Synthesis and pharmacological evaluation of 5-pyrrolidinylquinoxalines as a novel class of peripherally restricted κ-opioid receptor agonists

Article abstract of DOI:10.1021/jm500940q

5-Pyrrolidinyl substituted perhydroquinoxalines were designed as conformationally restricted κ-opioid receptor agonists restricted to the periphery. The additional N atom of the quinoxaline system located outside the ethylenediamine κ pharmacophore allows

Full text of DOI:10.1021/jm500940q

Article
pubs.acs.org/jmc
Synthesis and Pharmacological Evaluation of
5‑Pyrrolidinylquinoxalines as a Novel Class of Peripherally Restricted
κ‑Opioid Receptor Agonists
Christian Bourgeois,† Elena Werfel,† Fabian Galla,† Kirstin Lehmkuhl,† Hector Torres-Gomez,†
́
́
Dirk Schepmann,†,‡ Babette Kogel,§ Thomas Christoph,§ Wolfgang Straßburger,§ Werner Englberger,§
̈
Michael Soeberdt,∥ Sabine Huwel,⊥ Hans-Joachim Galla,⊥,‡ and Bernhard Wunsch*,†,‡
†Institut fur Pharmazeutische und Medizinische Chemie, Westfalische Wilhelms-Universitat Munster, Corrensstraße 48, D-48149
̈
̈
̈
̈
̈
̈
Munster, Germany
̈
‡Cells-in-Motion Cluster of Excellence (EXC 1003CiM), University Munster, D-48149 Munster, Germany
̈
̈
§Preclinical Research & Development, Grunenthal GmbH, Zieglerstraße 6, D-52078 Aachen, Germany
̈
∥Dr. August Wolff GmbH & Co. KG Arzneimittel, Sudbrackstraße 56, D-33611 Bielefeld, Germany
⊥Institut fur Biochemie, Westfalische Wilhelms-Universitat Munster, Wilhelm-Klemm-Straße 2, D-48149 D-48149 Munster, Germany
̈
̈
̈
̈
̈
S
* Supporting Information
ABSTRACT: 5-Pyrrolidinyl substituted perhydroquinoxalines were designed as conformationally restricted κ-opioid receptor
agonists restricted to the periphery. The additional N atom of the quinoxaline system located outside the ethylenediamine κ
pharmacophore allows the fine-tuning of the pharmacodynamic and pharmacokinetic properties. The perhydroquinoxalines were
synthesized stereoselectively using the concept of late stage diversification of the central building blocks 14. In addition to high κ-
opioid receptor affinity they demonstrate high selectivity over μ, δ, σ1, σ2, and NMDA receptors. In the [35S]GTPγS assay full
agonism was observed. Because of their high polarity, the secondary amines 14a (log D7.4 = 0.26) and 14b (log D7.4 = 0.21) did
not penetrate an artificial blood−brain barrier. 14b was able to inhibit the spontaneous pain reaction after rectal mustard oil
application to mice (ED50 = 2.35 mg/kg). This analgesic effect is attributed to activation of peripherally located κ receptors, since
14b did not affect centrally mediated referred allodynia and hyperalgesia.
The κ-opioid receptor consists of 380 amino acids and has a
molecular weight of 42.7 kDa. Very recently the structure of a
human κ-opioid receptor-T4 lysozyme construct in complex
with the selective κ antagonist JDTic ((R)-7-hydroxy-N-{(S)-2-
[(3R,4R)-4-(3-hydroxyphenyl)-1-isopropyl-3,4-dimethylpiperi-
din-1-yl]ethyl}-1,2,3,4-tetrahydroisoquinoline-3-carboxamide)
was determined by X-ray crystal structure analysis giving insight
into the binding pocket of the κ receptor. The polar
interactions are formed by Asp138 (salt bridge with both
protonated amino moieties) and Tyr312 (H-bond with
carbonyl O-atom). Whereas conserved Asp138 plays a role in
anchoring protonated κ ligands in the binding pocket, the
INTRODUCTION
■
In 1976 animal experiments with morphinoids led to a
subclassification of the opioid receptor into three subtypes,
which were termed according to their prototypical ligands μ-
opioid (from morphine), σ-opioid (from SKF-10,047), and κ-
opioid receptor (from ketocyclazocine).1 Today the group of
opioid receptors consists of four subtypes, μ (MOR), δ (DOR),
κ (KOR), and ORL1 (NOR) receptor; the σ receptor has been
removed from the class of opioid receptors.2 The four opioid
receptors belong to the peptide-binding γ subfamily of class A
(rhodopsin-like) G-protein-coupled receptors (GPCRs), which
are coupled predominantly to Gi/G0 proteins. In the early
1990s the four opioid receptors were cloned. They reveal a
sequence homology of more than 60%.3
Received: June 20, 2014
Published: July 25, 2014
© 2014 American Chemical Society
6845
dx.doi.org/10.1021/jm500940q | J. Med. Chem. 2014, 57, 6845−6860

Journal of Medicinal Chemistry
Article
Figure 1. Peripherally acting κ receptor agonists.
amino acids Val108, Val118, Ile294, and Tyr312 are unique for
the binding pocket of the κ receptor and are therefore
postulated to contribute to the subtype selectivity of ligands.4
Activation of κ-opioid receptors leads to strong analgesia,
which is not associated with the typical μ-mediated side effects,
such as euphoria, constipation, respiratory depression, and
development of dependency.5 However, κ agonists induce
different centrally mediated side effects including dysphoria,
sedation, and strong diuresis.6,7 Since the κ-opioid receptors are
located not only in the central nervous system but also in the
periphery, ligands activating predominantly peripherally located
κ receptors would avoid these side effects. The therapeutic
potential of κ agonists acting only in the periphery comprises
treatment of visceral pain, inflammation, and inflammatory and
pruritic skin diseases like atopic dermatitis and psoriasis.8−10
The known κ agonists belong to four structurally diverse
compound classes: peptides derived from the endogenous
agonist dynorphin A,6,11 morphinoids including morphinans
and benzomorphans,6,11 the nonbasic natural product salvinorin
A,12,13 and ethylenediamines (or arylacetamides) with the first
synthetic κ agonist U-50,488 (4)14 (Figure 2). The
pharmacophore of the fourth class of κ agonists is characterized
by an ethylene moiety connecting a pyrrolidine ring and an
arylacetamide moiety. On the basis of this key structure, several
ligands with high κ affinity have been developed.15,16
agonists 4 and 5 showing subnanomolar κ affinity (4, Ki = 0.34
nM; 5, Ki = 0.31 nM 25) were used as lead compounds (Figure
2). Their core structures should be combined in the
In order to develop κ agonists restricted to the periphery,
various approaches have been pursued (Figure 1). Tetrapep-
tides containing D-configured amino acids such as FE 200665
(1), also known as CR665, produce strong analgesia by
activating selectively κ receptors in the periphery.17−19
Nalfurafine (2) has been approved in Japan for the treatment
of hemodialysis-related uremic pruritus. However, centrally
mediated side effects were observed during the treatment of
patients with 2.20−22 Introduction of polar functional groups
such as hydroxy substituents of the pyrrolidine ring of
ethylenediamine-based κ agonists led to the discovery of
asimadoline (3), which has a preference for κ receptors in the
periphery. 3 successfully passed a phase II clinical trial in
irritable bowel syndrome (IBS). Recently, a phase III clinical
trial for the treatment of patients with diarrhea-predominant
IBS (D-IBS) was conducted by Tioga Pharmaceuticals, Inc.23,24
Herein we report on the synthesis and pharmacological
evaluation of novel κ agonists, which should selectively activate
κ receptors in the periphery. The ethylenediamine-based κ
Figure 2. Development of a novel class of peripherally acting κ
receptor agonists 6 based on the perhydroquinoxaline ring system.
perhydroquinoxaline ring system 6 while retaining the
pharmacophoric ethylenediamine substructure. The designed
perhydroquinoxalines 6a result from connection of either the
cyclohexane ring with the methyl moiety of 4 or the piperazine
ring with the methyl moiety of 5 by a two-atom moiety (see
arrows in Figure 2). The additional N-atom within the
quinoxaline ring system of 6 will allow the fine-tuning of the
pharmacodynamic and, moreover, the pharmacokinetic proper-
ties. In particular the introduction of polar substituents and/or
ionizable functional groups should increase the polarity and
inhibit the passage of the blood−brain barrier. The
introduction of a polar OH-moiety into the pyrrolidine ring
6846
dx.doi.org/10.1021/jm500940q | J. Med. Chem. 2014, 57, 6845−6860

Journal of Medicinal Chemistry
Article
a
Scheme 1. Synthesis of κ Agonists with Perhydroquinoxaline Framework
a
Reagents and reaction conditions: (a) CH3OH, NaOH, 80%;26,27 (b) four steps (OH → NHBn exchange, NO2 reduction, (CO2CH3)2,
NH4HCO2, Pd/C);29 (c) 1,4-diiodobutane or 1,4-dibromobutan-2-ol, CH3CN, NaHCO3, 80 °C, 76% (11a), 79% (11b); (d) AlCl3/LiAlH4 1:3,
THF, 0 °C, then rt, 96% (12a), 77% (12b); (e) (3,4-dichlorophenyl)acetyl chloride, CH2Cl2, NaOH, rt, 86% (13a), 90% (13b); (f) H2, 1 bar, Pd/C,
THF, H2O, HCl, rt, 30 min, 94% (14a), 89% (14b); (g) R′CHO, NaBH3CN or NaBH(OAc)3; for the definition of residues R (R = CH2R′) see
Table 1; (h) R′C(O)Cl or (R′CO)2O, CH2Cl2; for the definition of residues R (R = COR′) see Table 1.
as shown with 6b should further increase the polarity of the κ
agonists.
chloride to afford the amides 13a29 and 13b in 86% and 90%
yield, respectively. The acylation of the hydroxypyrrolidine 12b
was performed with exactly 1 equiv of acid chloride to avoid the
additional acylation of the OH moiety in the pyrrolidine ring.
To inhibit reductive dechlorination during the hydro-
genolytic removal of the second benzyl protective group of
13a,b with H2 and Pd/C, HCl was added to the reaction
mixture. The secondary amines 14a and 14b represent versatile
building blocks, which allow the introduction of diverse alkyl,
arylalkyl, acyl, and alkoxycarbonyl substituents at the 4-position
of the perhydroqinoxaline ring system at the end of the
synthesis according to the concept of late stage diversification.
Alkyl and arylalkyl residues were attached by reductive
alkylation using aldehydes and NaBH3CN or NaBH(OAc)3.
Acyl and alkoxycarbonyl residues were introduced by acylation
with acid chlorides or anhydrides.
SYNTHESIS
■
At first the cyclohexane ring of the perhydroquinoxaline system
with three adjacent functional groups in trans-orientation was
established by a double Henry reaction (nitro aldol reaction) of
glutaraldehyde (7) and nitromethane (8).26−28 The resulting
achiral (2r)-configured nitrocyclohexanediol 9 was transformed
via four reaction steps29 (substitution of OH by BnNH,
reduction of the nitro group, cyclization with dimethyl oxalate,
hydrogenolytic removal of the first benzyl protective group)
into racemic quinoxalinedione 10, which represents the key
intermediate of this project. (Scheme 1)
Alkylation of the primary amine 10 with 1,4-diiodobutane
and 1,4-dibromobutan-2-ol provided the pyrrolidine derivatives
11a29 and 11b in 76% and 79% yield, respectively. The
reduction of the quinoxalinediones 11a,b was performed with
AlH3, which was generated in situ by mixing AlCl3 and LiAlH4
in the ratio 1:3.30 The resulting secondary amines 12a and 12b
were subsequently acylated with 3,4-dichlorophenylacetyl
The N-residues were selected with the aim to achieve high κ
affinity and receptor selectivity. Additionally, the residues
should supply the compounds with high polarity to inhibit the
penetration into the central nervous system. Therefore, the
clogD7.4 values of all designed compounds were calculated
systematically prior to their synthesis.
6847
dx.doi.org/10.1021/jm500940q | J. Med. Chem. 2014, 57, 6845−6860

