Use of Molecular Modeling to Design Selective NTS2 Neurotensin Analogues

Article abstract of DOI:10.1021/acs.jmedchem.6b01848

Neurotensin exerts potent analgesia by acting at both NTS1 and NTS2 receptors, whereas NTS1 activation also results in other physiological effects such as hypotension and hypothermia. Here, we used molecular modeling approach to design highly selective NT

Full text of DOI:10.1021/acs.jmedchem.6b01848

Article
pubs.acs.org/jmc
Use of Molecular Modeling to Design Selective NTS2 Neurotensin
Analogues
‡
†
‡
Roberto Fanelli,† Nicolas Floquet,† Elie Besserer-Offroy, Bartholome Delort, Melanie Vivancos,
́
́
́
Jean-Michel Longpre,‡ Pedro Renault,†,§ Jean Martinez,† Philippe Sarret,‡,∥ and Florine Cavelier*,†,∥
́
†Institut des Biomolec
34095 Montpellier cedex 5, France
‡Department of Pharmacology and Physiology, Institut de Pharmacologie de Sherbrooke, Faculty of Medicine and Health Sciences,
́ ́ ̀
ules Max Mousseron, IBMM, UMR-5247, CNRS, Universite Montpellier, ENSCM, Place Eugene Bataillon,
́ ́
Universite de Sherbrooke, Sherbrooke J1H 5N4, Quebec, Canada
§Centre de Biochimie Structurale (CBS), UMR-5048, CNRS, Universite
34090 Montpellier, France
́
de Montpellier, INSERM U1054, 29 rue de Navacelles,
S
* Supporting Information
ABSTRACT: Neurotensin exerts potent analgesia by acting at both NTS1 and NTS2 receptors, whereas NTS1 activation also
results in other physiological effects such as hypotension and hypothermia. Here, we used molecular modeling approach to
design highly selective NTS2 ligands by investigating the docking of novel NT[8-13] compounds at both NTS1 and NTS2 sites.
Molecular dynamics simulations revealed an interaction of the Tyr11 residue of NT[8-13] with an acidic residue (Glu179) located
in the ECL2 of hNTS2 or with a basic residue (Arg212) at the same position in hNTS1. The importance of the residue at position
11 for NTS1/NTS2 selectivity was further demonstrated by the design of new NT analogues bearing basic (Lys, Orn) or acid
(Asp or Glu) function. As predicted by the molecular dynamics simulations, binding of NT[8-13] analogues harboring a Lys11
exhibited higher affinity toward the hNTS1-R212E mutant receptor, in which Arg212 was substituted by the negatively charged
Glu residue.
results in significant antinociception but also causes marked
hypotension and hypothermia.3 In sharp contrast, the neuro-
tensin receptor subtype 2 (NTS2) has emerged as an important
pain target because NTS2-selective analogues exhibit potent
analgesic activity in both acute and chronic pain conditions in
dose-dependent analgesic effects without inducing drop in
blood pressure or body temperature.7 Thus, the development
of new NTS2-selective agonists acting as nonopioid analgesics
could be useful for relieving chronic pain.
To develop new potent and selective NTS2 receptor ligands
as possible pain treatment, there is a crucial need to provide
valuable information on the structure−activity relationship
(SAR) of the ligand−receptor interactions, which serve as a
tool to the drug design process. To date, the design of
INTRODUCTION
■
Chronic pain is the most common reason for medical
consultations, and conventional opioid analgesics remain the
pharmacological cornerstone of pain therapy. However, despite
their proven analgesic efficacy, the management of the opioid
adverse effects still represents a major clinical challenge.1 There
is therefore an urgent need to develop new lead compounds for
improving the pain pharmacopeia.
Neurotensin (NT) is a tridecapeptide first isolated in 1973
by Carraway and Leeman from bovine hypothalamus extract.2
Following central administration, NT exerts a wide range of
biological functions including hypothermia,3 analgesia,4 and
antipsychotic-like properties.5 Importantly, the analgesic
activity of NT, mediated by its interaction with two G
protein-coupled receptors, namely NTS1 and NTS2, was found
to be as potent as that of opioids, such as morphine.6
Additionally, it was notably reported that NTS1 activation
Received: December 20, 2016
Published: April 3, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.jmedchem.6b01848
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 1. RMSD values computed on the backbone atoms of either receptor (orange) or peptide (blue) along the two trajectories obtained for
hNTS1 (A) and hNTS2 (B) receptors, respectively. In each case, the receptor was bound to the native NT[8-13] peptide.
Figure 2. Chi1 (in blue) and chi2 (in orange) values of the Tyr11 residue in the NT [8-13] peptide along the two trajectories computed for NTS2
(A) and NTS1 (C) receptors (180 ns of classical MD + 20 ns of accelerated MD). (B,D) Orientation of the Tyr11 residue after its rotation in both
receptors. In yellow is reported the reference position of the peptide in the rat NTS1 X-ray structure with an orientation of Tyr11 corresponding to
initial chi1 and chi2 values of 53 and −93 degrees, respectively.
minimum active fragments of the biologically active NT
tridecapeptide has led to the identification of the six carboxy-
terminal residues (Arg8-Arg9-Pro10-Tyr11-Ile12-Leu13, NT[8-
13]),8,9 therefore constituting a good starting point to provide
insight into the molecular determinants of the ligand−receptor
interactions. Interestingly, we recently demonstrated that the
replacement of the carboxy-terminal residue of NT[8-13] by a
(L)-(trimethylsilyl)alanine (TMSAla) enhanced its binding
affinity for both NTS1 and NTS2 by improving its interaction
with the hydrophobic ligand binding pocket of both
receptors.10 Additionally, position 11 has been found to be
critical for the selectivity toward NTS2 but no clear SAR has
emerged from these studies.11−14 Recently, the X-ray
crystallography structure of the rat NTS1 receptor was resolved
in an agonist-bound conformation.15 This crystal structure of
rNTS1 bound to its agonist peptide NT[8-13] has allowed
deciphering of the key features governing receptor−ligand
interactions and now gives the structural basis for investigating
the effects of bound ligands on receptor dynamics.
In the present study, starting from the high-resolution crystal
data of NTS1, we built structural models for both hNTS1 and
hNTS2 receptors bound to the NT[8-13] peptide. Our results
revealed the existence of some divergence at the ligand−
receptor interface, especially the NT receptor residues facing
the tyrosine residue in position 11 of the NT[8-13] peptide,
namely residue 212 in NTS1 and residue 179 in NTS2. On the
basis of these observations, we synthesized a series of NT[8-13]
analogues modified at position 11 to evaluate their selectivity
toward NTS1 and NTS2 receptors. As predicted by the
molecular dynamics simulations, our binding experiments were
able to demonstrate that the residue at position 11 in NT[8-13]
was crucial for the NTS1/NTS2 selectivity. Finally, to further
B
DOI: 10.1021/acs.jmedchem.6b01848
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
support these findings, we found that a single amino acid
mutation at position 212 in the second extracellular loop of the
NTS1 receptor (hNTS1-R212E) was able to restore affinity to
NTS2-selective analogues. Altogether, these results will help to
improve the design of the next-generation of NT[8-13]
analogues acting at the NTS2 receptor site.
RESULTS AND DISCUSSION
■
Molecular Modeling. Molecular dynamics simulations
were performed on both human NTS1 and NTS2 receptors,
each trajectory reaching 180 ns. In agreement with the known
NTS1 X-ray structure used as an initial template for homology
modeling (PDB 4BUO), both hNTS1 receptor conformation
and NT[8-13] ligand orientation inside the binding pocket
were remarkably stable all along the trajectory, with mean root
mean square deviations (RMSD) of 2.4 and 0.8 Å, respectively
(Figure 1A). For the hNTS2 model, we observed an increased
flexibility of both receptor (mean RMSD of 2.9 Å) and peptide,
with a RMSD oscillating between 2.0 and 3.0 Å (Figure 1B).
