Biological and conformational evaluation of bifunctional compounds for opioid receptor agonists and neurokinin 1 receptor antagonists possessing two penicillamines

Article abstract of DOI:10.1021/jm100157m

Neuropathic pain states and tolerance to opioids can result from system changes in the CNS, such as up-regulation of the NK1 receptor and substance P, lead to antiopioid effects in ascending or descending pain-signaling pathways. Bifunctional compounds, possessing both the NK1 antagonist pharmacophore and the opioid agonist pharmacophore with ?-selectivity, could counteract these system changes to have significant analgesic efficacy without undesirable side effects. As a result of the introduction of cyclic and topological constraints with penicillamines, 2 (Tyr-cyclo[d-Pen-Gly-Phe-Pen]-Pro-Leu-Trp-NH-[3?,5?-(CF ₃)₂-Bzl]) was found as the best bifunctional compound with effective NK1 antagonist and potent opioid agonist activities, and 1400-fold ?-selectivity over the ?-receptor. The NMR structural analysis of 2 revealed that the relative positioning of the two connected pharmacophores as well as its cyclic and topological constraints might be responsible for its excellent bifunctional activities as well as its significant ?-opioid selectivity. Together with the observed high metabolic stability, 2 could be considered as a valuable research tool and possibly a promising candidate for a novel analgesic drug.

Full text of DOI:10.1021/jm100157m

J. Med. Chem. 2010, 53, 5491–5501 5491
DOI: 10.1021/jm100157m
Biological and Conformational Evaluation of Bifunctional Compounds for Opioid Receptor Agonists
and Neurokinin 1 Receptor Antagonists Possessing Two Penicillamines
Takashi Yamamoto,†,§ Padma Nair,† Neil E. Jacobsen,† Vinod Kulkarni,† Peg Davis,‡ Shou-wu Ma,‡ Edita Navratilova,‡
Henry I. Yamamura,‡ Todd W. Vanderah,‡ Frank Porreca,‡ Josephine Lai,‡ and Victor J. Hruby*,†
†Departments of Chemistry and Biochemistry, 1306 East University Boulevard and ‡Pharmacology, 1501 North Campbell Avenue,
University of Arizona, Tucson, Arizona 85721. § Current Address: Exploratory Research Laboratory, Ajinomoto Pharmaceuticals Co., Ltd.,
1-1 Suzuki-cho, Kawasaki-ku, Kawasaki-shi, Japan 210-8681; E-mail: takashia_yamamoto@ajinomoto.com
Received February 4, 2010
Neuropathic pain states and tolerance to opioids can result from system changes in the CNS, such as
up-regulation of the NK1 receptor and substance P, lead to antiopioid effects in ascending or
descending pain-signaling pathways. Bifunctional compounds, possessing both the NK1 antagonist
pharmacophore and the opioid agonist pharmacophore with δ-selectivity, could counteract these
system changes to have significant analgesic efficacy without undesirable side effects. As a result of the
introduction of cyclic and topological constraints with penicillamines, 2 (Tyr-cyclo[D-Pen-Gly-Phe-
Pen]-Pro-Leu-Trp-NH-[30,50-(CF3)2-Bzl]) was found as the best bifunctional compound with effective
NK1 antagonist and potent opioid agonist activities, and 1400-fold δ-selectivity over the μ-receptor.
The NMR structural analysis of 2 revealed that the relative positioning of the two connected
pharmacophores as well as its cyclic and topological constraints might be responsible for its excellent
bifunctional activities as well as its significant δ-opioid selectivity. Together with the observed high
metabolic stability, 2 could be considered as a valuable research tool and possibly a promising candidate
for a novel analgesic drug.
Introduction
receptors for those newly released pain-causing ligands.1-3
Current treatment of prolonged pain generally can only
modulate pain and cannot counteract these induced neuro-
plastic changes. Thus, it is not surprising that current analge-
sic drugs do not work well in these pathological conditions.
To effectively address this problem, we have taken a new
approach to design bifunctional ligands that act as agonists at
opioid receptors and as antagonists at NK1a receptors to
block the signals mediated by substance P, which is one of
these newly expressed pain-causing neurotransmitters.4-7
Our developing ligands have synergistic biological activities
and overlapping pharmacophores (Figure 1) and would not
be expected to possess the undesirable side effects of current
opioids and other pain-killing drugs such as the development
of tolerance. Indeed, previous reports have indicated that
substance P plays a role as a mediator of signals induced by
opioid stimulation.1-3 Combination of the agonist effects at
opioid receptors together with the blocking of signals through
NK1 receptors has led to enhanced antinociceptive effects in
acute pain animal models and has prevented the opioid-
induced tolerance in chronic trials.8-12 NK1 knockout mice
did not show the rewarding properties of morphine.13 We also
emphasize that designing these bifunctional compounds with
multiple biological activities in a single ligand has additional
advantages over a cocktail of individual drugs including easy
administration, single biodistribution, no drug-drug interac-
tions, simple pharmacokinetic profiles, and higher local con-
centration in the synaptic cleft, which may lead to significant
synergies in potency and efficacy.14 However, obtaining a
bifunctional ligand with desired bioactivity is a difficult
The clinical treatment of pain, especially prolonged and
neuropathic pain, is still a major challenge, and the efficacy of
current analgesic drugs, such as opioids, is often reduced by
undesired dose-limiting side effects including development of
tolerance and physical dependence. The mechanisms for these
effects are still largely unclear, but it is clear that prolonged
pain states lead to neuroplastic changes in both ascending and
descending pathways in the spinal column in which there is
increased release of neurotransmitters (e.g., substance P) that
enhance pain and increased expression of the corresponding
*To whom correspondence should be addressed. Phone: (520)-621-
6332. Fax: (520)-621-8407. E-mail: hruby@email.arizona.edu.
aAbbreviations: (abbreviations used for amino acids and designation
of peptides follow the rules of the IUPAC-IUB Commission of Bio-
chemical Nomenclature in J. Biol. Chem. 1972, 247, 977-983); Boc, tert-
butyloxycarbonyl; Cl-HOBt, 1-hydroxy-6-chlorobenzotriazole; CHO, Chi-
nese hamster ovary; DCM, dichloromethane; DIEA, diisopropylethylamine;
DMF, N,N-dimethylformamide; DMSO, dimethylsulfoxide; DPC, dodecyl-
phosphocholine; DPDPE, Tyr-cyclo[D-Pen-Gly-Phe-D-Pen]-OH; DPLPE,
Tyr-cyclo[D-Pen-Gly-Phe-Pen]-OH; DQF-COSY, double quantum filtered
correlation spectroscopy; DAMGO, [D-Ala2, NMePhe4, Gly5-ol]-enkephalin;
Fmoc, fluorenylmethoxycarbonyl; GPI, guinea pig isolated ileum; HBTU,
1-[bis(dimethylamino)methylene]-1H-benzotriazolium 3-oxide hexafluoro-
phosphate; HCTU, 1H-benzotriazolium-1-[bis(dimethylamino)methylene]-
5-chloro-hexafluorophosphate-(1-),3-oxide; HRMS, high-resolution mass
spectroscopy; LMMP, longitudinal muscle with myenteric plexus; Mob,
p-methoxy benzyl; MVD, mouse vas deferens; NK1, neurokinin-1; NOESY,
nuclear Overhauser enhancement; rmsd, root-mean-square deviation; NMM,
N-methyl morpholine; RP-HPLC, reverse phase high performance liquid
chromatography; TFA, trifluoroacetic acid; TLC, thin-layer chromatography
TOCSY, total correlation spectroscopy; Trp-NH-[30,50-(CF3)2-Bzl], the 30,50-
(bistrifluoromethyl)-benzyl amide of tryptophan.
r
2010 American Chemical Society
Published on Web 07/09/2010
pubs.acs.org/jmc

5492 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15
Yamamoto et al.
Figure 1. Sequences of opioid and NK1 receptor peptides.
challenge because connected pharmacophores interfere with
each other to change or even to reduce their affinities at the
corresponding receptors.