Journal of Medicinal Chemistry
Article
Since the hydroxypyrrolidine derivatives (b-series) contain
an additional center of chirality, two diastereomers were formed
during the synthesis of 11b and its further transformations.
Exemplarily the diastereomers of the secondary amine 14b
were separated by preparative HPLC. Since it was not possible
to obtain crystals suitable for an X-ray crystal structure analysis,
a precise assignment of the relative configuration of the
diastereomeric derivatives 14bA (shorter retention time) and
14bB (longer retention time) was not possible.
effects are not expected. Moreover, zwitterions usually do not
readily pass the blood−brain barrier.
Exemplarily the diastereomers of the hydroxypyrrolidine 14b
were separated and tested. It was shown that 14bA, the
diastereomer with shorter retention time, has a considerably
higher κ affinity (Ki = 2.8 nM) than its diastereomer 14bB (Ki
= 16 nM). Since in 3 (see Figure 1) and similar
hydroxypyrrolidines (e.g., 2-(2-aminophenyl)-N-[2-(3-hydrox-
ypyrrolidin-1-yl)-1-phenylethyl]-N-methylacetamide (EMD
60400)) the (S,S)-configured hydroxypyrrolidine enantiomers
represent the eutomers,33,34 it can be speculated that the more
potent diastereomer 14bA has the relative configuration
(4aRS,8SR,8aSR-SR(pyrrolidine)).
PHARMACOLOGICAL EVALUATION
■
κ-Opioid Receptor Affinity. The κ receptor affinity of the
quinoxalines 13−27 was determined in competitive receptor
binding studies. Tritium labeled U-69,593 was employed as
radioligand. Guinea pig brain preparations were used as
receptor material, and the nonspecific binding was determined
in the presence of 10 μM U-69,593.31,32 The κ receptor affinity
of the reference compounds 4, 5, U-69,593, and naloxone
determined in this assay is included in Table 1.
The κ receptor affinity of the perhydroquinoxalines 13−27
was investigated in a cell-based assay. In this κ assay transfected
HEK-293 cell lines expressing the human κ-opioid receptor
served as receptor material and [3H]Cl-977 was used as
radioligand.35,36 In Table 1 the data recorded with guinea pig
brains/[3H]U-69,593 and HEK-293 cells/[3H]Cl-977 are
compared.
The Ki values obtained with the transfected HEK-293 cells
are generally 5- to 15-fold higher than the Ki values recorded
with guinea pig brain preparations. However, both assay
systems show the same tendencies; e.g., the Ki values of the
methyl derivatives 15a and 15b are 2.7 and 5.4 nM in the
guinea pig brain assay and 39 and 49 nM in the HEK-293 cell
assay.
The data in Table 1 indicate high κ affinity (i.e., Ki < 10 nM)
for several compounds of this novel class of pyrrolidinylqui-
noxalines. Generally, the κ affinity of hydroxypyrrolidine
derivatives (b-series) is slightly lower than the κ affinity of
pyrrolidine derivatives (a-series) as demonstrated with the pair
of secondary amines 14a (Ki = 2.1 nM) and 14b (Ki = 8.7 nM).
Only the benzyl derivatives 13a,b represent an exception of this
general trend.
Small (Me, 15) and medium alkyl groups (n-Bu, 16) as well
as benzyl moieties including the p-methoxybenzyl (17) and
heteroarylmethyl substituents such as the 2-pyridylmethyl (18)
and 5-imidazolylmethyl group (19) are well tolerated by the κ-
opioid receptor. This result is remarkable, since the ligands 13−
19 contain a second basic functional group within the
quinoxaline ring in addition to the basic center of the
pyrrolidine ring. These quinoxaline derivatives 13−19
represent another example of nonpeptidic κ agonists with two
basic functional groups, in addition to our previously reported
bicyclic κ agonists.32 This observation broadens the possibilities
of fine-tuning the pharmacodynamic and pharmacokinetic
properties of this type of ligands.
In the group of N-acyl and N-alkoxycarbonyl substituted
quinoxalines 20−27 the methoxycarbonyl derivatives 23a and
23b reveal the highest κ receptor affinity. The Ki values are 9.7
nM for the pyrrolidine derivative 23a and 11 nM for the
hydroxypyrrolidine derivative 23b. Enlargement of the
methoxycarbonyl moiety to an ethoxycarbonyl moiety (24a)
and its replacement by acyl residues (20−22) led to reduced κ
affinity. This observation correlates nicely with the high κ
receptor affinity of monocyclic (e.g., 5)24 and bridged
piperazines31 bearing an N-methoxycarbonyl moiety.
The carboxylic acids 25 and 26 were designed and
synthesized, since their high polarity (clogD7.4 = −1.1 to
+0.1) should not allow the passage of the blood−brain barrier.
However, the κ affinity of the acids 25 and 26 is very low, and
therefore, they are not suitable as peripherally restricted κ
agonists.
The amides 27a and 27b bearing an ester moiety within the
N-residue show promising κ receptor affinity of 11 and 16 nM.
Topical application of the esters 27 could lead to activation of
peripherally located κ receptors, e.g., in the skin. After systemic
absorption, the esters 27 will be hydrolyzed by esterases
present in the plasma. Since the hydrolysis products 25 show
only low κ receptor affinity, systemic or centrally mediated side
Discussion of the Dihedral Angle of the Ethylenedi-
amine Substructure. It has been postulated that the relative
orientation of the pyrrolidine ring and the dichlorophenylacet-
amide moiety is crucial for high κ receptor affinity. For the
rather rigid cyclohexane derivative 4 (Ki = 0.34 nM) a dihedral
angle N(pyrrolidine)−C−C−N(acyl) of 60° was postulated.37
Recently we have shown that a dihedral angle of 70° represents
the energetically most favored conformation of the piperazine 5
(Ki = 0.31 nM) with the flexible pyrrolidinylethyl side chain.31
Therefore, the dihedral angle of the ethylenediamine
substructure (N(pyrrolidine)−C8−C8a−N(acyl)) of the pyr-
rolidinyl substituted perhydroquinoxalines was determined by
AM1 calculations (Figure 3). Since the hybridization of the N
atom in 4-position outside the ethylenediamine pharmacophore
has an influence on the geometry of the perhydroquinoxaline
system, the dihedral angles were calculated exemplarily for the
secondary amine 14a (tetraedric N atom, Ki = 2.1 nM) and the
methoxycarbonyl derivative 23a (trigonal N atom, Ki = 9.7
nM). The calculations were performed with (4aR,8S,8aS)-
configured enantiomers, which correspond to the enantiomeri-
cally pure lead compounds 4 and 5.
AM1 calculations provided two types of energetically favored
conformers, respectively, which differ slightly in their dihedral
angle. The energetically most favored conformers 14aI and
23aI show dihedral angles of 55° and 54°, respectively (Table
2). In the second type conformers 14aII and 23aII with slightly
distorted geometry of the quinoxaline ring system resulting in
slightly lower heat of formation, the dihedral angles are 71° and
69°, respectively. These results clearly indicate that the
substituent at N-4 has only a small effect on the dihedral
angle of the ethylenediamine pharmacophore.
Since the dihedral angles of both types of conformations are
close to the dihedral angles reported for the lead compounds 4
and 5, we assume that the geometry along the pharmacophoric
ethylenediamine substructure of the conformationally con-
6848
dx.doi.org/10.1021/jm500940q | J. Med. Chem. 2014, 57, 6845−6860

Journal of Medicinal Chemistry
Article
6849
dx.doi.org/10.1021/jm500940q | J. Med. Chem. 2014, 57, 6845−6860

Journal of Medicinal Chemistry
Article
strained quinoxalines 13−27 is close to the bioactive
conformation of the flexible piperazine derivative 5.
Selectivity over μ- and δ-Opioid Receptors. The μ and
δ receptor affinities of the perhydroquinoxalines 13−27 were
investigated using transfected CHO-K1 cell lines expressing the
human μ- and δ-opioid receptors as receptor material. In the μ
assay and the δ assay, [3H]naloxone and [3H]deltorphine
served as radioligands, respectively.35,36 The μ and δ affinity
data of the quinoxalines 13−27 are summarized in Table 1.
The alkyl and arylalkyl substituted quinoxalines 13−19 show
very low affinity toward μ and δ receptors resulting in high κ/μ
and κ/δ selectivity of >100, respectively. With a κ/μ selectivity
of only 16 the p-methoxybenzyl substituted derivative 17a
represents the only exception. The comparatively low κ/μ
selectivity is due to a rather high submicromolar μ receptor
affinity of 740 nM.
Compared with the alkyl and arylalkyl substituted quinoxa-
lines 13−19, the acyl and alkoxycarbonyl substituted quinoxa-
lines 20−27 reveal reduced κ/μ and κ/δ selectivity values in the
range of 10−40. The reduced selectivity of these compounds is
explained with their lower κ receptor affinity.
It can be concluded that both types of compounds alkyl/
arylalkyl and acyl/alkoxycarbonyl substituted quinoxalines show
high selectivity for the κ-opioid over the related μ- and δ-opioid
receptors.
Selectivity over σ and NNDA Receptors (PCP Binding
Site). Although the structure of the κ receptor is not related to
the structure of σ and NMDA receptors, it has been shown that
small variations of κ agonists led to σ receptor ligands or
NMDA antagonists. For example the cis-configured analogue of
the trans-configured prototypical κ agonist 4 represents a high-
affinity σ ligand.38 Reduction of the phenylacetamide moiety of
4 to a phenylethylamino group increased the σ affinity and
decreased the κ affinity.39 In the class of benzomorphans the
absolute configuration and the nature of the N-substituent
control whether a ligand interacts preferably with κ-opioid, σ, or
NMDA receptors.40,41 Therefore, the affinity of this novel type
of κ agonists toward σ1, σ2,42,43 and NMDA receptors (PCP
binding site)44,45 was recorded in competitive radioligand
binding assays. (Table 1).
The data in Table 1 clearly indicate that the quinoxaline-
based κ agonists 13−27 do not interact considerably with σ1,
σ2, and NMDA receptors. Generally, the test compounds did
not compete with the radioligands up to a concentration of 1
μM, indicating IC50 values of at least greater than 1 μM. The n-
butyl derivative 16a possesses the highest σ1 affinity (KI = 518
nM) of these quinoxalines. Because of its high κ affinity (KI =
3.1 nM) the κ/σ1 selectivity of 16a is still >100.
In conclusion, the new quinoxaline-based κ agonists show
high selectivity against μ-opioid, δ-opioid, σ1, σ2, and NMDA
(PCP binding site) receptors.
Functional Activity. The intrinsic activity of the sub-
stituted quinoxalines 13−27 was investigated with the
[35S]GTPγS ([35S]guanosine 5′-3-O-(thio)triphosphate) bind-
ing assay using human HEK-293 (human embryonic kidney)
cells as source of human κ-opioid receptors.46,47
In Table 3 the EC50 values of the quinoxaline-based κ
agonists are summarized and compared with the EC50 value of
the prototypical full agonist U-69,593. Promising EC50 values
below 40 nM were obtained for the benzylamine 13a (EC50
=
29 nM), the secondary amines 14a (EC50 = 33 nM) and 14b
(EC50 = 35 nM), the methylamine 15a (EC50 = 20 nM), the
butylamine 16a (EC50 = 12 nM), and the p-methoxybenzyl-
6850
dx.doi.org/10.1021/jm500940q | J. Med. Chem. 2014, 57, 6845−6860

Journal of Medicinal Chemistry
Article
Figure 3. Left: Schematic view of the dihedral angle N(pyrrolidine)−C8−C8a−N(acyl). Middle: Energetically most favored conformation of 14a
with a dihedral angle of 55°. Right: Energetically most favored conformation of 23a with a dihedral angle of 54°.
position reveal higher agonistic activity than the N-4 acylated
and N-4 alkoxycarbonylated derivatives.
Table 2. Correlation of the Dihedral Angle N(yrrolidine)−
C−C−N(acyl) of the Pyrrolidinyl Substituted
Perhydroquinoxalines 14a and 23a and the Lead
Compounds 4 and 5 with Their κ Receptor Affinity
log D Value. Lipophilicity in terms of log P value is an
important physicochemical parameter indicating the ability of a
compound to penetrate lipophilic membranes. However, the
log P value is a descriptor for neutral compounds and therefore
does not take into account the ionization of compounds under
physiological conditions. The pH dependent log D value
(coefficient of distribution) describes the n-octanol/water
distribution of ionizable species at a given pH value. Therefore,
the calculation and measurement of the log D value at a
physiologically relevant pH value (e.g., pH 7.4) provide a more
realistic description of the lipophilicity properties of bioactive
compounds.48,49
In order to obtain κ agonists, which are not able to pass the
blood−brain barrier and stay in the periphery, we are interested
in rather polar compounds. Therefore, the log D7.4 value (log D
value at pH 7.4) was determined experimentally for the
benzylamines 13a and 13b, the secondary amines 14a and 14b,
and the methyl carbamates 23a and 23b. According to a
standardized protocol in our lab, the compounds were
partitioned between an aqueous buffer (3-morpholino-
propanesulfonic acid (MOPS), pH 7.4) and n-octanol. The
amount of the compound in the aqueous layer was determined
quantitatively by mass spectrometry. The results are
summarized in Table 4.
In addition to the experimentally determined log D7.4 values,
the log P and log D values were calculated using different
methods (see Table 4). The log P values calculated for neutral
compounds (ClogP and milogP values) do not correlate with
the experimentally determined log D7,4 values. However,
calculations taking the positive charge of the quinoxalines
into account, either with completely protonated species
(milogP(H+)) or at physiological pH value (clogD7.4), lead to
values close to the experimentally determined log D7,4 values.
Since the distribution between two layers at pH 7.4
characterizes better the behavior of compounds under
physiological conditions, the clogD7,4 values of all test
compounds are included in Table 3.
According to all different calculations, the log P values of the
pyrrolidine derivatives (a-series) are higher by 0.7−1.3 than the
log P values of the analogous hydroxypyrrolidine derivatives (b-
series). However, this difference in the polarity could not be
reproduced by the experiments. The experimentally determined
log D values of the pyrrolidine/hydroxypyrrolidine pairs 13a/b,
14a/b, and 23a/b do not differ, respectively.
a
b
Ki (nM)
(κ receptor)
dihedral angle
(deg)
ΔH0
compd
conformer
(kcal/mol)
14a
2.1
9.7
I
55
71
54
69
60
0.0
3.7
0.0
2.5
II
I
23a
II
4
5
0.34
0.31
c
70
a
b
Dihedral angle defined by N(pyrrolidine)−C−C−N(acyl). ΔH0 =
H0(conformer II) − H0(conformer I); H0 = heat of formation
calculated with AM1 (MOE). Dihedral angle of the energetically most
c
favored conformation of the flexible κ agonist 5.
amine 17a (EC50 = 26 nM). These EC50 values are close to the
EC50 value of the prototypical κ agonist U-69,593 (EC50 = 12
nM). Obviously, an additional basic functional group (N-4 of
the perhydroquinoxaline ring) is not detrimental for the κ
agonistic activity of these ligands. However, elimination of the
basicity the N-4-atom of the quinoxaline ring by acylation or
alkoxycarbonylation (compounds 20−27) resulted in consid-
erably reduced agonistic activity with EC50 values higher than
85 nM. Compounds 25a and 25b bearing an acyl moiety with
carboxylic acid substituent did not show a significant activation
of the κ opioid receptor at concentrations up to 1 μM.
However, these compounds also exhibited a reduced affinity
toward κ receptor in the binding assay.
In Figure 4 the binding curves of the secondary amines 14a
and 14b, the butylamine 16a, and the methyl carbamate 23a are
compared with the binding curve of the prototypical full agonist
U-69,593, respectively. All compounds behave as full agonists
reaching the same maximal effect as the full κ agonist U-69,593.
Even the methyl carbamate 23a with a rather high EC50 value of
110 nM is a full κ receptor agonist. The binding curve of the
butyl derivative 16a is remarkable, since it is almost identical
with the binding curve of U-69,593 resulting in the same EC50
value of 12 nM.
In conclusion, all perhydroquinoxaline derivatives but the
compounds bearing a carboxylic acid substituted acyl group in
4-position behave as full κ-opioid receptor agonists in the
[35S]GTPγS binding assay independently of the substituent at
N-4 (H, alkyl, arylalkyl, acyl, alkoxycarbonyl). However, the
compounds with an additional basic structural element in 4-
6851
dx.doi.org/10.1021/jm500940q | J. Med. Chem. 2014, 57, 6845−6860