Initially, this slight increase of the NTS2 receptor flexibility
was attributed to its reduced rate of sequence identity with the
rat NTS1 receptor (44% versus 88% for the hNTS1/rNTS1
couple). However, after inspection of both peptide and
receptor conformations, it appeared that this RMSD increase
was mainly due to a concerted motion of both the side-chain of
the Tyr11 residue in NT[8-13] and the extracellular loop 2
(ECL2) of the receptor. This rotation was clearly established,
showing a concerted shift of the two chi1 and chi2 angles of
Tyr11 after 40 ns (Figure 2A), leading to the interaction of
Tyr11 with Glu179 of NTS2 (KHE179LE) (Figure 2B). Such
rotation of the Tyr11 residue was not observed during the first
180 ns of the trajectory obtained for the hNTS1 receptor.
Therefore, to improve the sampling of the conformations of the
hNTS1 and hNTS2 peptide/receptor complexes, the two
trajectories were extended to ∼200 ns, performing 20 ns of
additional accelerated dynamics (aMD).
Accelerated molecular dynamics (aMD) is an enhanced-
sampling method that improves the conformational space
sampling by reducing energy barriers separating different states
of a same system. In this method, the potential energy
landscape is altered by adding a bias potential to the true
potential such that the escape rates from potential wells are
enhanced. This method has been recently applied successfully
to GPCRs, in particular to muscarinic receptors, to study the
dynamics of activation and the ligand binding.16 Using such a
protocol, the same rotation of the Tyr11 residue was also
observed in the hNTS1 receptor, orienting Tyr11 toward Arg212
(QNR212SA), as depicted in Figures 2C,D.
Figure 3. Localization of R212 and E179 residues in the extracellular
loop 2 of hNTS1 (in magenta) and hNTS2 (in green) receptors,
respectively. These models were both built by homology modeling
using the structure of the rat neurotensin receptor type 1 (PDB
4BUO) as a template.
hypothesized that this residue could significantly contribute
to the selectivity of peptides sharing the complementary charge
in position 11.
CHEMISTRY
■
All NT analogues prepared in this study were initially modified
to replace the two Arg residues of NT[8-13] (i.e., in positions 8
and 9) by Lys. Detailed SAR studies have revealed that the
presence of basic residues with positively charged side chains at
positions 8 and 9 is of critical importance for high-affinity
receptor binding and peptide activity.17 However, Lys is often
used in position 8, 9, or both because protection of amine side
chains is easier to handle in peptide synthesis. Importantly, this
modification does not affect the biological activities of the NT
analogues, as previously reported for compound 1
(JMV438).9,18 Therefore, the replacement of the two Arg
residues was done for all the newly generated analogues.
The molecular modeling results prompted us to synthesize a
series of NT[8-13] peptide analogues in which the aromatic
residue at position 11 was substituted with amino acids able to
provide ionic interactions with the corresponding residue on
the receptor at the entry of the binding pocket. The Tyr11
residue was therefore changed for amino acids bearing either a
basic or an acidic side chain. Tyr11 was successively replaced by
lysine to obtain compound 2 (H-Lys-Lys-Pro-Lys-Ile-Leu-OH,
JMV 5836), by ornithine in compound 3 (H-Lys-Lys-Pro-Orn-
Ile-Leu-OH, JMV 5837), and by histidine in compound 4 (H-
Lys-Lys-Pro-His-Ile-Leu-OH, JMV 5838), which presents a
charged side chain conserving the aromatic ring. Replacement
of Tyr11 by acid amino acids, such as aspartic and glutamic acid
residues, gave rise to compounds 5 (H-Lys-Lys-Pro-Asp-Ile-
Leu-OH, JMV 5839) and 6 (H-Lys-Lys-Pro-Glu-Ile-Leu-OH,
JMV 5963), respectively.
Interestingly, these two positively charged Arg212 and
negatively charged Glu179 residues were aligned on the initial
sequence alignment used for homology modeling step and were
therefore located at the same position of ECL2 in both
receptors at the entry of the binding site (Figure 3).
At this step of the work, our simulations suggested that the
Tyr11 residue of the native peptide could adopt a different
orientation than that observed in the available X-ray structure
of reference. This rotation was observed in both NTS1 and
NTS2 receptors. However, the residue promoting this rotation
was sharing an opposite charge according to the receptor
subtype: a glutamate in NTS2 and an arginine in NTS1.
Therefore, considering that this orientation could be another
binding mode of the peptide in its receptor, or at least a
transient conformation along its binding pathway, we
Compounds 2−6 were prepared by solid phase peptide
synthesis (SPPS) at a 0.25 mmol scale, using Fmoc-Leu
preloaded Wang resin, on a Liberty CEM microwave-assisted
peptide synthesizer. All Fmoc-amino acids were coupled by
using HATU as coupling reagent in the presence of DIPEA.
Deprotection cycles were carried out using 20% piperidine in
C
DOI: 10.1021/acs.jmedchem.6b01848
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
DMF. Final cleavage from the resin and removal of protecting
groups were performed by using a TFA/TIS mixture, leading to
the fully deprotected peptides, which were purified by
preparative HPLC, affording compounds 2−6 (Scheme 1).
was directly coupled with Boc-Ile-OH in the presence of BOP
and DIPEA, affording the dipeptide Boc-Ile-TMSAla-OMe in
very good yield. Removal of the Boc protecting group with
TFA afforded the free amine, which was directly coupled with
Z-Lys(Boc)-OH under the same experimental conditions (as
described above), affording the tripeptide Z-Lys(Boc)-Ile-
TMSAla-OMe in good yield. Cleavage of the Z protecting
group was achieved by catalytic hydrogenation in methanol in
the presence of one molar equivalent of 1 M hydrochloric acid
in order to avoid possible side reaction due to the
nucleophilicity of the primary amine. The tripeptide was then
coupled with Boc-Lys(Z)-Lys(Z)-Pro-OH, affording the fully
protected hexapeptide Boc-Lys(Z)-Lys(Z)-Pro-Lys(Boc)-Ile-
TMSAla-OMe in moderate yield. Total deprotection was
achieved in three steps: first, saponification of the methyl
ester moiety afforded the free carboxylic acid, followed by
removal of the Z protecting group of the lysine residues at
positions 8 and 9 by catalytic hydrogenation, and finally
removal of Boc in the presence of TFA and triisopropylsilane as
scavenger. Purification using preparative HPLC afforded
compound 7.
Scheme 1. Synthetic Strategy for Compounds 2−6
Binding Properties. The ability of this series of NT[8-13]
derivatives to inhibit the binding of 125I-[Tyr3]-NT on
membranes prepared from cells stably expressing either
hNTS1 or hNTS2 receptors was first evaluated and the results
are summarized in Table 1. Insertion of Lys in position 11
(compound 2) resulted in an important loss of affinity on
hNTS1 of more than 5000-fold compared to NT[8-13] (Ki of
5700 750 nM for compound 2 versus 1.07 0.05 nM for
NT[8-13]). This compound also showed a decreased affinity
for the hNTS2 receptor with a Ki value of 261.8 72.8 nM
when compared to NT[8-13]. Nevertheless, compound 2
showed more than 20-fold selectivity toward hNTS2, which
was consistent with the molecular modeling results described
above. When position 11 was modified for an ornithine residue,
a lysine analogue with a shorter side chain (compound 3), the
binding affinity was found to be decreased for both receptors
although some selectivity was conserved for hNTS2 (Ki of
Very recently, we also demonstrated that incorporation of
the silylated amino acid (L)-(trimethylsilyl)-alanine (TMSAla)
replacing the natural occurring leucine residue at the C-
terminal end of NT[8-13] led to a new analogue H-Lys-Lys-
Pro-Tyr-Ile-TMSAla-OH (JMV 2007),10 exhibiting high affinity
but no selectivity toward NT receptors. To exploit the gain in
affinity provided by this end terminal substitution, compound 7
(H-Lys-Lys-Pro-Lys-Ile-TMSAla-OH, JMV 5965) was also
synthesized in which Tyr11 and Leu13 were respectively replaced
by Lys and TMSAla residues.