In this report, we have focused on the δ/μ activity of targeted
opioid receptors for the bifunctional compounds. The μ-opiates
have been well-established to show strong analgesic potency
with several disadvantages such as constipation, respiratory
depression, dysphoria, opioid-induced tolerance development,
and addiction.15,16 Selective δ-opioid agonists have several
potential clinical advantages over their μ-counterparts, includ-
ing reduced respiratory depression17 andconstipation18 as well
as a decreased potential for the development of physical
dependence.19-21 However, selective δ-opioid agonists do
not attenuate the same level of pain as μ-agonists and no
δ-opioid selective agonist ligand has been found to have clinical
efficacy sufficient to be a useful agent for the treatment of
severe pain.22 Therefore, bifunctional ligands possessing se-
lective δ-opioid agonist activity but also some μ-opioid ago-
nist activity together with the NK1 antagonist activity could
be a promising strategy to find next-generation analgesics
which have powerful pain controlling effects without devel-
opment of tolerance, physical dependence, and other toxi-
cities.4,5 In fact, our lead bifunctional compounds, TY005
(Tyr1-D-Ala2-Gly3-Phe4-Met5-Pro6-Leu7-Trp8-O-[30,50-(CF3)2-
Bzl]) and TY027 (1: Tyr1-D-Ala2-Gly3-Phe4-Met5-Pro6-Leu7-
Trp8-NH-[30,50-(CF3)2-Bzl]) possessing moderately δ-selective
(up to 25-fold) δ/μ opioid agonist and potent NK1 antagonist
activities (Table 1) have shown the powerful attenuation of
neuropathic pain without producing the opioid-induced tol-
erancetoeffectivelyprove our opioid agonist/NK1antagonist
bifunctional concept in the animal models of neuropathic
pain.23,24
To perform further biological optimization of the bifunc-
tional ligands, we describe here the design, synthesis, structure-
activity relationships (SAR), and further biological character-
izations of bifunctional compounds which have significantly
improved selectivity both for affinity and agonist activity for
the δ-opioid over μ-opioid receptors, together with effective
NK1 antagonist activity. In fact, we previously reported the
finding of a selective δ-opioid activator, possessing two cys-
teine residues, in the conventional MVD and GPI assays, but
its δ-opioid selectivity in terms of the binding affinity was
relatively lower.25 While our earlier efforts led to the discovery
that introduction of two topologically constrained L- or
D-penicillamine (Pen or D-Pen) residues into several cyclic
pentapeptide enkephalin analogues produced compounds
with enhanced δ-selectivity through the reduction of opioid
affinity at the μ-opioid receptor: Tyr-cyclo[D-Pen-Gly-Phe-D-
Pen] (DPDPE) and Tyr-cyclo[D-Pen-Gly-Phe-Pen] (DPLPE)
are the most typical δ-selective opioid agonists (Tables 1
and 2).5,26-28 Previous conformation-activity relationship
studies suggested that the 14-membered ring structure with

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15 5493
Table 2. Functional Assay Results for Bifunctional Peptide Ligands at
Opioid and Substance P Receptors
groups) at 80 °C for 30 min to give the final linear crude pep-
tides, which were directly cyclized using the oxidizing agent
K3Fe(CN)6.25 Oxidation was performed by slowly adding a
solution of the peptide into a reaction vessel containing an
excess of the oxidizing agent in aqueous solution. The con-
centration of the reduced peptide was regulated by control-
ling the speed of addition of the peptide with the help of an
automated syringe pump.25 The crude oxidized peptides
were concentrated using a solid-phase extraction technique
with a C-18 reversed-phase silica gel column, and then puri-
fied with RP-HPLC followed by lyophilizion (g95%). The
final purified peptides were characterized by analytical
HPLC, 1H NMR, HRMS, and TLC.
opioid agonist
SP antagonist
MVD (δ)
GPI (μ)
GPI Ke
(nM)b
no.
IC50 (nM)a
IC50 (nM)a
1c
2
15 ( 2
9.6 ( 3.5
130 ( 25
250 ( 30
1100 ( 250
2.5 ( 0.03
2.2 ( 0.30
490 ( 29
10 ( 2
26% inh at 1 μMd
0.6% inh at 1 μMd
6% inh at 1 μMd
7% inh at 1 μMd
2720 ( 50
6.9 ( 1.8
8.2 ( 4.2
72 ( 7.2
66 ( 35
3
4
5
DPDPEe
DPLPEe
L-732,138
6930 ( 120
250 ( 87
a Concentration at 50% inhibition of muscle concentration at elec-
trically stimulated isolated tissues (n = 4). b Inhibitory activity against
the substance P induced muscle contraction in the presence of 1 μM
naloxone (n = 4). Ke: concentration of antagonist needed to inhibit
Substance P to half its activity. c Reference 5. d No antagonist activity was
Biological Activities. The evaluation of the synthesized
bifunctional ligands 2-5 was performed as previously descri-
bed.4-7 Binding affinities were tested on cell membranes
from cells that stably express the human δ-opioid, the rat
μ-opioid, the human NK1, or the rat NK1 receptors (Table 1).
The [35S]GTPγS binding assays were performed to charac-
terize the functional activities for the δ and μ opioid agonists
of 2-5 (Table 1). Tissue-based functional assays also were
performed to evaluate δ/μ opioid agonist as well as the
substance P antagonist activities in the guinea pig isolated
ileum (GPI) and mouse vas deferens (MVD) (Table 2).
At the δ-opioid receptor, the cyclic peptide possessing a
20-membered ring with D-Pen2 and Pen7 (4) had a Ki of
150 nM, which was 230-fold less binding affinity than that of
1 (Table 1). The δ-opioid affinity of its D-Pen7 derivative (5)
also showed a decrease in affinity of 720 nM, indicating that
introduction of D-Pen or Pen at the seventh residue in the
sequence of 1 decreases affinity at the δ-opioid receptor. The
Ki of the D-Pen2,D-Pen5 cyclic analogue 3 was 38 nM for the
δ receptor, demonstrating that the cyclic peptide with a
smaller 14-membered ring had enhanced affinity compared
to the 20-membered-ring analogues. Interestingly, the cyclic
D-Pen2,L-Pen5 analogue 2 had the best affinity among 2-5
for the δ-opioid receptor with Ki value of 1.7 nM. This is the
closest to the KD value of DPDPE (0.5 ( 0.1 nM), which was
used as a radioactive ligand of 1 nM in the competition assay
on the cells membrane expressing the δ-opioid receptors
(Table 1). Thus, the affinities of 2-5 for the δ-opioid recep-
tor were different depending on the ring-size and the chirality
of bridgehead penicillamine residues, and the combination
of D-Pen2 and Pen5 clearly showed the best results. Interest-
ingly, the affinity difference at the δ-opioid receptor between
pentapeptides DPDPE and DPLPE was only 1.6-fold and
DPDPE showed the better affinity,28 but in the case of the
bifunctional compounds with the longer peptide sequences,
the compound 2 (DPLPE-derivative) showed better affinity
than 3 (DPDPE-derivative), and their affinity difference was
22-fold. This result apparently indicates that the extended
sequence at the C-terminus (Pro-Leu-Trp-NH-3,5-Bzl(CF3)2)
plays a critical role for the δ-opioid selectivity, and 2 has a
preferable three-dimensional conformational motif for bind-
ing at the δ-opioid receptor compared to 3. For the linear
pentapeptide Tyr-D-Ala-Gly-Phe-Met-NH2, the C-terminal
elongation to the octapeptide derivative 1 resulted in a dec-
rease in μ-opioid binding affinity with no shift in the affinity
at the δ-receptor (Table 1). Consistent with the trend in the
affinities at the δ-opioid receptor, the EC50 values of 3-5
were more than 100 nM in the GTPγS binding assay (120,
530, and 120 nM for 3, 4, and 5, respectively), while only 2
showed significantly better efficacy with a nanomolar level
EC50 value (EC50 = 3.2 nM, Table 1). In accord with these
observed at the tested concentration. e Reference 28
.
their geminal-dimethyl group at the second position has special
importance to interfere with the binding at the μ-opioid
receptor so as to greatly favor binding to the δ-opioid
receptor.29,30 We have now applied this modification to our
lead bifunctional octapeptide analogue 1 in which Tyr1,
Gly3, Phe4, and Trp8 were considered as the major “message”
residues of each pharmacophore5 and Pro6 had a positive
effect on affinity at the NK1 receptors.31 Thus, the residues 2,
5, and 7 in 1 were selected as the sites to be cyclized to yield four
bifunctional peptide derivatives possessing a 14- or 20-mem-
bered ring structure with two Pen or D-Pen residues (2-5,
Figure 1). Subsequent conformational analysis was performed
using NMR experiments in the presence of membrane-
mimicking perdeuterated dodecylphosphocholine (DPC) mi-
celles because understanding membrane-bound structures of
ligands and ligand-membrane interactions is indispensable
for further evaluation of their diverse biological behaviors,5
and the elucidation of the conformational motif for δ-opioid
selective binding affinities could provide important informa-
tion in the discovery of novel drug candidates for next-
generation neuropathic pain treatment.
Results and Discussion
Peptide Synthesis. The peptide derivatives 2-5 were
synthesized using a convergent peptide synthesis strategy32
because the coupling of Tyr1 was difficult using the normal
sequential method. First, the N-terminal fragment, Boc-Tyr-
(tBu)-D-Pen(Mob)-Gly-OH, was synthesized with NR-Boc
chemistry using a solution peptide synthesis strategy. The
C-terminal fragments, H-Phe-Pen(Trt)-Pro-Leu-Trp(Boc),
H-Phe-D-Pen(Trt)-Pro-Leu-Trp(Boc), H-Phe-Nle-Pro-Pen-
(Trt)-Trp(Boc), and H-Phe-Nle-Pro-D-Pen(Trt)-Trp(Boc)
were prepared on 2-chlorotrityl resin using NR-Fmoc chem-
istry with HCTU as a coupling reagent. These two fragments
were coupled on the resin in the presence of an excess amount
of EDC and HOAt, followed by cleavage using 1% TFA in
DCM, to afford the complete sequence of the C-terminal-
free linear peptide with its protecting groups. The introduc-
tion of 30,50-bistrifluoromethyl benzyl amide at the C-termi-
nus was conducted by standard EDC/Cl-HOBt coupling
chemistry using 2 equiv of the amine with the protected
C-terminal-free peptide. All of the acid-labile protecting
groups were removed with the cleavage cocktail treatment
(6:1 mixture of TFA and anisole to quench the highly stabil-
ized carbocations released from the permanent protecting

5494 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15
Yamamoto et al.
results, 3-5 had relatively lower inhibitory activities in the
MVD assays (IC50 = 130, 250, and 1100 nM for 3, 4, and 5,
respectively), whereas 2 showed good activity (IC50 = 9.6 nM)
which is improved from that of 1 (Table 2).
