96
a
a
b
7a
217
1230
830
6%
0%
89
a
a
b
7b
1230
907
0%
0%
13%
111
106
94
a
a
b
7c
0%
12%
26%
0%
1%
8%
0%
a
a
a
a
b
7d
7%
0%
0%
a
a
a
b
b
7e
12%
681
237
0%
4%
0%
104
100
a
a
a
7f
0%
0%
ACS Paragon Plus Environment

Page 15 of 113
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
a
a
b
7
g
0.31 ± 0.19
91
1090
0%
0%
0%
3%
5%
0%
1.5
2.2
1.8
5.9
8.0
6.9
1.4
5.3
ꢀ
106
105
106
99
a
a
a
7h
0.95 ± 0.72 21 ± 8.1 11 ± 0.5
0%
2500
a
b
8a
0.63 ± 0.14
1.7 ± 0.40
2.1 ± 0.23
0.98 ± 0.54
0.82 ± 0.10
0.97 ± 0.40
336
367
575
394
880
1380
ꢀ
15%
0%
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
a
b
8b
0%
3%
6%
a
a
a
b
8c
308
12%
24%
0%
16%
105
103
105
100
ꢀ
a
a
b
8d
0%
6%
ꢀ
0%
0%
a
a
a
b
8
e
0%
7%
13%
Uꢀ69,593
naloxone
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
7.3 ± 0.40 2.3 ± 1.1
103
morphine
ꢀ
ꢀ
ꢀ
ꢀ
5.2 ± 1.6 2.0 ± 0.3
ꢀ
ꢀ
(+)ꢀpentazocine
haloperidol
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
5.4 ± 0.50
6.6 ± 0.9 78 ± 2.3
ꢀ
ꢀ
ꢀ
ꢀ
(
+)ꢀMKꢀ801
ꢀ
ꢀ
2.8 ± 1.8
a
Inhibition of the radioligand binding (%) at a concentration of 1 µM of the test compound.
Inhibition of the radioligand binding (%) at a concentration of 10 µM of the test compound.
b
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 113
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The methyl carbamate 5a representing the starting point of this study shows a subnanomolar κ
affinity with a K value of 0.25 nM. The κ affinity of the prepared analogs 5ꢀ8 with variations of
i
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
the pyrrolidine ring, the substituents at the phenyl ring and at Nꢀ1 did not alter the κ affinity
dramatically. The K values are between 0.16 nM (6) and 2.5 nM (7f). In particular the tolerance
i
of diverse substituents at Nꢀ1 supports the hypothesis that substituents at that position are
oriented outside the binding pocket.
Introduction of a hydroxy moiety into the pyrrolidine ring results in more polar compounds. The
stereoisomer 5b with (S)ꢀconfiguration at Cꢀ3 of the pyrrolidine ring shows a slightly higher κ
affinity (K = 0.45 nM) than its diastereomer 5i with (R)ꢀconfiguration at this position (K = 1.3
i
i
nM). Therefore, (S)ꢀconfigured hydroxypyrrolidine is the preferred substituent in this class of
polar κ agonists. The highest κ affinity was found for 6 (K = 0.16 nM) bearing a dimethylamino
i
moiety instead of the pyrrolidine ring. A K value higher than 2 nM was found only for two
i
compounds, 7f (K = 2.5 nM) and 8c (K = 2.1 nM).
i
i
κ
agonistic activity and intrinsic activity
The κ agonistic activity and the intrinsic activity of the decahydroquinoxalines 5ꢀ8 were
3
5
determined in the [ S]GTPγS assay (Table 1). In this assay all compounds behaved as very
potent κ agonists. With exception of 5e, 5f, and 5g, the EC values of the compounds are below
5
0
1
0 nM. In 5f and 5g one of the Clꢀatoms of the dichlorophenylacetyl derivative 5b (EC = 0.9
50
nM) is missing, in 5e both Clꢀatoms of 5b are replaced by Fꢀatoms. On the other hand the
decahydroquinoxalines 5b, 5h and 5j show subnanomolar activity. The dichlorophenylacetyl
derivative 5b bearing the (S)ꢀconfigured hydroxypyrrolidine ring has an EC value of 0.9 nM.