Compound 7 was prepared in solution exploiting a 3 + 3
fragment condensation between the TMSAla-containing
tripeptide and the Boc-Lys(Z)-Lys(Z)-Pro-OH (Scheme 2).
The methyl ester of H-TMSAla-OH19 was prepared in
quantitative yield by using thionyl chloride in methanol and
>10000 nM toward hNTS1 and 619.5
198 nM toward
hNTS2). Therefore, these results suggest that the length of the
side chain exerts only a minor impact on the ionic interaction
stability with Glu179 facing residue in the NTS2 receptor. We
Scheme 2. Synthetic Strategy for Compound 7
D
DOI: 10.1021/acs.jmedchem.6b01848
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 1. Binding Affinities of the Reference Compound NT[8-13] and NT Analogues toward the hNTS1 and hNTS2 Receptors
a
Ki binding (nM)
compd
sequence
hNTS1
hNTS2
selectivity hNTS2/hNTS1
0.16
NT[8-13]
H-Arg-Arg-Pro-Tyr-Ile-Leu-OH
H-Lys-Lys-Pro-Tyr-Ile-Leu-OH
H-Lys-Lys-Pro-Lys-Ile-Leu-OH
H-Lys-Lys-Pro-Orn-Ile-Leu-OH
H-Lys-Lys-Pro-His-Ile-Leu-OH
H-Lys-Lys-Pro-Asp-Ile-Leu-OH
H-Lys-Lys-Pro-Glu-Ile-Leu-OH
H-Lys-Lys-Pro-Lys-Ile-TMSAla-OH
1.07 0.05
0.33 0.002
5700 750
>10000
6.57 2.18
0.95 0.003
261.8 72.8
619.5 198
474.1 82.0
4200 740
1600 370
67.0 26.0
c
1
2
3
4
5
6
7
0.35
21.8
b
nd
455.6 78.4
>10000
0.96
b
nd
b
>10000
nd
662.7 80.1
9.90
a
Ki refers to the equilibrium dissociation constant of a ligand determined by competitive ligand binding assay using 125I-[Tyr3]-NT as reference
b
c
probe. Not determinable. IC50 values taken from ref 18.
Figure 4. Snapshots extracted from MD simulations performed on compound 2 in complex with hNTS2 (A), hNTS1 (B), or hNTS1-R212E (C). In
each case, two distances were also computed between the nitrogen atom of Lys11 and the two oxygen (Glu) or nitrogen (Arg) atoms of the facing
residue in each receptor. This highlights the interaction of the Lys11 residue with Glu179 (A) and Glu212 (C) and the absence of this interaction in the
hNTS1 receptor.
next evaluated the influence of both positive charge and
aromaticity with compound 4, where Tyr11 was replaced by a
histidine. Binding results indicated a complete loss of selectivity
due to an increased affinity toward hNTS1 compared to
compounds 2 and 3 (Ki of 455.6 78.4 and 474.1 82.0 nM
toward hNTS1 and hNTS2, respectively). The aromaticity of
the imidazole ring seems therefore to be responsible of the
restored affinity toward hNTS1. Furthermore, these results
reinforce the importance of the complementary positively
charged residue in position 11 for the binding toward hNTS2.
Replacement of Tyr11 by negatively charged amino acids, i.e.,
aspartic (compound 5) or glutamic (compound 6), resulted in
an important loss of affinity for hNTS2, as predicted by the
presence of the negatively charged Glu179 in the ECL2 of
hNTS2 (Ki of 4200 740 and 1600 370 nM for compounds
5 and 6, respectively). It was, however, not expected that this
mutation also induced a complete loss in binding affinity for
hNTS1 (Ki of >10000 nM for both analogues). On the basis of
the binding data with compound 4 carrying an aromatic residue
in position 11, we can hypothesize that this loss in binding
affinity of compounds 5 and 6 for NTS1 may be attributed in
part to the absence of aromaticity. We can also speculate that
compounds 5 and 6 did not bind efficiently to the hNTS1
receptor due to an intramolecular interaction between the
negatively charged incorporated residue and one of the
positively charged residues of the hexapeptide, thus leading to
a drastic structural change on the active NT peptide
conformation. Finally, we can suppose that both hNTS1 and
hNTS2 receptors share several negatively charged residues in
their extracellular surface, impacting on the binding of
compounds 5 and 6 to both hNTS1 and hNTS2 receptors.
Finally, insertion of a TMSAla at position 13, resulting in
compound 7, induced a gain of binding affinity on both
receptor subtypes compared to compound 2 (Ki for hNTS1 of
662.7 80.1 versus 67.0 26.0 nM for hNTS2). However, this
compound still exhibited a 10-fold selectivity toward the
hNTS2 receptor. Altogether, as predicted by the molecular
dynamics simulations, these binding experiments reinforce that
Tyr11 is a key residue for maintaining receptor binding and that
the residue at position 11 in NT[8-13] is crucial for NTS1/
NTS2 selectivity.
E
DOI: 10.1021/acs.jmedchem.6b01848
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 5. Displacement curves of 125I-[Tyr3]-NT with NT[8-13], (A) compound 2, and (B) compound 7 on the wild-type (WT) form of the
hNTS1 receptor (dotted line) and on the hNTS1-R212E mutant (solid line).
Table 2. Binding Affinities of NT Analogues toward the Wild-Type hNTS1 and Mutated hNTS1-R212E Receptors
a
Ki binding (nM)
compd
sequence
hNTS1-WT
hNTS1-R212E
affinity gain R212E/WT
NT[8-13]
H-Arg-Arg-Pro-Tyr-Ile-Leu-OH
H-Lys-Lys-Pro-Lys-Ile-Leu-OH
H-Lys-Lys-Pro-Asp-Ile-Leu-OH
H-Lys-Lys-Pro-Glu-Ile-Leu-OH
H-Lys-Lys-Pro-Lys-Ile-TMSAla-OH
1.07 0.05
5700 750
>10000
4.69 0.75
199.2 31.6
>10000
−4.4
2
5
6
7
28.6
b
nd
b
>10000
>10000
nd
662.7 80.1
126.9 17.1
5.22
a
Ki refers to the equilibrium dissociation constant of a ligand determined by competitive ligand binding assay using 125I-[Tyr3]-NT as reference
b
probe. Not determinable.
Figure 6. Docking results of compounds 2 (A), and 7 (B) in the hNTS2 receptor, showing (A) the interaction of the Lys11 residue of compound 2
with Glu179 of hNTS2, and (B) the position taken by the TMSAla group of compound 7 in the bottom of the pocket (in white), compared to the
adamantane group of SR142948A (in rust) and SR48692 (in magenta). The native NT[8-13] peptide is reported in yellow in (A).
As observed by molecular modeling, the selectivity of NT
ligands toward hNTS2 seemed to be driven by the interaction
between the residues in position 11 of NT[8-13] and Glu179 of
hNTS2. To confirm this, we performed molecular dynamics
simulations of compound 2 bound either to hNTS1 or hNTS2.