For the μ-opioid receptor, the cyclic peptides with a
14-membered ring (2 and 3) showed drastically reduced
affinities (Ki = 2300 and 2100 nM for 2 and 3, respectively;
Table 1), as expected from the results of DPDPE and
DPLPE. The compounds with the larger 20-membered ring
(4 and 5) also showed reduced affinity at the μ-opioid re-
ceptor (Ki = 2000 and 1000 nM for 4 and 5, respectively),
confirming that the introduction of a disulfide ring posses-
sing D-Pen2 is detrimental to the affinities at the μ-opioid re-
ceptor, regardless of the ring size of the octapeptide deriva-
tives. In the GTPγS binding assays, the cyclic peptides 2-5
stimulate the μ-opioid receptor poorly with the Emax values
less than 28% (Table 1). In agreement with their low Ki and
Emax values, 2-5 showed very weak μ-opioid-related agonist
activities in the GPI assays (26, 0.6, 6, and 7% inhibition at
1 μM for 2, 3, 4, and 5, respectively; Table 2). Therefore, only
2 showed significant δ-opioid selectivity over the μ-receptor
(1400-fold), which is 58 times more δ-selective than 1 (24-
fold) and 3.7 and 7.8 times more selective compared to DPDPE
(380-fold) and DPLPE (180-fold), respectively (Table 1). The
δ-selectivity of 3 was 55-fold, which is better than that of 1.
These results suggest that three important structural fac-
tors have an influence on the opioid affinity and bioactivity
for these novel bifunctional compounds: (1) the 14-mem-
bered cyclic structure with D-Pen2, (2) the chirality of the fifth
residue, and (3) the existence of the C-terminal NK1 pharma-
cophore. Apparently, introduction of a 14-membered ring
with a D-Pen2 resulted in a limited shift in their δ-opioid
binding affinities but a more significant reduction in affi-
nities at the μ-receptor, leading to the enhanced δ-receptor
selectivity compared to DPDPE and DPLPE. Addition of
the C-terminal NK1 pharmacophore to the cyclic opioid
sequence in 2 resulted in improved affinity at the δ-receptor,
while the combination with the N-terminus in 3 possessing a
D-Pen5 resulted in a reduction in the δ-opioid binding
affinity. These results clearly indicate that the NK1 phar-
macophore works as an address region for the opioid agonist
activities, and an appropriate combination of two connected
pharmacophores can yield improved affinities and selecti-
vities of the ligand. This is an important synergistic effect of
our bifunctional concept.
Biological evaluations were also conducted at the NK1
receptors. Analogue 4, possessing a D-Pen2 and Pen7, showed
reduced affinities at both the hNK1 and rNK1 receptors
compared to those of 1 (Ki = 59 and 160 nM, respectively;
Table 1). The Ki value of peptide 5 with a D-Pen7 was 1.9 nM
at the hNK1 receptor and 26 nM at the rNK1 receptor,
indicating that the D-Pen7 provided better affinities than
L-Pen7 introduction. These affinities of compounds with
20-membered rings were drastically decreased from those
of 1, and their antagonist activities on the GPI tissue were
consistent with the observed affinities (Ke = 72 and 66 nM
for 4 and 5, respectively; Table 2). However, the introduction
of a ring structure between the second and seventh residues
resulted in a reduction of biological activities at the NK1
receptors, suggesting an unfavorable topographical struc-
ture for the NK1R. Analogue 3, possessing a smaller 14-
membered ring with the DPDPE sequence, has 28 times
lower affinity for the hNK1 receptor than that of 1 but its Ki
value at the rNK1 receptor was slightly improved from that
Figure 2. Comparison of the in vitro stability of peptide derivatives
for 1 and 2 incubated in rat plasma at 37 °C (*, p < 0.05).
of 1 (Ki = 0.18 and 4.5 nM for hNK1 and rNK1 receptors,
respectively). The antagonist activity against substance P
stimulation of 3 in the GPI assay was nearly comparable to
that of 1 (Ke = 8.2 nM). The DPLPE derivative 2 showed
similar antagonist activity against substance P stimulation at
the GPI, with a Ke value of 6.9 nM (Table 2). 2 also showed
the best affinity at the hNK1 receptor among the synthesized
cyclic peptide derivatives 2-5 with a picomolar Ki value
(Ki = 0.0053 nM).
For further biological characterization of 2, its metabolic
stability was tested by incubation in rat plasma at 37 °C.6
Analogue 2 possesses significantly improved stability com-
pared to the linear peptide derivative 1. The calculated half-
lives of the peptide derivatives (T1/2) were 4.8 h for 1 and 5.8 h
for 2 (Figure 2). Consequently, c[D-Pen2, Pen5]TY027 (2) is a
promising ligand with significant δ-selective opioid agonist
activity, improved metabolic stability, and nearly equivalent
antagonist activities for NK1 receptors compared to those of
the lead compound 1. The important conformational motif
of the two pharmacophores for the bioactivities of 2 was next
examined using NMR-based three-dimensional structural
analysis.
Conformational Analysis of Cyclic Peptide 2 Based on Assig-
ned 1H NMR Resonances
The ligand binding sites for opioid and NK1 receptors have
been found in their transmembrane domain,33,34 which is
surrounded by the cell membrane, suggesting that the ligand
binding event at these receptors occur near the membrane.
The potential importance of membrane-bound conforma-
tions of ligands has increased recently.5,6 The transfer from
an aqueous environment to a cell membrane environment is
an important step in receptor-ligand binding and may be
accompanied with a conformational change to a more biolo-
gically relevant form.17,38 Thus, NMR structural analysis of 2
was performed to clarify the relationship between its biologi-
cal activity and its conformation in a membrane-mimicking
environment.
The NMR experiments and analysis of 2 were conducted
using previously reported procedures.5,6,35,36 The two-dimen-
sional 1H NMR studies included TOCSY, DQF-COSY,
and NOESY and were performed in pH 4.5 buffer (45 mM
CD3CO2Na/HCl, 1 mM NaN3, 90% H2O/10% D2O) solu-
tions in the presence of a 40-fold excess of DPC micelles,
which is a commonly used membrane-like environment to
determine the solution NMR structures of proteins and
1
peptides.35-40 All H chemical shift assignments in aqueous

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15 5495
Table 4. Number of β-Turn Structural Elements and the Distance
between R Carbons of ith and (i þ 3)rd Residuesa
residuesb
number of β-turns
distance (A)
˚
C2R-C5R
10
10
5.36 ( 0.00
6.43 ( 0.04
C6R-Bzl
a Out of the best 10 calculated structures. The distance is the mean
distance between two R carbons ( standard deviation (SD). The
˚
sequences with less than 7 A distance between R carbons of ith and
(i þ 3)rd residues without helical structure were considered as a β-turn.57
Bzl stands for the benzyl moiety at the C-terminus. b Only residues
possessing β-turn structural element were displayed.
the conformation of the N-terminal half, where the two
penicillamines were introduced, but also the C-terminus,
whose amino acid sequence is identical to 1, showed reduced
rmsd values as well. The rmsd values of 2 for the alignment
with all non-hydrogen atoms were also significantly reduced
Figure 3. Ensembles of the best 10 calculated structures of 2 in 40-
fold DPC micelle/pH 4.5 buffer with the lowest restraint energy,
aligned on backbone atoms of residues 1-8. (A) Only backbone
atoms in the aligned structures are illustrated with C-terminal
benzyl moiety and disulfide bond. (B) All non-hydrogen atoms
were displayed.
ꢀ
˚
(0.52 ( 0.12 A, Table 3), indicating well-defined three-dimen-
sional positions of the side-chains in 2 under the applied time
scale in the NMR experimental (Figure 3B). Different from
the observed well-defined side-chainconformation of2, rather
poorly defined side-chains were previously observed for the
linear compound 1. Thus, the constraints inligand 2 due to the
incorporation of two penicillamines provided improved con-
formation stability for the backbone atoms as well as for the
side-chain atoms, including the two aromatic amino acid re-
sidues Tyr1 and Phe4 and the protonated N-terminal nitrogen,
all of which have been reported as important pharmacophores
in the binding of enkephalins to opioid receptors.41 The
observed well-defined conformation of 2 suggests it has a
structured conformation inside or close to the cell membrane,
in close relationship with the improvement in activity and
selectivity at the tested membrane-bound receptors.
˚
Table 3. Atomic rmsd Values (A) for the Final Nine Conformers
Compared to the Most Stable Conformer of Bifunctional Peptide
Derivatives 2
aligned
residues
whole
molecule
1-4
residues
5-8 residues
and C-terminus
backbone atoms (N,CR,C0) 0.15 ( 0.06 0.02 ( 0.04
all non-hydrogen atoms 0.52 ( 0.12 0.35 ( 0.21
0.14 ( 0.05
0.58 ( 0.11
media and other NMR data are found in the Supporting
Information.
