Superpotent [Dmt1]dermorphin tetrapeptides containing the 4-aminotetrahydro-2-benzazepin-3-one scaffold with mixed μ/δ opioid receptor agonistic properties

Article abstract of DOI:10.1021/jm200894e

Novel dermorphin tetrapeptides are described in which Tyr¹ is replaced by Dmt¹, where d-Ala² and Gly⁴ are N-methylated, and where Phe³-Gly⁴ residue is substituted by the constrained Aba³-Gly⁴ peptidomimetic. Most of these peptidic ligands displayed binding affinities in the nanomolar range for both - and -opioid receptors but no detectable affinity for the -opioid receptor. Measurements of cAMP accumulation, phosphorylation of extracellular signal-regulated kinase (ERK1/2) in HEK293 cells stably expressing each of these receptors individually, and functional screening in primary neuronal cultures confirmed the potent agonistic properties of these peptides. The most potent ligand H-Dmt-NMe-d-Ala-Aba-Gly-NH₂ (BVD03) displayed mixed / opioid agonist properties with picomolar functional potencies. Functional electrophysiological in vitro assays using primary cortical and spinal cord networks showed that this analogue possessed electrophysiological similarity toward gabapentin and sufentanil, which makes it an interesting candidate for further study as an analgesic for neuropathic pain.

Full text of DOI:10.1021/jm200894e

Article
pubs.acs.org/jmc
Superpotent [Dmt1]Dermorphin Tetrapeptides Containing the
4‑Aminotetrahydro-2-benzazepin-3-one Scaffold with Mixed μ/δ
Opioid Receptor Agonistic Properties
Bart Vandormael,† Danai-Dionysia Fourla,‡ Alexandra Gramowski-Voß,§ Piotr Kosson,∥ Dieter G. Weiss,§
Olaf H.-U. Schroder,⊥ Andrzej Lipkowski,∥ Zafiroula Georgoussi,‡ and Dirk Tourwe*,†
́
̈
†Department of Organic Chemistry, Vrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussels, Belgium
‡Cellular Signaling and Molecular Pharmacology, Institute of Biology, National Centre for Scientific Research “Demokritos”, Ag.
Paraskevi 15310, Athens, Greece
§Cell Biology and Biosystems Technology, Institute of Biological Sciences, University of Rostock, 18059 Rostock, Germany
∥Medical Research Center, Polish Academy of Sciences, 02-106 Warsaw, Poland
⊥NeuroProof GmbH, 18119 Rostock, Germany
S
* Supporting Information
ABSTRACT: Novel dermorphin tetrapeptides are described in which Tyr1 is
replaced by Dmt1, where D-Ala2 and Gly4 are N-methylated, and where Phe3-Gly4
residue is substituted by the constrained Aba3-Gly4 peptidomimetic. Most of these
peptidic ligands displayed binding affinities in the nanomolar range for both μ- and
δ-opioid receptors but no detectable affinity for the κ-opioid receptor.
Measurements of cAMP accumulation, phosphorylation of extracellular signal-
regulated kinase (ERK1/2) in HEK293 cells stably expressing each of these
receptors individually, and functional screening in primary neuronal cultures
confirmed the potent agonistic properties of these peptides. The most potent ligand H-Dmt-NMe-D-Ala-Aba-Gly-NH2 (BVD03)
displayed mixed μ/δ opioid agonist properties with picomolar functional potencies. Functional electrophysiological in vitro assays
using primary cortical and spinal cord networks showed that this analogue possessed electrophysiological similarity toward
gabapentin and sufentanil, which makes it an interesting candidate for further study as an analgesic for neuropathic pain.
the development of mixed μ/δ opioid ligands with low
propensity to induce tolerance and physical dependence.7,18,19
Dermorphin (H-Tyr1-D-Ala2-Phe3-Gly4-Tyr5-Pro6-Ser7-NH2),
an opioid heptapeptide isolated from the skin of the South
American frog Phyllomedusa sauvagei, is a selective μ-opioid
agonist that produces intense opioid analgesia.20,21 Besides its
higher antinociceptive efficacy and potency compared to
morphine, dermorphin is also less likely to produce tolerance,
dependence, and opiate side effects than morphine.22 SAR
studies showed that the N-terminal tetrapeptide (H-Tyr1-D-
Ala2-Phe3-Gly4-NH2) is the minimal sequence required for
opiate-like activity in vivo. Despite the reduced potency of the
tetrapeptide compared to the parent heptapeptide, the μ-
selectivity was retained. It was also shown that amidation of the
C-terminus increased the μ-receptor affinity and potency.22
The amine and phenolic group of Tyr1 and the aromatic ring of
Phe3 in a specific relative orientation are key requirements for
receptor recognition of these peptides.23 Inversion of the
configuration at residue 2 to the L-Ala2 isomer produced a 100-
fold decrease in μ-receptor affinity and GPI activity. Therefore,
INTRODUCTION
■
Opioid drugs exert their biological effects through their binding
to the three types of opioid receptor subtypes μ, δ, and κ (μ-
OR, δ-OR, and κ-OR), which belong to the superfamily of G-
protein-coupled receptors.1 The binding of these ligands to the
opioid receptors, however, also results in other effects like
tolerance, dependence, respiratory depression, and inhibition of
gastrointestinal motility, which often limit their use. Therefore,
the design and discovery of new analgesics with a different
pharmacological profile and with diminished adverse side
effects are a major challenge for medicinal chemists.
An emerging approach is the use of mixed μ/δ opioid
ligands, e.g., μ agonist/δ agonist and μ agonist/δ antagonist.2−7
Various studies proved the existence of physical and functional
interactions between the μ and δ receptors.8−14 Studies have
shown that the μ-OR−δ-OR heterodimer affects ligand binding
and receptor signaling and that this heterodimer could depict a
functional unit distinct from the μ-OR and δ-OR.10,15,16 Indeed,
activation of this heterodimer will result in different down-
stream effects such as inhibition of adenylyl cyclase activity,
phosphorylation of extracellular signal-regulated kinase 1/2
(ERK), activation of K+ currents, and inhibition of Ca2+
channels compared to the activation of isolated μ-OR and δ-
OR.17 The existence of these intermodulary effects stimulated
Received: July 8, 2011
Published: October 6, 2011
© 2011 American Chemical Society
7848
dx.doi.org/10.1021/jm200894e|J. Med. Chem. 2011, 54, 7848−7859

Journal of Medicinal Chemistry
Article
Figure 1. Design of N-methylated and constrained [Dmt1]dermorphin tetrapeptide analogues.
the second amino acid residue must play a significant role in
orienting these functional groups.
rewarding effects of drugs of abuse and adaptive responses
(tolerance),35,36 the μ-OR and δ-OR mediated ERK1/2
phosphorylation was also examined.
The use of opioid peptides as therapeutics has severe
limitations due to the rapid degradation in vivo by peptidases
and the poor bioavailability that is mainly due to their inability
to pass the blood−brain barrier (BBB). These problems can be
circumvented by the introduction of (i) unnatural amino acids,
(ii) N-methylation of amide bonds, and (iii) conformational
constraints in the native peptide chain.24−27 Although it was
reported that the dermorphin tetrapeptides were highly stable
against enzymatic degradation, their oral bioavailability
remained low. Several studies have shown that methylation at
the fourth position increased lipophilicity and resulted in
compounds such as H-Tyr1-D-Arg2-Phe3-NMe-Gly4-OH or Nα-
amidino-Tyr1-D-Arg2-Phe3-NMe-βAla4-OH with improved anti-
nociceptive activity after peripheral administration.28,29 In
contrast, methylation of the third amide bond resulted in a
detrimental loss in μ- and δ-receptor affinity.30,31 Our group
reported a highly active, double N-methylated dermorphin
The electrophysiological properties of these peptides can be
evaluated using their ability for induction of activity changes in
primary cortical and spinal cord networks grown on micro-
electrode array neurochips. Such network cultures can remain
spontaneously active and pharmacologically responsive for
more than a year.37−40 We have previously demonstrated the
interculture repeatability of the networks and that the relevant
receptors in the tissue of origin are also expressed in
culture.37,41 The spontaneous neuronal activity is specific to
the brain-region culture and responds in a substance-specific
manner to the compounds applied.39,42 This enables the
pattern recognition and similarity analysis of the activity profiles
of those substances.42,43 Thereby the effect of new drug
candidates on the CNS and possible neurotoxic effects can be
predicted early by studying changes in a functional network;
those possess the temporal resolution of cell-assembly
cooperation. Thus, it is possible to clarify the mode of action
and considerably accelerate the preclinical development of CNS
drugs by using this non-animal testing model.
tetrapeptide H-Tyr1-NMe-D-Ala2-Phe3-NMe-Gly4-NH2
1
(SB0306) with a strong analgesic effect after iv administra-
tion.32 This peptide was equipotent to the constrained
dermorphin tetrapeptide H-Hba1-D-Ala2-Aba-Gly4-NH2
2
(SB0304)).32,33 In this analogue the Tyr1 and Phe3 residues
were conformationally constrained by using the 4-amino-
tetrahydro-2-benzazepin-3-on (Aba) scaffold (Figure 1).
RESULTS
■
Chemistry. Two new dermorphin tetrapeptide analogues 3
and 4 were designed from the reference peptides 1 and 2.32,33
The Tyr1 residue in 1 or the Hba residue in 2 was substituted
by 2′,6′-dimethyltyrosine (Dmt1) (Figure 1), which is known
for its enhancement of the μ and δ receptor interactions and
significant increase in opioid agonist potency.44,45 The
conformationally constrained Aba3 residue, present in 2, was
retained in 4. The effect of N-methylation of the second residue
(Figure 1) on the receptor affinity and biological activity was
investigated by comparing 2 to 5 (Figure 1).
In this manuscript we report the preparation of two
dermorphin tetrapeptide analogues 3 and 4 that were designed
from two reference peptides by substituting the Tyr1 residue in
1 with 2′,6′-dimethyltyrosine (Dmt1) or the Hba1-D-Ala residue
in 2 with Dmt-NMe-D-Ala in order to enhance bioactivity at the
μ- and δ-ORs. To determine receptor affinity, specificity, and
agonist efficacy, these opioid tetrapeptides were evaluated using
a variety of biochemical approaches in intact HEK293 cells
stably expressing μ-, δ-, or κ-ORs or in primary neuronal cell
cultures. On the basis of the hypothesis that opioid-mediated
ERK1/2 activation plays an important role in pain regulation
and analgesia34 and ERK1/2 signaling is involved in the
The peptides 1, 2, and 5 were used as references for the
novel analogues and were prepared as previously de-
scribed.32,33,46 Analogue 3 was prepared using the Fmoc-
protected amino acids on Rink amide resin as solid support. For
7849
dx.doi.org/10.1021/jm200894e|J. Med. Chem. 2011, 54, 7848−7859