For each receptor, we started from the conformation obtained
at the end of the trajectories described in Figures 1 and 2, e.g.,
after rotation of the Tyr11 residue of the native NT[8-13]
peptide. The later was then replaced by compound 2 by
replacing the two N-terminal Arg8 and Arg9 as well as the Tyr11
residue into lysine. For each peptide:receptor complex, a new
MD trajectory of 50 ns was then computed. As shown in Figure
4A,B, these trajectories confirmed that the Lys11 residue of
compound 2 was able to stably interact with Glu179 of hNTS2,
whereas no residue was able to play this role in hNTS1. On the
contrary, the charge:charge repulsion between Lys11 and Arg212
in the hNTS1 was leading to a rotation of this last residue.
To confirm these findings, a mutant of hNTS1 was expressed
in which the positively charged Arg212 residue was replaced by
the negatively charged Glu residue (hNTS1-R212E).
As shown in Figure 4C, MD simulations reinforced our
hypothesis suggesting that compound 2 could bind to the
hNTS1-R212E mutant and then interacts with the receptor’s
Glu residue, as shown for hNTS2. Affinities of compounds 2, 5,
6, and 7 were evaluated on the hNTS1-R212E mutant by
competition binding experiments using 125I-[Tyr3]-NT. As
shown in Figure 5 and Table 2, the binding affinity increased
for compounds 2 and 7 when tested on the hNTS1-R212E
mutant receptor. Compound 2 showed a more important gain
in binding affinity with a 28-fold increase, whereas compound 7
displayed a moderate affinity gain of 5-fold. As expected,
compounds 5 and 6 did not display any gain in binding affinity
on the R212E mutant, as compared to the wild type NTS1
receptor. These results indicated that the ionic interaction
F
DOI: 10.1021/acs.jmedchem.6b01848
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
receptors shared 88% and 44% sequence identity rate with the initial
template, respectively. Twenty models were built for each receptor, the
one sharing the best score being selected for further calculations. The
CHARMM-gui interface30 was then used to insert the NTS1 and
NTS2 human receptors in a 100 nm2 membrane bilayer composed of
100% POPC (1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine).
Water molecules and ions were then added to bring this model to
completion and neutralize the global charge of the system. Molecular
dynamics simulations were all performed with NAMD31 and the
CHARMM forces field.32 After energy minimization, equilibration was
completed in six several successive steps, as provided by CHARMM-
gui in NAMD scripts, using NVT dynamics for the two first
equilibration steps and then shifting to NPT for the four following
ones. Harmonic restraints applied to heavy atoms of the protein or
used to hold the position of lipid head groups were progressively
decreased for these six successive steps.33 For the production phase, an
integration step of 2 fs was used. Particle Mesh Ewald (PME) was
employed for the treatment of long-range interactions. For all
calculations, a classical cutoff scheme was used, electrostatic
interactions being truncated at 12 Å, applying a switching function
between 10 and 12 Å. After a total of 675 ps for equilibration, a
production trajectory of 180 ns was obtained for each receptor.
To improve sampling of the possible conformations of each
peptide:receptor complex, 20 ns of accelerated molecular dynamics
(aMD)34 were also performed. In this method, whenever the potential
decreased below a threshold value, a boost potential was added,
increasing the rate of conformational transitions. The modified
potential was V*(r) = V(r) + ΔV,
directly influences the binding affinity and plays a significant
role for the selectivity of NT ligands toward NTS1.
Docking of the different peptides was then performed in the
most representative conformations of the hNTS2 (20
conformations) binding pocket. First, the docking method
was validated on the NT[8-13] peptide, confirming the ability
of the PLANTS docking software in properly predicting the
binding mode of the native peptide into the native structure.
Our docking results confirmed that the Lys11 residue of
compounds 2 or 7 was able to form the expected salt bridge
with Glu179 in the hNTS2 receptor, while the conformation of
the peptide was very close to that of the native NT[8-13]
peptide (reported in yellow) (Figure 6A).
Our calculations also showed how the TMSAla moiety in
compound 7 fits quite nicely at the bottom of the binding
pocket. This was also confirmed by comparison with the
adamantane groups of SR142948A and SR48692, two non-
peptide compounds that showed binding affinity toward
NTS219−21 when docked using the same experimental protocol.
Again, in this model we also observed the Lys11:Glu179
interaction (Figure 6B).
CONCLUSION
■
In conclusion, we report here that the Tyr11 residue of the NT
peptide plays a crucial role for both affinity and selectivity
toward the hNTS1 and hNTS2 receptor subtypes. These
results are in accordance with previous studies supporting the
importance of aromatic residues in the structure−activity
relationship for the affinity toward NT receptors.11−14 The
extension of the aromaticity and changes in side chain
orientation seem to favor hNTS2 selectivity. Indeed, sub-
stitution of the L-Tyr in position 11 by unnatural D-amino acids
such as D-α-naphthylalanine is well-tolerated by NTS2 and
leads to substantial loss in NTS1 binding.22−24 Furthermore,
high NTS2 selectivity has also been observed with a peptide−
peptoid hybrid resulting from replacement of Tyr with N-
homotyramine12,25 and more recently with tetrahydrofuran
amino acids.26
⎧
⎪
⎨
⎪
⎩
0V(r) ≥ E
(E − V(r))2
ΔV =
V(r) < E
α + E − V(r)
E is the threshold, and α is the acceleration parameter that shaped the
modified potential.
We used a dual-boost scheme, with the application of boost
potentials to dihedral angles and to the total potential energy of the
entire system. Boost parameters were obtained according to the
protocol described by Miao et al., which is adequate for protein−
16
membrane systems:
E
= Vdihed_avg + 0.3 × Vdihed_avg, αdihed = 0.3 ×
dihed
Vdihed_avg/5, Etotal = Vtotal_avg + 0.2 × total number of atoms, αtotal = 0.2 ×
total number of atoms. The average values Vdihed_avg and Vtotal_avg were
obtained from the previous conventional MD runs (180 ns) and
restarting from the last structure of the respective conventional MD
simulation.
Peptide:Receptor Docking. Representative conformations of NTS1
and NTS2 receptor binding sites were obtained by using a clustering
procedure. In each case, after concatenation of the trajectories
(classical and accelerated dynamics), rotations and translations were
removed by least-squares fitting to the backbone atoms of the receptor.
Clustering of the fitted trajectory was based on receptor atoms closer
than 15 Å to the bound peptide in the reference crystal structure. It
was performed with the g_cluster tool of the Gromacs package35 using
the algorithm of Daura et. al36 and a 2.5 Å cutoff. The different
Finally, hydrophobicity of the C-terminal end is essential for
receptor binding. Indeed, as demonstrated by alanine scan
strategy, which sequentially replaced each amino acid of NT[8-
13] with alanine, the isopropyl group of leucine at position 13
has been found to be particularly important for the binding and
biological activity of the peptide.27 Accordingly, replacement of
Ile or/and Leu with more hydrophobic non-natural amino
acids, such as trimethylsilylalanine, results in improved
affinity.10,14 However, changing the L-configuration in position
13 and introducing β-amino acid to prevent exoprotease
cleavages appear to induce an important loss in receptor
binding.17 Altogether, these molecular modeling and binding
results will help to improve the design of the next-generation of
NTS2-selective pain-reliever drugs.
peptides were built with the Shrodinger Maestro software (http://
̈
www.schrodinger.com/) and converted to the.mol2 file format for
further docking calculations. Docking of the different peptides was
then performed in the two sets of representative conformations with
PLANTS37 using the ChemPLP scoring function. A sphere of 25 Å
around the residues Arg212 (NTS1) or Glu179 (NTS2) was used for
docking site definition, encompassing both the whole binding site of
the receptor and its whole extracellular surface. The 10 best poses
obtained for each peptide in each receptor conformation were then
conserved for further analyses.
EXPERIMENTAL SECTION
■
Computational Methods. Molecular Dynamics Dimulations.
