Covalently induced activation of the δ opioid receptor by a fluorogenic affinity label, 7′-(phthalaldehydecarboxamido) naltrindole (PNTI) [1]

Full text of DOI:10.1021/jm010004u

© Copyright 2001 by the American Chemical Society
Volume 44, Number 7
March 29, 2001
Letters
Cova len tly In d u ced Activa tion of th e δ
Op ioid Recep tor by a F lu or ogen ic
Affin ity La bel, 7′-(P h th a la ld eh yd e-
ca r boxa m id o)n a ltr in d ole (P NTI)
Bertrand Le Bourdonnec,† Rachid El Kouhen,‡
Gennady Poda,† Ping Y. Law,‡ Horace H. Loh,‡
David M. Ferguson,† and Philip S. Portoghese*,†
Department of Medicinal Chemistry, College of Pharmacy,
and Department of Pharmacology, School of Medicine,
University of Minnesota, Minneapolis, Minnesota 55455
Received J anuary 4, 2001
The prototypical nonpeptide δ opioid receptor antago-
nist, naltrindole1 (1, NTI), is widely employed as a
pharmacologic tool in opioid research. It is one of the
most potent δ antagonists, with potency in the sub-
nanomolar range.2 Here we report that modification of
NTI (1) with an electrophilic o-phthalaldehyde (OPA)
moiety to afford 2 (PNTI) results in an unprecedented
change in activity from antagonist to a highly potent
agonist upon covalent binding to the δ opioid receptor.
We have previously reported on a new approach to
the design of an affinity label 3 whose OPA moiety binds
covalently to opioid receptors.3 The covalent binding
involves the generation of an isoindole fluorophore,4
presumably as a consequence of cross-linking neighbor-
ing lysine and cysteine residues by the OPA moiety.
This class of fluorogenic ligands has been termed
“reporter affinity labels” because the generation of
fluorescence implicates covalent cross-linking to specific
amino acid residues.3
group may be in the vicinity of Lys214 and Cys216
residues located on TM 5 close to the extracellular
surface of the δ opioid receptor (Figure 1A).
PNTI (2) and its nonelectrophilic analogue 4 were
synthesized from 7′-aminonaltrindole7 (5) as illustrated
in Scheme 1. Coupling of 5 with benzoic acid or 3,4-bis-
[2-(1,3-dioxolanyl)]benzoic acid3 in the presence of 1-hy-
droxybenzotriazole and dicyclohexylcarbodiimide gave
the corresponding amido esters 6 or 7 which were
selectively hydrolyzed with potassium carbonate to
afford the amides 4 or 8. Target compound 2 was then
obtained by hydrolysis of 8 under acidic conditions.
Receptor binding of 2 to membranes from CHO cells
that possessed stably expressed µ (83 fmol/mg protein),
δ (51 fmol/mg protein), and κ (36 fmol/mg protein) opioid
receptors revealed that it was δ-selective, with apparent
Ki8 values of 14.9 ( 0.8, 1.9 ( 0.1, and 20.1 ( 0.8 nM
for µ, δ, and κ receptors, respectively. Pretreatment of
δ receptors with low (20 nM) or high (1 µM) concentra-
tions of 2 (37°C, 15 min incubation in HEPES buffer,
pH ) 7.5) followed by extensive washing reduced the
binding of [3H]diprenorphine in a concentration-depend-
ent fashion (Figure 2). In contrast, pretreatment of δ
The rationale for the design of PNTI (2) was based
on prior structure-activity relationship (SAR) studies
which revealed that 7′-substitution on the indole ring
of NTI is well-tolerated.2,5,6 Also, computer-aided dock-
ing of PNTI onto the δ opioid receptor revealed that an
OPA moiety attached to NTI through a 7′-carboxamide
* To whom correspondence should be addressed. Tel: 612-624-9174.
Fax: 612-626-6891. E-mail: porto001@tc.umn.edu.
† Department of Medicinal Chemistry, College of Pharmacy.
‡ Department of Pharmacology, School of Medicine.
10.1021/jm010004u CCC: $20.00 © 2001 American Chemical Society
Published on Web 02/28/2001