Journal of Medicinal Chemistry
Article
a
Table 1. Receptor Binding of Dermorphin Tetrapeptide Analogues
b
c
Ki (nM)
IC50 (nM)
compd
sequence
μ-OR
δ-OR
κ-OR
Kiδ/Kiμ
μ-OR
δ-OR
1
3
4
5
2
H-Tyr1-NMe-D-Ala2-Phe3-NMe-Gly4-NH2
H-Dmt1-NMe-D-Ala2-Phe3-NMe-Gly4-NH2
H-Dmt1-NMe-D-Ala2-Aba3-Gly4-NH2
H-Dmt1-D-Ala2-Aba3-Gly4-NH2
1100 ± 160
60 ± 3
14.8 ± 2.0
170 ± 3
>104
>106
>106
>106
>106
>106
N/D
2.2
28.1 ± 6.132
0.30 ± 0.01
0.95 ± 0.06
ND
7.1 ± 5.932
4.90 ± 0.24
0.95 ± 0.07
ND
130 ± 5
5 ± 0.4
18 ± 3
160 ± 45
0.34
0.106
0.059
H-Hba1-D-Ala2-Aba3-Gly4-NH2
2700 ± 300
20.8 ± 3.632
17.8 ± 5.032
a
b
c
ND: not determined. Displacement of [3H]diprenorphine binding in HEK293 cells stably expressing μ-, δ-, or κ-ORs, Displacement of
[3H]naloxone for μ-OR and [3H]deltorphin II for δ-OR from rat brain binding sites.32
Table 2. Functional Activity of Dermorphin Tetrapeptide Analogues
a
b
μ-OR
δ-OR
EC50 (nM)
μ
compd
sequence
EC50 (nM)
3.28 ± 0.25
0.0790 ± 0.0065
0.00174 ± 0.00034
0.0832 ± 0.0085
5.17 ± 0.31
Emax
Emax
EC50δ/EC50
1
3
4
5
2
H-Tyr1-NMe-D-Ala2-Phe3-NMe-Gly4-NH2
H-Dmt1-NMe-D-Ala2-Phe3-NMe-Gly4-NH2
H-Dmt1-NMe-D-Ala2-Aba3-Gly4-NH2
H-Dmt1-D-Ala2-Aba3-Gly4-NH2
H-Hba1-D-Ala2-Aba3-Gly4-NH2
H-Tyr1-D-Ala2-Gly3-NMe-Phe4-Gly5-ol
H-Tyr1-c(D-Pen2-Gly3-Phe4-D-Pen5)-OH
77 ± 3.4
71 ± 3.25
72 ± 2.3
78 ± 1.55
78 ± 0.3
73 ± 0.3
960 ± 220
4400 ± 500
0.016 ± 0.009
0.17 ± 0.005
N/D
74 ± 0.5
72 ± 3
61 ± 3.9
66 ± 1
293
55700
9.20
2.04
79 ± 1.7
DAMGO
DPDPE
3.18 ± 0.30
0.14 ± 0.03
69 ± 0.7
a
Agonistic properties of peptides on forskolin-stimulated cAMP accumulation by μ- opioid receptor. The inhibition of cAMP accumulation was
b
measured as described in the Experimental Section. Agonistic properties of peptides on forskolin-stimulated cAMP accumulation by δ-opioid
receptor. The inhibition of cAMP accumulation was measured as described in the Experimental Section.
the synthesis of 4 the Boc-Aba-Gly building block was prepared
using a published procedure using acyliminium chemistry.33
The Aba-constrained dermorphin tetrapeptide 2 was then
synthesized on 4-methylbenzhydrylamine (MBHA) resin. In
both cases Fmoc-Dmt-OH was used in the synthesis, and the
Fmoc protection was removed prior to cleavage of the peptides
from the resin. The peptides were cleaved from the Rink amide
or MBHA resin by using respectively trifluoroacetic acid or
HFliq. Subsequent purification by preparative high pressure
liquid chromatography (HPLC) yielded the pure peptides
(>99%).
Structure−Affinity/Activity Study. Opioid binding affin-
ities of the new ligands (1−5) for the μ-OR, δ-OR, or κ-OR
were determined by radioligand competition binding assays
using [3H]diprenorphine in membrane preparations from
HEK293 cells stably expressing each of the respective receptor
subtype tested. Since the binding affinities for the reference
compounds 1 and 2 have previously been determined in rat
brain homogenates, using [3H]naloxone and [3H]deltorphin
II,32 the affinities of the new analogues 3 and 4 were also
determined in these conditions. The functional activity data of
all dermorphin tetrapeptide analogues were obtained by
measuring the inhibition of forskolin-stimulated cAMP
accumulation in HEK293 cells stably expressing the μ- or the
δ-opioid receptor and MAP kinase phosphorylation mediated
by these receptor ligands.
tested in the individual receptors expressed in HEK293 is
similar to that detected in rat brain membranes.
As indicated in Table 1, the analogue 4 displayed a high
affinity for the μ- and δ-opioid receptor but had no affinity for
the κ-opioid receptor. As expected, incorporation of Dmt1 in
lieu of Tyr1 resulted in analogues with enhanced μ- and δ-
opioid receptor affinity, with low nanomolar affinities, and with
functional potencies in the picomolar range for the μ-OR and
δ-OR with higher potency for the μ-opioid receptor (Table 2).
Changing 1 to 3 resulted in a somewhat higher increase in δ-
affinity than in μ-affinity, resulting in a compound that is almost
nonselective (Kiδ/ Kiμ = 2.2) when tested in HEK293 cells and
μ
moderately selective (IC50δ/IC50 = 16) in rat brain
membranes. It is worth noting that (remarkably) in the
functional assays on adenylyl cyclase, compound 3 showed a 41
times higher potency versus 1 at the μ-receptor but a 5 times
decreased potency at the δ-receptor. On the other hand,
compound 3 phosphorylated ERK1,2 at much higher levels for
the δ-OR than that detected for the μ-opioid receptor. These
apparent differences in effector potency could be due to the
conformation the receptor adopts upon binding to each ligand,
and the population of the Gα, RGS proteins, or other signaling
intermediates1,14 is complexed. It is already evident that ligand-
specific receptor states differentially modulate heptahelical
receptor signaling.48,49 The incorporation of the Phe3 side chain
in 3 into an azepinone ring resulted in analogue 4, with mainly
an improved δ-affinity. This is in agreement with our previous
findings in dermorphin and its N-terminal tetrapeptide that the
Aba3 residue enhances selectivity toward the δ-OR,33,50 but in
this case also improved μ-OR affinity. However in the
functional assays performed measuring cAMP accumulation,
the potency of 4 toward the μ-OR was further enhanced
compared to that of 3, resulting in an extremely potent
compound. However, remarkably enough this analogue also
displayed a picomolar potency for the δ-OR.
As shown in Table 1, the Ki values determined in the
HEK293 cells expressing the human μ- and rat δ-opioid
receptors using [3H]diprenorphine are much larger than the
IC50 values determined in rat brain using different radioligands.
This is not only due to the fact that the rat brain expresses a
different population of (all three) receptors but also because
[3H]diprenorphine recognizes both cell surface and internalized
receptors, resulting in much higher Bmax and lower Kd values.47
Nevertheless, the general trend in potency of these ligands
7850
dx.doi.org/10.1021/jm200894e|J. Med. Chem. 2011, 54, 7848−7859

Journal of Medicinal Chemistry
Article
Figure 2. Effect of 1 (A), 3 (B), 4 (C), 5 (D), and 2 (E) on μ-opioid-receptor-mediated ERK phosphorylation. Stably transformed HEK293 cells
expressing the μ-opioid receptor were challenged with the appropriate ligands for the indicated time, and cell lysates were resolved in SDS−PAGE
(10%) as described by Morou and Georgoussi.47 Phosphorylated ERK1/2 was visualized by immunoblotting with a phospho-ERK antibody (top
panels). Equal loading was verified by stripping and reprobing using an anti-ERK antibody (bottom panels). Results are representative of three
independent experiments.
Figure 3. Effect of 1 (A), 3 (B), 4 (C), 5 (D), and 2 (E) on δ-opioid-receptor-mediated ERK phosphorylation. Stably transformed HEK293 cells
expressing the δ-opioid receptor were challenged with the appropriate ligands for the indicated time, and cell lysates were resolved in SDS−PAGE
(10%) as described by Morou and Georgoussi.47 Phosphorylated ERK1/2 was visualized by immunoblotting with a phospho-ERK antibody (top
panels). Equal loading was verified by stripping and reprobing using an anti-ERK antibody (bottom panels). Results are representative of three
independent experiments.
7851
dx.doi.org/10.1021/jm200894e|J. Med. Chem. 2011, 54, 7848−7859