Both the human NTS1 and NTS2 receptors were built from their
Swissprot sequences (sequences IDs P30989 and O95665) using the
MODELER software28 and the 4BUO PDB structure as a template.29
This structure describes the rat NTS1 receptor complexed to the
NT[8-13] agonist peptide (H-Lys-Lys-Pro-Tyr-Ile-Leu-OH) and was
chosen among of others because of a better resolution and a higher
level of completeness. The bound hexapeptide was included during the
whole homology-modeling step. The human NTS1 and NTS2
Chemistry. All solvents and reagents for the synthesis of
compounds 1−4 were purchased from Sigma-Aldrich, Fluka, and
Alfa Aesar in gradient grade or reagent quality. All reactions involving
air-sensitive reagents were performed under nitrogen or argon.
Purifications were performed with column chromatography using
G
DOI: 10.1021/acs.jmedchem.6b01848
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
silica gel (Merck 60, 230−400 mesh) or with a Biotage instrument
Isolera 4 using SNAP KP-SIL flash cartridges. Nuclear magnetic
resonance NMR spectra were recorded on a Bruker spectrometer
Avance 300 at 600 MHz. Chemicals shifts (δ) are reported from
tetramethylsilane with the solvent resonance as internal standard. LC/
MS system consisted of a Waters Alliance 2690 HPLC, coupled to a
ZQ spectrometer (Manchester, UK), fitted with an electrospray source
operated in the positive ionization mode (ESI+). All the analyses were
carried out using a C18 Chromolith Flash 25 mm × 4.6 mm column
operated at a flow rate of 3 mL/min. A gradient of 0% or 0.1%
aqueous TFA (solvent A) to 100% of acetonitrile containing 0.1%
TFA (solvent B) was developed over 3 min. Positive-ion electrospray
mass spectra were acquired at a solvent flow rate of 100−200 μL/min.
Nitrogen was used for both the nebulizing and drying gas. The data
were obtained in a scan mode ranging from 200 to 1700 m/z in 0.1 s
intervals. A total of 10 scans was summed up to get the final spectrum.
All the fully unprotected peptides were purified using a gradient
composed of water/acetonitrile with 0.1% TFA at 50 mL/min flow
rate performed on a PLC2020 Gilson using a 75 mm × 21.2 mm
Phenomenex Luna 5μ C18(2) column. Purity was determined by RP-
Analytic HPLC performed on a Agilent 1220 using a 50 mm × 4.6 mm
Chromolith high resolution column. Compounds were separated using
a linear gradient system (0−100% solvent B in 10 min) using a
constant flow rate of 3 mL·min−1. All peptides were obtained with
purity ≥95%. High-resolution mass spectra (HRMS) were performed
by the “Laboratoire de Mesures Physiques” of Montpellier University
on a Micromass Q-Tof spectrometer equipped with electrospray
source ionization (ESI), using phosphoric acid as internal standard.
General Procedures. General Procedure for Coupling Reaction
in Solution (Procedure 1). To a solution of the NH2-free amino ester
or peptide in DMF, was added the suitable N-protected amino acid or
peptide followed by addition of BOP (1 equiv) and N,N-
diisopropyethylamine (3 equiv). The mixture was stirred 24 h at
room temperature. The solvent was evaporated in vacuo and the
residue diluted with EtOAc and consecutively extracted with 1 M
KHSO4 (2×), aqueous NaHCO3 (2×), and brine (2×), dried over
Na2SO4, and the solvent evaporated under reduced pressure to afford
the crude product, which was then purified by chromatography on a
silica gel column and characterized by LC-MS.
DMF (3 × 15 mL). Two deprotection cycles were carried out using 10
mL of a 20% piperidine solution in DMF under microwave irradiation
at 40 W for 30 s at 75 °C. After each deprotection cycle, the solvent
was removed by filtration and the resin washed with DMF. For the
cleavage from the resin and final deprotection, the resin was suspended
in a mixture of TFA/TIS (99/1) and stirred for 3 h at room
temperature. The solution was filtered and the peptide precipitated
with ice-cold diethyl ether. After centrifugation and elimination of the
supernatant, the crude was purified by preparative HPLC.
H-Lys-Lys-Pro-Lys-Ile-Leu-OH (2). H-Lys-Lys-Pro-Lys-Ile-Leu-OH
was prepared on SPPS following procedure 5. ESI-MS: 726.7 [M +
H]+, 363.8 [M + 2H]2+. RP-LC: tR = 2.60 min. HRMS: calcd For
C35H68N9O7 [M + H]+, 726.5244; found, 726.5242.
H-Lys-Lys-Pro-Orn-Ile-Leu-OH (3). H-Lys-Lys-Pro-Orn-Ile-Leu-OH
was prepared on SPPS following procedure 5. ESI-MS: 712.5 [M +
H]+, 356.7 [M + 2H]2+. RP-LC: tR = 2.57 min. HRMS: calcd For
C34H66N9O7 [M + H]+, 712.5085; found, 712.5081.
H-Lys-Lys-Pro-His-Ile-Leu-OH (4). H-Lys-Lys-Pro-His-Ile-Leu-OH
was prepared on SPPS following procedure 5. ESI-MS: 735.4 [M +
H]+, 368.2 [M + 2H]2+. RP-LC: tR = 2.98 min. HRMS: calcd For
C35H63N10O7 [M + H]+, 735.4881; found, 735.4875.
H-Lys-Lys-Pro-Asp-Ile-Leu-OH (5). H-Lys-Lys-Pro-Asp-Ile-Leu-OH
was prepared on SPPS following procedure 5. ESI-MS: 713.4 [M +
H]+. RP-LC: tR = 2.74 min. HRMS: calcd For C33H61N8O9 [M +
H]+, 713.4568; found, 713.4562.
H-Lys-Lys-Pro-Glu-Ile-Leu-OH (6). H-Lys-Lys-Pro-Glu-Ile-Leu-OH
was prepared on SPPS following procedure 5. ESI-MS: 727.3 [M +
H]+. RP-LC: tR = 2.71 min. HRMS: calcd For C34H63N8O9 [M +
H]+, 727.4722; found, 727.4726.
H-Lys-Lys-Pro-Lys-Ile-TMSAla-OH (7). Boc-Ile-TMSAla-OMe. TFA·
H-TMSAla-OH (223 mg, 0.67 mmol) was dissolved in methanol (6
mL), thionyl chloride was added (245 μL, 5 equiv), and the mixture
stirred at reflux 4 h 30 min. The solvent was removed under reduced
pressure to afford H-TMSAla-OMe·HCl, which was used without
further purification for the coupling with Boc-Ile-OH·1/2H2O (160
mg, 1 equiv) following the general procedure 1. The dipeptide Boc-Ile-
TMSAla-OMe was obtained as a white foam (220 mg, 83%). Rf: 0.4
(cyclohexane/ethyl acetate 7:3). ESI-MS: 389.4 [M + H]+. RP-LC: tR
= 1.92 min.
Z-Lys(Boc)-Ile-TMSAla-OMe. Boc-Ile-TMSAla-OMe (350 mg, 0.9
mmol) was deprotected following general procedure 3. The TFA salt
was directly coupled with Z-Lys(Boc)-OH following the general
procedure 1. The tripeptide Z-Lys(Boc)-Ile-TMSAla-OMe was
obtained as a white foam (375 mg, 64%). Rf: 0.2 (cyclohexane/ethyl
acetate 6:4). ESI-MS: 651.4 [M + H]+. RP-LC: tR = 1.98 min.