50
ACS Paragon Plus Environment

Page 17 of 113
Journal of Medicinal Chemistry
1
2
3
Replacement of the 4ꢀCl substituent by a CF group resulted in 5h with an EC value of 0.3 nM.
3
50
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
A similar EC value was found for the fluoropyrrolidinyl substituted derivative 5j bearing a Clꢀ
50
and Fꢀatom at the phenyl ring. It can be concluded that two electronegative Clꢀatoms at the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
phenyl ring are required for high κꢀagonistic activity. It is possible to replace one of the Clꢀatoms
by a Fꢀatom or a CF group, but replacing one of the Clꢀatoms by a proton or replacement of
3
both Clꢀatoms by Fꢀatoms resulted in decreased agonistic activity.
At a test compound concentration of 1 µM all decahydroquinoxalines 5ꢀ8 showed >90%
activation of the κꢀopioid receptor, i.e. all compounds of this series behave as full agonists at the
κꢀopioid receptor.
Selectivity over µꢀ and δꢀopioid receptors
Next the selectivity of the κ agonists 5ꢀ8 over related µꢀopioid and δꢀopioid receptors was
investigated in competitive radioligand receptor binding studies. In these assays the same
receptor preparation from guinea pig brains as in the κ assay was used and the radioligands
3
3
[
H]DAMGO and [ H]DPDPE were employed to selectively label µ and δ receptors. At first the
decahydroquinoxalines were tested at a concentration of 1 µM. The complete competition curves
were recorded only when the radioligand binding was more than 50% reduced. Otherwise the
reduction of bound radioactivity is given in % at 1 µM (Table 1).
In general the µꢀopioid receptor affinity of all prepared decahydroquinoxalines is higher than 100
nM (K > 100 nM) resulting in excellent selectivity for the κꢀ over the µꢀopioid receptor. Only
i
the K values of 5a, 7g, and 7h for the µꢀopioid receptor are lower than 100 nM. Nevertheless,
i
the κ/µꢀselectivity of the methoxycarbonyl derivative 5a and the methylsulfonyl derivative 7g is
160ꢀfold and 300ꢀfold, respectively. The dimethylaminosulfonyl derivative 7h reveals the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 113
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
highest µ affinity of this set of ligands (K = 29 nM). Although, a 30ꢀfold preference for the κ
i
over the µ receptor was recorded, 7h could be regarded as dual ligand interacting with both κ
and µ receptors. In addition to the high κ/µꢀselectivity, the decahydroquinoxalines 5ꢀ8 show high
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
κ/δꢀselectivity as well. The K (δ) values of most of the test compounds are higher than 100 nM,
i
only 5a, 5c, 5j and 6 have K values below 100 nM. Due to the high κ affinity, the κ/δꢀselectivity
i
is still 200ꢀfold (5a), 80ꢀfold (5c) and 100ꢀfold (5j). Even the most potent δꢀligand, the
dimethylamine 6 (K = 7.9 nM), shows more than 50ꢀfold selectivity for the κ receptor (K = 0.16
i
i
nM).
Selectivity over related receptors
4
9
Originally, the σ receptor was regarded as opioid receptor subtype. Moreover, some examples
exist, where rather small variations of the stereochemistry and/or the substitution pattern shifted
5
0,51
the receptor profile from a κ agonist to a σ ligand
or an antagonist of the NMDA (Nꢀmethylꢀ
52,53
Dꢀaspartate) receptor at the PCP (phencyclidine binding site).
Therefore, the affinity of
decahydroquinoxalines 5ꢀ8 towards σ and σ receptors as well as the PCP binding of the
1
2
NMDA receptor was recorded in radioligand receptor binding studies. At test compound
concentrations of 1 µM (σ , σ ) and 10 µM, the test compounds did not compete significantly
1
2
with the corresponding radioligands, indicating excellent κ/σ ꢀ, κ/σ ꢀ and κ/PCP selectivity. The
1
2
only exception is the methoxycarbonyl derivative 5a revealing a σ affinity of 252 nM. However,
2
the κ/σ ꢀselectivity of 5a is still 1000ꢀfold.