Journal of Medicinal Chemistry
Article
Figure 4. Comparison of the acute effects of the four tetrapeptides 1, 3, 4, and 2 on the cortical (A, C) and spinal cord (B, D) network activity in
vitro. Displayed are the activity describing parameters spike rate (general activity) and burst duration (burst structure) for treatment of nine
accumulating concentrations in the range of 1 fM to 100 μM (mean ± standard error).
The removal of the N-methyl group at D-Ala2 in 5 resulted in
a decreased affinity for the μ- and δ-OR. In the functional assay
5 was 48 times less potent than 4, whereas at the δ-receptor a
10-fold loss of potency was observed. Despite this, 5 remains a
very potent μ- and δ-opioid agonist.
signaling cascade mediates a variety of cellular functions like cell
differentiation and proliferation, and it has been implicated in
synaptic plasticity, memory formation, and long-term gene
expression changes underlying drug tolerance and depend-
ence.52,53 The fact that all dermorphin tetrapeptide analogues
(1−5) enhanced ERK1/2 phosphorylation is a clear indication
of the similar agonistic properties displayed by them on the two
opioid receptor subtypes (μ-OR and δ-OR).
The incorporation of the Tyr1 side chain into an azepinone
ring to give the previously reported analogue 232,33 resulted in a
large decrease in μ- and δ-affinity compared to 4, indicating that
this conformational constraint is not optimal for recognition at
both receptors. The functional assay of 4 displayed
subnanomolar potency for both μ- and δ-opioid receptors,
thus suggesting that analogue 4 behaves as full agonist for both
μ, δ-opioid receptors.
Functional Screening in Neuronal Cell Cultures. As an
alternative approach to define in more detail the pharmaco-
logical profile of tetrapeptide analogues 1, 3, 4, and 2,
functional screening of these compounds in neuronal cell
cultures was performed using brain-region specific primary
network cocultures and microelectrode neurochip technology.
The coculture system consists of both neuronal and glial cell
populations forming an intricate network with in vivo like
properties. Neuronal activity in the system is monitored using
modern multichannel recording techniques and characterized
using multiparametric data analysis. To clarify the mode of
action of the new tetrapeptide analogues (1−4), their acute
dose-dependent effects on the activity of cortical and spinal
cord networks were studied. We further analyzed the frontal
cortex experiments using methods of pattern recognition. We
carried out a classification against 77 substances in the
NeuroProof database (Table S1).42,43 Proprietary procedures
of pattern recognition are applied for a multiparametric data
analysis. Thus, the effects of CNS drug candidates can be
examined directly, which was not possible before. A high
reproducibility of data is ensured, for example, the same
Effect of Dermorphin Tetrapeptide Analogues on
ERK1/2 Activation by the μ- and δ-Opioid Receptors. It
has previously been demonstrated that μ- and δ-opioid
receptors expressed in HEK293 cells or rat C6 glioma cells
stimulate ERK1/2 activity via pertussis toxin-sensitive G
protein signaling mechanisms.47,51
To examine whether stimulation of the μ- and δ-opioid
receptors by the tetrapeptides (1−5) has a downstream effect
for μ- and δ-opioid receptors, we measured alteration of ERK1/
2 phosphorylation mediated by these compounds upon
activation of the δ- and μ-opioid receptor. Serum-starved
HEK293 cells stably expressing the μ- or the δ-opioid receptor
were challenged with the opioid tetrapeptide analogues 1, 3, 4,
5, and 2. As shown in Figures 2 and 3, Western blotting with a
specific phospho-ERK1/2 antibody revealed an increase in
ERK1/2 phosphorylation by μ- or δ-opioid receptor within 2
min of the respective agonist stimulation. The MAPK/ERK1,2
7852
dx.doi.org/10.1021/jm200894e|J. Med. Chem. 2011, 54, 7848−7859

Journal of Medicinal Chemistry
Article
substance can be identified by its inducing specific activity
changes even after years.
the analgesic potency in general of the whole variety of
reference compounds. Neither in vitro binding affinities nor in
vitro EC50 effective concentration correlated well to the
analgesic potency of all the different compounds. As result of
this situation, we performed correlations by comparing the
activity pattern “fingerprint” of the four tetrapeptides to those
of 77 substances in the NeuroProof database (Supporting
Information).
Electrophysiological Screening Results. To character-
ize the electrophysiological properties of the four tetrapeptides
(1−4), we first measured well-characterized reference com-
pounds to establish a database with substance-specific activity
“fingerprints” for reference compounds. Besides compounds
characterizing major receptor systems in neurotransmission
(e.g., glutamatergic, GABAergic, gap junction) and specific
clinical applications (e.g., antipsychotic, anticonvulsive, anti-
depressive, analgesic, anesthetic), we especially focused on
those related to the opioid analgesic system. Here we
characterized the μ-, δ-, and κ-opioid receptors and compounds
related to nociception such as neurotensin and substance P. As
a control series, we measured untreated network activity for 12
h and analyzed the activity pattern on a hourly basis, related to
the accumulating compound concentrations.
For the acute electrophysiological evaluation of the
dermorphin tetrapeptide analogues (1−4) primary neuronal
networks from either frontal cortex or spinal cord were acutely
treated with accumulating concentrations of the peptides in the
range from 1 fM to 100 μM.
Of the four tetrapeptides, 4 was the most potent in frontal
cortex in terms of activity inhibition and the effects on the burst
structure, followed by 3, 2, and 1 (Figure 4, Table 3). Analogue
Similarity Analysis and Classification. To specify the
anticonvulsive and analgesic properties of compounds 1−4, we
used the similarity of the activity pattern changes of these
compounds to those induced in a concentration-dependent
manner by sufentanil (SUF, for analgesic properties) and
gabapentin (GBP, for anticonvulsive properties against seizure
and antineuroleptic properties against neuropathic pain). The
results of the similarity analysis to these activity profiles are
shown in Table 3. Analogue 4 showed the highest similarities to
sufentanil and to gabapentin. The second best similarities were
found for 2, followed by 3 and 1. Analogue 4 also showed a
high similarity profile to the neuroleptic drugs clozapine and
haloperidol and the antidepressant fluvoxamine. Analogue 2
also showed a high similarity to sufentanil and gabapentin. In
the top five compounds in the similarity analysis for 2 were
endogenous neuropeptide orexin A, the δ-opioid receptor
agonist enkephalin, and the analgesic and μ-opioid receptor
agonist L-Polamidone. For analogue 3 the top five similarities
were found to be the GABAAR agonist muscimol, the selective
δ-opioid receptor agonist DPDPE, the selective μ-opioid
receptor agonist DAMGO, the selective δ-opioid receptor
agonist deltorphin, and the antipsychotic haloperidol. Analogue
1 showed the highest similarity to nalorphine, the analgesic and
μ-opioid receptor agonist fentanyl, the antidepressant ami-
triptyline, the immunosuppressant tacrolimus, and the un-
specific κ-opioid receptor agonist dynorphin A.
Table 3. Electrophysiological Activity and Similarity Analysis
a
Data
EC50 (M)
compd
N
frontal cortex
spinal cord
GBP
SUF
1
3
4
2
8
5.00 × 10−12
1.29 × 10−13
1.61 × 10−15
5.67 × 10−7
1.06 × 10−6
2.87 × 10−10
1.30 × 10−9
6.70 × 10−9
3
1
2
11
20
12
8
10
7
21
14
a
DISCUSSION
EC50 for frontal cortex: effective concentration to induce a 50%
reduction in the frontal cortex activity. EC50 for spinal cord: effective
concentration to induce a 50% reduction in the spinal cord network
activity. GBP, similarity to gabapentin effects: The numbers give the
percentage of activity data sets of a given newly synthesized substance
that have the highest similarity to gabapentin activity patterns. SUF,
similarity to sufentanil effects: The numbers give the percentage of
activity data sets of a given newly synthesized substance that have the
highest similarity to the frontal cortex activity pattern of sufentanil.
■
The highly constrained H-Hba-D-Ala-Aba-Gly-NH2 2 was
reported to be very potent in the GPI and MVD assay and
to display in vivo antinociception after intravenous admin-
istration.32 The latter property was not related to the
conformational constraint but rather to the double N-alkylation
at the D-Ala and Gly residues. Indeed, the “ring opened”
analogue H-Tyr-NMe-D-Ala-Phe-Sar-NH2 1 displayed similar in
vitro and in vivo potencies.32 Changing Tyr1 in 1 by Dmt
resulted in the expected increase in μ- and δ-opioid affinity and
in μ-, but not δ-, functional potency in 3. As a consequence,
ligand 3 becomes a highly selective μ-OR agonist.
4 induced biphasic changes in the cortical activity, where the
first minor activity decrease (≤1 μM) was accompanied with an
increase in the burst duration (Figure 4A,C). A higher
concentration caused a strong activity decrease, accompanied
by a decrease in the burst duration. A similar effect on the
frontal cortex network activity was induced by 1, 2, and 3. The
activity decrease combined with an increase in the burst
duration seemed to be characteristic for these four compounds.
In spinal cord, 3 induced the largest inhibition in spike rate
activity and burst duration, followed by 4, 2, and 1. The
different tissue specific effects induced by the tetrapeptides are
ascribed to the different receptor composition and colocalized
distribution in the cortex and spinal cord.
Constraining the Phe3 side chain into the Aba3 ring in 4
resulted in the expected increase in δ-affinity, giving a
compound with extremely high μ- and δ-agonist potency.
The Aba3 ring fixes the phenylalanine side chain into the
gauche(+) or trans rotameric state, the latter being proposed as
required for opioid affinity.30,33 In dermorphin it is responsible
for the enhancement in selectivity toward the δ-OR without
dramatically affecting μ-OR affinity. This is reflected in the very
high potencies of analogue 4 toward both opioid receptors.
N-Methylation, such as present in 1−4, restricts the
conformations accessible to the Cα−C(O) bond of the
preceding amino acid residue, increases metabolic stability,
and decreases the capacity to form hydrogen bonds with water
molecules, which results in an improved bioavailability.54,55
Moreover, in an N-methylamide, the cis-amide bond con-
formation is energetically more favored than in the non-
With the data derived from pattern recognition we tried to fit
the 200 activity-describing parameters and combinations
thereof to the analgesic properties of the reference compounds.
However, we were unable to define a set of activity parameters,
which change in a specific way, so that they can be correlated to
7853
dx.doi.org/10.1021/jm200894e|J. Med. Chem. 2011, 54, 7848−7859