Boc-Lys(Z)-Lys(Z)-Pro-Lys(Boc)-Ile-TMSAla-OMe. Z-Lys(Boc)-Ile-
TMSAla-OMe (170 mg, 0.26 mmol) was deprotected following
general procedure 2. Without further purification it was directly
coupled with Boc-Lys(Z)-Lys(Z)-Pro-OH (1 equiv) following the
general procedure 1. The hexapeptide was obtained as a white foam
(56 mg, 17%). Rf: 0.47 (dichloromethane/methanol 9:1). ESI-MS:
1238.8 [M + H]+. RP-LC: tR = 2.20 min.
General Procedure for Z Deprotection of Peptides in Solution
(Procedure 2). The protected peptide was dissolved in methanol (5
mL/mmol) in the presence of 1 M HCl (1 equiv) and 10% Pd/C
under hydrogen atmosphere. The mixture was stirred for 3 h at room
temperature. The mixture was filtered on a pad of Celite, which was
washed with methanol. The solvent was evaporated under reduced
pressure to afford the free amine derivative.
General Procedure for Boc Deprotection of Peptides in Solution
(Procedure 3). The protected peptide was dissolved in TFA (4 mL/
mmol). The mixture was stirred for 2.5 h at room temperature.
Volatiles were removed under reduced pressure. To precipitate the
compound, diethyl ether was added and evaporated to afford the
corresponding TFA salt.
General Procedure for Saponification of Methyl Esters
(Procedure 4). The peptide methyl ester was dissolved in methanol
(10 mL/mmol), and a solution of 4N KOH (5 equiv) was added. The
reaction was stirred until disappearance of the starting material. The
solvent was removed under reduced pressure and the residue dissolved
in water (10 mL/mmol) and acidified with citric acid 15% to pH 4−5.
The solution was extracted with ethyl acetate (4×), and the combined
organic phases were dried over Mg2SO4 and volatiles removed under
reduced pressure to afford the desired carboxylic acid derivative, which
was used without further purification.
Boc-Lys(Z)-Lys(Z)-Pro-Lys(Boc)-Ile-TMSAla-OH. Starting from Boc-
Lys(Z)-Lys(Z)-Pro-Lys(Boc)-Ile-TMSAla-OMe (56 mg, 0.045 mmol),
Boc-Lys(Z)-Lys(Z)-Pro-Lys(Boc)-Ile-TMSAla-OH (55 mg, 100%)
was obtained from following general procedure 4 as a white foam.
ESI-MS: 1224.9 [M + H]+. RP-LC: tR = 2.07 min.
Boc-Lys-Lys-Pro-Lys(Boc)-Ile-TMSAla-OH. Starting from Boc-Lys-
(Z)-Lys(Z)-Pro-Lys(Boc)-Ile-TMSAla-OH (55 mg, 0.045 mmol),
Boc-Lys-Lys-Pro-Lys(Boc)-Ile-TMSAla-OH (43 mg, 99%) was
obtained from following general procedure 2. ESI-MS: 956.6 [M +
H]+. RP-LC: tR = 1.18 min.
General Procedure for Solid Phase Peptide Synthesis (Procedure
5). All peptides were prepared using a Liberty CEM microwave-
assisted peptide synthesizer at a 0.25 mmol scale on a Wang resin
preloaded with Fmoc-leucine (loading 0.8 mmol/g). All Fmoc-amino
acids (4 equiv) were coupled by using a solution of HATU in DMF (4
equiv) as coupling reagent and a solution of DIPEA (10 equiv) in
DMF under microwave irradiation at 40 W for 300 s at 70 °C. The
reactant and solvent were then filtered and the resin washed with
H-Lys-Lys-Pro-Lys-Ile-TMSAla-OH (7). Compound 4 was obtained
starting from Boc-Lys-Lys-Pro-Lys(Boc)-Ile-TMSAla-OH (43 mg,
0.044 mmol) by following the general procedure 3 and purified by
preparative HPLC. ESI-MS: 756.6 [M + H]+. RP-LC: tR = 2.87 min.
HRMS: calcd For C35H70N9O7 [M + H]+, 756.5158; found,
756.5167.
In Vitro Procedures. Generation of HA-hNTS1-R212E by Site-
Directed Mutagenesis. The cDNA clone of the human HA-tagged
H
DOI: 10.1021/acs.jmedchem.6b01848
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
NTS1 receptor was purchased from cdna.org (St-Louis, MO). NTS1
mutant R212E was generated using the Q5 Site-directed mutagenesis
kit from New England Biolabs (Ipswich, MA) following manufac-
turer’s recommendations. Primers used to introduce the mutation
were synthesized by Integrated DNA Technologies (Coralville, IA)
and are described in Table 3 (changed bases in primer’s sequence are
in lower case). Construct was verified by DNA sequencing.
where L refers to the concentration of radiolabeled tracer (125I-[Tyr3]-
NT) and Kd refers to the equilibrium dissociation constant of the
radioligand.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jmed-
chem.6b01848.
Table 3
1
primer
sequence
HPLC analyses, HRMS spectra for compounds 2−7, H
NMR and 13C NMR spectra of compounds 2 and 7, and
saturation binding curves of hNTS1-WT and hNTS1-
R212E receptors (PDF)
forward
reverse
5′-C GAG CAG AAC gaa AGC GCC GAC G-3′
5′-C CCA TGG TGA ACA GCA TAG-3′
Molecular formula strings (CSV)
NTS1 (PDB)
NTS2 (PDB)
Cell Culture and Transfections. CHO-K1 cells stably expressing
hNTS1 and 1321N1 cells stably expressing hNTS2 were cultured
respectively in Ham’s F12 and DMEM. Culture media were
supplemented with 10% FBS, 100 U/mL penicillin, 100 μg/mL
streptomycin, 20 mM HEPES, and 0.4 mg/mL G418 at 37 °C in a
humidified chamber at 5% CO2.
HEK293 cells were cultured in DMEM supplemented with 10%
FBS, 100 U/mL penicillin, 100 μg/mL streptomycin, and 20 mM
HEPES. HEK293 were seeded at 2.2 × 106 cells in a 10 cm Petri dish
and transfected 24 h after plating with 10 μg of DNA coding for the
human HA-tagged NTS1-R212E receptor using PEI as a transfection
agent. Membrane preparation was conducted 48 h after transfection, as
described below.
Competitive Radioligand Binding Assay on WT Neurotensin
Receptors. Binding experiments were carried out on freshly prepared
membranes, as previously described.10 Briefly, competitive radioligand
binding experiments were performed by incubating 15 μg of cell
membranes expressing the hNTS1 receptor with 45 pM of 125I-[Tyr3]-
NT (2200 Ci/mmol) or 50 μg of cell membranes expressing the
hNTS2 receptor with 255 pM of 125I-[Tyr3]-NT (2200 Ci/mmol) in
binding buffer (50 mM Tris-HCl, pH7.5, 0.2% BSA) in the presence of
increasing concentrations of the analogues ranging from 10−11 to 10−5
M for 30 min at 25 °C. After incubation, the binding reaction mixture
was transferred in polyethylenimine-coated 96-well filter plates (glass
fiber filters GF/B, Millipore, Billerica, MA). Reaction was terminated
by filtration, and plates were washed three times with 200 μL of ice-
cold binding buffer. Glass filters were then counted in a γ-counter
(2470 Wizard2, PerkinElmer, Missisauga, Ontario, Canada). Non-
specific binding was measured in the presence of 10−5 M unlabeled
NT [8-13] and represented less than 5% of total binding. IC50 values
were determined from the competition curves as the unlabeled ligand
concentration inhibiting half of the 125I-[Tyr3]-NT-specific binding.
Data were plotted using GraphPad Prism 6 using the One-site-Fit
Log(IC50) and represented the mean SEM of at least three separate
experiments.
AUTHOR INFORMATION
Corresponding Author
*Phone: +33 467 143 765. Fax: +33 467 144 866. E-mail:
florine.cavelier@umontpellier.fr.