As for the observed secondary structural elements, the
NMR structure of 1 has two type IV β-turns at residues 2 to
5 and 6 to the C-terminus.5 Although the peptide sequence
was different, the obtained NMR structure of ligand 2 had
type IV β-turns at the same positions in all 10 of the best
structures (Table 4), indicating that the introduction of con-
straints by D-Pen2 and Pen5 made the two β-turns more
structured without changing the secondary structural ele-
ments of ligand. DPDPE also has a β-turn at residues 2-5
in its X-ray crystal structure,42 in the solution NMR structure,43
and in the docking study at the δ opioid receptor,44 and thus,
this turn has been hypothesized as a key structure to enhance
the δ-selectivity of enkephalin analogues, through the lock-in
of flexible pharmacophores at the appropriate positions
(Figure 4A).
Conformational Calculations
The total number of NOE restraints for structural calcula-
tion of 2 was 184, including 63 intraresidual NOEs, 67 seq-
uential NOEs, 49 medium-range NOEs (2-4 residues), and 5
3
long-range NOEs (>4 residues). For the JHN-HR values,
only those greater than 8 Hz or less than 6 Hz were used as
jdihedral angle constraints. The 1HNMRspectrain2showed
a negligible amount of the minor cis/trans rotamer at the Pro6
residue. For conformational calculations, the Pen5-Pro6 bond
of the major rotamer was fixed in the trans configuration based
on the observation of Pen5 HR to Pro6 Hδ sequential NOEs,
together with the absence of sequential Pen5 HR-Pro6 HR
NOEs.
Analysis and statistics of the calculated conformation for 2
were performed on the 10 structures with the lowest total
energies after restrained molecular dynamics (rMD) refine-
ment. The number of total NOE restraint violations was 25
Although 1 and 2 had the same secondary structural
elements, their NMR structures did not overlap completely
on alignment with the backbone atoms of the whole molecule
˚
˚
with a maximum NOE violation of 0.15 A, and no j dihedral
(rmsd = 2.34 A, Figure 4B). However, the rmsd value was
˚
decreased to 1.41 A when the alignment was made only for the
angle violations were found. The restraints energy derived
from the amber force field was only 3.9 kcal mol-1 in the most
stable 10 structures.
N-terminal opioid pharmacophore where the common β-turn
structural elements for 1 and 2 were observed (Figure 4C). It
should be mentioned that when the alignment was made on
the N-terminus of 1 and 2, their corresponding C-terminal
halves were directed differently and showed almost no over-
lap. As expected, when the alignment was made with the
As shown in Figure 3A, the best 10 structures of 2 showed
well-aligned superimposed conformations whose rmsd values
ꢀ
˚
were 0.15 ( 0.06 A in an alignment with the backbone atoms
(Table 3). Because these rmsd values were significantly im-
˚
˚
proved from those of 1 (1.03 ( 0.57 A for the best 10
C-terminus (rmsd = 0.65 A), the N-terminal halves of 1 and 2
structures; Supporting Information),5 the extensively struc-
tured backbone of 2 could be attributed to the cyclic and
topological constraints induced by the disulfide bridge with
the D-Pen2 and Pen5 residues. It should be noted that not only
showed almost no conformational overlap (Figure 4D). These
results clearly indicated that the opioid and NK1 pharmaco-
phores of ligands 1 and 2 independently possess nearly the
same conformations in terms of their backbone atoms, but the

5496 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15
Yamamoto et al.
Figure 4. (A) Superimposed image of 2 (red) with the X-ray crystal structure of highly selective δ opioid agonist DPDPE (green).42
˚
Superimposition was performed on all the backbone atoms of residues 1-5 (rmsd = 1.81 A). Only backbone atoms are illustrated with
C-terminal benzyl moiety and disulfide bond (yellow). Superimposed image of obtained NMR structure of 2 with the NMR structure of 15
˚
(blue) at the lowest restraint energies. Superimposition was performed on all the backbone atoms of: (B) residues 1-8, rmsd = 2.34 A;
˚
˚
(C) residues 1-5, rmsd = 1.40 A; (D) residues 5-8, rmsd = 0.65 A.
Table 5. j and ψ Angle Values of the Bifunctional Peptide Derivative 2
in the Observed NMR Structurea
angle
Tyr1
D-Pen2
Gly3
Phe4
j
ψ
-154.9 ( 2.3
-173.6 ( 0.2
-85.0 ( 1.4
-73.6 ( 0.9
-105.9 ( 0.6
128.3 ( 0.3
-70.9 ( 0.5
angle
Pen5
Pro6
Leu7
Trp8
j
-91.6 ( 0.6
-86.6 ( 0.8
-120.9 ( 4.7 -146.1 ( 1.3
ψ
118.5 ( 0.6 -102.7 ( 138.5
57.0 ( 1.1
-41.7 ( 0.9
a Average ( SD values out of the best 10 calculated structures were
listed. The C-terminus was considered as residue 9.
three-dimensional positionings of the NK1 pharmacophore
relative to the opioid pharmacophore are different. This could
have an important connection with the different address effect
of the C-terminus of 1 and 2 for the observed binding affinities
at δ and μ opioid receptors.
Further structural examination was made on the torsional
angles of the fifth residue, where the opioid and NK1 phar-
macophores were overlapped (Figure 1), to examine the
relative positioning of two pharmacophores. The NMR con-
former of 1 had the Met5 residue with positive j and ψ angles
in all 10 of the best structures.5 Because the positive j angle in
an L-amino acid residue is energetically unfavorable, the back-
bone conformation of 1 tends to bend or to twist near this
“distorted” site at the connecting position of two interacting
pharmacophores. While Pen5 in all 10 structures of 2 had
negative j angles (Table 5). This difference in torsional angles
for the fifth residue leads to conformational differences near
the overlapping site and modifies the positional relation-
ship of the two connecting pharmacophores which may
possibly account for the shift in biological activities at the
tested receptors.
Figure 5. Ramachandran j, ψ plots for 2 for residues 2-8 of 10
final structures.
a 62.2 ms mixing time (Figure 5).5 Interestingly, most of the
cross-peaks derived from the backbone NH protons of 2 were
eliminated with Mn2þ addition. The only exception was
the NH proton of Pen5, whose cross-peak was retained in the
presence of Mn2þ ions, suggesting that the backbone of 2 was
mostly exposed to the surface of micelles and only the back-
bone atoms of Pen5 were buried inside of micelles (Figure 6).
The side-chains of 2, most of which were lipophilic groups,
were unchanged after the Mn2þ addition, implying that these
side-chain groups interact with DPC micelles very well and are
buried inside of them. Therefore, these results revealed that 2
locates in DPC micelles with the amino acid side-chain groups
and backbone atoms of Pen5 buried inside of the micelles,
indicating its strong interaction with the membrane. This
observed interacting mode of 2 with membrane-like micelles
Paramagnetic Broadening Studies on 1H NMR
To evaluate the interacting mode of 2 in a membrane-like
environment, Mn2þ was used as a paramagnetic ion to
eliminate the resonance intensities of solvent-exposed protons
of 2 in the DPC micelles. The effects of the reagent were
observed as an ensemble of cross-peaks belonging to the
same residue spin system in TOCSY spectra measured with

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15 5497
Figure 6. Typical effect of Mn2þ ions on TOCSY Spectra. 2 with DPC micelles (top row) and with 200 μM Mn2þ (bottom), for HN-HR region
(left column), aliphatic side-chain region (middle), and aromatic region (right). Preserved resonances (labeled) are shown by the addition of
Mn2þ. Spectra were compared from the same noise level.
was similar to that of 1,5 suggesting that the incorporation of a
14-membered ring structure with D-Pen2 and Pen5 did not
affect the strong interaction of these peptides with membranes
but provides a more structured conformation which could
modulate the affinity and selectivity of the ligand for the
receptor in biological membranes.
International (Wood Dale, IL). 2-Chlorotrityl resin was ac-
quired from Iris Biotech GmbH (Marktredwitz, Germany).
Perdeuterated DPC was purchased from C/D/N Isotopes
(Quebec, Canada). ACS grade organic solvents were purchased
from VWR Scientific (West Chester, PA), and other reagents
were obtained from Sigma-Aldrich (St. Louis, MO) and used as
obtained. The polypropylene reaction vessels (syringes with
frits) were purchased from Torviq (Niles, MI). Myo-[2-3H-
(N)]-inositol, [tyrosyl-3,5-3H(N)]-D-Ala2-Mephe4-glyol5-enke-
phalin (DAMGO),[tyrosyl-2,6-3H(N)]-(2-D-penicillamine,5-D-
penicillamine)enkephalin (DPDPE), [3H]-Substance P, and
[ S]-guanosine 5-(γ-thio) triphosphate (GTPγS) were purchased
from Perkin-Elmer (Wellesley, MA). Bovine serum albumin
(BSA), protease inhibitors, Tris, and other buffer reagents were
obtained from Sigma (St. Louis, MO). Culture medium, peni-
cillin/streptomycin, and fetal calf serum (FCS) were purchased
from Invitrogen (Carlsbad, CA).