Journal of Medicinal Chemistry
Article
methylated one, leading to the discussion whether the cis- or
the trans-amide isomer represents the bioactive conforma-
tion.55,56 The incorporation of NMe-D-Ala2 at the second
residue of dermorphin (H-Tyr1-D-Ala2-Phe3-Gly4-Tyr5-Pro6-
Ser7-NH2) and endomorphin-2 (H-Tyr1-Pro2-Phe3-Phe4-NH2)
was reported to result in an increase of μ-OR affinity by a factor
of 6 and 3, respectively, and an increase of 13 and 354 times in
μ-OR mediated biological activity.57,58
Therefore, analogue 5 was compared to its N-methylated
analogue 4. The removal of the N-Me group at the second
position resulted in a somewhat larger drop in μ-affinity (11-
fold) than in δ-affinity (4 fold), making 5 somewhat more δ-
selective (Kiδ/Kiμ = 0.106) than 4 (Kiδ/Kiμ = 0.34) in the
binding assay. In the functional assay 4 and 5 show μ-
selectivity. Despite a 48-fold drop in functional potency at the
μ-receptor, 5 remains a very potent μ- and δ-agonist. This
suggests that the Dmt-D-Ala peptide bond adopts a trans
bioactive conformation.
The double Aba constrained reference tetrapeptide 2 is a
somewhat lower potent analogue for both receptors compared
to the other dermorphin tetrapeptide analogues. Hence, the
introduction of the Tyr1 constraint Hba1 is not the optimal
replacement for recognition by the μ- and δ-ORs.
In the electrophysiological studies, 4 was the most potent in
frontal cortex, followed by 3, 2, and 1 (Figure 4), whereas in
spinal cord 3 induced the largest inhibition in spike rate activity
and burst duration, followed by 4, 2, and 1. The different tissue
specific effects induced by the tetrapeptides are ascribed to the
different receptor composition and colocalized distribution in
the cortex and spinal cord. These data confirm that 4 behaves
as an extremely potent opioid agonist.
reduce the use of animal testing for the determination of the
analgesic effect. The four peptides 1, 3, 4, and 2 were studied
by a functional electrophysiological in vitro screening using
primary cortical and spinal cord networks, which indicated that
analogue 4 is the most potent compound in the series followed
by 3, 2, and 1. The classification of the electrophysiological
cortical network changes induced by these compounds revealed
the highest similarity of 4 to sufentanil and gabapentin.
Considering the best five similarity hits for 4 and since
gabapentin is known as an effective drug against neuropathic
pain, these results emphasize its potential application as a drug
candidate as analgesic for neuropathic pain.
These highly potent compounds will be pursued further in
the testing of their antinociceptive effects in vivo after systemic
administration.
EXPERIMENTAL SECTION
■
General Methods. The following amino acids, bases, solvents,
and reagents were purchased from Aldrich (Bornem, Belgium): Fmoc-
N-Me-D-Ala-OH, Fmoc-N-Me-Gly-OH, Boc-N-Me-D-Ala-OH, Boc-D-
Ala-OH, triethylamine, piperidine, iPrOH, acetonitrile, anisole,
triethylsilane, acetic acid, and hydrazine hydrate. Fmoc-Dmt-OH was
obtained from RSP Amino Acids (Shirley, MA, U.S.). Fmoc-
Tyr(OtBu)-OH, diisopropylcarbodiimide (DIC), 1-hydroxybenzotria-
zole (HOBt), trifluoroacetic acid (TFA), and dichloromethane were
provided by Fluka (Bornem, Belgium). The Rink amide resin and
Fmoc-Phe-OH were obtained from NovaBiochem (Laufelfingen,
̈
Switzerland). N,N-Diisopropylethylamine (DIPEA), trifluoromethane-
sulfonic acid (TFMSA), and dimethylformamide (DMF) were
obtained from Acros (Geel, Belgium). The MBHA resin was
purchased from Neosystem (Strasbourg, France). 2-1H-(benzotria-
zol-1-yl)-1,1,3,3-tetramethyluronium tetrafluoroborate (TBTU) was
purchased from Senn Chemicals (Dielsdorf, Switzerland).
The ranking in the electrophysiological potency in the
induced activity changes were also confirmed with the similarity
analysis (Table 3). Here, 4 showed the highest similarities to
sufentanil and to gabapentin. The second best similarities were
found for 2, followed by 3 and 1. Compound 4 also showed a
high similarity profile to the neuroleptic drugs clozapine and
haloperidol and the antidepressant fluvoxamine. These
similarities suggest that 4 has a very interesting profile as a
potential analgesic for neuropathic pain. Compound 2 also
showed a high similarity to sufentanil and gabapentin. In the
top five compounds in the similarity analysis for 2 were
endogenous neuropeptide orexin A, the δ-opioid receptor
agonist enkephalin, and the analgesic and μ-opioid receptor
agonist L-Polamidone. The similarity to orexin A could reflect
the interplay of the hypocretic/orexin and nociceptin/orphanin
peptidergic system in the regulation of stress-induced
analgesia.59
All the compounds were analyzed using a Waters Breeze analytical
HPLC on a reversed phase C-18 column DiscoveryBIO SUPELCO
wide pore C18 column (25 cm × 4.6 mm, 5 μm) with a linear gradient
(3−100% CH3CN in H2O, containing 0.1% TFA, in 20 min) and a
Grace Vydac, Everest C18 column (25 cm × 4.6 mm, 5 μm) with a
linear gradient (10−100% MeOH in H2O, containing 0.1% TFA, in 20
min) at a flow rate of 1 mL/min with UV detection at 215 nm.
Purification was carried out on a Gilson semipreparative HPLC
system with Gilson 322 pumps on a reversed phase C-18 column
(DiscoveryBIO SUPELCO wide pore C18 column, 25 cm × 2.21 cm,
5 μm) with a linear gradient (20−80% CH3CN in H2O, containing
0.1% TFA, in 20 min) at a flow rate of 20 mL/min with UV detection
at 215 nm.
Thin layer chromatography (TLC) was performed on a plastic sheet
precoated with silica gel 60 F254 (Merck, Darmstadt, Germany).
Mass spectrometry (MS) data were recorded on a VG Quattro II
spectrometer (Micromass, Manchester, U.K.) using electrospray
(ESP) ionization (cone voltage, 70 V). Data collection was performed
with Masslynx, version 2.22, software.
Synthetic Procedures. Synthesis of H-Dmt-NMe-D-Ala-Phe-
NMe-Gly-NH2 3. Solid phase peptide synthesis was performed in a
10 mL polypropylene syringe with a PE frit (MultiSynTech, Germany)
which was shaken during coupling, deprotection, washing, and
cleavage steps. Rink amide polystyrene resin (0.18 mmol, 0.60
mmol/g) was used. The resin was swollen in CH2Cl2 for 1 h. After
filtration of the resin, the Fmoc protecting group on the Rink linker
was removed by 20% piperidine/DMF (2 × 10 min). Subsequently the
CONCLUSION
■
Novel highly potent dermorphin tetrapeptides 3, 4, and 5 were
obtained by combining Dmt1 substitution, N-alkylation of D-
Ala2, and conformational constraint of Phe3. The most potent
analogue H-Dmt1-NMe-D-Ala2-Aba3-Gly4-NH2 4 behaves as
mixed μ-/δ-agonist with subnanomolar potency. In the cAMP
assay, analogue H-Dmt1-NMe-D-Ala2-Phe3-NMe-Gly4-NH2 3 is
a selective μ-opioid agonist with subnanomolar potency. All
compounds stimulated ERK1/2 phosphorylation, and these
analogues might be of great importance in future studies on the
role of MKPs in the development of opioid dependence and
tolerance.
i
resin was filtered and washed (three times with DMF, PrOH, and
DMF again). Couplings were carried out by adding 3.0 equiv of the
Nα-Fmoc-protected amino acid, together with 3.0 equiv of DIC and
3.0 equiv of HOBt in DMF ([Fmoc-AA-OH] = 0.5 M) and overnight
reaction. The completeness of the couplings was checked with the NF-
31 test60 except for the coupling of Fmoc-Dmt-OH, which was
checked with the chloranil test.61 If incomplete, the coupling was
repeated. When the coupling was completed, the resin was washed,
A novel electrophysiological model assay with similarity
analysis was applied at this stage, which could contribute to
7854
dx.doi.org/10.1021/jm200894e|J. Med. Chem. 2011, 54, 7848−7859