Journal of Medicinal Chemistry
Article
Table 3. Agonistic Activity (EC50) in the [35S]GTPγS Assay and clogD7.4 of Quinoxaline-Based κ Agonists
a
compd
13a
R
X
κ affinity, Ki (nM)
κ agonistic activity, EC50 (nM)
clogD7.4
Bn
H
9.4
6.6
2.1
8.7
2.7
5.4
3.1
6.8
4.2
4.3
13%
15
29
64
33
35
20
79
12
26
41
56
3.2
2.5
1.1
0.4
1.4
0.7
2.7
3.0
2.2
1.1
0.5
3.2
2.5
1.3
2.0
2.0
1.3
2.4
0.1
−1.1
0.1
1.3
0.7
13b
14a
14b
15a
15b
16a
17a
18a
19a
19b
20a
20b
21a
22a
23a
23b
24a
25a
25b
26a
27a
27b
U-69,593
Bn
OH
H
H
H
OH
H
CH3
CH3
n-Bu
OH
H
p-H3COBn
H
CH2-2-pyridyl
CH2-imidazol-5-yl
CH2-imidazol-5-yl
COPh
H
H
b
OH
H
nd
260
470
330
600
110
85
COPh
OH
H
22
COCH3
24
COEt
H
26
CO2CH3
H
9.7
11
CO2CH3
OH
H
CO2Et
15
150
c
COCH2CO2H
COCH2CO2H
CO(CH2)2CO2H
COCH2CO2CH3
COCH2CO2CH3
H
169
482
136
11
na
c
OH
H
na
b
nd
H
220
170
OH
18
0.97
12
a
b
c
clogD7.4 = calculated log D value at pH 7.4 (MarvinSketch 6.2.3, ChemAxon). nd = not determined. na = not active; i.e., at a test compound
concentration of 1 μM the agonistic activity was lower than 50%.
High polarity was a criterion for the selection of the
substituents at the additional N-atom in 4-position of the
quinoxaline ring system outside the κ pharmacophore. Low
clogD7.4 values were calculated for the secondary amines 14, the
methylamines 15, the pyridylmethyl and imidazolylmethyl
substituted derivatives 18 and 19, the methyl carbamates 23 as
well as the compounds 25 and 26 with a carboxy moiety in the
side chain.
barrier model, only passive diffusion of compounds had to be
considered.50−52
Prior to the penetration test, the potential toxicity of the
compounds toward the capillary endothelial cells has to be
determined. For this purpose various concentrations of the test
compounds (1, 100, 500, 1000 μM) were added to the apical
compartment of the model together with 14C-labeled sucrose.
The amount of [14C]sucrose reaching the basolateral compart-
ment was determined by a β counter over an incubation period
of 180 min. Since sucrose cannot penetrate an intact
endothelial monolayer, the amount of [14C]sucrose in the
basolateral compartment indicates a damage of the integrity of
the endothelial monolayer.
For the further tests the secondary amines 14a and 14b were
selected because of their high κ receptor affinity (Ki = 2.1 and
8.7 nM), high selectivity over related receptors, full κ agonistic
activity (EC50 = 33 and 35 nM) and high polarity (log D7.4
=
0.26 and 0.21).
Passage of an Artificial Blood−Brain Barrier. A model
of the blood−brain barrier (Figure 5) was used to determine
the penetration of the potent and polar κ agonists 14a and 14b
into the brain. For the purpose of comparison the potent but
more lipophilic benzylamine 13a was selected for being tested
in this model as well. In this model pig cerebral capillary
endothelial cells were grown on a filter forming a tight
monolayer. The polarity of the monolayer is improved by
coating the filter with collagen. After addition of a compound
solution to the apical compartment, which corresponds to the
blood side of the blood−brain barrier, its penetration into the
basolateral compartment corresponding to the brain side was
measured. Since active transporters for the compounds under
consideration are shown to be not present in this blood−brain
Concentrations of 100 μM and higher of the benzylamine
13a increased the permeability of the endothelial monolayer
considerably, indicating nontolerable toxicity. A concentration
of 1 μM was tolerated by the endothelial cells, but this
concentration is too low to quantify 13a in the apical and
basolateral compartments. Therefore, this compound was no
longer considered for this assay.
The secondary amines 14a and 14b were well tolerated by
the endothelial cells up to a concentration of 500 μM. Only a
concentration of 1000 μM led to increased permeability of
[14C]sucrose. Therefore, the permeability test was performed
with 500 μM solutions of the secondary amines 14a and 14b.
After incubation of the apical compartment with 500 μM
solutions of 14a and 14b for 3 h, the amounts of 14a and 14b
6852
dx.doi.org/10.1021/jm500940q | J. Med. Chem. 2014, 57, 6845−6860

Journal of Medicinal Chemistry
Article
Figure 4. Binding curves of the secondary amines 14a (top left) and 14b (top right), the butylamine 16a (bottom left), and the methyl carbamate
23a (bottom right) at human κ receptors in the [35S]GTPγS assay. The binding curves of the test compounds are compared with the binding curves
of the prototypical κ agonist U-69,593, respectively.
Table 4. Correlation of Variously Calculated log P/log D and Experimentally Determined log D Values of Selected
Perhydroquinoxalines
a
b
c
d
e
compd
Ki (nM) (κ receptor)
ClogP
milogP
milogP(H+)
clogD7.4
log D7.4, exptl
13a
13b
14a
14b
23a
23b
9.4
6.6
2.1
8.7
9.7
11
6.4
5.1
4.1
2.8
4.9
3.7
5.3
4.4
3.7
2.8
3.8
2.9
2.2
1.3
3.2
2.5
1.1
0.4
2.0
1.3
2.38 0.09
2.51 0.09
0.26 0.10
0.21 0.09
1.29 0.08
1.34 0.08
0.5
−0.4
0.6
−0.3
a
b
c
ClogP = calculated log P value (ChemDraw). milogP = calculated log P value (www.molinspiration.com). milogP(H+) = calculated log P value of
the monoprotonated species (www.molinspiration.com). clogD7.4 = calculated log D value at pH 7.4 (MarvinSketch 6.2.3, ChemAxon) (www.
chemaxon.com). log D7.4 = experimentally determined log D value at pH 7.4, n ≥ 8.
d
e
in the basolateral compartment were recorded by RP-HPLC
analysis. Table 5 shows that 2.1%/cm2 and 2.2%/cm2 of the
secondary amines 14a and 14b passed the cerebral capillary
endothelial monolayer, respectively. The very low penetration
of the endothelial monolayer by the secondary amines 14a and
14b, which is only slightly higher than the penetration of
sucrose, can be attributed to their low log D7.4 values
determined experimentally.
Analgesic Activity in Animal Models of Visceral Pain.
The secondary amines 14a and 14b demonstrate high κ-opioid
agonistic activity and low permeability of the endothelial
monolayer as artificial model of the blood−brain barrier.
Therefore, these compounds were selected to investigate their
analgesic activity in animal models of visceral pain. For this
purpose the phenylquinone writhing assay and the mustard oil
assay in mice were performed.
The phenylquinone writhing assay is an unspecific pain test
comprising both visceral and nonvisceral afferent structures.53
Various doses of the test compounds 14a and 14b were iv
administered to male mice. After 10 min, an aqueous solution
of 2-phenyl-1,4-benzoquinone (phenylquinone) was given
intraperitoneally and the resulting pain-induced writhing
reactions, such as stretching, twisting a hind leg inward, or
contraction of abdomen, were counted for 15 min (from 5 to
20 min after phenylquinone administration).
Whereas doses of 10 and 21.5 mg/kg body weight of the
pyrrolidine 14a did not lead to a decreased number of writhing
reactions, a dose of 46.4 mg/kg induced lethality. In contrast,
6853
dx.doi.org/10.1021/jm500940q | J. Med. Chem. 2014, 57, 6845−6860

Journal of Medicinal Chemistry
Article
Figure 5. Cell culture model of the blood−brain barrier. A monolayer of pig cerebral capillary endothelial cells is grown on top of a permeable
collagen coated filter. According to the polarity of the cells on the filter, the medium in the upper (apical) compartment of the cell monolayer
corresponds to the blood side of the capillary; the compartment below the filter (basolateral) corresponds to the extravasal (brain) part.
behavior but did not affect referred allodynia and referred
hyperalgesia. Since only one dose was investigated, ED50 values
were not determined for the analgesic activity of 14a.
Table 5. Penetration of the Secondary Amines 14a and 14b
via the Endothelial Monolayer
a
b
c
compd
14a
permeability (%/cm2)
log D7,4/log P
In Figure 6 the effect of the hydroxypyrrolidine 14b is
displayed. The spontaneous pain reaction of mice was inhibited
dose-dependently leading to an ED50 value of 2.35 (1.87−2.87)
mg/kg. Obviously, the hydroxypyrrolidine 14b represents a
potent analgesic for the treatment of visceral pain. The efficacy
of 14b for the treatment of referred allodynia and hyperalgesia
was recorded 20−40 min after mustard oil application using
von Frey filaments of 1 and 16 mN, respectively. As shown in
Figure 6B and Figure 6C, 14b did not have any effect on these
readouts. According to the low passage of the artificial blood−
brain barrier, a low passage of the real blood barrier is expected,
which might explain the lack of efficacy of 14b in these centrally
mediated readouts.
These findings were further substantiated by intracerebro-
ventricular (icv) administration of the hydroxypyrrolidine 14b
in the same experimental models. After icv administration of
14b spontaneous pain response, referred allodynia, and referred
hyperalgesia were reduced dose-dependently with correspond-
ing ED50 values of 2.72 (1.01−5.76) μg/animal, 3.31 (2.53−
4.39) μg/animal, and 2.16 (1.63−2.73) μg/animal, respectively.
These results led to the conclusion that the hydroxypyrro-
lidine 14b is a highly potent peripherally acting analgesic.
Systemic (iv) application of 14b only reduced the spontaneous
pain reactions but did not affect the centrally mediated
allodynia and hyperalgesia. However, after icv application
these readouts were also inhibited with similar potency.
2.1 0.9
2.2 0.5
57
0.26 0.10
0.21 0.09
2.84
14b
diazepam
sucrose
0.63
−3.18
a
b
The permeability is correlated with log D7.4/log P values. The
permeability was standardized taking the size of the filter (1.13 cm2)
into account. Values are the mean of three independent experiments
c
(n = 3). log D7.4 values of 14a and 14b were experimentally
determined. log P values of the reference compounds diazepam and
sucrose were calculated with ChemDraw.
the hydroxypyrrolidine 14b showed a dose dependent
inhibition of the writhing reactions indicating analgesic activity
(Table 6). However, increasing occurrence of side effects
(decreased locomotion, flat posture, ptosis) was observed when
increasing doses were administered.
Table 6. Analgesic Activity of the Hydroxypyrrolidine 14b in
the Phenylquinone Writhing Assay
inhibition of writhing reactions
dose (mg/kg) no. of writhing reactions
(%)
0 (control)
3.16
98
93
59
7
0
5.1
40
93
10.0
21.5
CONCLUSION
■
In the second assay visceral pain was induced in mice by
rectal administration of mustard oil.54,55 The pain induced by a
chemical stimulation of a defined visceral organ, i.e., the colon,
was quantified in three parameters. Spontaneous visceral pain
behavior (e.g., abdominal licking, stretching, squashing,
contraction of flank muscles) occurs during the first minutes
after mustard oil administration. This period of spontaneous
pain is followed by referred allodynia and referred hyperalgesia
measured by assessing the animals reaction toward mechanical
stimulation (von Frey filaments) at the abdominal wall, a
dermatom corresponding to the colon. Allodynia is charac-
terized by pain induced by a non-nociceptive stimulus, which
can be measured using a 1 mN von Frey filament (non-
nociceptive). Hyperalgesia is defined as increased pain reaction
induced by a nociceptive stimulus. In this model hyperalgesia
was evoked by a 16 mN von Frey filament.
Following the concept of late stage diversification a large set of
differently substituted perhydroquinoxalines has been prepared.
The κ receptor affinity, κ-agonistic activity, receptor selectivity,
and polarity of the κ agonists were modulated by the additional
substituent at N-4 outside the ethylenediamine κ pharmaco-
phore. The very polar secondary amine 14b (log DpH7,4 = 0.21)
did not penetrate an artificial blood−brain barrier but was
efficacious (ED50 = 2.35 mg/kg) in mustard oil induced
spontaneous visceral pain measures. Differential efficacy on
centrally mediated referred allodynia and hyperalgesia depend-
ing on peripheral/systemic (iv) or central (icv) administration
supports the concept of peripherally acting κ-opioid analgesics.
EXPERIMENTAL SECTION
Chemistry. General. For flash chromatography (fc), silica gel 60,
40−64 μm (Merck) was used. Data presented in parentheses include
diameter of the column, length of the column, eluent, fraction size, and
■
Intravenous application of a dose of 21.5 mg/kg of the
pyrrolidine 14a led to complete inhibition of visceral pain
6854
dx.doi.org/10.1021/jm500940q | J. Med. Chem. 2014, 57, 6845−6860