1018 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 7
Letters
F igu r e 1. Modeled interaction between PNTI (2) (green) and the δ opioid receptor (orange) before (A) and after (B) covalent
binding. (A) The OPA moiety of 2 is hydrogen-bonded to the ꢀ-NΗ3(+) group of Lys214 as a reversibly bound complex. (B) Formation
of the fluorescent isoindole after covalent binding of the OPA moiety to the side chains of Lys214 and Cys216 (magenta) in TM
5 leads to receptor activation triggered by clockwise rotation (when viewed from the extracellular side) of TM 5.
Sch em e 1a
a
Reagents and conditions: (a) benzoic acid or 3,4-bis[2-(1,3-dioxolanyl)]benzoic acid, DCC, HOBt, DMF, rt, 5 days, 72-67%; (b) K2CO3,
CH3OH, rt, 1 h, 52-48%; (c) 1 N HCl, N2, acetone, rt, 5 days, 67%.
opioid receptors with naloxone (1 µM) followed by
washing did not reduce radioligand binding. The non-
electrophilic analogue 4 could be partially washed out
from the membrane preparation.9
µM) with nontransfected CHO cells. This permitted us
to visualize the specific fluorescent staining of δ opioid
receptors (Figure 3).
The in vitro pharmacological profile of 2 was inves-
tigated on the electrically stimulated mouse vas defer-
ens10 (MVD) and the guinea-pig ileal longitudinal
muscle11(GPI) preparations as previously described.12
Compound 2 behaved as an extremely potent full
agonist in the MVD assay, with an IC50 of 0.12 ( 0.03
nM. This represented a 2-fold greater potency than the
prototypical δ agonist [D-Ala2,D-Leu5]enkephalin13 (DA-
DLE) when tested in the same tissue.14 This potent
agonist effect of 2 could not be reversed by extensive
washing. The finding that pretreatment with NTI (100
nM) in the MVD afforded a 160-fold shift of the
Covalent binding of 2 to δ opioid receptors via the
formation of a fluorescent isoindole was established
using flow cytometry. This was accomplished with a
Becton Dickinson Facs Vantage equipped with a mul-
tiwavelength UV laser for excitation using a band-pass
filter of 530 ( 30 nm as described previously.3 When
CHO cells with stably expressed δ opioid receptors were
incubated with 2 (1 µM) in HEPES buffer (pH ) 7.5),
the fluorescence intensity rapidly (50 s) increased rela-
tive to the autofluorescence of untreated cells. Nonspe-
cific fluorescence was evaluated by incubation of 2 (1