Journal of Medicinal Chemistry
Article
Table 4. Sequences and Physicochemical Properties of Dermorphin Tetrapeptide Analogues
a
c
m/z (M + H)+
HPLC tR (min)
yield
(%)
b
compd
sequence
calcd
obsd
TLC Rf
1
2
3
4
H-Dmt1-NMe-D-Ala2-Phe3-NMe-Gly4-NH2
H-Dmt1-NMe-D-Ala2-Aba3-Gly4-NH2
512.28
510.26
512.41
510.42
0.70
0.67
12.1
12.0
13.6
13.4
38
43
a
b
c
Determined by ESI MS+ ionization. TLC system: EtOAc/n-BuOH/AcOH/H2O, 1:1:1:1. HPLC system 1: RP C-18 column, Supelco Discovery
BIO wide pore, 25 cm × 4.6 mm, 5 μm; solvent A, 0.1% TFA in water; solvent B, 0.1% TFA in acetonitrile; gradient, 3−97% B in A over 20 min;
flow rate, 1.0 mL/min. HPLC system 2: RP C-18 column, Grace Vydac, Everest C18 column (25 cm × 4.6 mm, 5 μm); solvent A, 0.1% TFA in
water; solvent B, 0.1% TFA in MeOH; gradient, 10−100% B in A over 20 min; flow rate, 1.0 mL/min.
followed by Fmoc deprotection (20% piperidine/DMF (2 × 10 min)),
with washings as described above, and coupling with the next amino
acid. After completion of the synthesis and final Fmoc-deprotection,
the resin was washed (three times with DMF, PrOH, DMF again,
CH2Cl2, and Et2O), dried under reduced pressure (8 h), and stored at
4 °C.
Receptor Binding Experiments. In the experiments designed to
define peptide specificity for μ- and δ-opioid receptors, membranes
expressing either μ-OR or δ-OR (20 μg) were incubated at 30 °C for
45 min in buffer containing 50 mM Tris-HCl, pH 7.5, 100 mM NaCl,
and 10 mM MgCl2 as described by Megaritis et al.62 The ability of
peptides 1−5 to displace [3H]diprenorphine (3.75 nM) binding to the
opioid receptors was assessed. Nonspecific binding was measured in
the presence of 10 μM naloxone. The reaction was stopped by rapid
filtration and three washes in ice-cold 50 mM Tris-HCl, pH 7.4,
through GF/C filters (Whatman, Maidstone, U.K.) using an
automated cell harvester (Brandel Inc., Gaithersburg, MD). The
radioactivity was measured by liquid scintillation counting (liquid
scintillation analyzer, Packard). Analysis of the binding data was
performed using the Origin, version 7.5, software (OriginLab
Corporation, Northampton, MA, U.S.).
Biological Activity Assays. Cell Culture. HEK293 cells stably
expressing either the EE-tagged-μ-opioid receptor or the δ-opioid
receptor were grown in Dulbecco’s modified Eagle’s medium
(DMEM) supplemented with 10% fetal calf serum, 2 mM glutamine,
100 units/mL penicillin, and 0.1 mg/mL streptomycin, under 5% CO2
at 37 °C as described by Megaritis et al.62
Measurements of cAMP Accumulation. Measurements of
adenylyl cyclase activity were performed as described by Merkouris
et al.63 Briefly, HEK293 cells stably expressing the μ- or the δ-opioid
receptor were seeded in 12-well plates and incubated for 24 h in a
medium containing [3H]adenine (GE Healthcare) (1.5 μCi/well). The
generation of [3H]cAMP was assessed in response to treatment of the
cells with various concentrations of the appropriate ligand (10 fM to
10 μM or 1 pM to 10 μM) using 50 μM forskolin at 37 °C for 30 min.
Results are calculated as the ratio of levels of [3H]cAMP to total
[3H]adenine nucleotides (×1000), and the data are presented as
percentage of forskolin-stimulated cAMP accumulation upon agonist
treatment. The radioactivity was measured by liquid scintillation
counting (liquid scintillation analyzer, Packard). Analysis of the data
was performed using the Origin, version 7.5, software (OriginLab
Corporation, Northampton, MA, U.S.).
Cell Membrane Preparations. Confluent monolayers of
HEK293 cells stably expressing either μ-OR or δ-OR were harvested,
collected by centrifugation at 1500 rpm for 5 min, and washed once
with phosphate buffered saline (PBS) at pH 7.5. Membranes from cell
pellets were carried out as described by Georgoussi and Zioudrou.64
Briefly cell pellets were resuspended in ice-cold membrane buffer (10
mM Tris-HCl, pH 7.5, 0.1 mM EDTA), homogenized, and centrifuged
at 2000 rpm for 3 min at 4 °C. Supernatants were further centrifuged
at 45 000 rpm for 30 min at 4 °C. The membrane pellet was
resuspended in ice-cold membrane buffer at a protein concentration of
approximately 1 mg/mL and stored in aliquots at −70 °C. Protein
concentration was determined according to the Bradford assay.
Detection of MAPK Phosphorylation. HEK293 cells stably
expressing the μ- or the δ-opioid receptor were seeded in 10 cm plates.
Sixteen hours before the addition of the ligands, the culture medium
was removed and replaced by fresh serum-free medium. The ligands
were added to the cells and allowed incubate in a time-dependent
manner at 37 °C. Cell monolayers were rinsed with PBS, and whole
lysates were prepared by the addition of lysis buffer containing 1%
Triton X-100, 10 mM Tris, pH 7.6, 5 mM Na2EDTA, 50 mM NaCl,
50 mM NaF, and 30 mM Na4P2O7 supplemented with 2 μg/mL
i
The peptide was then cleaved from the resin by adding the dry resin
to a mixture of 2.5% TES, 2.5% water in TFA (10 mL/g resin). After 3
h, the resin was removed by filtration and the filtrate was added
dropwise to dry, cold diethyl ether. The precipitated peptide was
isolated by centrifugation, and the ether was decanted. Cold ether was
added two more times to the peptide, and it was decanted again after
centrifugation. After this isolation process, the peptide was purified by
preparative HPLC. The purified peptide 3 was isolated in 38% overall
yield and was >99% pure as determined by analytical RP-HPLC. The
molecular weight of the pure peptide was confirmed by electrospray
ionization (ESI) (positive ion mode) mass spectrometry (Table 4).
Synthesis of H-Dmt-NMe-D-Ala-Aba-Gly-NH2 4. Analogue 4
was prepared on a MBHA polystyrene resin (0.14 mmol, 0.95 mmol/
g) using the Boc peptide strategy. The resin was swollen in CH2Cl2 for
1 h, filtered, and a 20% DIPEA/CH2Cl2 solution was added and
subsequently filtered (3 × 1 min). After filtration and washing of the
resin three times with CH2Cl2, couplings were carried out by adding
4.0 equiv of the αN-Boc-protected amino acid, together with 4.0 equiv
of TBTU and 8.0 equiv of DIPEA in DMF/CH2Cl2, 1/1 ([Boc-AA-
OH] = 0.5 M), for 4 h. The completeness of the couplings was
checked with the NF-31 test.60 All couplings were performed two
times because of positive color tests. When the coupling was
i
completed, the resin was washed (three times with DMF, PrOH,
and CH2Cl2), followed by Boc deprotection which was performed in
50% TFA/2% anisole/CH2Cl2 (1 × 10 min and 1 × 20 min). After
that, the resin was washed with CH2Cl2 three times and filtered. A 20%
DIPEA/CH2Cl2 solution was then added, and the resin was filtered (3
× 1 min). After the resin was washed three times with CH2Cl2, the
coupling with the next amino acid was carried out.
Fmoc-Dmt-OH (2.5 equiv) was coupled overnight using HOBt (2.5
equiv) and DIC (2.5 equiv) in DMF ([Fmoc-Dmt-OH] = 0.5 M). The
completeness of the couplings was checked with the chloranil test.61
When the coupling was completed, the resin was washed (three times
i
with DMF, PrOH, and DMF again), followed by Fmoc deprotection
and filtration (20% piperidine/DMF (2 × 10 min)). The resin was
then washed (three times with DMF, iPrOH, DMF again, CH2Cl2, and
Et2O), dried under reduced pressure (8 h), and stored at 4 °C.
Cleavage of the peptide from the resin was performed using
anhydrous HF in a Teflon made HF line (Asti, France) under vacuum.
Anisole (2 mL) and HFliq (10 mL) were used for 1 g of peptide−resin.
After the sample was stirred for 1 h in an ice bath, the HF was
removed by evaporation under vacuum. The reaction vessel was
disconnected from the HF line, filled with dry ether (30 mL), and
filtered on a glass filter. The precipitated peptide and the resin beads
were washed with pure acetic acid. After the filtrate was lyophilized,
the peptide was purified by preparative HPLC. The purified peptide 4
was isolated in 43% overall yield and was >99% pure as determined by
analytical RP-HPLC. The molecular weight of the pure peptide was
confirmed by electrospray ionization (ESI) (positive ion mode) mass
spectrometry (Table 4).
7855
dx.doi.org/10.1021/jm200894e|J. Med. Chem. 2011, 54, 7848−7859