Journal of Medicinal Chemistry
Article
Figure 6. Analgesic activity after iv administration of various doses of hydroxypyrrolidine 14b in the mustard oil assay: (A) spontaneous pain
behavior; (B) referred allodynia; (C) referred hyperalgesia.
Rf value. Melting points were determined with an SMP 3 (Stuart
Scientific) melting point apparatus and are uncorrected.
21 mL, 21 mmol) was added dropwise. The suspension was warmed to
rt and stirred for 20 min. A solution of dilactam 11b (1.29 g, 3.8
mmol) in THF (65 mL) was added to the freshly prepared solution of
AlH3 at 0 °C. The mixture was stirred at 0 °C for 45 min and at rt for
20 min. Then 2 M NaOH (13 mL) was added dropwise under cooling
and the mixture was extracted with CH2Cl2 (5 × 50 mL). The organic
layer was dried (Na2SO4), filtered, and concentrated in vacuo. Rf =
0.26 (CH2Cl2/MeOH/NH3 = 9:1:0.1), Rf = 0.19 (CH3OH). Pale
yellow solid, mp 142 °C, yield 908 mg (77%). C19H29N3O (315.5).
1H NMR (400 MHz) and 13C NMR (100 MHz) data were
obtained using a Mercury 400BB spectrometer (Varian). δ in ppm is
referenced to tetramethylsilane, and coupling constants are given with
0.5 Hz resolution.
According to HPLC methods A and B, the purity of all test
compounds was greater than 95%. For details of the methods, see
Supporting Information.
Synthetic Procedures. (4aRS,5SR,8aRS)-1-Benzyl-5-(3-hy-
droxypyrrolidin-1-yl)perhydroquinoxaline-2,3-dione (11b). A
mixture of primary amine 10 (144 mg, 0.53 mmol), racemic 1,4-
dibromobutan-2-ol (purity 85%, 1.15 g, 4.20 mmol, 0.57 mL),
NaHCO3 (300 mg, 3.57 mmol), and CH3CN (16 mL) was heated to
reflux for 24 h. NaHCO3 was filtered off, and the solvent was removed
in vacuo. The residue was dissolved in CH2Cl2, and the solution was
extracted with 2 M HCl (3×). Then 2 M NaOH was added to the
combined aqueous layers until pH 8 was obtained and the alkaline
solution was extracted with CH2Cl2 (3×). The combined organic layer
was dried (Na2SO4), filtered, concentrated in vacuo, and the residue
was purified by fc (2 cm, 17 cm, acetone/MeOH/Et2NH 9.5:0.5:0.1, 5
mL, Rf = 0.63 (CH3OH)). Pale yellow solid, mp 151 °C, yield 143 mg
(79%). C19H25N3O3 (343.4). MS (EI): m/z (%) = 343 (M+, 18), 325
((M − H2O)+, 52), 258 ((M − C4H7NO)+, 19), 252 ((M − C7H7)+,
MS (ESI): m/z (%) = 316 (MH+, 100). MS (EI): m/z (%) = 315 (M+,
+
8), 297 ((M − H20)+, 100), 91 (C7H7 , 93). IR: ν
̃
(cm−1) = 3400−
3100 (m, ν (O−H)), 3276 (m, ν (N−H)), 1336 (w, δ (OH)), 1132
(w, ν (C−OH)). H NMR (CDCl3): δ (ppm) = 1.08−1.35 (m, 3 H,
1
6-Ha, 7-Ha, 8-Ha), 1.62−1.81 (m, 2 H, 6-He, N(CH2CH2)), 1.81−1.90
(m, 1 H, 7-He), 1.96−2.09 (m, 2 H, 8a-H, N(CH2CH2)), 2.14 (td, 2J =
3J = 11.3 Hz, 3J = 3.2 Hz, 1 H, 2-Ha), 2.18−2.24 (m, 1 H, 8-He), 2.40
(t, 3J = 10.6 Hz, 1 H, 4a-H), 2.58−2.95 (m, 8 H, 2-H, 3-H (2 H), 5-H,
N(CH2CHOH) (2 H), N(CH2CH2) (2 H)), 3.13 (d, 3J = 13.3 Hz, 1
H, Ph-CH2), 4.13 (d, 3J = 13.2 Hz, 1 H, Ph-CH2), 4.24−4.31 (m, 1 H,
N(CH2CHOH)), 7.24−7.32 (m, 5 H, Ph-H). Signals for the NH and
OH protons are not visible in the spectrum. 13C NMR (CDCl3): δ
(ppm) = 21.9/22.2 (C-6), 22.8/22.9 (C-7), 28.9 (C-8), 34.8/34.9
(N(CH2CH2)), 46.2 (C-3), 47.1 (N(CH2CH2)), 53.2 (C-2), 55.7
(N(CH2CHOH)), 58.1 (Ph-CH2), 60.5/60.7 (C-5), 62.5/62.6 (C-4a),
66.0/66.1 (C-8a), 71.2/71.3 (N(CH2CHOH)), 127.0 (Ph-C), 128.4
(2 C, Ph-C), 129.4 (2 C, Ph-C), 139.3 (quart. Ph-C). HPLC (method
A): purity 95.4%, tR = 7.1 min.
21), 234 ((M − H2O − C7H7)+, 100), 167 ((M − C7H7 − C4H7NO)+,
+
33), 91 (C7H7 , 36). IR: ν
̃
(cm−1) = 3458 (m, ν (O−H)), 3202 (m, ν
(N−H)), 1695 (s, ν (CO), tert amide), 1667 (s, ν (CO), sec
amide), 1327 (w, δ (OH)), 1142 (w, ν (C−OH)). 1H NMR (CDCl3):
δ (ppm) = 1.18−1.40 (m, 3 H, 6-Ha, 7-Ha, 8-Ha), 1.66−1.88 (m, 3 H,
N(CH2CH2), 6-He, 7-He), 2.05−2.16 (m, 2 H, N(CH2CH2), 8-He),
1-{(4aRS,8SR,8aSR)-4-Benzyl-8-[(3SR)- and (3RS)-3-hydroxy-
pyrrolidin-1-yl]perhydroquinoxalin-1-yl}-2-(3,4-
dichlorophenyl)ethan-1-one (13b). 2-(3,4-Dichlorophenyl)acetyl
chloride (1.8 g, 8.1 mmol) was added dropwise to a solution of 12b
(2.6 g, 8.1 mmol) in absolute CH2Cl2 (200 mL), and the mixture was
stirred at rt. After 30 min, 2 M NaOH (200 mL) was added, and the
mixture was stirred overnight. The aqueous layer was separated, and
the organic layer was extracted with 1 M HCl (3×). The pH value of
the aqueous layer was adjusted to pH 8 by addition of 2 M NaOH.
The alkaline aqueous layer was extracted with CH2Cl2 (3×). The
combined organic layer was dried (Na2SO4), filtered, and concentrated
in vacuo. Rf = 0.14 (CH3OH); Rf = 0.55 and 0.62 (CH2Cl2/MeOH/
NH3 = 9:1:0.1). Pale yellow solid, mp 78 °C, yield 3.7 g (90%).
C27H33CI2N3O2 (502.5) MS (ESI): m/z (%) = 502 (MH+, 2 × 35Cl,
3
2.50 (q, J = 8.1 Hz, 1 H, N(CH2CH2)), 2.57−2.78 (m, 3 H, 5-H,
N(CH2CHOH), N(CH2CH2)), 2.80−2.88 (m, 1 H, N(CH2CHOH)),
3
3.42/3.43 (2 t, J = 10.8 Hz, 1 H, 4a-H), 3.54−3.64 (m, 1 H, 8a-H),
4.24−4.32 (m, 1 H, N(CH2CHOH)), 4.59 (d, 2J = 15.7 Hz, 1 H, Ph-
CH2), 5.06 (d, 2J = 15.7 Hz, 1 H, Ph-CH2), 7.22−7.28 (m, 3 H, ortho-
Ph-H, para-Ph-H), 7.30−7.37 (m, 2 H, meta-Ph-H). Signals for the
NH and OH protons are not visible in the spectrum. 13C NMR
(CDCl3): δ (ppm) = 19.9/20.2 (C-6), 21.8/21.9 (C-7), 27.9 (C-8),
33.7/33.9 (N(CH2CH2)), 43.5 (N(CH2CH2)), 45.6 (Ph-CH2), 55.9/
56.2 (C-4a), 57.1 (N(CH2CHOH)), 58.9 (C-8a), 59.0/59.1 (C-5),
70.3/70.6 (N(CH2CHOH)), 126.8 (2 C, Ph-C), 127.2 (Ph-C), 128.8
(2 C, Ph-C), 137.1 (quart. Ph-C), 158.2/158.3 (C-3), 160.2/160.3 (C-
2). HPLC (method A): purity 97.7%, tR = 12.8 min. HPLC (method
B): purity 97.0%, tR = 11.3 min.
100), 504 (MH+, 35Cl/37Cl, 65), 506 (MH+, 2 × 37Cl, 12). IR: ν
̃
(cm−1) = 3100−3500 (m, ν (O−H)), 1636 (s, ν (CO)), 875 (w,
out-of-plane (Ar−H)), 814 (w, out-of-plane (Ar−H)). 1H NMR
(toluene-d8, 100 °C): δ (ppm) = 0.86−1.01 (m, 2 H, 5-Ha, 7-Ha),
1.01−1.12 (m, 1 H, 6-Ha), 1.45−1.65 (m, 3 H, 6-He, 5-He,
N(CH2CH2)), 1.67−1.86 (m, 2 H, 7-He, N(CH2CH2)), 1.86−1.92
(m, 1 H, 3-H), 2.23−2.32 (m, 1 H, 8-H), 2.36−2.52 (m, 3 H, 2-H, 3-
(4aRS,5RS,8aSR)-1-Benyzl-5-[(3SR)- and (3RS)-3-hydroxypyr-
rolidin-1-yl]perhydroquinoxaline (12b). Under N2, dry AlCl3
(940 mg, 6.8 mmol) was dissolved in absolute THF (52 mL), and
the solution was cooled to 0 °C. A solution of LiAlH4 (1.0 M in THF,
6855
dx.doi.org/10.1021/jm500940q | J. Med. Chem. 2014, 57, 6845−6860