ORCID
Nicolas Floquet: 0000-0002-3883-2852
Jean Martinez: 0000-0002-4551-4254
Florine Cavelier: 0000-0001-5308-6416
■
Author Contributions
∥These authors contributed equally to directing this study.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was granted access to the HPC resources of CINES
under the allocation no. c2016077530. This work was also
realized with the support of HPC@LR, a Center of
Competence in High-Performance Computing from the
Languedoc-Roussillon region, funded by the Languedoc-
́
Roussillon region, the Europe and the Universite de
Montpellier. The HPC@LR Center is equipped with an IBM
hybrid supercomputer. P.S. is the recipient of the Canada
Research Chair in Neurophysiopharmacology of Chronic Pain.
This work was supported by research funding from Montpellier
University (F.C. for postdoctoral fellowship to R.F., N.F. for
Ph.D. fellowship to B.D.) and the Canadian Institutes of Health
Research (P.S.). The ANR BIP-BIP is acknowledged for
financial support of P.R.
Competitive Radioligand Binding Assay on the hNTS1-R212E
Receptor. Binding experiments were carried out on freshly prepared
membranes 48 h after transfection. Competitive radioligand binding
experiments were performed by incubating 50 μg of cell membranes
expressing the hNTS1-R212E receptor with 130 pM of 125I-[Tyr3]-NT
(2200 Ci/mmol) in binding buffer (50 mM Tris-HCl, pH7.5, 0.2%
BSA). All other steps in binding experiments were unchanged
comparing to the one using the WT receptors.
Data Analysis. Competitive radioligand binding data were plotted
using Prism 7 (GraphPad, La Jolla, CA) using the One-site-Fit
Log(IC50) and represented the mean SEM of at least three separate
experiments.
ABBREVIATIONS USED
■
BOP, (benzotriazol-1-yloxy)tris(dimethylamino)phosphonium
hexafluorophosphate; DMF, N,N-dimethylformamide; TMSA-
la, (L)-(trimethylsilyl)alanine; NT, neurotensin; NTS1, neuro-
tensin receptor subtype 1; NTS2, neurotensin receptor subtype
2; TFA, trifluoroacetic acid; DIPEA, N,N-diisopropylethyl-
amine; Z, benzyloxycarbonyl; Boc, tert-butyloxycarbonyl; Fmoc,
9-fluorenyl-methoxycarbonyl; HATU, 1-[bis(dimethylamino)-
methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium 3-oxide hexa-
fluorophosphate
IC50 calculated from the competitive radioligand binding assays
were then transformed into Ki using the Cheng−Prusoff equation:38
REFERENCES
■
IC50
(1) Trang, T.; Al-Hasani, R.; Salvemini, D.; Salter, M. W.; Gutstein,
H.; Cahill, C. M. Pain and Poppies: The good, the bad, and the ugly of
opioid analgesics. J. Neurosci. 2015, 35, 13879−13888.
Ki =
[L]
1 +
(
)
Kd
I
DOI: 10.1021/acs.jmedchem.6b01848
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
(2) Carraway, R.; Leeman, S. E. The isolation of a new hypotensive
peptide, neurotensin, from bovine hypothalami. J. Biol. Chem. 1973,
248, 6854−6861.
(3) Dubuc, I.; Costentin, J.; Doulut, S.; Rodriguez, M.; Martinez, J.;
Kitabgi, P. JMV 449: a pseudopeptide analogue of neurotensin-(8-13)
with highly potent and long-lasting hypothermic and analgesic effects
in the mouse. Eur. J. Pharmacol. 1992, 219, 327−9.
(4) Nemeroff, C. B.; Osbahr, A. J., 3rd; Manberg, P. J.; Ervin, G. N.;
Prange, A. J., Jr. Alterations in nociception and body temperature after
intracisternal administration of neurotensin, beta-endorphin, other
endogenous peptides, and morphine. Proc. Natl. Acad. Sci. U. S. A.
1979, 76, 5368−5371.
(19) Rene, A.; Vanthuyne, N.; Martinez, J.; Cavelier, F. (L)-
(Trimethylsilyl)alanine synthesis exploiting hydroxypinanone-induced
diastereoselective alkylation. Amino Acids 2013, 45, 301−317.
(20) Gully, D.; Canton, M.; Boigegrain, R.; Jeanjean, F.; Molimard, J.
C.; Poncelet, M.; Gueudet, C.; Heaulme, M.; Leyris, R.; Brouard, A.
Biochemical and pharmacological profile of a potent and selective
nonpeptide antagonist of the neurotensin receptor. Proc. Natl. Acad.
Sci. U. S. A. 1993, 90, 65−69.
(21) Schaeffer, P.; Laplace, M. C.; Bernat, A.; Prabonnaud, V.; Gully,
D.; Lespy, L.; Herbert, J. M. SR142948A is a potent antagonist of the
cardiovascular effects of neurotensin. J. Cardiovasc. Pharmacol. 1998,
31, 545−550.
(22) Richelson, E.; Mccormick, D. J.; Pang, Y. P.; Phillips, K. S.
Peptide analogs that are potent and selective for human neurotensin
receptor subtype 2. U.S. Pat. Appl. Publ. WO 2008137720, November
13, 2008.
(5) Skoog, K. M.; Cain, S. T.; Nemeroff, C. B. Centrally administered
neurotensin suppresses locomotor hyperactivity induced by d-
amphetamine but not by scopolamine or caffeine. Neuropharmacology
1986, 25, 777−782.
(23) Tyler, B. M.; Douglas, C. L.; Fauq, A.; Pang, Y.-P.; Stewart, J. A.;
Cusack, B.; McCormick, D. J.; Richelson, E. In vitro binding and CNS
effects of novel neurotensin agonists that cross the blood−brain
barrier. Neuropharmacology 1999, 38, 1027−1034.
(6) Kleczkowska, P.; Lipkowski, A. W. Neurotensin and neurotensin
receptors: characteristic, structure-activity relationship and pain
modulation-a review. Eur. J. Pharmacol. 2013, 716, 54−60.
(7) Tetreault, P.; Beaudet, N.; Perron, A.; Belleville, K.; Rene, A.;
Cavelier, F.; Martinez, J.; Stroh, T.; Jacobi, A. M.; Rose, S. D.; Behlke,
M. A.; Sarret, P. Spinal NTS2 receptor activation reverses signs of
neuropathic pain. FASEB J. 2013, 27, 3741−3752.
(8) Coutant, J.; Curmi, P. A.; Toma, F.; Monti, J. P. NMR solution
structure of neurotensin in membrane-mimetic environments:
molecular basis for neurotensin receptor recognition. Biochemistry
2007, 46, 5656−5663.
(9) Granier, C.; Van Rietschoten, J.; Kitabgi, P.; Poustis, C.; Freychet,
P. Synthesis and characterization of neurotensin analogues for
structure/activity relationship studies. Eur. J. Biochem. 1982, 124,
(24) Dubuc, I.; Sarret, P.; Labbe-Jullie, C.; Botto, J. M.; Honore, E.;
Bourdel, E.; Martinez, J.; Costentin, J.; Vincent, J. P.; Kitabgi, P.;
Mazella, J. Identification of the receptor subtype involved in the
analgesic effect of neurotensin. J. Neurosci. 1999, 19, 503−510.
(25) Held, C.; Plomer, M.; Hubner, H.; Meltretter, J.; Pischetsrieder,
̈
M.; Gmeiner, P. Development of a metabolically stable neurotensin
receptor 2 (NTS2) ligand. ChemMedChem 2013, 8, 75−81.
(26) Simeth, N. A.; Bause, M.; Dobmeier, M.; Kling, R. C.;
Lachmann, D.; Hubner, H.; Einsiedel, J.; Gmeiner, P.; Konig, B.