2
ACS Paragon Plus Environment

Page 19 of 113
Journal of Medicinal Chemistry
1
2
3
In conclusion, the decahydroquinoxalines 5ꢀ8 represent potent κ ligands with full agonistic
activity. They show excellent selectivity over µꢀopioid, δꢀopioid, σ and σ receptors and the
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2
PCP binding site of the NMDA receptor.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Physicochemical properties
One aim of this study was the synthesis of potent and selective κ agonists devoid of central
nervous system effects. Nevertheless, the compounds should still be able to penetrate into skin.
The passive transcellular diffusion permeation of compounds into the brain is largely influenced
by their physicochemical properties. Hydrogen bond donors and acceptors, lipophilicity, polar
5
4
surface area (PSA), molecular weight (MW) and acidity/basicity are key structural properties.
In a recent viewpoint it was reported that MW < 500 Da, PSA < 140, and < 10 rotatable bonds
have been associated with good oral absorption while MW < 450 and PSA < 70 are required for
5
5
good penetration into the CNS. Potts and Guy developed an equation that correlates skin
56
permeation of a drug with its MW and clogP. During the development of an in silico model for
human skin permeation logD6.4 (pH 6.4 was chosen as it is halfway between blood and skin pH),
clog P, topological polar surface area (TPSA) and the number of rotatable bonds were correlated
5
7
with permeability. A weak correlation of permeability with TPSA, clogD, and clogP was
found, but a correlation with the number of rotatable bonds of the molecule could not be
detected. It was concluded that inter alia hydrogen bond donors and acceptors and molecular
shape (globularity) are important factors influencing skin permeation. In an Organisation for
Economic Coꢀoperation and Development (OECD) guideline for dermal absorption it is stated
that many authorities consider a compound with MW > 500 and clogP < ꢀ1 or clogP > 4 as a
5
8
compound with low dermal absorption.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 113
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In order to learn whether the calculated clogD7.4 values are valid, the logD7.4 values of four test
compounds were determined experimentally. For this purpose the recently developed and
validated micro shake flask method was applied. Quantification of the κ agonist in the nꢀoctanol
and/or buffer layer by means of LCꢀMS allowed the minimization of volumes and thus only
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
4
minimal amounts (< 0.8 mg) of valuable test compounds were consumed. In Table 2 the
experimentally determined logD7.4 values are correlated with calculated clogD values at pH 7.4.
With exception of the difluorophenyl derivative 5e, a nice correlation between the calculated and
experimentally determined logD7.4 values was shown, indicating that the method for the
54
calculation of logD values provided valid data.
7
.4
Table 2: Correlation of calculated clogD7.4 values and experimentally determined logD values
7.4
of selected decahydroquinoxalines.
a)
b)
clogD7.4
calcd.
logD7.4
compd.
experim. determd.
5
e
1.8
2.6
1.5
1.9
1.4
0.77 ± 0.02
1.64 ± 0.04
1.71 ± 0.01
2.41 ± 0.08
1.40 ± 0.02
7b
7
e
8a
8
e
a)
b)
clogD = calculated logD value at pH 7.4 (ACD/Percepta)
7
.4
logD = experimentally determined logD value at pH 7.4, n ≥ 8.
7
.4
ACS Paragon Plus Environment

Page 21 of 113
Journal of Medicinal Chemistry
1
2
3
In order to favor restriction from the CNS while allowing diffusion into skin we have tried to
synthesize compounds with ꢀ1 < clogD5.5 < 4, 450 < MW < 500 and PSA > 70. In addition to
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
9
clogD7.4 the clogD5.5 value was considered, since normal skin exhibits a pH of around 5.5. All
the calculated physicochemical properties are shown in Table 3.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Starting from 5a an increased TPSA was obtained by introduction of a hydroxy group in 3ꢀ
position of the pyrrolidine ring (5b-i). All these compounds except 5h (due to MW slightly
above 500) also fulfill the other criteria. Alternatively, compounds with increased TPSA were
obtained upon replacement of the methyl carbamate in 5a by an oxalic acid amide (7e), a
methanesulfonamide (7g) and a sulfuric diamide (7h). The TPSA of 7g was calculated as 69.