Journal of Medicinal Chemistry
Article
H, N(CH2CHOH)), 2.60−2.66 (m, 1 H, N(CH2CHOH)), 2.67−2.76
(m, 2 H, N(CH2CH2)), 2.87/2.96 (2 d, 2J = 13.7 Hz, 1 H, Ph-CH2-N),
2.93−3.01 (m, 1 H, 2-H), 3.20−3.29 (m, 3 H, Ph-CH2-CO (2 H),
CH2-CO), 3.62−3.71 (m, 1 H, 4a-H), 6.87 (d broad, 3J = 8.1 Hz, 1
H, Ph-6-H), 7.02 (d, 3J = 8.4 Hz, 1 H, Ph-5-H), 7.24 (d, 4J = 1.9 Hz, 1
H, Ph-2-H). HPLC (method A): purity 97.4%, tR = 15.0 min, HPLC
(method B): purity 97.7%, tR = 12.0 min.
2
8a-H), 3.49−3.58 (m, 1 H, 4a-H), 3.62/3.70 (2 d, J = 13.4 Hz, 1 H,
Ph-CH2-N), 3.96−4.02 (m, 1 H, N(CH2CHOH)), 6.85−6.89 (m, 2
H, Ph-H), 7.02−7.08 (3 H, Ph-H), 7.09−7.13 (m, 2 H, Ph-H), 7.21−
7.23 (m, 1 H, Ph-H). A signal for the OH proton is not visible in the
spectrum. 13C NMR (toluene-d8, 100 °C): δ (ppm) = 23.4/23.7 (C-
6), 25.7 (C-5), 31.0/31.2 (C-7), 35.3/35.5 (N(CH2CH2)), 41.6/41.7
(Ph-CH2-CO), 46.4 (N(CH2CH2), 48.1 (C-2), 53.2/53.4 (N-
(CH2CHOH)), 56.9 (C-3), 57.1/57.4 (Ph-CH2-N), 58.2 (C-4a),
62.1/62.3 (C-8), 67.1 (C-8a), 71.4/71.5 (N(CH2CHOH)), 127.3 (2
C, Ph-C), 128.4 (Ph-C), 128.7 (2 C, Ph-C), 129.0 (Ph-C), 129.3 (Ph-
C), 128.9 (Ph-C), 130.7 (Ph-C), 131.5/131.6 (Ph-C), 137.1/137.3
(quart. C), 140.3 (quart. C), 170.4/170.7 (CO). HPLC (method
A): purity 98.0%, tR = 16.6 min. HPLC (method B): purity 96.6%, tR =
13.7 min.
2-(3,4-Dichlorophenyl)-1-{(4aRS,8SR,8aSR)-8-[(3SR)- and
(3RS)-3-hydroxypyrrolidin-1-yl]perhydroquinoxalin-1-yl}-
ethan-1-one (14b). A mixture of 13b (373 mg, 0.74 mmol), conc
HCl (7.4 mL), Pd/C (158 mg), and THF/H2O (1:1, 74 mL) was
stirred at rt under H2 (1 bar) for 30 min. The suspension was filtered,
and the organic solvent was removed in vacuo. The pH value of the
aqueous layer was adjusted to pH 8 by addition of 2 M NaOH. The
aqueous layer was extracted with CH2Cl2 (5×). The combined organic
layer was dried (Na2SO4), filtered, concentrated in vacuo, and the
2-(3,4-Dichlorophenyl)-1-{(4aRS,8SR,8aSR)-4-methyl-8-
[ ( 3 S R ) - a n d ( 3 R S ) - 3 - h y d r o x y p y r r o l i d i n - 1 - y l ] -
perhydroquinoxalin-1-yl}ethan-1-one (15b). Formalin (37%,
170 mg, 2.1 mmol) was dissolved in MeOH (5 mL), and NaBH3CN
(132 mg, 2.1 mmol) was added. The pH value of the mixture was
adjusted with conc acetic acid to pH 5. Then a solution of 14b (86 mg,
0.21 mmol) in MeOH (15 mL) was added and the mixture was stirred
at rt for 15 min. A saturated Na2CO3 solution (15 mL) was added, and
the mixture was stirred at rt for 2 h. The precipitate was filtered off,
and MeOH was evaporated in vacuo. The remaining aqueous layer was
extracted with CH2Cl2 (5×). The combined organic layer was dried
(Na2SO4), filtered, concentrated in vacuo, and the residue was purified
by fc (2 cm, 16 cm, CH2Cl2/MeOH/NH3 = 9:1:0.1, 5 mL, Rf = 0.34
and 0.36 (CH2Cl2/MeOH/NH3 = 9:1:0.1), Rf = 0.05 (CH3OH)). Pale
yellow resin, yield 54 mg (60%). C21H29CI2N3O2 (426.4). MS (ESI):
m/z (%) = 426 (MH+, 2 × 35Cl, 100), 428 (MH+, 35Cl/37Cl, 77), 430
̃
(MH+, 2 × 37Cl, 12). IR: ν (cm−1) = 3600−3000 (m, ν (O−H)), 1636
(s, ν (CO)), 874 (w, out-of-plane (Ar−H)), 812 (w, out-of-plane
(Ar−H)). 1H NMR (toluene-d8, 100 °C): δ (ppm) = 0.82 (q broad, 2J
= 3J = 12.8 Hz, 1 H, 7-Ha), 0.90 (qd, 2J = 3J = 12.3 Hz, 3J = 3.1 Hz, 1
H, 5-Ha), 0.98−1.10 (m, 1 H, 6-Ha), 1.44−1.63 (m, 4 H, N(CH2CH2)
(2 H), 6-He, 5-He), 1.65−1.77 (m, 1 H, 7-He), 1.79−1.88 (m, 1 H, 3-
H), 1.91/1.92 (s, 3 H, CH3), 1.93−1.99 (m, 1 H, 8-H), 2.32−2.45 (m,
1 H, N(CH2CHOH)), 2.45−2.52 (m, 1 H, 3-H), 2.52−2.60 (m, 1 H,
N(CH2CHOH)), 2.62−2.68 (m, 1 H, N(CH2CH2)), 2.68−2.74 (m, 1
residue was purified by fc (3 cm, 18 cm, CH2Cl2/MeOH/NH3
=
9:1:0.1, 10 mL, Rf = 0.25 and 0.29 (CH2Cl2/MeOH/NH3 = 9:1:0.1),
Rf = 0.04 (CH3OH)). Pale yellow resin, yield 272 mg (89%).
C20H27CI2N3O2 (412.4 g). MS (ESI): m/z (%) = 412 (MH+, 2 × 35Cl,
3
H, N(CH2CH2)), 2.80 (t, J = 10.3 Hz, 1 H, H-8a), 2.95−3.11 (m, 2
100), 414 (MH+, 35Cl/37Cl, 62), 416 (MH+, 2 × 37Cl, 10). IR: ν
̃
H, 2-H (2 H)), 3.22/3.25 (d, 2J = 15.4 Hz, 1 H, Ph-CH2-CO), 3.32
(cm−1) = 3100−3500 (m, ν (O−H, N−H)), 1633 (s, ν (CO)), 875
2
(d, J = 15.3 Hz, 1 H, Ph-CH2-CO), 3.61−3.70 (m, 1 H, 4a-H),
1
3.94−4.00 (m, 1 H, N(CH2CHOH)), 6.87 (d broad, 3J = 8.6 Hz, 1 H,
Ph-6-H), 7.06 (d, 3J = 8.2 Hz, 1 H, Ph-5-H), 7.22 (d, 4J = 1.9 Hz, 1 H,
Ph-2-H). A signal for the OH proton is not visible in the spectrum.
HPLC (method A): purity 98.8%, tR = 14.1 min. HPLC (method B):
purity 98.2%, tR = 11.4 min.
(w, out-of-plane (Ar−H)), 828 (w, out-of-plane (Ar−H)). H NMR
(DMSO-d6, 100 °C): δ (ppm) = 1.11−1.33 (m, 3 H, 7-Ha, 5-Ha, 6-
Ha), 1.45−1.55 (m, 1 H, N(CH2CH2)), 1.71−1.80 (m, 2 H, 6-He, 7-
He), 1.81−1.91 (m, 2 H, 5-He, N(CH2CH2)), 2.50−2.57 (m, 2 H, 3-H,
8-H), 2.61−2.85 (m, 5 H, 2-H (2 H), N(CH2CHOH), N(CH2CH2)
(2 H)), 2.86−2.93 (m, 2 H, 3-H, N(CH2CHOH),), 3.18−3.27 (m, 1
1-[(4aRS,8SR,8aRS)-4-Butyl-8-(pyrrolidin-1-yl)-
perhydroquinoxalin-1-yl]-2-(3,4-dichlorophenyl)ethan-1-one
(16a). Butyraldehyde (93 mg, 1.3 mmol) was dissolved in MeOH (5
mL), and NaBH3CN (82 mg, 1.3 mmol) was added. The pH value of
the mixture was adjusted with conc acetic acid to pH 5. Then a
solution of 14a (101 mg, 0.25 mmol) in MeOH (15 mL) was added
and the mixture was stirred at rt overnight. A saturated Na2CO3
solution (15 mL) was added, and the mixture was stirred at rt for 15
min. The precipitate was filtered off, and MeOH was evaporated in
vacuo. The remaining aqueous layer was extracted with CH2Cl2 (3×).
The combined organic layer was dried (Na2SO4), filtered, concen-
trated in vacuo, and the residue was purified by fc (2 cm, 15 cm,
CH2Cl2/MeOH/NH3 = 9.5:0.5:0.05, 5 mL, Rf = 0.67 (CH2Cl2/
MeOH/NH3 = 9:1:0.1), Rf = 0.09 (MeOH)). Colorless resin, yield 45
mg (40%). C24H35CI2N3O (452.5). MS (ESI): m/z (%) = 452 (MH+,
2 × 35Cl, 100), 454 (MH+, 35Cl/37Cl, 61), 456 (MH+, 2 × 37Cl, 11).
2
H, 8a-H), 3.31−3.42 (m, 1 H, 4a-H), 3.72−3.79 (d broad, J = 16.2
Hz, 1 H, Ph-CH2-CO), 3.87/3.89 (2 d, 2J = 15.5 Hz, 1 H, Ph-CH2-
CO), 4.07−4.15 (m, 1 H, N(CH2CHOH)), 7.23−7.28 (m, 1 H, Ph-
6-H), 7.45−7.49 (m, 2 H, Ph-5-H, Ph-2-H). Signals for the NH and
OH protons are not visible in the spectrum. HPLC (method A): purity
97.3%, tR = 13.8 min. HPLC (method B): purity 96.6%, tR = 11.2 min.
2-(3,4-Dichlorophenyl)-1-[(4aRS,8SR,8aRS)-4-methyl-8-(pyr-
rolidin-1-yl)perhydroquinoxalin-1-yl]ethan-1-one (15a). For-
malin (37%, 223 mg, 2.7 mmol) was dissolved in MeOH (5 mL),
and NaBH3CN (17.2 mg, 0.27 mmol) was added. The pH value of the
mixture was adjusted with conc acetic acid to pH 5. Then a solution of
14a (109 mg, 0.27 mmol) in MeOH (15 mL) was added and the
mixture was stirred at rt for 1.5 h. A saturated Na2CO3 solution (12
mL) was added, and the mixture was stirred at rt for 15 min. The
precipitate was filtered off, and MeOH was evaporated in vacuo. The
remaining aqueous layer was extracted with CH2Cl2 (5×). The
combined organic layer was dried (Na2SO4), filtered, concentrated in
vacuo, and the residue was purified by fc (2 cm, 16 cm, CH2Cl2/
MeOH/NH3 = 9.5:0.5:0.05, 5 mL, Rf = 0.48 (CH2Cl2/MeOH/NH3 =
9:1:0.1), Rf = 0.04 (MeOH)). Pale yellow resin, yield 34 mg (31%).
C21H29CI2N3O (410.4). MS (ESI): m/z (%) = 410 (MH+, 2 × 35Cl,
IR: ν
̃
(cm−1) = 1641 (s, ν (CO)), 876 (w, out-of-plane (Ar−H)),
790 (w, out-of-plane (Ar−H)). 1H NMR (toluene-d8, 100 °C): δ
(ppm) = 077−0.82 (t, 3J = 8.0 Hz, 3 H, N(CH2CH2CH2CH3)), 0.83−
2
3
3
0.92 (m, 1 H, 7-Ha), 0.98 (qd, J = J = 12.1 Hz, J = 3.0 Hz, 1 H, 5-
Ha), 1.07 (qt, 2J = 3J = 13.3 Hz, 3J = 3.3 Hz, 1 H, 6-Ha), 1.10−1.22 (m,
4 H, N(CH2CH2CH2CH3)), 1.47−1.55 (m, 5 H, 6-He, N-
(CH2CH2)2), 1.62−1.68 (m, 1 H, 5-He), 1.72−1.78 (m, 1 H, 7-He),
1.94−2.00 (m, 1 H, 3-H), 2.20 (td, 3J = 9.9 Hz, 3J = 4.1 Hz, 1 H, 8-H),
2.38−2.52 (m, 5 H, 3-H, N(CH2CH2)2 (4 H)), 2.61 (m, 1 H, 2-H),
100), 412 (MH+, 35Cl/37Cl, 62), 414 (MH+, 2 × 37Cl, 10). IR: ν
̃
(cm−1) = 1639 (s, ν (CO)), 873 (w, out-of-plane (Ar−H)), 789 (w,
out-of-plane (Ar−H)). 1H NMR (toluene-d8, 100 °C): δ (ppm) = 0.84
(qd, 2J = 3J = 12.7 Hz, 3J = 3.9 Hz, 1 H, 7-Ha), 0.94 (qd, 2J = 3J = 12.3
Hz, 3J = 3.2 Hz, 1 H, 5-Ha), 1.07 (qt, 2J = 3J = 13.3 Hz, 3J = 3.1 Hz, 1
H, 6-Ha), 1.47−1.52 (m, 5 H, 6-He, N(CH2CH2)2), 1.61−1.67 (m, 1
H, 5-H), 1.71−1.78 (m, 1 H, 7-He), 1.81−1.88 (m, 1 H, 3-H), 1.94 (s,
3
2.97 (t, J = 8.7 Hz, 3 H, 8a-H, N(CH2CH2CH2CH3) (2 H)), 3.18−
3.27 (m, 1 H, 4a-H), 3.32−3.39 (m, 3 H, Ph-CH2-CO (2 H), 2-H),
6.89−6.92 (m, 1 H, Ph-6-H), 7.03 (d, 3J = 8.2 Hz, 1 H, Ph-5-H), 7.27
(d, 4J = 1.9 Hz, 1 H, Ph-2-H). HPLC (method A): purity 97.5%, tR =
17.5 min.
Methyl (4aRS,8SR,8aRS)-1-[2-(3,4-Dichlorophenyl)acetyl]-8-
(pyrrolidin-1-yl)perhydroquinoxaline-4-carboxylate (23a).
Under N2, 14a (100.9 mg, 0.25 mmol) was dissolved in CH2Cl2 (13
3
3
3 H, CH3), 1.99 (td, J = 10.0 Hz, J = 4.3 Hz, 1 H, 8-H), 2.41−2.48
(m, 3 H, 3-H, N(CH2CH2)2), 2.48−2.54 (m, 2 H, N(CH2CH2)2),
2.84 (t, 3J = 10.5 Hz, 1 H, 8a-H), 3.01−3.13 (m, 2 H, 2-H (2 H)), 3.27
(d, 2J = 15.6 Hz, 1 H, Ph-CH2-CO), 3.35 (d, 2J = 15.5 Hz, 1 H, Ph-
6856
dx.doi.org/10.1021/jm500940q | J. Med. Chem. 2014, 57, 6845−6860