̈
̈
NTS2-selective neurotensin mimetics with tetrahydrofuran amino
acids. Bioorg. Med. Chem. 2017, 25, 350−359.
117−125.
́
(27) Heyl, D. L.; Sefler, A. M.; He, J. X.; Sawyer, T. K.; Wustrow, D.
J.; Akunne, H. C.; Davis, M. D.; Pugsley, T. A.; Heffner, T. G.; Corbin,
A. E.; et al. Structure-activity and conformational studies of a series of
modified C-terminal hexapeptide neurotensin analogues. Int. J. Pept.
Protein Res. 1994, 44, 233−238.
(10) Fanelli, R.; Besserer-Offroy, E.; Rene,
́
A.; Co
̂
te,
́
J.; Tet
́
reault, P.;
́
Collerette-Tremblay, J.; Longpre, J.-M.; Leduc, R.; Martinez, J.; Sarret,
P.; Cavelier, F. Synthesis and characterization in vitro and in vivo of
(L)-(trimethylsilyl)alanine containing neurotensin analogues. J. Med.
Chem. 2015, 58, 7785−7795.
(28) Webb, B.; Sali, A. Comparative protein structure modeling using
MODELLER. Curr. Protoc. Bioinf. 2014, 47, 5.6.1−5.6.32.
(29) Egloff, P.; Hillenbrand, M.; Klenk, C.; Batyuk, A.; Heine, P.;
(11) Hapau
̆
, D.; Rem
́
ond, E.; Fanelli, R.; Vivancos, M.; Rene,
́
A.;
́
Cote, J.; Besserer-Offroy, E.; Longpre,
̂
́
́
J.-M.; Martinez, J.; Zaharia, V.;
Sarret, P.; Cavelier, F. Stereoselective synthesis of β-(5-arylthiazolyl)
α-amino acids and use in neurotensin analogues. Eur. J. Org. Chem.
2016, 2016, 1017−1024.
Balada, S.; Schlinkmann, K. M.; Scott, D. J.; Schutz, M.; Pluckthun, A.
̈
̈
Structure of signaling-competent neurotensin receptor 1 obtained by
directed evolution in Escherichia coli. Proc. Natl. Acad. Sci. U. S. A.
2014, 111, E655−E662.
(12) Einsiedel, J.; Held, C.; Hervet, M.; Plomer, M.; Tschammer, N.;
Hubner, H.; Gmeiner, P. Discovery of highly potent and neurotensin
̈
(30) Lee, J.; Cheng, X.; Swails, J. M.; Yeom, M. S.; Eastman, P. K.;
Lemkul, J. A.; Wei, S.; Buckner, J.; Jeong, J. C.; Qi, Y.; Jo, S.; Pande, V.
S.; Case, D. A.; Brooks, C. L.; MacKerell, A. D.; Klauda, J. B.; Im, W.
CHARMM-GUI input generator for NAMD, GROMACS, AMBER,
OpenMM, and CHARMM/OpenMM simulations using the
CHARMM36 additive force field. J. Chem. Theory Comput. 2016, 12,
405−413.
receptor 2 selective neurotensin mimetics. J. Med. Chem. 2011, 54,
2915−2923.
(13) Boules, M.; Johnston, H.; Tozy, J.; Smith, K.; Li, Z.; Richelson,
E. Analgesic synergy of neurotensin receptor subtype 2 agonist NT79
and morphine. Behav. Pharmacol. 2011, 22, 573−581.
(14) Boules, M.; Shaw, A.; Liang, Y.; Barbut, D.; Richelson, E.
NT69L, a novel analgesic, shows synergy with morphine. Brain Res.
2009, 1294, 22−38.
(31) Phillips, J. C.; Braun, R.; Wang, W.; Gumbart, J.; Tajkhorshid,
́
E.; Villa, E.; Chipot, C.; Skeel, R. D.; Kale, L.; Schulten, K. Scalable
molecular dynamics with NAMD. J. Comput. Chem. 2005, 26, 1781−
1802.
(15) White, J. F.; Noinaj, N.; Shibata, Y.; Love, J.; Kloss, B.; Xu, F.;
Gvozdenovic-Jeremic, J.; Shah, P.; Shiloach, J.; Tate, C. G.;
Grisshammer, R. Structure of the agonist-bound neurotensin receptor.
Nature 2012, 490, 508−513.
(16) Miao, Y.; Nichols, S. E.; Gasper, P. M.; Metzger, V. T.;
McCammon, J. A. Activation and dynamic network of the M2
muscarinic receptor. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 10982−
10987.
(32) Brooks, B. R.; Brooks, C. L.; Mackerell, A. D.; Nilsson, L.;
Petrella, R. J.; Roux, B.; Won, Y.; Archontis, G.; Bartels, C.; Boresch,
S.; Caflisch, A.; Caves, L.; Cui, Q.; Dinner, A. R.; Feig, M.; Fischer, S.;
Gao, J.; Hodoscek, M.; Im, W.; Kuczera, K.; Lazaridis, T.; Ma, J.;
Ovchinnikov, V.; Paci, E.; Pastor, R. W.; Post, C. B.; Pu, J. Z.; Schaefer,
M.; Tidor, B.; Venable, R. M.; Woodcock, H. L.; Wu, X.; Yang, W.;
York, D. M.; Karplus, M. CHARMM: the biomolecular simulation
program. J. Comput. Chem. 2009, 30, 1545−1614.
(17) Kitabgi, P.; Poustis, C.; Granier, C.; Van Rietschoten, J.; Rivier,
J.; Morgat, J. L.; Freychet, P. Neurotensin binding to extraneural and
neural receptors: comparison with biological activity and structure−
activity relationships. Mol. Pharmacol. 1980, 18, 11−19.
́
(33) Wu, E. L.; Cheng, X.; Jo, S.; Rui, H.; Song, K. C.; Davila-
(18) Rem
́
ond, E.; Martin, C.; Martinez, J.; Cavelier, F. Silicon-
Contreras, E. M.; Qi, Y.; Lee, J.; Monje-Galvan, V.; Venable, R. M.;
Klauda, J. B.; Im, W. CHARMM-GUI membrane builder toward
realistic biological membrane simulations. J. Comput. Chem. 2014, 35,
1997−2004.
containing amino acids: synthetic aspects, conformational studies, and
applications to bioactive peptides. Chem. Rev. 2016, 116, 11654−
11684.
J
DOI: 10.1021/acs.jmedchem.6b01848
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
(34) Hamelberg, D.; Mongan, J.; McCammon, J. A. Accelerated
molecular dynamics: a promising and efficient simulation method for
biomolecules. J. Chem. Phys. 2004, 120, 11919−11929.
(35) Hess, B.; Kutzner, C.; van der Spoel, D.; Lindahl, E. GROMACS
4: algorithms for highly efficient, load-balanced, and scalable molecular
simulation. J. Chem. Theory Comput. 2008, 4, 435−447.
(36) Daura, X.; Jaun, B.; Seebach, D.; van Gunsteren, W. F.; Mark, A.
E. Reversible peptide folding in solution by molecular dynamics
simulation1. J. Mol. Biol. 1998, 280, 925−932.
(37) Korb, O.; Stutzle, T.; Exner, T. E. Empirical scoring functions
̈
for advanced protein−ligand docking with PLANTS. J. Chem. Inf.
Model. 2009, 49, 84−96.
(38) Cheng, Y.; Prusoff, W. H. Relationship between the inhibition
constant (K1) and the concentration of inhibitor which causes 50%
inhibition (I50) of an enzymatic reaction. Biochem. Pharmacol. 1973,
22, 3099−3108.
K
DOI: 10.1021/acs.jmedchem.6b01848
J. Med. Chem. XXXX, XXX, XXX−XXX