Journal of Medicinal Chemistry
Article
antipain, 2 μg/mL leupeptin, 1 μg/mL benzamidine, 1 mM
phenylmethylsulfonyl fluoride, 1 mM Na3VO4, protease (Complete
Mini, EDTA-free; Roche), and phosphatase inhibitor cocktails (P5726,
Sigma-Aldrich) for 30 min. Soluble proteins were separated by
centrifugation at 3000 rpm for 10 min. Protein concentration was
determined according to the method of Bradford.65 The proteins were
prepared for SDS−polyacrylamide gel electrophoresis and electro-
blotted onto polyvinylidene difluoride membranes. Membranes were
exposed to mouse monoclonal antibodies that are specific to
phosphorylated MAPK (1:1000 dilutions in blocking solution, p-
ERK1, Santa Cruz Biotechnology Inc.). Immunoreactive proteins were
visualized using a horseradish-peroxidase-sensitive ECL chemilumi-
nescent Western blotting kit (Pierce, Thermo Scientific, U.S.).
Functional Screening of Dermorphin Tetrapeptide Ana-
logues with Neuronal Cell Cultures. Primary Cortical Cell
Cultures. Frontal cortex tissue were harvested from embryonic day
15 and day 14 crl:NMRI mice (Charles River, Sulzfeld, Germany).
After ethyl ether anesthesia, mice were sacrificed by cervical dislocation
in accordance with the German Animal Protection Act, Section 4.
Frontal cortex tissue was dissociated enzymatically in DMEM 10/10
(10% horse and 10% fetal calf serum), including papain (10 U/mL)
and DNase I (8000 units/mL), and mechanically with transfer
pipettes. The cells were resuspended in DMEM 10/10 at a density of
1.0 × 106 cells/mL, and an amount of 300 μL was seeded onto
microelectrode array neurochip (MEA) surfaces. Cultures were
incubated at 37 °C in a 10% CO2 atmosphere until used, typically 4
weeks to 3 months after seeding. Culture medium was replenished
three times a week with DMEM containing 10% horse serum. The
neuronal networks are composed of a mixture of neurons and glial
cells comparable to the tissue of origin. The developing cocultures
were initially treated with 5-fluoro-2′-deoxyuridine (25 μM) and
uridine (63 μM) for 48 h to prevent further glial proliferation and
overgrowth. Electrical activity starts spontaneously after approximately
3−4 days in vitro in the form of random spiking. After 4 weeks in
culture, the activity pattern stabilizes and is composed of one
coordinated main burst pattern with several coordinated subpat-
terns.37−39 For this study, cultures between 28 and 53 days in vitro
were used.
Microelectrode Array Neurochips. MEAs were provided by the
Centre for Network Neuroscience (CNNS) at the University of North
Texas. These 5 × 5 cm2 glass chips have a central recording matrix
with 64 passive electrodes and indium tin oxide conductors. The
hydrophobic insulation material surface was activated by a brief butane
flaming pulse through a stainless steel mask and coated with poly-D-
lysine (25 μg/mL, 30−70 kDa) and laminin (16 μg/mL) to ensure cell
attachment within a confined adhesive region (5 mm diameter
centered on the electrode array).
Multichannel Recording and Data Analysis. For extracellular
recording, MEAs were placed into sterilized constant-bath recording
chambers and maintained at 37 °C. Recordings were made in DMEM/
10% horse serum. The pH was maintained at 7.4 with a continuous
stream of filtered, humidified airflow with 10% CO2. Recording was
performed with a computer-controlled 64-channel MEA workstation
acquisition system (Plexon, Inc., Dallas, TX, U.S.) providing
amplification, filtering, and digital signal processing of microelectrode
array signals. The total system gain used was 10K with a simultaneous
40 kHz sampling rate. The signals routinely recorded by these
neurochips are located in the range of 15−1800 μV.
The action potentials, or “spikes”, were recorded as spike trains;
they are clustered in so-called bursts. Bursts were quantitatively
described via direct spike train analysis using the program Neuro-
EXplorer (Plexon Inc., Dallas, TX, U.S.) and in-house programs.
Bursts were defined by the beginning and end of short spike events.
Maximum spike intervals defining the start of a burst were adjusted
from 50 to 150 ms, and maximum intervals to end a burst were from
100 to 300 ms.
High content analysis of the network activity patterns provides a
multiparametric description characterizing the changes in four
categories: overall activity, burst structure, synchronicity, and
oscillatory behavior. We quantify the substance-specific activity
changes by extracting a total of 35 activity-describing spike train
parameters for these four categories as described previously37−39
Synchronicity and oscillatory behavior were captured through the
temporal and network coefficients of variation (CVTIME and
CVNETWORK) of the burst rate, burst period, and spike rate parameters.
The variation coefficients therefore quantify the spatiotemporal
behavior, reflecting temporal dynamics and the fundamental
interactions within the networks. Here, CVTIME characterizes the
periodic behavior of a single neuron activity pattern. CVNETWORK
describes the degree of coordination between different neurons in the
activity patterns and is a measure of firing synchronicity. Other
activity-describing parameters quantify the dose response kinetics in
their course: number of phases, slope (Hill coefficient), and 10%, 50%,
and 90% effective concentrations. For direct comparison, all
parameters were normalized for each experiment and each
experimental treatment with regard to the corresponding values of
the native reference activity. Values were derived from 60 s bin data
from 30 min after the stabilization of activity. From each network 21−
158 separate neurons were simultaneously recorded.
Pattern Recognition and Classification. To clarify the mode of
action of the dermorphin tetrapeptide analogues 1−4 on the activity of
cortical networks, we further analyzed these experiments using
methods of pattern recognition. For each stable activity phase after
substance application, we calculated 200 spike train parameters
normalized by the native activity using Squid (NeuroProof GmbH,
Rostock, Germany). These data records were computed for the
tetrapeptides 1−4 and the reference substances. Using feature
selection algorithms as a widely accepted method in bioinformatics
and a wide range of different approaches proposed,66 we calculated
rankings of features using various score methods based on the
respective class decomposition. The 40 most suitable parameters of all
200 spike train parameters were selected, and their total correct
predictions were compared. In this manner we obtained the best result
for a MDL (minimal description length) modified algorithm. A train
data set with these 40 spike train parameters was established in the
form using the data records from the reference substances. We then
trained an artificial neuronal network, a multilayer feed forward
network, and back-propagation algorithm without hidden units. The
respective data records of the tetrapeptides were all subsequently
classified using PatternExpert (NeuroProof GmbH, Rostock, Ger-
many).
The database was used for training of an artificial neuronal network.
The machine learning algorithm is implemented in our pattern
recognition software platform. We use a multilayer feed forward
perceptron and a resilient-propagation learning algorithm. We do not
use a hidden layer. We use as many input knots as we have features
and one output knot for each class which has to be classified. The
nonusage of hidden layers is justified by the relatively high variation of
our data.
Validation of Classification. To validate our machine learning
algorithms, we performed the validation in the form of cross-
validation. For this we left out the data records of one experiment (one
culture on one MEA neurochip) and trained the artificial neuronal
network with all other data records. Then the left out data record was
classified with the trained artificial network. This process was repeated
until each data record was used once as a left out data record for
classification.
The neuronal networks were acutely treated with accumulating
concentrations of the new dermorphin tetrapeptide analogues 1−4 in
the range from 1 fM to 100 μM (1 fM, 1 pM, 100 pM, 10 nM, 100
nM, 1 μM, 10 μM, 30 μM, 100 μM). The network response (spike
rate) was observed online. To obtain quantitative data from the 1 h
recordings, we analyzed for each of the nine concentrations a stable
activity phase of the last 30 min.
The multichannel signal acquisition system delivered single neuron
spike data including action potential waveforms. Spike identification
and separation were accomplished using a template-matching
algorithm in real time. This permitted the simultaneous extracellular
recording of action potentials from a maximum of 256 neurons.
7856
dx.doi.org/10.1021/jm200894e|J. Med. Chem. 2011, 54, 7848−7859