Journal of Medicinal Chemistry
Article
mL) and methyl chloroformate (28.9 mg, 0.31 mmol) was added. The
mixture was stirred at rt for 2 h. The solvent was removed in vacuo,
and the residue was purified by fc (2 cm, 16 cm, CH2Cl2/MeOH/NH3
= 9.5:0.5:0.05, 5 mL, Rf = 0.76 (CH2Cl2/MeOH/NH3 = 9:1:0.1), Rf =
0.17 (CH3OH)). Pale yellow resin, yield 67 mg (59%).
C22H29CI2N3O3 (454.4). MS (ESI): m/z (%) = 454 (MH+, 2 ×
device with adjustable temperature and tumbling speed (scientific
workshop of the institute). Vortex: Vortex Genie 2 (Thermo Fisher
Scientific, Langenselbold, Germany). Harvester: MicroBeta Filter-
Mate-96 harvester. Filter: Printed Filtermat types A and B. Scintillator:
Meltilex (type A or B) solid state scintillator. Scintillation analyzer:
MicroBeta Trilux (all PerkinElmer LAS, Rodgau-Jugesheim, Ger-
̈
35Cl, 100), 456 (MH+, 35Cl/37Cl, 66), 458 (MH+, 2 × 37Cl, 10). IR: ν
̃
many). Chemicals and reagents were purchased from different
commercial sources and of analytical grade.
(cm−1) = 1700 (s, ν (N-CO-O)), 1646 (s, ν (N-CO)), 873 (w,
out-of-plane (Ar−H)). 1H NMR (CD2Cl2, −40 °C): δ (ppm) = 0.74−
0.85 (m, 1 H, 5-Ha), 1.10−1.40 (m, 4 H, 5-He, 6-Ha, 7-Ha, 7-He),
1.55−1.62 (m, 1 H, 6-He), 1.63−1.75 (m, 4 H, N(CH2CH2)2), 1.82−
1.92 (m, 2 H, 2-H, 3-H), 2.00−2.07 (m, 1 H, 8a-H), 2.56−2.71 (m, 4
H, N(CH2CH2)2), 3.00−3.07 (m, 1 H, 2-H), 3.44−3.54 (m, 3 H,
COOCH3), 3.55−3.61 (m, 2 H, Ph-CH2-CO), 3.76−3.82 (m, 1 H,
8-H), 3.89−3.97 (m, 1 H, 4a-H), 4.00−4.12 (m, 1 H, 3-H), 6.98−7.03
(m, 1 H, Ph-6-H), 7.28−7.32 (m, 1 H, Ph-2-H), 7.33−7.38 (m, 1 H,
Ph-5-H). HPLC (method A): purity 98.4%, tR = 18.6 min. HPLC
(method B): purity 97.3%, tR = 15.4 min.
Preparation of Membrane Homogenates from Guinea Pig
Brain.31,32 Five guinea pig brains were homogenized with the potter
(500−800 rpm, 10 up-and-down strokes) in 6 volumes of cold 0.32 M
sucrose. The suspension was centrifuged at 1200g for 10 min at 4 °C.
The supernatant was separated and centrifuged at 23 500g for 20 min
at 4 °C. The pellet was resuspended in 5−6 volumes of buffer (50 mM
Tris, pH 7.4) and centrifuged again at 23 500g (20 min, 4 °C). This
procedure was repeated twice. The final pellet was resuspended in 5−6
volumes of buffer and frozen (−80 °C) in 1.5 mL portions containing
about 1.5 mg protein/mL.
Methyl (4aRS,8SR,8aRS)-1-[2-(3,4-Dichlorophenyl)acetyl]-8-
[ ( 3 S R ) - a n d ( 3 R S ) - 3 - h y d r o x y p y r r o l i d i n - 1 - y l ] -
perhydroquinoxaline-4-carboxylate (23b). Under N2, 14b (132
mg, 0.32 mmol) was dissolved in CH2Cl2 (20 mL) and methyl
chloroformate (30 mg, 0.32 mmol) was added dropwise. The mixture
was stirred at rt for 3 h. Then it was concentrated in vacuo, and the
residue was purified by fc (2 cm, 15 cm, CH2Cl2/MeOH/NH3
Protein Determination. The protein concentration was determined
by the method of Bradford,56 modified by Stoscheck.57 The Bradford
solution was prepared by dissolving 5 mg of Coomassie Brilliant Blue
G 250 in 2.5 mL of EtOH (95%, v/v). Then 10 mL of deionized H2O
and 5 mL of phosphoric acid (85%, m/v) were added to this solution.
The mixture was stirred and filled to a total volume of 50.0 mL with
deionized water. The calibration was carried out using bovine serum
albumin as a standard in nine concentrations (0.1, 0.2, 0.4, 0.6, 0.8, 1.0,
1.5, 2.0, and 4.0 mg/mL). In a 96-well standard multiplate, an amount
of 10 μL of the calibration solution or an amount of 10 μL of the
membrane receptor preparation was mixed with 190 μL of the
Bradford solution, respectively. After 5 min, the UV absorption of the
protein−dye complex at λ = 595 nm was measured with a plate reader
(Tecan Genios, Tecan, Crailsheim, Germany).
General Protocol for the Binding Assays. The test compound
solutions were prepared by dissolving approximately 10 μmol (usually
2−4 mg) of test compound in DMSO so that a 10 mM stock solution
was obtained. To obtain the required test solutions for the assay, the
DMSO stock solution was diluted with the respective assay buffer. The
Filtermats were presoaked in 0.5% aqueous polyethylenimine solution
for 2 h at room temperature before use. All binding experiments were
carried out in duplicate in 96-well multiplates. The concentrations
given are the final concentrations in the assay. Generally, the assays
were performed by addition of 50 μL of the respective assay buffer, 50
μL of test compound solution in various concentrations (10−5, 10−6,
10−7, 10−8, 10−9, and 10−10 mol/L), 50 μL of corresponding
radioligand solution, and 50 μL of the respective receptor preparation
into each well of the multiplate (total volume 200 μL). The receptor
preparation was always added last. During the incubation, the
multiplates were shaken at a speed of 500−600 rpm at the specified
temperature. Unless otherwise noted, the assays were terminated after
120 min by rapid filtration using the harvester. During the filtration
each well was washed five times with 300 μL of water. Subsequently,
the Filtermats were dried at 95 °C. The solid scintillator was melted on
the dried Filtermats at a temperature of 95 °C for 5 min. After
solidification of the scintillator at room temperature, the trapped
radioactivity in the Filtermats was measured with the scintillation
analyzer. Each position on the Filtermat corresponding to one well of
the multiplate was measured for 5 min with the 3H-counting protocol.
The overall counting efficiency was 20%. The IC50 values were
calculated with the program GraphPad Prism 3.0 (GraphPad Software,
San Diego, CA, USA) by nonlinear regression analysis. Subsequently,
the IC50 values were transformed into Ki values using the equation of
Cheng and Prusoff.58 The Ki values are given as mean value SEM
from three independent experiments.
9:1:0.05, 5 mL, Rf = 0.36 and 0.41 (CH2Cl2/MeOH/NH3
=
9:1:0.1), Rf = 0.22 (MeOH)). The solvent was removed in vacuo.
The solid residue was dissolved in CH2Cl2, and the solution was
extracted with 1 M HCl (3×). The pH value of the aqueous layer was
adjusted with 2 M NaOH to pH 8, and the solution was extracted with
CH2Cl2 (3×). The combined organic layer was dried (Na2SO4),
filtered, concentrated in vacuo, and the residue was purified once more
by fc (2 cm, 15 cm, CH2Cl2/MeOH/NH3 = 9.5:0.5:0.05, 5 mL). The
solvent was removed in vacuo, and the residue was purified by
preparative HPLC method C (eluent, MeOH/H2O/Et2NH
70:30:0.1). The organic solvent was removed in vacuo, and the
aqueous layer was extracted with CH2Cl2 (3×). The combined organic
layer was dried (Na2SO4), filtered, and concentrated in vacuo. Pale
yellow resin, yield 93 mg (62%). C22H29CI2N3O4 (470.4). MS (ESI):
m/z (%) = 470 (MH+, 2 × 35Cl, 100), 472 (MH+, 35Cl/37Cl, 61), 474
̃
(MH+, 2 × 37Cl, 11). IR: ν (cm−1) = 3600−3100 (m, ν (O−H)), 1693
(s, ν (MeO-CO)), 1642 (s, ν (N-CO)), 875 (w, out-of-plane
1
(Ar−H)), 792 (w, out-of-plane (Ar−H)). H NMR (CD2Cl2, −50
°C): δ (ppm) = 1.05−1.19 (m, 3 H, 5-Ha, 6-Ha, 7-Ha), 1.23−1.35 (m,
2 H, 5-He, 7-He), 1.57−1.74 (m, 2 H, 6-He, N(CH2CH2)), 1.76−1.93
(m, 3 H, 2-H, 3-H, N(CH2CH2)), 1.97−2.06 (m, 1 H, 8a-H), 2.56−
2.68 (m, 2 H, N(CH2CH2)), 2.71−2.87 (m, 2 H, N(CH2CHOH)),
3.00−3.08 (m, 1 H, 2-H), 3.49−3.54 (m, 1 H, Ph-CH2-CO), 3.58
(s, 3 H, COOCH3), 3.65 (d, 2J = 15.9 Hz, 1 H, Ph-CH2-CO), 3.73−
3.80 (m, 1 H, 8-H), 3.85−3.98 (m, 1 H, 4a-H), 4.01−4.06 (m, 1 H, 3-
H), 4.14−4.19 (m, 1 H, N(CH2CHOH)), 7.04−7.08 (m, 1 H, Ph-6-
H), 7.26−7.30 (m, 1 H, Ph-2-H), 7.32−7.39 (m, 1 H, Ph-5-H). A
signal for the OH proton is not visible in the spectrum. HPLC
(method A): purity 99.9%, tR = 17.4 min. HPLC (method B): purity
99.3%, tR = 14.8 min.
Molecular Modeling. The conformational analysis was performed
with force field MMFF94x of the molecular modeling program MOE
(molecular operating environment), version 2013.08 (Chemical
Computing Group AG).
Receptor Affinity Using Animal Sources as Receptor
Material. Materials. The guinea pig brains and rat liver for the κ,
σ1, and σ2 receptor binding assays were commercially available
(Harlan-Winkelmann, Borchen, Germany). Pig brains for the NMDA
assay were obtained from the local slaughterhouse. The following
pieces of equipment were used. Homogenizer: Elvehjem Potter (B.
Braun Biotech International, Melsungen, Germany). Cooling
centrifuge model Rotina 35R (Hettich, Tuttlingen, Germany) and
high-speed cooling centrifuge model Sorvall RC-5C Plus (Thermo
Fisher Scientific, Langenselbold, Germany). Multiplates: standard 96-
well multiplates (Diagonal, Muenster, Germany). Shaker: self-made
Protocol of the κ Receptor Binding Assay (Guinea Pig Brain
Homogenates). The assay was performed with the radioligand [3H]U-
69,593 (55 Ci/mmol, Amersham, Little Chalfont, U.K.). The thawed
guinea pig brain membrane preparation (about 100 μg of the protein)
was incubated with various concentrations of test compounds, 1 nM
[3H]U-69,593, and Tris-MgCl2 buffer (50 mM, 8 mM MgCl2, pH 7.4)
6857
dx.doi.org/10.1021/jm500940q | J. Med. Chem. 2014, 57, 6845−6860