Conclusions
Cyclic and topographical constraints with two penicilla-
mine residues were introduced and optimized based on our
developing bifunctional drug-design concept. As a result, the
14-membered ring structure with a Pen5 and a C-terminal
NK1 pharmacophore in 2 was found as the best combination
to show potent antagonist actions at NK1 receptors and
potent opioid agonist actions with significantly improved
δ-opioid affinity selectivity over the μ-opioid (1400-fold) than
those of DPDPE (380-fold), DPLPE (180-fold), and the pre-
viously reported linear peptide derivative 1 (24-fold). NMR
structural analysis showed that significant δ-opioid selectivity
in 2 could be attributed to its well-defined conformation with
cyclic and topographical constraints and appropriate posi-
tioning of opioid and NK1 pharmacophores. The torsional
angle of the fifth residue plays a key role to modulate the
address effect of the C-terminal NK1 pharmacophore for the
δ/μ opioid selectivity of the ligand by adjusting the three-
dimensional positioning of the two connected pharmaco-
phores. Together with its observed good metabolic stability,
2 could be considered as a valuable research tool and possibly
a promising candidate for a novel analgesic drug.
35
ꢁ
Nr-Boc-Tyr(tBu)-D-Pen(Mob)-Gly-OH. Boc-D-Pen(Mob)-
OH (1.50 g, 4.06 mmol) and H-Gly-OEt HCl (677 mg, 4.87 mmol)
were dissolved in DMF (5 mL). HOBt (657 mg, 4.87 mmol),
HBTU (1.85 g, 4.87 mmol), and NMM (904 mg, 8.95 mmol) were
added to the solution at 0 °C. After stirring overnight at room
temperature, saturated aqueous sodium bicarbonate was added
to the solution and most of the organic solvent was removed
under reduced pressure. The residue was extracted with ethyl
acetate three times followed by washing with 5% aqueous citrate
and saturated aqueous sodium chloride. The solution was dried
over sodium sulfate. The solvent was evaporated, and the crude
peptide was precipitated in cold petroleum ether, centrifuged,
and dried under reduced pressure. The obtained solid was
dissolved in TFA (6 mL) at 0 °C. After stirring for 3 h, the solu-
tion was concentrated. Ether (50 mL) and 4 M HCl in 1,4-
dioxane (2 mL) were added to the solution to obtain 1.33 g
(3.42 mmol, 84.4%) of crude H-D-Pen(Mob)-Gly-OEt HCl. The
peptide was used for the next reaction without further purifica-
tion. Boc-Tyr(tBu)-OH (1.38 g, 4.10 mmol) was next coupled in
the presence of Cl-HOBt (694 mg, 4.10 mmol), HCTU (1.69 g,
4.10 mmol), and NMM (760 mg, 8.20 mmol) in DMF (5 mL).
Experimental Section
Materials. All amino acid derivatives and coupling reagents
were purchased from EMD Biosciences (Madison, WI), Bachem
(Torrance, CA), SynPep (Dublin, CA), and Chem Impex

5498 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15
Yamamoto et al.
The reaction mixture was stirred overnight at room tempera-
ture. Saturated aqueous sodium bicarbonate was added to the
residue and extracted with ethyl acetate three times, followed by
washing the combined solution with saturated aqueous sodium
chloride. The solution was dried over sodium sulfate and con-
centrated under reduced pressure. The obtained residue was
purified using silica gel chromatography (hexane:EtOAc = 10:1
to 2:1). The residue was dissolved in methanol (8 mL) and THF
(8 mL), then 1 M NaOH (9 mL) was added at 0 °C. After stirring
for 1 h at room temperature, 1 M HCl (10 mL) was added at 0 °C
and then evaporated. The obtained residue was extracted with
ethyl acetate three times, followed by washing with 5% aqueous
citrate and saturated aqueous sodium chloride. The solution
was dried over sodium sulfate. The solvent was evaporated, and
the crude peptide was precipitated in cold ether. The precipitate
was centrifuged and then dried under reduced pressure to obtain
the title compound (1.53 g, 84.1%).
peptide was used for the subsequent reactions without further
purification.
The protected peptide with a free C-terminal (300 mg,
0.173 mmol) and Cl-HOBt (33.4 mg, 0.208 mmol) were dis-
solved in DMF (3 mL). 3,5-Bistrifluoromethylbenzyl amine
(84.1 mg, 0.346 mmol) and EDC (39.7 mg, 0.208 mmol) were
added to the solution at RT and stirred until the starting
material (the protected peptide with free C-terminal) was not
detected by TLC, then saturated aqueous sodium bicarbonate
(100 mL) was added. The reaction mixture was extracted with
ethyl acetate (100 mL) three times. The combined organic
phases were washed with 5% aqueous citrate and saturated
aqueous sodium chloride (100 mL each) and then dried over
sodium sulfate. The solvent was evaporated off and the crude
protected peptide was precipitated in cold petroleum ether
(45 mL). The product was twice dispersed in cold petroleum
ether, centrifuged and decanted, and then dried under reduced
pressure. The obtained protected peptide was treated with a 6:1
mixture of TFA and anisole at 80 °C (3 mL, 30 min). The crude
peptide was precipitated out by the addition of chilled petroleum
ether (45 mL) to give a dark oil. The resulting peptide suspen-
sions were centrifuged for 20 min at 7000 rpm, and the liquid was
decanted. The crude peptides were washed with diethyl ether
(2 ꢀ 45 mL) to give a precipitate, and after a final centrifugation,
the peptides were dried under vacuum (2 h). The resulting pep-
tides were directly used for cyclization.
1H NMR (DMSO-d6) δ: 1.10-1.32 (21H, m), 1.45 (3H, s),
1.99 (2H, s), 2.28-3.10 (2H, m), 3.68-3.92 (5H, m), 4.03 (1H, q,
J = 7.0 Hz), 4.80 (1H, d, J = 9.5 Hz), 6.84 (4H, d, J = 5.5 Hz),
7.00 (1H, d, J = 8 .5 Hz), 7.18-7.26 (4H, m), 8.22 (1H, d, J =
9.5 Hz), 8.65-8.72 (1H, m). MS (ESI) 646 (MH)þ.
Linear Peptide Synthesis. The peptide was synthesized manu-
ally by the NR-Fmoc solid-phase methodology using HCTU as
the coupling reagents as previously reported.5,7 2-Chlorotrityl
resin (500 mg, 0.78 mmol/g) was placed into a 50 mL poly-
propylene syringe with the frit on the bottom and swollen in
DMF (10 mL) for 1 h. The resin was washed with DMF (3 ꢀ
15 mL) and then with DCM (3 ꢀ 15 mL). Fmoc-Trp(Boc)-OH
(1.2 equiv) was dissolved in 30 mL of DCM, and then DIEA
(5 equiv) was added. The reaction mixture was transferred into
the syringe with the resin and then shaken for 2 h. The resin was
washed three times with DMF (15 mL) and three times with
DCM (15 mL) and then with DMF (3 ꢀ 15 mL). The NR-Fmoc
protecting group was removed by 20% piperidine in DMF
(10 mL, 1 ꢀ 2 min and 1 ꢀ 20 min). The deprotected resin was
washed with DMF (3 ꢀ 15 mL), DCM (3 ꢀ 15 mL), and then
with DMF (3 ꢀ 15 mL). The protected amino acid (3 equiv) and
HCTU (2.9 equiv) were dissolved in 30 mL of DMF, and then
DIEA (6 equiv) was added. Fmoc-D-Pen(Trt)-OH, Fmoc-Pen-
(Trt)-OH, Fmoc-Pro-OH, Fmoc-Nle-OH, and Fmoc-Phe-OH
were used for respective coupling as protected amino acids.
The following coupling with BocTyr(tBu)-D-Pen(Mob)-Gly-
OH (1.5 equiv) were performed using EDC (2.25 equiv), HOAt
(2.25 equiv), and NMM (3 equiv) in DMF (10 mL). The
coupling mixture was transferred into the syringe with the
resin and then shaken for 2 h. All the other amino acids were
consecutively coupled using the procedure described above,
using the TNBS test or chloranil test to check the extent of
coupling. In case of a positive test result, the coupling was
repeated until a negative test result was obtained. The resulting
batch of the resin-bound protected peptide was carefully
washed with DMF (3 ꢀ 15 mL), DCM (3 ꢀ 15 mL), DMF
(3 ꢀ 15 mL), and DCM (3 ꢀ 15 mL) and dried under reduced
pressure. The peptide was cleaved off the solid support with
1% v/v TFA in DCM (10 mL) for 30 min, and most of the
organic solvent was removed under reduced pressure. The
obtained protected peptides with free C-terminal were precipi-
tated out by the addition of chilled petroleum ether (30 mL) to give
a white precipitate. The suspensions were centrifuged for 20 min
at 7000 rpm, and then the liquid was decanted off. The crude
peptides were washed with petroleum ether (2 ꢀ 50 mL), and
after another centrifugation, the peptides were dried under
vacuum (2 h) to obtain the title compound. The purity of the
protected peptides with a free C-termini (>90%) were checked
by analytical RP-HPLC using a Hewlett-Packard 1100 system
(230 nm) on a reverse phase column (Waters NOVA-Pak C-18
Oxidative Cyclization to Disulfides.25,45 A solution of K3Fe-
(CN)6 was prepared as follows: 1 mmol (330 mg) of K3Fe(CN)6
was dissolved in a mixture of water (100 mL) and CH3CN
(100 mL). A saturated solution of ammonium acetate (20 mL)
was added to it, and the pH was adjusted to 6.0 with glacial
acetic acid. A solution of the linear peptide (0.173 mmol) in the
mixture of CH3CN (40 mL), DMSO (5 mL), and H2O (5 mL)
was added to the above solution dropwise overnight with the
help of a syringe pump.25 After the overnight reaction, glacial
acetic acid was added to the reaction mixture to obtain pH 5.0,
followed by 20 mL of Amberlite IRA-68 anion-exchange resin
(pre-equilibrated with 1 M HCl and extensively washed with dis-
tilled water), and the suspension was stirred for 30 min until the
solution turned colorless and the resin turned yellow. The resin
was suction-filtered and the filtrate rotoevaporated to remove
most of the organic solvent. The remaining solution was con-
centrated on a Sep-Pak C18 cartridge (10 g, Waters, Milford,
MA) and then eluted with CH3CN. The obtained yellow solu-
tion was concentrated under reduced pressure for the final
purification by preparative RP-HPLC and then lyophilized.