Again, MW and clogD5.5 are in the desired range. Lastly we synthesized compounds in which
both the amine part and the carbamate part of 5a were modified. Compounds 8a-c bearing a
hydroxypyrrolidinyl and methylsulfonyl group have a TPSA of 90, an acceptable clogD5.5 and
MW < 500. TPSA can be further increased to 116 by introduction of a 1,2ꢀoxazolꢀ3ꢀ
ylmethylsulfonyl group (8d,e), but this rather large substituent increased the MW (523 and 541
g/mol, respectively).
Table 3: Calculated physicochemical properties.
MW
a)
b)
c)
compound
clogP
clogD5.5
clogD7.4
TPSA
Hd
Ha
(g/mol)
5
a
454.39
470.39
453.93
453.93
4.5
3.6
3.1
3.1
1.6
1.2
0.7
0.7
3.0
2.9
2.5
2.5
53
73
73
73
0
1
1
1
6
7
7
7
5b
5c
5d
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 113
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
e
437.48
435.94
435.94
503.94
470.39
455.18
428.35
396.15
468.42
482.44
496.47
481.42
481.46
474.44
502.16
490.44
473.99
473.99
523.04
541.04
2.4
3.0
3.0
3.8
3.6
3.7
4.0
3.8
4.0
4.4
4.7
2.5
4.3
3.6
2.8
2.4
2.1
2.1
1.9
2.0
0.0
0.7
0.7
1.4
1.2
2.7
1.5
ꢀ0.2
1.1
1.5
1.8
ꢀ0.1
1.4
0.7
ꢀ0.1
0.2
ꢀ0.1
ꢀ0.1
ꢀ0.4
ꢀ0.3
1.8
2.4
2.4
3.1
2.9
3.7
3.2
2.2
2.6
2.8
3.3
1.5
3.0
2.2
1.5
1.9
1.6
1.6
1.3
1.4
73
73
73
73
73
53
53
36
53
53
53
73
56
69
73
90
90
90
116
116
1
1
1
1
1
0
0
1
0
0
0
1
1
0
0
1
1
1
1
1
7
7
7
7
7
6
6
4
6
6
6
7
6
6
7
7
7
7
9
9
5
f
5g
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5h
5i
5
j
6
7a
7b
7c
7d
7e
7
f
7g
7h
8a
8b
8c
8d
8e
a)
b)
c)
clogD = calculated logD value at pH 7.4 (ACD/Percepta)
7.4
Number of hydrogen bond donors.
Number of hydrogen bond acceptors.
ACS Paragon Plus Environment

Page 23 of 113
Journal of Medicinal Chemistry
1
2
3
Pharmacokinetics and CNS effects
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
When determining the peripheral selectivity of κ agonists with analgesic activity, therapeutic
indices are often obtained by dividing the ED value from the central activity assay by the ED
50
5
0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
value from the peripheral antinociception assay. This can be achieved by administering the
compound icv and using a peripheral route like sc or iv application in the pain model.
Alternatively, a rotarod assay or a platform stepꢀdown latency assay can be applied to determine
3
behavioral CNS effects. Herein, a different approach to assess the peripheral selectivity of the
novel κ agonists was used. Compounds were administered to rats by the iv route. Blood samples
were taken 15 min and 1 h after compound administration. In parallel monitoring of clinical
symptoms was performed 15 min and 1 h after test compound application. At the end of the
study perfused brains were collected. Test compound concentrations in plasma and brain were
measured and the brain/plasma ratio was calculated for the 1 h time point.