Journal of Medicinal Chemistry
Article
at 37 °C. The nonspecific binding was determined with 10 μM
unlabeled U-69,593. The Kd value of U-69,593 is 0.69 nM.
vehicle (50 μL of PEG200). Compound or vehicle was given iv or icv
5 min before mustard oil. Seven animals were tested per group.
At 2−12 min after mustard oil administration the spontaneous pain
score was determined by counting and scoring of visceral pain
behaviors (score 1 = licking of abdominal wall; score 2 = stretching,
squashing, mounting, backward movement, or contraction of the flank
muscles.
At 20−40 min after mustard oil administration referred allodynia
and hyperalgesia were determined by tactile stimulations with von
Frey filaments. In the referred allodynia test the number of withdrawal
reactions against 10 stimulations with a 1 mN von Frey filament was
determined. In the hyperalgesia assay the withdrawal reactions against
10 stimulations with a 16 mN von Frey filament were counted and
scored (score 1 = lifting of abdomen, licking, movement; score 2 =
extrusion of flinching of hind paws, slight jumping, strong licking;
score 3 = strong jumping, vocalization).
Statistical Analysis. Data were analyzed by means of two-factor
analysis of variance (ANOVA) with repeated measures. Significance of
effects was analyzed by means of Wilks’ Lambda statistics. In the case
of a significant treatment effect, pairwise comparison was performed at
every test time point on raw data by Fischer’s least significant
difference test. Results were considered statistically significant if p <
0.5. ED50 values and 95% confidence intervals were calculated by linear
regression.
[35S]GTPγS Binding Assay. Agonistic Activity at the κ-Opioid
Receptor. The [35S]guanosine 5′-3-O-(thio)triphosphate (GTPγS)
assay was carried out as described in ref 45. The receptor material was
obtained from human HEK 293 (human embryonic kidney) cells.
Parameters were the following. Vehicle, 1.00% DMSO. Incubation
time, 30 min. Incubation temperature, 30 °C. Incubation buffer, 20
mM HEPES, pH 7.4, 100 mM NaCl, 10 mM MgCl2, 1 mM DTT, 1
mM EDTA. Quantification of bound [35S]GTPγS. Significance
criterion for agonists: >50% increase of bound [35S]GTPγS relative
to U-69,593 response. The EC50 values were determined by a
nonlinear, least-squares regression analysis using MathIQTM (ID
Business Solutions Ltd., U.K.). Reference standards were run as an
integral part of each assay to ensure the validity of the results obtained.
Passage of an Artificial Blood−Brain Barrier. Blood−Brain
Barrier Cell Culture Model. Pig brain cortex endothelial cells (BCECs)
were gently thawed and seeded (250 000/cm2) on rat tail collagen-
coated (0.54 mg/mL) Transwell microporous polycarbonate filter
inserts (Corning, Wiesbaden, Germany, 1.12 cm2 growth area, 0.4 μm
pore size) on DIV2 in plating medium (medium 199 Earle
supplemented with 10% newborn calf serum, 0.7 mM L-glutamine,
100 μg/mL gentamycin, 100 U/mL penicillin, 100 μg/mL
streptomycin (all Biochrom, Berlin, Germany)) in the apical
compartment at 37 °C with 5% CO2 and 100% humidity. After
PBCECs reached confluence (DIV4), the plating medium was
replaced by serum-free culture medium (Dulbecco’s modified Eagle
medium/Ham’s F 12 (1:1) containing 4.1 mM L-glutamine, 100 μg/
mL gentamycin, 100 U/mL penicillin, 100 μg/mL streptomycin (all
Biochrom, Berlin, Germany)) and 550 nM hydrocortisone (Sigma-
Aldrich, Deisenhofen, Germany) in order to induce differentiation.
Permeability ([14C]sucrose, test compounds) and barrier integrity
studies were started on day DIV6. For quantification of the barrier
integrity, the transendothelial electrical resistance (TER) device, a
cellZscope (nanoAnalytics, Muenster, Germany), was used. Only wells
with TER values of >600 Ω·cm2 and capacitance values between 0.45
and 0.6 μF/cm2 on day DIV6 indicate a confluent PBCEC monolayer
with sufficient barrier properties and were thus used for the respective
experiments.
ASSOCIATED CONTENT
* Supporting Information
■
S
Physical, spectroscopic, and purity data of all compounds;
synthetic methods; calculation and determination of log P and
log D7.4 values; details of the receptor binding assays and the
[35S]GTPγS binding assay; passage of an artificial blood−brain
barrier; experimental details of the phenylquinone writhing
assay and the mustard oil assay. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*Phone: +49-251-8333311. Fax: +49-251-8332144. E-mail:
wuensch@uni-muenster.de. Address: Bernhard Wunsch In-
■
Transfer over the Barrier Model. To study transfer across the in
vitro barrier, the model was exposed to the test compounds on the
apical side. Thereby test compounds were applied by replacing 10% of
the apical volume to reach finally concentration (500 μM). After 3 h
the amount of the test compound in the basolateral compartment was
recorded by RP-HPLC analysis.
̈
stitute of Pharmaceutical and Medicinal Chemistry, University
of Munster, Corrensstaße 48, D-48149 Munster, Germany.
̈
̈
Notes
Analgesic Activity in Animal Models of Visceral Pain.
Phenylquinone Writhing Assay.53 Male NMRI mice with a weight
of 25−35 g were used. Ten animals were used per compound dose.
Ten minutes after intravenous injection of 0.3 mL/mouse of a solution
ot the test compound, an aqueous solution of phenylquinone (2-
phenyl-1,4-benzoquinone, 0.02% in water with 5% ethanol, 0.35 mL)
was injected intraperitoneally. The phenylquinone solution was kept in
a water bath of 45 °C until use. Each animal was put into a single cage
and observed. The pain-induced writhing reactions (=stretching,
twisting a hind leg inward, contraction of abdomen) were counted for
15 min (5−20 min after ip injection of phenylquinone). Animals
treated with vehicle (iv) and animals treated with phenylquinone (ip)
served as negative and positive controls, respectively. Test compounds
were used in the standard dose of 10 mg/kg. If necessary, higher or
lower doses were employed.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the Deutsche Forschungsgemein-
schaft (DFG), SFB 656 MoBil, University of Munster,
̈
Germany, which is gratefully acknowledged.
ABBREVIATIONS USED
■
GPCR, G-protein-coupled receptor; JDTic, (R)-7-hydroxy-N-
{(S)-2-[(3R,4R)-4-(3-hydroxyphenyl)-1-isopropyl-3,4-dime-
thylpiperidin-1-yl]ethyl}-1,2,3,4-tetrahydroisoquinoline-3-car-
boxamide; DTG, di(o-tolyl)guanidine; MOE, molecular operat-
ing environment; HEK, human embryonic kidney; GTPγS,
guanosine 5′-3-O-(thio)triphosphate; APCI, atmospheric pres-
sure chemical ionization; DTG, di-o-tolylguanidine; SEM,
standard error of the mean; MOPS, 3-morpholinopropanesul-
fonic acid
Mustard Oil Assay.54,55 Male NMRI mice (20−35 g) were
habituated on a lattice in a cage of plexiglas (14.5 cm × 14.5 cm ×
10 cm) for 30 min (test conditions). The mice were stimulated with
von Frey filaments onto the abdominal wall. Ten stimulations with von
Frey filaments of 1, 4, 8, 16, 32 mN were applied in ascending order
(i.e., 10 × 1 mN, 10 × 4 mN, etc.). Animals with more than 25 positive
reactions during this phase were excluded. Vaseline was applied in the
perianal area to avoid the stimulation of somatic areas with the irritant
chemical. Colitis was induced by rectal administration of 50 μL of
mustard oil (3.5% in PEG200). Control animals were treated with
REFERENCES
■
(1) Gilbert, P. E.; Martin, W. R. The effects of morphine and
nalorphine like drugs in the nondependent and morphine dependent
chronic spinal dog. J. Pharmacol. Exp. Ther. 1976, 197, 517−532.
6858
dx.doi.org/10.1021/jm500940q | J. Med. Chem. 2014, 57, 6845−6860