Preparative RP-HPLC was performed on a Waters Delta Prep
4000 with a Waters XTerra C-18 column (19 mm ꢀ 250 mm,
10 μm, a linear gradient of 33-53% or 40-60% acetonitrile/
0.1% TFA at a flow rate of 15.0 mL/min).
The purified final peptides (g95%) were characterized by
HRMS, TLC, analytical HPLC, and H NMR (all the data,
1
charts, and experimental details are available in the Supporting
Information).
NMR Spectroscopy in DPC Amphipathic Media and Confor-
mational Structure Determination. All of the conformational
determinations were performed as follows,5,6,35,36 based on the
NMR spectra using a Bruker DRX600 600 MHz spectrometer.
The samples were prepared by dissolving the peptide
(4.2 mM) in 0.5 mL of 45 mM sodium acetate-d3 buffer (pH
4.5) containing 40 equiv of dodecylphosphocholine-d38 and
1 mM sodium azide (90% H2O/10% D2O), followed by sonica-
tion for 5 min. Two-dimensional double quantum filtered correl-
ation (DQF-COSY), nuclear Overhauser enhancement spec-
troscopy46 (NOESY, mixing time = 450 ms), rotating frame
Overhauser effect spectroscopy (ROESY, mixing time = 150 ms),
and total correlation spectra47 (TOCSY, MLEV-17 mixing
time = 62.2 ms, spin-lock field = 8.33 kHz) were acquired
using standard pulse sequences at 310 K. Coupling constants
(3JNH-HR) were measured from 2D DQF-COSY spectra by
˚
column, 3.9 mm ꢀ 150 mm, 5 μm, 60 A). The peptide was eluted
with a linear gradient of aqueous CH3CN/0.1% CF3CO2H
(10-90% in 40 min) at a flow rate of 1.0 mL/min. The crude

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15 5499
analysis of the fingerprint region with a curve-fitting using
5-parameter Levenberg-Marquardt nonlinear least-squares
protocol to a general antiphase doublet.
For the conformational structure determination, the volumes
of the assigned cross-peaks in the 2D NOESY spectrum were
˚
converted into upper distance bounds of 3.0, 3.8, 4.8, or 5.8 A.
previously described.4-7 For competition analysis, 10 concen-
trations of a test compound were each incubated with 50 μg of
membranes from δ- or μ-opioid receptors expressing cells
labeled with [3H]DPDPE (1.0 nM) or [3H]DAMGO (1.0 nM),
respectively. Naloxone at 10 μM was used to define the non-
specific binding of the radioligands in all assays. For the affinity
at the human or rat NK1 receptors, binding assays utilized
membranes from transfected CHO cells that stably express
corresponding receptors, using [3H]-substance P (135 Ci/mmol,
0.3-0.4 nM) as the standard radioligand. The [3H] substance
P concentration was selected based on the saturation binding
experiments which showed a high affinity binding with KD
values of 0.40 ( 0.17 for hNK1 and 0.16 ( 0.03 nM for rNK1,
respectively. Substance P at 10 μM was used to define the
nonspecific binding. All data were determined from a mini-
mum of 2 independent analyses. The data were analyzed by
nonlinear least-squares regression analysis using GraphPad
Prism.
For overlapping cross-peaks, the distance categories were in-
creased by one or two levels, depending on the qualitative esti-
mate of the extent of overlap. Pseudoatoms were created for
nonstereospecifically assigned methylene protons with a correc-
tion of 1.0 A applied to their upper bound distances.48 In addi-
˚
tion to the distance constraints, j dihedral angle constraints
3
derived from JHN-HR coupling constants were set to between
-90° and 40° for 3JHN-HR < 6 Hz and to between -150° and
-90° for 3JHN-HR > 8 Hz. Dihedral angle constraints of 180 (
5° for peptide bonds (ω) were also used to maintain the planarity
of these bonds.
Structures were calculated using a hybrid distance geometry-
dynamical simulated annealing protocol, followed by energy
refinement using restrained molecular dynamics (rMD), using
the NOE-derived distance constraints and dihedral angle (j)
constraints. Distance geometry calculations were performed using
the DGII49 program within the INSIGHT II package (Accelrys
Inc., San Diego, CA). Solvent was not explicitly included in the
calculations. The bounds for the interatomic distances were
embedded in four dimensional spaces, followed by the optimi-
zation using a simulated annealing protocol with sigmoidal
cooling schedule from a maximum temperature of 200 K in
10000 steps at simulation steps of 2.5 ps. The fail level was set to
2.00. All the embedded structures successfully passed the simu-
lated annealing step and were minimized using the consistent
valency force field (CVFF) (Accelrys Inc.). The entire process
was repeated for 100 times to generate 100 independently optim-
ized structures.
The [35S]GTPγS binding assays were used to estimate the
functional activities of the test compounds at the δ and μ opioid
receptors expressed in the transfected cells. Data were analyzed
by nonlinear least-squares analysis using GraphPad Prism to
determine the EC50 and Emax values for each test compound.
The isolated tissue-based functional assays also were used to
evaluate opioid agonist activities in the GPI (μ) and MVD (δ).
Dose effects of the test compounds against substance P stimula-
tion were determined in isolated GPI in the presence of 1 μM
naloxone.
In Vitro Stability of Peptide Derivatives in Rat Plasma.6 Stock
solution of compounds (50 mg/mL in DMSO) were diluted
1000-fold into rat plasma (lot 24927, Pel-Freez Biologicals,
Rogers, AK) to give an incubation concentration of 50 μg/
mL. All samples were incubated at 37 °C and 200 μL of aliquots
were withdrawn at 1, 2, 4, and 6 h. Then 300 μL of acetonitrile
was added and the proteins were removed by centrifugation.
The supernatant was analyzed for the amount of remaining
parent compound by HPLC (Hewlett-Packard1090 m with Vydac
Of those, the 50 structures with the lowest penalty function
were further refined by two rounds of rMD using the all-atom
AMBER force field with additional parameters for fluorine
atom,44,50-52 using the standalone DISCOVER version 2.98
˚
218TP104 C-18 column; 4.6 mm ꢀ 250 mm, 10 μm, 300 A). The
program (Accelrys Inc.).35,53-55 A 12.0 A cutoff for nonbonded
˚
samples were tested in three independent experiments (n = 3),
and the mean values ( SD were used for the analysis. The
statistical significances were evaluated with the Student t test.
interactions and a distance-dependent dielectric constant (4r)
were used. All amide bonds were constrained to trans conforma-
tion by a 100 kcal mol-1 rad-2 energy penalty. The distance
constraints and dihedral angles (j) constraints were applied
Acknowledgment. The work was supported by grants from
the USDHS, National Institute on Drug Abuse, DA-13449
and DA-06284. We thank Dr. Guangxin Lin for kind assis-
tance and advice with the NMR measurements, Magdalena
Kaczmarska for culturing cells, and the University of Arizona
Mass Spectrometry Facility for the mass spectral measure-
ments. We express appreciation to Margie Colie for assistance
with the manuscript.
with a force constant of 25 kcal mol-1
A
-2 and 100 kcal mol-1
˚
rad-2, respectively. After 100 steps of steepest descents mini-
mization and 1000 steps of conjugate gradient minimization on
the initial structures, an rMD equilibration at 500 K was
performed for 1.5 ps, during which a scale factor of 0.1 was
applied to the experimental restraint force constants. During the
next 2 ps, full values of the experimental restraint force con-
stants were applied. A further 1 ps rMD simulation was run at
500 K, and the system was then cooled to 0 K over 3 ps. After an-
other 1 ps at 0 K, 100 cycles of steepest descents and 2000 steps of
conjugate gradient minimization were performed. The final 10
structures with the lowest energies were used for the analysis.
All calculations were performed on a Silicon Graphics Octane
computer.
1
Supporting Information Available: H NMR, HPLC, and MS
data of the cyclic peptide derivatives 2-5 and their intermedi-
ates. This material is available free of charge via the Internet at
http://pubs.acs.org.