Journal of Medicinal Chemistry
Article
Since we were interested in the mode of action, we did not use the
concentration as an input parameter for the classification. Although
this decreases the correctness of our classification experiments, we
obtained assignments dependent solely on the mode of action. A given
substance can be identified correctly by its changes of activity patterns
if most of the data records are classified correctly.
With methods of pattern recognition and classification algorithms
the electrical activity patterns of all newly synthesized compounds that
we received from partners were compared to those of the reference
compounds in the database and the similarities were calculated.
The new compounds underwent a further specific classification
protocol to determine modes of actions for these compounds such as
unspecific binding, agonist vs antagonist behavior, analgesic potency,
and anticonvulsive properties.
Statistical Analysis. Results are expressed as series mean ± SEM.
The distributions of the absolute parameters were tested for normality.
By use of the SPSS statistical software, significant changes induced by
substance application were tested by analysis of variance (ANOVA)
followed by Dunnett’s multiple comparison post hoc test with the
native activity as the common control. P < 0.05 was considered
statistically significant.
To determine the half effective concentration (EC50), standard
logistic dose−response curves (either one or the sum of two,
depending on the data) were fitted to the data points using the sum of
least squares algorithm of the Solver module in Microsoft Excel.
REFERENCES
■
(1) Georgoussi, Z. Novel Interactive Partners Regulating Opioid
Receptor Signalling beyond the G Protein Paradigm. In Molecular
Aspects of G Protein-Coupled Receptors: Interacting Proteins and Function;
Ciruela, F., Lujan, R., Eds.; Nova Biomedical Books: New York, 2008;
pp 169−206.
(2) Lee, Y. S.; Petrov, R.; Park, C. K.; Ma, S. W.; Davis, P.; Lai, J.;
Porreca, F.; Vardanyan, R.; Hruby, V. J. Development of novel
enkephalin analogues that have enhanced opioid activities at both mu
and delta opioid receptors. J. Med. Chem. 2007, 50, 5528−5532.
(3) Ananthan, S. Opioid ligands with mixed mu/delta opioid
receptor interactions: an emerging approach to novel analgesics. AAPS
J. 2006, 8, E118−E125.
(4) Mollica, A.; Davis, P.; Ma, S. W.; Porreca, F.; Lai, J.; Hruby, V. J.
Synthesis and biological activity of the first cyclic biphalin analogues.
Bioorg. Med. Chem. Lett. 2006, 16, 367−372.
(5) Balboni, G.; Guerrini, R.; Salvadori, S.; Bianchi, C.; Rizzi, D.;
Bryant, S. D.; Lazarus, L. H. Evaluation of the Dmt-Tic
pharmacophore: conversion of a potent delta-opioid receptor
antagonist into a potent delta agonist and ligands with mixed
properties. J. Med. Chem. 2002, 45, 713−720.
(6) Balboni, G.; Salvadori, S.; Guerrini, R.; Negri, L.; Giannini, E.;
Jinsmaa, Y.; Bryant, S. D.; Lazarus, L. H. Potent delta-opioid receptor
agonists containing the Dmt-Tic pharmacophore. J. Med. Chem. 2002,
45, 5556−5563.
(7) Yamazaki, M.; Suzuki, T.; Narita, M.; Lipkowski, A. W. The
opioid peptide analogue biphalin induces less physical dependence
than morphine. Life Sci. 2001, 69, 1023−1028.
ASSOCIATED CONTENT
■
S
* Supporting Information
(8) Qi, J.; Mosberg, H. I.; Porreca, F. Modulation of the potency and
efficacy of mu-mediated antinociception by delta-agonists in the
mouse. J. Pharmacol. Exp. Ther. 1990, 254, 683−689.
List of substances in the NeuroProof database used in the
similarity analysis. This material is available free of charge via
the Internet at http://pubs.acs.org.
(9) Zhang, X.; Bao, L.; Guan, J. S. Role of delivery and trafficking of
delta-opioid peptide receptors in opioid analgesia and tolerance.
Trends Pharmacol. Sci. 2006, 27, 324−329.
AUTHOR INFORMATION
■
(10) Gomes, I.; Gupta, A.; Filipovska, J.; Szeto, H. H.; Pintar, J. E.;
Devi, L. A. A role for heterodimerization of mu and delta opiate
receptors in enhancing morphine analgesia. Proc. Natl. Acad. Sci. U.S.A.
2004, 101, 5135−5139.
Corresponding Author
*Phone: +31 2 629 3295. E-mail: datourwe@vub.ac.be.
ACKNOWLEDGMENTS
■
(11) He, L.; Fong, J.; von Zastrow, M.; Whistler, J. L. Regulation of
opioid receptor trafficking and morphine tolerance by receptor
oligomerization. Cell 2002, 108, 271−282.
(12) Kieffer, B. L. Opioids: first lessons from knockout mice. Trends
Pharmacol. Sci. 1999, 20, 19−26.
We are thankful to M.-P. Papakonstantinou for her help with
the receptor binding experiments. This work was supported by
the Research FoundationFlanders (FWO-Vlaanderen, Bel-
gium, Grant G.0008.08) and the European grant “Normolife”
(Grant LSHC-CT-2006-037733).
(13) Fan, T.; Varghese, G.; Nguyen, T.; Tse, R.; O’Dowd, B. F.;
George, S. R. A role for the distal carboxyl tails in generating the novel
pharmacology and G protein activation profile of mu and delta opioid
receptor hetero-oligomers. J. Biol. Chem. 2005, 280, 38478−38488.
(14) Leontiadis, L. J.; Papakonstantinou, M. P.; Georgoussi, Z.
Regulator of G protein signaling 4 confers selectivity to specific G
proteins to modulate mu- and delta-opioid receptor signaling. Cell.
Signalling 2009, 21, 1218−1228.
(15) Gomes, I.; Jordan, B. A.; Gupta, A.; Trapaidze, N.; Nagy, V.;
Devi, L. A. Heterodimerization of mu and delta opioid receptors: a
role in opiate synergy. J. Neurosci. 2000, 20, No. RC110.
(16) Wang, D. X.; Sun, X. C.; Bohn, L. M.; Sadee, W. Opioid
receptor homo- and heterodimerization in living cells by quantitative
bioluminescence resonance energy transfer. Mol. Pharmacol. 2005, 67,
2173−2184.
(17) Rozenfeld, R.; Devi, L. A. Receptor heterodimerization leads to
a switch in signaling: beta-arrestin2-mediated ERK activation by mu-
delta opioid receptor heterodimers. FASEB J. 2007, 21, 2455−2465.
(18) He, L.; Whistler, J. L. An opiate cocktail that reduces morphine
tolerance and dependence. Curr. Biol. 2005, 15, 1028−1033.
(19) Bishop, M. J.; Garrido, D. M.; Boswell, G. E.; Collins, M. A.;
Harris, P. A.; Mcnutt, R. W.; O’Neill, S. J.; Wei, K.; Chang, K. J. 3-
(alpha R)-alpha-((2S,5R)-4-Allyl-2,5-dimethyl-1-piperazinyl)-3-hy-
droxybenzyl)-N-alkyl-N-arylbenzamides: potent, non-peptidic agonists
of both mu and delta opioid receptors. J. Med. Chem. 2003, 46, 623−
633.
ABBREVIATIONS USED
■
Aba, 4-amino-1,2,4,5-tetrahydro-2-benzazepin-3-one; BBB,
blood−brain barrier; Boc, tert-butyloxycarbonyl; BVD03, H-
Dmt-NMe-D-Ala-Aba-Gly-NH2; cAMP, cyclic adenosine mono-
phosphate; CNS, central nervous system; DAMGO, H-Tyr-D-
Ala-Gly-NMe-Phe-Gly-ol; DIC, 1,3-diisopropylcarbodiimide;
DIPEA, N,N′-diisopropylethylamine; DMF, N,N′-dimethylfor-
mamide; Dmt, 2′,6′-dimethyltyrosine; DPDPE, c[D-Pen2,D-
Pen5]enkephalin; ERK, extracellular regulated kinases; ESI,
electrospray ionization; FC, frontal cortex; Fmoc, 9-fluorenyl-
methyloxycarbonyl; GPI, guinea pig ileum; HEK293, human
embryonic kidney; GBP, gabapentin; HOBt, 1-hydroxybenzo-
triazole; HPLC, high performance liquid chromatography;
MAPK, mitogen activated protein kinases; MBHA, 4-
methylbenzhydrylamine; MVD, mouse vas deferens; OR,
opioid receptor; PAGE, polyacrylamide gel electrophoresis;
RP, reversed phase; SAR, structure−activity relationship; SC,
spinal cord; SDS, sodium dodecyl sulfate; SUF, sufentanil;
TBTU, O-(benzotriazol-1-yl)-N,N,N′,N′-tetramethyluronium
tetrafluoroborate; TFA, trifluoroacetic acid; TLC, thin layer
chromatography
7857
dx.doi.org/10.1021/jm200894e|J. Med. Chem. 2011, 54, 7848−7859