Journal of Medicinal Chemistry
Article
(2) Dhawan, B. N.; Cesselin, F.; Raghubir, R.; Reisine, T.; Bradley, P.
B.; Portoghese, P. S.; Hamon, M. International Union of
Pharamcology. XII. Classification of opioid receptors. Pharmacol.
Rev. 1996, 48, 567−592.
aminomorphinones: effect of substitution in the aromatic ring of
cinnamoylaminomorphines and codeinones. J. Med. Chem. 2006, 49,
5333−5338.
(21) Kumagai, H.; Ebata, T.; Takamori, K.; et al. Effect of a novel
kappa-receptor agonist, nalfurafine hydrochloride, on severe itch in
337 haemodialysis patients: a phase III, randomized, double-blind,
placebo-controlled study. Nephrol., Dial., Transplant. 2010, 25, 1251−
1257.
(22) European Medicines Agency. http://www.ema.europa.eu/ema/
index.jsp?curl=pages/medicines/human/medicines/002683/smops/
Negative/human_smop_000631.jsp&mid=WC0b01ac058001d127
(2013).
(3) Calo, G.; Guerrini, R.; Rizzi, A.; Salvadori, S.; Regoli, D.
Pharmacology of nociception and its receptor: a novel therapeutic
target. Br. J. Pharmacol. 2000, 129, 1261−1283.
(4) Wu, H.; Wacker, D.; Mileni, M.; Katritch, V.; Han, G. W.; Vardy,
E.; Liu, W.; Thompson, A. A.; Huang, X. P.; Carroll, F. I.; Mascarella,
S. W.; Westkaemper, R. B.; Mosier, P. D.; Roth, B. L.; Cherezov, V.;
Stevens, R. C. Structure of the human κ-opioid receptor in complex
with JDTic. Nature 2012, 485, 327−332.
(5) Vanderah, T. W. Delta and kappa opioid receptors as suitable
drug targets for pain. Clin. J. Pain 2010, 26, S10−S15.
(6) Scopes, D. I. C. Combination of cyclohexane and piperazine
based κ-opioid receptor agonists: Synthesis and pharmacological
evaluation of trans,trans-configured perhydroquinoxalines. Drugs
Future 1993, 18, 933−947.
(23) Camilleri, M. D. Novel pharmacology: asimadoline, a κ-opioid
agonist, and visceral sensation. Neurogastroenterol. Motil. 2008, 20,
971−979.
(24) Barber, A.; Bartoszyk, G. D.; Bender, H. M.; Gottschlich, r.;
Greiner, H. E.; Harting, J.; Mauler, F.; Minck, K.-O.; Murray, R. D.;
Simon, M. A pharmacological profile of the novel, peripherally-
selective κ-opioid receptor agonist, EMD 61753. Br. J. Pharmacol.
1994, 113, 1317−1327.
(7) Holzgrabe, U.; Nachtsheim, C.; Siener, T.; Drohsin, S.; Brandt,
W. Opioid-agonisten und -antagonisten, opioid-rezeptoren. Pharmazie
1997, 52, 4−22.
(25) Soukara, S.; Maier, C. A.; Predoiu, U.; Ehret, A.; Jackisch, R.;
́
(8) Riviere, P. J.-M. Peripheral kappa-opioid agonists for visceral pain.
Br. J. Pharmacol. 2004, 141, 1331−1334.
Wunsch, B. Methylated analogues of methyl (R)-4-(3,4-dichlorophe-
̈
nylacetyl)-3-(pyrrolidin-1-ylmethyl)piperazine-1-carboxylate (GR-
89,696) as highly potent κ-receptor agonists: stereoselective synthesis,
opioid-receptor affinity, receptor selectivity, and functional studies. J.
Med. Chem. 2001, 44, 2814−2826.
(9) Togashi, Y.; Umeuchi, H.; Okano, K.; Ando, N.; Yoshizawa, Y.;
Honda, T.; Kawamura, K.; Endoh, T.; Utsumi, J.; Kamei, J.; Tanaka,
T.; Nagase, H. Antipruritic activity of the kappa-opioid receptor
agonist, TRK-820. Eur. J. Pharmacol. 2002, 435, 259−264.
(10) Nakao, K.; Ikeda, K.; Kurokawa, T.; Togashi, Y.; Umeuchi, H.;
Honda, T.; Okano, K.; Mochizuki, H. Effect of TRK-820, a selective
kappa opioid receptor agonist, on scratching behavior in an animal
model of atopic dermatitis. Nihon Shinkei Seishin Yakurigaku Zasshi
2008, 28, 75−83.
(26) Lichtenthaler, F. W.; Nakagawa, T.; El-Scherbiney, A. Synthese
cyclischer 1,2,3-triamine durch umsetzung von dialdehyden mit
nitromethan und benzylamin. Angew. Chem. 1967, 79, 530−531.
(27) Lichtenthaler, F. W.; Nakagawa, T.; El-Scherbiney, A.
Nitromethan-kondensation mit dialdehyden, XIV. Synthese cyclischer
nitrodiamine durch bis-aminoalkylierung von nitromethan. Chem. Ber.
1968, 101, 1837−1845.
(11) Horwell, D. C. Kappa opioid analgesics. Drugs Future 1988, 13,
1061−1071.
(12) Roth, B. L.; Baner, K.; Westkaemper, R.; Siebert, D.; Rice, K. C.;
Steinberg, S.; Ernsberger, P.; Rothman, R. B. Salvinorin A: a potent
naturally occurring nonnitrogenous kappa opioid selective agonist.
Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 11934−11939.
(13) Siebert, D. Salvia divinorum and salvinorin A: new
pharmacologic findings. J. Ethnopharmacol. 1994, 43, 53−56.
(14) de Costa, B. R.; Bowen, W. D.; Hellewell, S. B.; George, C.;
Rothman, R. B.; Reid, A. A.; Walker, J. M.; Jacobson, A. E.; Rice, K. C.
Alterations in the stereochemistry of the kappa-selective opioid agonist
U50,488 result in high-affinity sigma ligands. J. Med. Chem. 1989, 32,
1996−2002.
(15) Naylor, A.; Judd, D. B.; Lloyd, J. E.; Scopes, D. I. C.; Hayes, A.
G.; birch, P. J. A potent new class of κ-receptor agonists: 4-substituted
1-(arylacetyl)-2-[(dialkylamino)methyl]piperazines. J. Med. Chem.
1993, 36, 2075−2083.
(16) Scopes, D. I. C. Recent developments in non-peptide kappa
receptor agonists. Drugs Future 1993, 18, 933−947.
(17) Arendt-Nielsen, L.; Olesen, A. E.; Staahl, C.; Menzaghi, F.; Kell,
S.; Wong, G. Y. Analgesic efficacy of peripheral kappa-opioid receptor
agonist CR665 compared to oxycodone in a multi-modal, multi-tissue
experimental human pain model: selective effect on visceral pain.
Anesthesiology 2009, 111, 616−624.
(28) Eberle, M.; Missbach, M.; Seebach, D. Enantioselective
saponification with pig liver esterase (PLE): (1S,2S,3R)-3-hydroxy-2-
nitrocyclohexyl acetate. Org. Synth. Collect. 2003, 8, 332−340.
(29) Bourgeois, C.; Werfel, E.; Schepmann, D.; Wunsch, B.
̈
Combination of cyclohexane and piperazine based κ-opioid receptor
agonists: synthesis and pharmacological evaluation of trans,trans-
configured perhydroquinoxalines. Bioorg. Med. Chem. 2014, 22, 3316−
3324.
(30) Wunsch, B.; Geiger, C.; Enders, D.; Schaumann, E. Reduction of
̈
carbonic and carboxylic acid serivatives. Sci. Synth. 2009, 40a, 23−64.
(31) Kracht, D.; Rack, E.; Schepmann, D.; Frohlich, R.; Wunsch, B.
̈
̈
Stereoselective synthesis and structure−affinity relationships of bicyclic
κ receptor agonists. Org. Biomol. Chem. 2010, 8, 212−225.
(32) Geiger, C.; Zelenka, C.; Lehmkuhl, K.; Schepmann, D.;
Englberger, W.; Wunsch, B. Conformationally constrained κ receptor
̈
agonists: stereoselective synthesis and pharmacological evaluation of
6,8-diazabicyclo[3.2.2]nonane derivatives. J. Med. Chem. 2010, 53,
4212−4222.
(33) Gottschlich, R.; Ackermann, K. A.; Barber, A.; Bartoszyk, G. D.;
Greiner, H. E. EMD 61 753 as a favourable representative of
structurally novel arylacetamido-type K opiate receptor agonists.
Bioorg. Med. Chem. Lett. 1994, 4, 677−682.
(34) Gottschlich, R.; Krug, M.; Barber, A.; Devant, R. M. κ-Opioid
activity of the four stereoisomers of the peripherally selective κ-
agonists, EMD 60 400 and EMD 61 753. Chirality 1994, 6, 685−689.
(18) Vanderah, T. W.; Schteingart, C. D.; Trojnar, J.; Junien, J.-L.;
Lai, J.; Riviere, P. J.-M. FE200041 (D-Phe-D-Phe-D-NIe-D-Arg-NH2):
a peripheral efficacious κ opioid agonist with unprecedented selectivity.
J. Pharmacol. Exp. Ther. 2004, 310, 326−333.
(19) Vanderah, T. W.; Largent-Milnes, T.; Lai, J.; Porreca, F.;
Houghten, R. A.; Menzaghi, F.; Wisniewski, K.; Stalewski, J.; Sueiras-
(35) Jung, B.; Englberger, W.; Frohlich, R.; Schepmann, D.;
̈
Lehmkuhl, K.; Wunsch, B. Synthesis and pharmacological evaluation
̈
of bicyclic SNC80 analogues with separated benzhydryl moiety. Bioorg.
Med. Chem. 2008, 16, 2870−2886.
(36) Holl, R.; Jung, B.; Schepmann, D.; Englberger, W.; Grunert, R.;
̀
Diaz, J.; Galyean, R.; Schteingart, C.; Junien, J.-L.; Trojnar, J.; Riviere,
̈
P. J.-M. Novel D-amino acid tetrapeptides produce potent anti-
nociception by selectively acting at peripheral κ-opioid receptors. Eur.
J. Pharmacol. 2008, 583, 62−72.
Humpf, H.-U.; Bednarski, P. A.; Wunsch, B. Synthesis and
̈
pharmacological evaluation of novel SNC 80 analogues with a bridged
piperazine ring. ChemMedChem 2009, 4, 2111−2122.
(37) Subramanian, G.; Paterlini, G.; Larson, D. L.; Portoghese, P. S.;
Ferguson, D. M. Conformational analysis and automated receptor
(20) Nieland, N. P. R.; Moynihan, H. A.; Carrington, S.; Broadbear,
J.; Woods, J. H.; Traynor, J. R.; Husbands, S. M.; Lewis, J. W.
Structural determinants of opioid activity in derivatives of 14-
6859
dx.doi.org/10.1021/jm500940q | J. Med. Chem. 2014, 57, 6845−6860

Journal of Medicinal Chemistry
Article
docking of selective arylacetamide-based κ-opioid agonists. J. Med.
Chem. 1998, 41, 4777−4789.
(38) de Costa, B. R.; Bowen, W. D.; Hellewell, S. B.; George, C.;
Rothman, R. B.; Reid, A. A.; Walker, J. M.; Jacobson, A. E.; Rice, K. C.
Alterations in the stereochemistry of the κ-selective opioid agonist
U50,488 result in high-affinity σ ligands. J. Med. Chem. 1989, 32,
1996−2002.
(39) Radesca, L.; Bowen, W. D.; Di Paolo, L.; de Costa, B. R.
Synthesis and receptor binding of enantiomeric N-substituted cis-N-[2-
(3,4-dichlorophenyl)ethyl]-2-(1-pyrrolidinyl)cyclohexylamines as
high-affinity σ receptor ligands. J. Med. Chem. 1991, 34, 3058−3065.
(40) Carroll, F. I.; Parham, A. K.; Bai, X.; Zhang, X.; Brine, G. A.;
Mascarella, S. W.; Martin, B. R.; May, E. L.; Sauss, C.; Di Paolo, L.;
Wallace, P.; Walker, J. M.; Bowen, W. D. Enantiomeric N-substituted
N-normetazocines: a comparative study of affinities at σ, PCP, and μ
opioid receptors. J. Med. Chem. 1992, 35, 2812−2818.
(56) Bradford, M. M. A rapid and sensitive method for the
quantitation of microgram quantities of protein utilizing the principle
of protein−dye binding. Anal. Biochem. 1976, 72, 248−254.
(57) Stoscheck, C. Quantitation of proteins. Methods Enzymol. 1990,
182, 50−68.
(58) Cheng, Y.; Prusoff, H. W. Relationship between the inhibition
constant (K1) and the concentration of inhibitor which causes 50 per
cent inhibition (I50) of an enzymatic reaction. Biochem. Pharmacol.
1973, 22, 3099−3108.
(41) May, E. L.; Aceto, M. D.; Bowman, E. R.; Bentley, C.; Martin, B.
R.; Harris, L. S.; Medzihradsky, F.; Mattson, M. V.; Jacobson, A. E.
Antipodal α-N-(methyl through decyl)-N-normetazocines (5,9α-
dimethyl-2′-hydroxy-6,7-benzomorphans): in vitro and in vivo proper-
ties. J. Med. Chem. 1994, 37, 3408−3418.
(42) Maier, C. A.; Wunsch, B. Novel σ receptor ligands. Part 2. SAR
̈
of spiro[[2]benzopyran-1,4′-piperidines] and spiro[[2]benzofuran-
1,4′-piperidines] with carbon substituents in position 3. J. Med.
Chem. 2002, 45, 4923−4930.
(43) Meyer, C.; Neue, B.; Schepmann, D.; Yanagisawa, S.;
Yamaguchi, J.; Wurthwein, E.-U.; Itami, K.; Wunsch, B. Improvement
̈
̈
of σ1 receptor affinity by late-stage C-H-bond arylation of spirocyclic
lactones. Bioorg. Med. Chem. 2013, 21, 1844−1856.
(44) Kohler, J.; Bergander, K.; Fabian, J.; Schepmann, D.; Wunsch, B.
̈
̈
Enantiomerically pure 1,3-dioxanes as highly selective NMDA and σ1
receptor ligands. J. Med. Chem. 2012, 55, 8953−8957.
(45) Banerjee, A.; Schepmann, D.; Kohler, J.; Wurthwein, E.-U.;
Wunsch, B. Synthesis and SAR studies of chiral non-racemic
dexoxadrol analogues as uncompetitive NMDA receptor antagonists.
Bioorg. Med. Chem. 2010, 18, 7855−7867.
̈
̈
̈
(46) Zhu, J.; Luo, L. Y.; Li, J. G.; Chen, C.; Liu-Chen, L. Y. Activation
of the cloned human kappa opioid receptor by agonists enhances
[35S]GTPγS binding to membranes: determination of potencies and
efficacies of ligands. J. Pharmacol. Exp. Ther. 1987, 282, 676−684.
(47) Strange, P. G. Use of the GTPγS ([35S]GTPγS and Eu-GTPγS)
binding assay for analysis of ligand potency and efficacy at G protein-
coupled receptors. Br. J. Pharmacol. 2010, 16, 1238−1249.
(48) Bhal, S. K.; Kassam, K.; Peirson, I. G.; Pearl, G. M. The rule of
five revisited: applying log D in place of log P in drug-likeness filters.
Mol. Pharmaceutics 2007, 4, 556−660.
(49) Kokate, A.; Li, X.; Jasti, B. Effect of drug lipophilicity and
ionization on permeability across the buccal mucosa: a technical note.
AAPS PharmSciTech 2008, 9, 501−504.
(50) Franke, H.; Galla, H.-J.; Beuckmann, C. An improved low-
permeability in vitro model of the blood brain barrier. Brain Res. 1999,
818, 65−571.
(51) Franke, H.; Galla, H.-J.; Beuckmann, C. Primary cultures of
brain microvessel endothelial cells: a valid and flexible model to study
drug transport through the blood−brain-barrier in vitro. Brain Res.
Protoc. 2000, 5, 248−256.
(52) Bornhorst, J.; Wehe, C. A.; Huwel, S.; Karst, U.; Galla, H.-J.;
̈
Schwerdtle, T. Impact of manganese on and transfer across the blood−
brain and blood−csf barrier in vitro. J. Biol. Chem. 2012, 287, 17140−
17151.
(53) Hendershof, L. C.; Forsaith, J. Antagonism of the frequency of
phenylquinone-induced writhing in the mouse by weak analgesics and
nonanalgesics. J. Pharmacol. Exp. Ther. 1959, 125, 237−240.
(54) Laird, J. M.; Martinez, C. L.; Garcia-Nicas, E.; Cervero, F. A new
model of visceral pain and referred hyperalgesia in the mouse. Pain
2001, 92, 335−342.
́
(55) Christoph, T.; Kogel, B.; Schiene, K.; Meen, M.; De Vry, J.;
̈
Friderichs, E. Broad analgesic profile of buprenorphine in rodent
models of acute and chronic pain. Eur. J. Pharmacol. 2005, 507, 87−98.
6860
dx.doi.org/10.1021/jm500940q | J. Med. Chem. 2014, 57, 6845−6860