6
0
Initially a pharmacokinetic study and a modified Irwin test were performed with 5a. It was
shown that 5a readily penetrated into the brain following iv administration. 15 min and 1 h after
test compound application, brain/plasma ratios were 0.50 and 0.65, respectively. CNS effects
like slight lethargy and hunched or abnormal posture were noted already at a dose of 0.3 mg/kg,
iv and lasted up to 2 h after dosing (Table 4). Introduction of a hydroxy group in 3ꢀposition of
the pyrrolidine ring with (S)ꢀconfiguration (5b) and (R)ꢀconfiguration (5i) resulted in compounds
with reduced ability to cross the bloodꢀbrain barrier. Both compounds did not exhibit any CNS
effects following dosing at 1 mg/kg iv. In the acetic acid writhing model in mice, a model for
6
1,62
acute visceral nociception,
5a, b and i were shown to have ED values of 0.26 mg/kg, 2.2
50
mg/kg and 9.6 mg/kg, resp., after sc application (data not shown), which nicely correlates with
the observed functional activity in the GTPγS assay. However, all compounds produced side
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 113
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
effects often observed with centrally active κ opioids in this model. Side effects were observed at
lowest doses of 0.06 mg/kg (5a), 5.4 mg/kg (5b) and 16.2 mg/kg (5i). Thus, the more active
compound 5b was tested again at 5 mg/kg, iv in a PK study combined with monitoring of CNS
effects. As expected for a linear PK, plasma concentrations at the 1 h time point were fivefold
higher. The brain/plasma ratio was found to be 0.11 and CNS effects were not observed. All
other analogs 5c-h and 5i were subsequently injected at 5 mg/kg. For most of the compounds
brain concentrations were in the range of 41.2 to 74.1 ng/mg with brain/plasma ratios of 0.09 to
0.19 and no observed CNS effects. For the very potent κ agonists 5h and 5j, however, centrally
mediated effects were observed. Slight lethargy was seen throughout the observation period for
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5h which was detected in the brain with 95.8 ng/mg and which exhibited a brain/plasma ratio of
0.17. The strongest CNS effects in the complete series were observed for 5j, an analog of
hydroxypyrrolidine 5c in which the hydroxy group was replaced by a fluorine atom. The effects
were much stronger as compared to 5h although brain concentrations were in the same range.
We can only speculate about the reasons for this discrepancy. It is possible that the extent of
brain tissue binding differs for 5h and 5j. Since PK data is only a measure for total brain
concentration a huge difference in brain tissue binding might result in a marked difference of
pharmacologically relevant concentrations in the brain. Not unexpectedly, 6, the analog of 5a in
which the pyrrolidine ring was replaced by a dimethylamino group, readily penetrated into the
brain and showed CNS effects similar to 5a. Compounds 7b, f and g were all dosed at 1 mg/kg,
iv, since they had been synthesized in an early stage of the study. The esters 7bꢀd were designed
with a metabolic soft spot, namely the ester group, in the side chain. The idea was to generate
compounds that are rapidly metabolized once they penetrate the skin and reach the plasma.
Cleavage of the esters would lead to zwitterionic metabolites which should not cross the blood
ACS Paragon Plus Environment

Page 25 of 113
Journal of Medicinal Chemistry
1
2
3
brain barrier. Indeed, the methyl ester 7b was found to be unstable in plasma. Moreover, its brain
concentration was also very low. Unfortunately, 7b-d were also not stable in buffered solutions
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(data not shown) which prevented their further development as topical drugs.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Methanesulfonamide 7g was further optimized to 8a-c by introduction of a hydroxy moiety at the
pyrrolidine ring and modification of the dichlorophenyl substitution pattern. At a dose of 5
mg/kg 8a-c exhibited limited brain penetration with brain/plasma ratios of 0.02 to 0.11. In
addition they did not show any CNS effects during the observation period.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 113
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 4: PK data and central activity.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
plasma plasma
0.25 h 1 h
brain
1 h
brain/
plasma
ratio
compoun dose, iv
CNS effects
d
(mg/kg)
(ng/ml) (ng/ml) (ng/mg)
Slight lethargy and hunched or
abnormal posture, observed up to
a
b
5a
0.3
117
37.3
24.2
0.65
2
h after dosing.
c
d
5
5
b
1
5
5
5
5
5
5
599
1445
970
128
669
465
546
392
374
414
<LLOQ
74.1
n/a
No clinical signs.