Journal of Medicinal Chemistry
Article
(20) Broccardo, M.; Erspamer, V.; Falconierierspamer, G.; Improta,
G.; Linari, G.; Melchiorri, P.; Montecucchi, P. C. Pharmacological data
on dermorphins, a new class of potent opioid-peptides from amphibian
skin. Br. J. Pharmacol. 1981, 73, 625−631.
(37) Gramowski, A.; Stuwe, S.; Jugelt, K.; Loock, J.; Schroder, O.;
Gross, G. W.; Weiss, D. G. Detecting neurotoxicity through electrical
activity changes of neuronal networks on multielectrode neurochips.
ALTEX 2006, 23, 410−415.
(38) Gramowski, A.; Jugelt, K.; Stuwe, S.; Schulze, R.; McGregor, G.
P.; Wartenberg-Demand, A.; Loock, J.; Schroder, O.; Weiss, D. G.
Functional screening of traditional antidepressants with primary
cortical neuronal networks grown on multielectrode neurochips. Eur.
J. Neurosci. 2006, 24, 455−465.
(39) Gramowski, A.; Jugelt, K.; Weiss, D. G.; Gross, G. W. Substance
identification by quantitative characterization of oscillatory activity in
murine spinal cord networks on microelectrode arrays. Eur. J. Neurosci.
2004, 19, 2815−2825.
(40) Gross, G. W.; Harsch, A.; Rhoades, B. K.; Gopel, W. Odor, drug
and toxin analysis with neuronal networks in vitro: extracellular array
recording of network responses. Biosens. Bioelectron. 1997, 12, 373−
393.
(41) Gramowski, A.; Schiffmann, D.; Gross, G. W. Quantification of
acute neurotoxic effects of trimethyltin using neuronal networks
cultured on microelectrode arrays. Neurotoxicology 2000, 21, 331−342.
(42) Johnstone, A. F. M.; Gross, G. W.; Weiss, D. G.; Schroeder, O.
H. U.; Gramowski, A.; Shafer, T. J. Microelectrode arrays: a
physiologically based neurotoxicity testing platform for the 21st
century. Neurotoxicology 2010, 31, 331−350.
(43) Jugelt, K. Functional Screening of Neuroactive Compounds:
Multielectrode-Array Neurochips as a Versatile Tool in Drug Development;
Institute of Biological Sciences, University of Rostock: Rostock,
Germany, 2007.
(21) Montecucchi, P. C.; Decastiglione, R.; Piani, S.; Gozzini, L.;
Erspamer, V. Amino-acid-composition and sequence of dermorphin, a
novel opiate-like peptide from the skin of Phyllomedusa sauvagei. Int. J.
Pept. Protein Res. 1981, 17, 275−283.
(22) Melchiorri, P.; Negri, L. The dermorphin peptide family. Gen.
Pharmacol. 1996, 27, 1099−1107.
(23) Borics, A.; Toth, G. Structural comparison of mu-opioid
receptor selective peptides confirmed four parameters of bioactivity. J.
Mol. Graphics Modell. 2010, 28, 495−505.
(24) Hitchcock, S. A.; Pennington, L. D. Structure−brain exposure
relationships. J. Med. Chem. 2006, 49, 7559−7583.
(25) Doan, K. M. M.; Humphreys, J. E.; Webster, L. O.; Wring, S. A.;
Shampine, L. J.; Serabjit-Singh, C. J.; Adkison, K. K.; Polli, J. W.
Passive permeability and P-glycoprotein-mediated efflux differentiate
central nervous system (CNS) and non-CNS marketed drugs. J.
Pharmacol. Exp. Ther. 2002, 303, 1029−1037.
(26) Malakoutikhah, M.; Prades, R.; Teixido, M.; Giralt, E. N-Methyl
phenylalanine-rich peptides as highly versatile blood−brain barrier
shuttles. J. Med. Chem. 2010, 53, 2354−2363.
(27) Brasnjevic, I.; Steinbusch, H. W. M.; Schmitz, C.; Martinez-
Martinez, P. Delivery of peptide and protein drugs over the blood−
brain barrier. Prog. Neurobiol. 2009, 87, 212−251.
(28) Sasaki, Y.; Matsui, M.; Taguchi, M.; Suzuki, K.; Sakurada, S.;
Sato, T.; Sakurada, T.; Kisara, K. D-Arg2-Dermorphin tetrapeptide
analogsa potent and long-lasting analgesic activity after subcuta-
neous administration. Biochem. Biophys. Res. Commun. 1984, 120,
214−218.
(29) Ogawa, T.; Miyamae, T.; Murayama, K.; Okuyama, K.;
Okayama, T.; Hagiwara, M.; Sakurada, S.; Morikawa, T. Synthesis
and structure−activity relationships of an orally available and long-
acting analgesic peptide, N-alpha-amidino-Tyr-D-Arg-Phe-Me-beta-
Ala-OH (ADAMB). J. Med. Chem. 2002, 45, 5081−5089.
(30) Tourwe, D.; Verschueren, K.; Frycia, A.; Davis, P.; Porreca, F.;
Hruby, V. J.; Toth, G.; Jaspers, H.; Verheyden, P.; VanBinst, G.
Conformational restriction of Tyr and Phe side chains in opioid
peptides: information about preferred and bioactive side-chain
topology. Biopolymers 1996, 38, 1−12.
(31) Calderan, A.; Ruzza, P.; Ancona, B.; Cima, L.; Giusti, P.; Borin,
G. Synthesis, conformational and pharmacological studies on
dermorphin N-terminal tetrapeptide analogues. Lett. Pept. Sci. 1998,
5, 71−73.
(32) Ballet, S.; Misicka, A.; Kosson, P.; Lemieux, C.; Chung, N. N.;
Schiller, P. W.; Lipkowski, A. W.; Tourwe, D. Blood−brain barrier
penetration by two dermorphin tetrapeptide analogues: role of
lipophilicity vs structural flexibility. J. Med. Chem. 2008, 51, 2571−
2574.
(33) Ballet, S.; Frycia, A.; Piron, J.; Chung, N. N.; Schiller, P. W.;
Kosson, P.; Lipkowski, A. W.; Tourwe, D. Synthesis and biological
evaluation of constrained analogues of the opioid peptide H-Tyr-D-
Ala-Phe-Gly-NH2 using the 4-amino-2-benzazepin-3-one scaffold. J.
Pept. Res. 2005, 66, 222−230.
(34) Dai, Y.; Iwata, K.; Fukuoka, T.; Kondo, E.; Tokunaga, A.;
Yamanaka, H.; Tachibana, T.; Liu, Y.; Noguchi, K. Phosphorylation of
extracellular signal-regulated kinase in primary afferent neurons by
noxious stimuli and its involvement in peripheral sensitization. J.
Neurosci. 2002, 22, 7737−7745.
(44) Schiller, P. W.; Nguyen, T. M. D.; Berezowska, I.; Dupuis, S.;
Weltrowska, G.; Chung, N. N.; Lemieux, C. Synthesis and in vitro
opioid activity profiles of DALDA analogues. Eur. J. Med. Chem. 2000,
35, 895−901.
(45) Bryant, S. D.; Jinsmaa, Y.; Salvadori, S.; Okada, Y.; Lazarus, L.
H. Dmt and opioid peptides: a potent alliance. Biopolymers 2003, 71,
86−102.
(46) De Wachter, R.; de Graaf, C.; Keresztes, A.; Vandormael, B.;
́
́
Ballet, S.; Tοth, G.; Rognan, D.; Tourwe, D. Synthesis, biological
evaluation and automated docking of constrained analogs of the opioid
peptide H-Dmt-D-Ala-Phe-Gly-NH2 using the 4- or 5-methyl
susbstituted 4-amino-1,2,3,4-tetrahydro-2-benzazepin-3-one scaffold.
J. Med. Chem. 2011, 54, 6538−6547.
(47) Morou, E.; Georgoussi, Z. Expression of the third intracellular
loop of the δ-opioid receptor inhibits signaling by opioid receptors and
other G protein-coupled receptors. J. Pharmacol. Exp. Ther. 2005, 315,
1368−1379.
(48) Pineyro, G.; Archer-Lahlou, E. Ligand-specific receptor states:
implications for opiate receptor signaling and regulation. Cell.
Signalling 2007, 19, 8−19.
(49) Pineyro, G. Membrane signaling complexes: implications for
development of functionally selective ligands modulating heptahelical
receptor signaling. Cell. Signalling 2009, 21, 179−185.
(50) Tourwe, D.; Verschueren, K.; Van Binst, G.; Davis, P.; Porreca,
F.; Hruby, V. J. Dermorphin sequence with high delta-affinity by fixing
the phe side-chain to trans at chi-1. Bioorg. Med. Chem. Lett. 1992, 2,
1305−1308.
(51) Belcheva, M. M.; Haas, P. D.; Tan, Y.; Heaton, V. M.; Coscia, C.
J. The fibroblast growth factor receptor is at the site of convergence
between mu-opioid receptor and growth factor signaling pathways in
rat C6 glioma cells. J. Pharmacol. Exp. Ther. 2002, 303, 909−918.
(52) Asensio, V. J.; Miralles, A.; Garcia-Sevilla, J. A. Stimulation of
mitogen-activated protein kinase kinases (MEK1/2) by mu-, delta- and
kappa-opioid receptor agonists in the rat brain: regulation by chronic
morphine and opioid withdrawal. Eur. J. Pharmacol. 2006, 539, 49−56.
(53) Ferrer-Alcon, M.; Garcia-Fuster, M. J.; La Harpe, R.; Garcia-
Sevilla, J. A. Long-term regulation of signalling components of adenylyl
cyclase and mitogen-activated protein kinase in the pre-frontal cortex
of human opiate addicts. J. Neurochem. 2004, 90, 220−230.
(54) Sagan, S.; Karoyan, P.; Lequin, O.; Chassaing, G.; Lavielle, S. N-
and C alpha-methylation in biologically active peptides: synthesis,
(35) Valjent, E.; Pages, C.; Rogard, M.; Besson, M. J.; Maldonado,
R.; Caboche, J. Delta 9-tetrahydrocannabinol-induced MAPK/ERK
and Elk-1 activation in vivo depends on dopaminergic transmission.
Eur. J. Neurosci. 2001, 14, 342−352.
(36) Narita, M.; Ioka, M.; Suzuki, M.; Narita, M.; Suzuki, T. Effect of
repeated administration of morphine on the activity of extracellular
signal regulated kinase in the mouse brain. Neurosci. Lett. 2002, 324,
97−100.
7858
dx.doi.org/10.1021/jm200894e|J. Med. Chem. 2011, 54, 7848−7859

Journal of Medicinal Chemistry
Article
structural and functional aspects. Curr. Med. Chem. 2004, 11, 2799−
2822.
(55) Kessler, H.; Chatterjee, J.; Doedens, L.; Opperer, F.; Gilon, C.;
Hruby, V. J.; Mierke, D. New perspectives in peptide chemistry by N-
alkylation. Biopolymers 2007, 88, 519.
(56) Wilkes, B. C.; Schiller, P. W. Comparative-analysis of various
proposed models of the receptor-bound conformation of H-Tyr-Tic-
Phe-Oh related delta-opioid antagonists. Biopolymers 1995, 37, 391−
400.
(57) Schmidt, R.; Kalman, A.; Chung, N. N.; Lemieux, C.; Horvath,
C.; Schiller, P. W. Structure−activity-relationships of dermorphin
analogs containing N-substituted amino-acids in the 2-position of the
peptide sequence. Int. J. Pept. Protein Res. 1995, 46, 47−55.
(58) Perlikowska, R.; Fichna, J.; Wyrebska, A.; Poels, J.; Broeck, J. V.;
Toth, G.; Storr, M.; do Rego, J. C.; Janecka, A. Design, synthesis and
pharmacological characterization of endomorphin analogues with non-
cyclic amino acid residues in position 2. Basic Clin. Pharmacol. Toxicol.
2010, 106, 106−113.
(59) Xie, X. M.; Wisor, J. P.; Hara, J.; Crowder, T. L.; LeWinter, R.;
Khroyan, T. V.; Yamanaka, A.; Diano, S.; Horvath, T. L.; Sakurai, T.;
Toll, L.; Kilduff, T. S. Hypocretin/orexin and nociceptin/orphanin FQ
coordinately regulate analgesia in a mouse model of stress-induced
analgesia. J. Clin. Invest. 2008, 118, 2471−2481.
(60) Madder, A.; Farcy, N.; Hosten, N. G. C.; De Muynck, H.; De
Clercq, P. J.; Barry, J.; Davis, A. P. A novel sensitive colorimetric assay
for visual detection of solid-phase bound amines. Eur. J. Org. Chem.
1999, 2787−2791.
(61) Christensen, T. Qualitative test for monitoring coupling
completeness in solid-phase peptide-synthesis using chloranil. Acta
Chem. Scand., Ser. B 1979, 33, 763−766.
(62) Megaritis, G.; Merkouris, M.; Georgoussi, Z. Functional
domains of delta- and mu-opioid receptors responsible for adenylyl
cyclase inhibition. Recept. Channels 2000, 7, 199−212.
(63) Merkouris, M.; Mullaney, I.; Georgoussi, Z.; Milligan, G.
Regulation of spontaneous activity of the delta-opioid receptor: studies
of inverse agonism in intact cells. J. Neurochem. 1997, 69, 2115−2122.
(64) Georgoussi, Z.; Zioudrou, C. Effect of delta-opioid antagonists
on the functional coupling between opioid receptors and G-proteins in
rat-brain membranes. Biochem. Pharmacol. 1993, 45, 2405−2410.
(65) Bradford, M. M. Rapid and sensitive method for quantitation of
microgram quantities of protein utilizing principle of protein−dye
binding. Anal. Biochem. 1976, 72, 248−254.
(66) Guyon, I.; Elisseeff, A. An introduction to variable and feature
selection. J. Mach. Learn. Res. 2003, 3 (7−8), 1157−1182.
7859
dx.doi.org/10.1021/jm200894e|J. Med. Chem. 2011, 54, 7848−7859