Radioligand Labeled Binding Assay, [35S]GTP-γ-S Binding
Assay, GPI, and MVD in Vitro Bioassay. The methods were car-
ried out according to that previously described.4-7 Briefly, the
evaluation of the binding affinities of the synthesized bi-
functional peptide derivatives at the human δ-opioid receptors
and rat μ-opioid receptors were performed on the cell (HN9.10)
membranes that stably express these corresponding receptors
using [3H]-c[D-Pen2,D-Pen5]-enkephalin ([3H]DPDPE, 44 Ci/
mmol) and [3H]-[D-Ala2,NMePhe4,Gly5-ol]-enkephalin ([3H]-
DAMGO, 47.2 Ci/mmol) as radioligands, respectively. The
affinity of [3H]DAMGO at the μ-opioid receptor and that of
[3H]DPDPE for the δ-opioid receptor was defined by the dis-
sociation constant (KD) determined by saturation analysis as
References
(1) Mantyh, P. W.; Allen, C. J.; Ghilardi, J. R.; Rogers, S. D.; Mantyh,
C. R.; Liu, H.; Basbaum, A. I.; Vigna, S. R.; Maggio, J. E. Rapid
endocytosis of a G protein-coupled receptor: substance P evoked
internalization of its receptor in the rat striatum in vivo. Proc. Natl.
Acad. Sci. U.S.A. 1995, 92, 2622–2626.
(2) Kalso, E. Improving opioid effectiveness: from ideas to evidence.
Eur. J. Pain 2005, 9, 131–135.
(3) King, T.; Ossipov, M. H.; Vanderah, T. W.; Porreca, F.; Lai, J. Is
paradoxical pain induced by sustained opioid exposure an under-
lying mechanism of opioid antinociceptive tolerance? Neurosignals
2005, 14, 194–205.
(4) Yamamoto, T.; Nair, P.; Davis, P.; Ma, S. W.; Navratilova, E.;
Moye, M.; Tumati, S.; Vanderah, T. W.; Lai, J.; Porreca, F.;

5500 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15
Yamamoto et al.
Yamamura, H. I.; Hruby, V. J. Design, Synthesis and Biological
Evaluation of Novel Bifunctional C-terminal Modified Peptides
for δ/μ Opioid Receptor Agonists and Neurokinin-1 Receptor
Antagonists. J. Med. Chem. 2007, 50, 2779–2786.
(23) Largent-Milnes, T. M.; Yamamoto, T.; Nair, P.; Navratrilova, E.;
Davis, P.; Ma, S.-W.; Hruby, V. J.; Yamamura, H. I.; Lai, J.;
Porreca, F.; Vanderah, T. W. Dual acting opioid agonist/NK1
antagonist peptide reverses neuropathic pain in an animal model
without demonstrating common opioid unwanted side effects. In
International Association for the Study of Pain/12th World Con-
gress on Pain, Glasgow, Scotland, 2008.
(24) Largent-Milnes, T. M.; Yamamoto, T.; Nair, P.; Hruby, V. J.;
Yamamura, H. I.; Lai, J.; Porreca, F.; Vanderah, T. W. Spinal or
systemic TY005, a peptidic opioid agonist/neurokinin 1 antagonist,
attenuates pain with reduced tolerance Br. J. Pharmacol. 2010,
accepted with revision.
(25) Yamamoto, T.; Nair, P.; Davis, P.; Ma, S.-W.; Yamamura, H. I.;
Vanderah, T. W.; Porreca, F.; Lai, J.; Hruby, V. J. The biological
activity and metabolic stability of peptidic bifunctional compounds
that are opioid receptor agonists and neurokinin 1 receptor
antagonists with a cystine moiety. Bioorg. Med. Chem. 2009, 17,
7337–7343.
(26) Haaseth, R. C.; Horan, P. J.; Bilsky, E. J.; Davis, P.; Zalewska, T.;
Slaninova, J.; Yamamura, H. I.; Weber, S. J.; Davis, T. P.; Porreca,
F.; Hruby, V. J. [L-Ala3]DPDPE: a new enkephalin analog with a
unique opioid receptor activity profile. Further evidence of delta-
opioid receptor multiplicity. J. Med. Chem. 1994, 37, 1572–1577.
(27) Toth, G.; Russell, K. C.; Landis, G.; Kramer, T. H.; Fang, L.;
Knapp, R.; Davis, P.; Burks, T. F.; Yamamura, H. I.; Hruby, V. J.
Ring substituted and other conformationally constrained tyrosine
analogues of [D-Pen2,D-Pen5]enkephalin with delta opioid recep-
tor selectivity. J. Med. Chem. 1992, 35, 2384–2391.
(28) Mosberg, H. I.; Hurst, R.; Hruby, V. J.; Gee, K.; Yamamura, H. I.;
Galligan, J. J.; Burks, T. F. Bis-penicillamine enkephalins possess
highly improved specificity toward delta opioid receptors. Proc.
Natl. Acad. Sci. U.S.A. 1983, 80, 5871–5874.
(29) Flippen-Anderson, J. L.; Hruby, V. J.; Collins, N.; George, C.;
Cudney, B. X-ray Structure of [D-Pen2,D-Pen5]enkephain, a
Highly Potent, Delta Opioid Receptor-Selective Compound: Com-
parisons with Proposed Solution Conformations. J. Am. Chem.
Soc. 1994, 116, 7523–7531.
(30) Mosberg, H. I.; Omnaas, J. R.; Goldstein, A. Structural require-
ments for delta opioid receptor binding. Mol. Pharmacol. 1987, 31,
599–602.
(31) Millet, R.; Goossens, L.; Bertrand-Caumont, K.; Chavatte, P.;
Houssin, R.; Henichart, J. P. Synthesis and biological evaluation of
tripeptide derivatives of Cbz-Gly-Leu-Trp-OBzl(CF3)2 as NK1/
NK2 ligands. Lett. Pept. Sci. 1999, 6, 255–262.
(32) Barlos, K.; Gatos, D. Convergent Peptide Synthesis. In Fmoc Solid
Phase Peptide Synthesis: A Practical Approach; Chan, W. C.; White,
P. D., Eds.; Oxford University Press: New York, 2000; pp 214-228.
(33) D’Alagni, M.; Delfini, M.; Di Nola, A.; Eisenberg, M.; Paci, M.;
Roda, L. G.; Veglia, G. Conformational study of
[Met5]enkephalin-Arg-Phe in the presence of phosphatidylserine
vesicles. Eur. J. Biochem. 1996, 240, 540–549.
(34) Deber, C. M.; Behnam, B. A. Role of membrane lipids in peptide
hormone function: binding of enkephalins to micelles. Proc. Natl.
Acad. Sci. U.S.A. 1984, 81, 61–65.
(35) Jacobsen, N. E.; Abadi, N.; Sliwkowski, M. X.; Reilly, D.; Skelton,
N. J.; Fairbrother, W. J. High-resolution solution structure of the
EGF-like domain of heregulin-alpha. Biochemistry 1996, 35, 3402–
3417.
(36) Ying, J.; Ahn, J. M.; Jacobsen, N. E.; Brown, M. F.; Hruby, V. J.
NMR solution structure of the glucagon antagonist [desHis1,
desPhe6,Glu9]glucagon amide in the presence of perdeuterated
dodecylphosphocholine micelles. Biochemistry 2003, 42, 2825–
2835.
(5) Yamamoto, T.; Nair, P.; Jacobsen, N. E.; Davis, P.; Ma, S. W.;
Navratilova, E.; Lai, J.; Yamamura, H. I.; Vanderah, T. W.;
Porreca, F.; Hruby, V. J. The Importance of Micelle-Bound States
for the Bioactivities of Bifunctional Peptide Derivatives for δ/μ
Opioid Receptor Agonists and Neurokinin 1 Receptor Antago-
nists. J. Med. Chem. 2008, 51, 6334–6347.
(6) Yamamoto, T.; Nair, P.; Jacobsen, N. E.; Vagner, J.; Kulkarni, V.;
Davis, P.; Ma, S.-W.; Navratilova, E.; Yamamura, H. I.; Vander-
ah, T. W.; Porreca, F.; Lai, J.; Hruby, V. J. Improving metabolic
stability by glycosylation: bifunctional peptide derivatives that are
opioid receptor agonists and neurokinin 1 receptor antagonists.
J. Med. Chem. 2009, 52, 5164–5175.
(7) Yamamoto, T.; Nair, P.; Vagner, J.; Davis, P.; Ma, S. W.; Navratilova,
E.; Moye, M.; Tumati, S.; Vanderah, T. W.; Lai, J.; Porreca, F.;
Yamamura, H. I.; Hruby, V. J. A Structure-Activity Relationship
Study and Combinatorial Synthetic Approach of C-Terminal
Modified Bifunctional Peptides That Are δ/μ Opioid Receptor
Agonists and Neurokinin 1 Receptor Antagonists. J. Med. Chem.
2008, 51, 1369–1376.
(8) King, T.; Gardell, L. R.; Wang, R.; Vardanyan, A.; Ossipov,
M. H.; Malan, T. P., Jr.; Vanderah, T. W.; Hunt, S. P.; Hruby,
V. J.; Lai, J.; Porreca, F. Role of NK-1 neurotransmission in
opioid-induced hyperalgesia. Pain 2005, 116, 276–288.
(9) Ma, W.; Zheng, W. H.; Kar, S.; Quirion, R. Morphine treatment
induced calcitonin gene-related peptide and substance P increases
in cultured dorsal root ganglion neurons. Neuroscience 2000, 99,
529–539.
(10) Powell, K. J.; Quirion, R.; Jhamandas, K. Inhibition of neurokinin-
1-substance P receptor and prostanoid activity prevents and re-
verses the development of morphine tolerance in vivo and the
morphine-induced increase in CGRP expression in cultured dorsal
root ganglion neurons. Eur. J. Neurosci. 2003, 18, 1572–1583.