No clinical signs.
No clinical signs.
No clinical signs.
No clinical signs.
No clinical signs.
No clinical signs.
.25 h: Slight lethargy in all
animals.
b
0.11
0.09
0.12
0.12
0.19
0.10
5
c
44.0
5d
1350
1414
1341
1503
66.2
5
e
46.4
5
f
70.6
5g
41.2
0
5
h
5
1
1857
454
573
115
95.8
7.9
0.17
0.07
1 h: Slight lethargy in all animals.
No clinical signs.
5
i
ACS Paragon Plus Environment

Page 27 of 113
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
0
.25 h: Severe lethargy in all
animals. Moderate tonic spasms in
one animal. Moderate extension
spasms of the tail in all animals.
Severe cramped posture in all
animals. Flat posture in all
animals.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
j
5
3071
811
86.6
0.11
1 h: Severe lethargy in all animals.
Moderate tonic spasms in one
animal. Moderate extension
spasms of the tail in all animals.
Severe cramped posture in all
animals. Flat posture in all
animals.
0.25 h: All animals displayed
slight lethargy.
6
1
355
186
101
0.54
1
h: All animals displayed slight
lethargy.
e
7b
1
1
1
477
199
130
124
4.6
2.9
2.9
0.02
0.02
0.02
No clinical signs.
No clinical signs.
No clinical signs.
7
f
290
246
7
g
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 113
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
.25 h: Moderate lethargy in all
7
h
5
1084
560
541
0.97
animals and lateral recumbency.
1
h: Slight lethargy in all animals.
No clinical signs.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
8
a
5
5
5
4650
1214
938
2173
356
52.3
38.7
37.5
0.02
0.11
8b
No clinical signs.
8
c
437
0.09
No clinical signs.
a
b
c
Data were obtained in two independent experiments. 0.50 at 0.25 h. Lower level of
d
e
quantification. Not applicable. Plasma values are indicative. Compound is not stable in
plasma.
ACS Paragon Plus Environment

Page 29 of 113
Journal of Medicinal Chemistry
1
2
3
4
5
Antiꢀinflammatory activity in animal models of acute and chronic inflammation
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
For the investigation of antiꢀinflammatory activity in vivo two well established mouse models
were used. The first model of arachidonic acidꢀinduced ear inflammation is an established tool
for a rapid screening of numerous compounds in an acute setting in vivo following topical or
systemic administration. Application of arachidonic acid on mouse ears induces a fast
inflammatory response including vasodilation and erythema followed by edema formation with
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
3
maximum amplitude at about 1 h following treatment with arachidonic acid. The inflammatory
response is due to arachidonic acid metabolites formed by cyclooxygenase and lipoxygenase.
Under appropriate conditions it allows not only for screening cyclooxygenase and lipoxygenase
inhibitors, but also for other compound classes like histamine antagonists, phosphodiesterase
64
inhibitors, and free radical scavengers that can show activity, too. Moreover, arachidonic acid
treated mice show a robust scratching response which also allows not only monitoring of
potential antipruritic effects, but also of centrally mediated side effects like sedation, which is of
great interest for κ agonist screening. The drawback of this model, however, is that due to its
short duration compounds have to be administered in a prophylactic setting in order to show
activity.
At first the potent and selective κ agonist 5a was tested in a mouse model of arachidonic acidꢀ
induced ear inflammation. In the model 5a was applied topically in solution. Since there is a
rapid onset of inflammation with a short duration in the mouse model of arachidonic acidꢀ
induced dermatitis, compounds have to be administered in a prophylactic setting, namely prior to
6
3
arachidonic acid. Therefore, 5a and the positive control, the cyclooxygenase inhibitor
ACS Paragon Plus Environment