(11) Misterek, K.; Maszczynska, I.; Dorociak, A.; Gumulka, S. W.;
Carr, D. B.; Szyfelbein, S. K.; Lipkowski, A. W. Spinal co-admin-
istration of peptide substance P antagonist increases antinocicep-
tive effect of the opioid peptide biphalin. Life Sci. 1994, 54, 939–
944.
(12) Gu, G.; Kondo, I.; Hua, X. Y.; Yaksh, T. L. Resting and evoked
spinal substance P release during chronic intrathecal morphine
infusion: parallels with tolerance and dependence. J. Pharmacol.
Exp. Ther. 2005, 314, 1362–1369.
(13) Ripley, T. L.; Gadd, C. A.; De Felipe, C.; Hunt, S. P.; Stephens,
D. N. Lack of self-administration and behavioural sensitisation to
morphine, but not cocaine, in mice lacking NK1 receptors. Neuro-
pharmacology 2002, 43, 1258–1268.
(14) Morphy, R.; Rankovic, Z. Designed multiple ligands. An emerging
drug discovery paradigm. J. Med. Chem. 2005, 48, 6523–6543.
(15) Ling, G. S.; Spiegel, K.; Lockhart, S. H.; Pasternak, G. W.
Separation of opioid analgesia from respiratory depression: evi-
dence for different receptor mechanisms. J. Pharmacol. Exp. Ther.
1985, 232, 149–155.
(16) Paakkari, P.; Paakkari, I.; Vonhof, S.; Feuerstein, G.; Siren, A. L.
Dermorphin analog Tyr-D-Arg2-Phe-sarcosine-induced opioid an-
algesia and respiratory stimulation: the role of mu 1-receptors?
J. Pharmacol. Exp. Ther. 1993, 266, 544–550.
(17) Chang, K. J.; Rigdon, G. C.; Howard, J. L.; McNutt, R. W. A novel
potent and selective nonpeptidic delta opioid receptor agonist
BW373U86. J. Pharmacol. Exp. Ther. 1993, 267, 852–857.
(18) Sheldon, R. J.; Riviere, P. J.; Malarchik, M. E.; Mosberg, H. I.;
Burks, T. F.; Porreca, F. Opioid regulation of mucosal ion trans-
port in the mouse isolated jejunum. J. Pharmacol. Exp. Ther. 1990,
253, 144–151.
(37) Braun, W.; Wider, G.; Lee, K. H.; Wuthrich, K. Conformation of
glucagon in a lipid-water interphase by 1H nuclear magnetic
resonance. J. Mol. Biol. 1983, 169, 921–948.
(19) Cowan, A.; Zhu, X. Z.; Mosberg, H. I.; Omnaas, J. R.; Porreca, F.
Direct dependence studies in rats with agents selective for different
types of opioid receptor. J. Pharmacol. Exp. Ther. 1988, 246, 950–
955.
(20) Quock, R. M.; Burkey, T. H.; Varga, E.; Hosohata, Y.; Hosohata,
K.; Cowell, S. M.; Slate, C. A.; Ehlert, F. J.; Roeske, W. R.;
Yamamura, H. I. The delta-opioid receptor: molecular pharma-
cology, signal transduction, and the determination of drug efficacy.
Pharmacol. Rev. 1999, 51, 503–532.
(21) Rew, Y.; Malkmus, S.; Svensson, C.; Yaksh, T. L.; Chung, N. N.;
Schiller, P. W.; Cassel, J. A.; DeHaven, R. N.; Taulane, J. P.;
Goodman, M. Synthesis and biological activities of cyclic lanthio-
nine enkephalin analogues: delta-opioid receptor selective ligands.
J. Med. Chem. 2002, 45, 3746–3754.
(38) Thornton, K.; Gorenstein, D. G. Structure of glucagon-like pep-
tide (7-36) amide in a dodecylphosphocholine micelle as deter-
mined by 2D NMR. Biochemistry 1994, 33, 3532–3539.
(39) Karslake, C.; Piotto, M. E.; Pak, Y. K.; Weiner, H.; Gorenstein,
D. G. 2D NMR and structural model for a mitochondrial signal
peptide bound to a micelle. Biochemistry 1990, 29, 9872–9878.
(40) Arora, A.; Abildgaard, F.; Bushweller, J. H.; Tamm, L. K. Struc-
ture of outer membrane protein A transmembrane domain by
NMR spectroscopy. Nature Struct. Biol. 2001, 8, 334–338.
(41) Hruby, V. J.; Gehrig, C. A. Recent developments in the design of
receptor specific opioid peptides. Med. Res. Rev. 1989, 9, 343–401.
(42) Flippen-Anderson, J. L.; Hruby, V. J.; Collins, N.; George, C.;
Cudney, B. X-ray Structure of [D-Pen2,D-Pen5]enkephain, a
Highly Potent, delta Opioid Receptor-Selective Compound: Com-
parisons with Proposed Solution Conformations. J. Am. Chem.
Soc. 1994, 116, 7523–31.
(22) Coop, A.; MacKerell, A. D. J. The future of opioid analgesics. Am.
J. Pharm. Educ. 2002, 66, 153–156.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15 5501
(43) Shenderovich, M. D.; Kover, K. E.; Nikiforovich, G. V.; Jiao, D.;
Hruby, V. J. Conformational analysis of beta-methyl-para-nitro-
phenylalanine stereoisomers of cyclo[D-Pen2,D-Pen5]enkephalin
by NMR spectroscopy and conformational energy calculations.
Biopolymers 1996, 38, 141–156.
(44) Fluorine Parameters; http://amber.scripps.edu/Questions/fluorine.
html (accessed February 2007).
(50) Weiner, S. J.; Kollman, P. A.; Case, D. A.; Singh, U. C.; Ghio, C.;
Alagona, G. S.; Profeta, J.; Weiner, P. A New Force Field for
Molecular Mechanical Simulation of Nucleic Acids and Proteins.
J. Am. Chem. Soc. 1984, 106, 765–784.
(51) Weiner, S. J.; Kollman, P. A.; Case, D. A. An all atom force field
for simulations of proteins and nucleic acids. J. Comput. Chem.
1986, 7, 230–252.
(45) Balse-Srinivasan, P.; Grieco, P.; Cai, M.; Trivedi, D.; Hruby, V. J.
Structure-activity relationships of novel cyclic alpha-MSH/beta-
MSH hybrid analogues that lead to potent and selective ligands for
the human MC3R and human MC5R. J. Med. Chem. 2003, 46,
3728–3733.
(52) Gough, C. A.; DeBolt, S. E.; Kollman, P. A. Derivation of fluorine
and hydrogen atom parameters using liquid simulations. J. Com-
put. Chem. 1992, 13, 963–970.
(53) Milbradt, A. G.; Kerek, F.; Moroder, L.; Renner, C. Structural
characterization of hellethionins from Helleborus purpurascens.
Biochemistry 2003, 42, 2404–2411.
€
(46) Kumar, A.; Ernst, R. R.; Wuthrich, K. A two-dimensional nuclear
Overhauser enhancement (2D NOE) experiment for the elucida-
tion of complete proton-proton cross-relaxation networks in
biological macromolecules. Biochem. Biophys. Res. Commun.
1980, 95, 1–6.
(54) Renner, C.; Behrendt, R.; Sporlein, S.; Wachtveitl, J.; Moroder, L.
Photomodulation of conformational states. I. Mono- and bicyclic
peptides with (4-amino)phenylazobenzoic acid as backbone con-
stituent. Biopolymers 2000, 54, 489–500.
(55) Fuhrmans, M.; Milbradt, A. G.; Renner, C. Comparison of Pro-
tocols for Calculation of Peptide Structures from Experimental
NMR Data. J. Chem. Theory Comput. 2006, 2, 201–208.
(56) Lipkowski, A. W.; Misicka, A.; Davis, P.; Stropova, D.; Janders,
J.; Lachwa, M.; Porreca, F.; Yamamura, H. I.; Hruby, V. J.
Biological activity of fragments and analogues of the potent
dimeric opioid peptide, biphalin. Bioorg. Med. Chem. Lett. 1999,
9, 2763–2766.
(47) Davis, D. G.; Bax, A. Assignment of complex proton NMR spectra
via two-dimensional homonuclear Hartmann-Hahn spectroscopy.
J. Am. Chem. Soc. 1985, 107, 2820–2821.
€
(48) Wuthrich, K.; B., M.; Braun, W. Pseudostructures for the 20
common amino acids for use in studies of protein conformations
by measurements of intramolecular proton-proton distance con-
straints with nuclear magnetic resonance. J. Mol. Biol. 1983, 169,
949–961.
(49) Havel, T. F. An evaluation of computational strategies for use in
the determination of protein structure from distance constraints
obtained by nuclear magnetic resonance. Prog. Biophys. Mol. Biol.
1991, 56, 43–78.
(57) Wagner, G.; Neuhaus, D.; Worgotter, E.; Vasak, M.; Kagi, J. H.;
Wuthrich, K. Nuclear magnetic resonance identification of “half-
turn” and 3(10)-helix secondary structure in rabbit liver metal-
lothionein-2. J. Mol. Biol. 1986, 187, 131–135.