Letters
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 7 1019
amine17 (â-FNA) (20 µmol/kg), and norbinaltorphimine18
(nor-BNI) (12 µmol/kg), which are selective antagonists
for δ, µ, and κ opioid receptors, respectively. The
antinociception produced by 2 was antagonized by
pretreatment with NTI [ED50 ratio ) 13.9919 (10.60-
18.64)] but only marginally by â-FNA [ED50 ratio ) 2.67
(1.65-4.09)] or nor-BNI [ED50 ratio ) 2.14 (1.19-3.64)].
These data confirmed that the antinociceptive action of
2 was mediated primarily through δ opioid receptors.
Antinociceptive activity was undetectable after 90
min. Given that continued exposure of the δ opioid
receptor20 to agonist gives rise to rapid, reduced respon-
siveness as a consequence of desensitization and down-
regulation, antinociceptive testing of the δ agonist,
[D-Pen2,D-Pen5]enkephalin21 (DPDPE), was conducted
2 h after administration of 2 (2 nmol icv), when the tail-
flick latency time had returned to normal values. Under
these conditions, the antinociceptive effect of DPDPE
was diminished, as indicated by its ED50 ratio of 4.98
(3.99-6.30). We believe that this “antagonism” was a
consequence of desensitization and down-regulation
caused by the persistent agonism due to covalent
binding of 2 to the δ opioid receptor recognition site.
F igu r e 2. Irreversible binding of 2 to the δ opioid receptor.
Membranes of CHO cells stably transfected with the δ opioid
receptor were pretreated with 2 (20 nM or 1 µM) or 4 (20 nM
or 1 µM) at 37 °C for 15 min. Free receptor sites were
determined in the presence of [3H]diprenorphine (1 nM) before
and after wash. Data are reported as percent of the total [3H]-
diprenorphine binding of the treated sample. The values
represent mean ( SE of three independent experiments
performed in triplicate.
In conclusion, PNTI (2) has been shown to bind
covalently to cloned δ opioid receptors, as evidenced by
the generation of fluorescence that is characteristic of
isoindole cross-linked lysine and cysteine residues. That
PNTI acts as a potent δ opioid receptor agonist is
counterintuitive, given that affinity labels derived from
potent antagonists generally exhibit irreversible an-
tagonist activity.1
Since its nonelectrophilic analogue 4 did not display
agonist activity in the MVD preparation, it appears
likely that covalent binding to the δ receptor by PNTI
(2) may be responsible for promoting a conformational
change of the receptor from an inactive to an active
state. A possible explanation for this unusual transfor-
mation is that the isoindole that is formed from covalent
binding of PNTI to the neighboring Lys214 and Cys216
residues (Figure 1B) promotes a conformational change
of TM 5. Recently, evidence for a rigid body motion of
TM 6 has been observed in light-activated bacterior-
hodopsin22 and constitutively activated G protein-
coupled receptors.23 The involvement of TM 5 in the
activation of the R2-adrenergic receptor has been also
reported.24 If perturbation of TM 5 or TM 6 leads to a
conformational change of intracellular loop 3, which is
known to be involved in the activation of G proteins, it
is possible that the δ agonism of 2 is initiated through
a similar mechanism. Accordingly, cross-linking of the
neighboring Lys214 and Cys216 residues may lead to
axial rotation of TM 5 due to the torsion created by the
formation of the isoindole fluorophore, as illustrated in
Figure 1B. PNTI should be a useful tool to investigate
the δ receptor recognition site and the conformational
transitions that take place in receptor activation.
F igu r e 3. Representative flow cytometric analysis of fluores-
cent opioid labeling of CHO cells. Untransfected CHO cells
(-DOR) or CHO cells transfected with the δ opioid receptor
(+DOR) were incubated with (+2) or without (-2) compound
2 (1 µM) at 25 °C for 50 s. The median fluorescence intensity
values for each curve are as follows: green (2.71), blue (4.26),
red (7.84).
concentration-effect curve of 2 suggested that it acted
through δ receptors. Compound 4 did not display
significant agonist activity (3.4% inhibition at 100 nM)
and was a weak antagonist (DADLE IC50 ratio ) 2.6)
in the MVD.9,15 In the GPI, PNTI (2) (100 nM) did not
display any agonist activity and only weak antagonist
activity at µ (morphine IC50 ratio ) 2.5) and κ (eth-
ylketazocine IC50 ratio ) 2.1) receptors. Thus, the
bioassay experiments have shown that 2 prefers δ
receptors and possesses potent agonist potency in the
MVD.
In the mouse tail-flick antinociceptive assay16 PNTI
(2) produced a dose- and time-dependent antinociception
after icv administration. The antinociceptive peak effect
was reached 20 min after icv administration, with an
ED50 value of 2.06 (1.59-2.64) nmol/mouse. To deter-
mine the selectivity of 2 at peak activity (20 min), mice
were pretreated sc with NTI2 (4 µmol/kg), â-funaltrex-
Ack n ow led gm en t. We thank Mary M. Lunzer,
J anet Peller, and Michael Powers for capable technical
assistance. We are indebted to Dr. Shiv Kumar Sharma
for the supply of 7′-aminonaltrindole employed in the
synthesis of PNTI. This work was supported by the
National Institute on Drug Abuse.

1020 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 7
Letters
(14) DADLE: IC50 ) 0.28 nM in the same tissue.
(15) The IC50 ratio represents the IC50 of DADLE in the presence of
4 (100 nM) divided by the control IC50
(16) Ward, S. J .; Portoghese, P. S.; Takemori, A. E. Pharmacological
Characterization in Vivo of the Novel Opiate, â-Funaltrexamine.
J . Pharmacol. Exp. Ther. 1982, 220, 494-498.
(17) Takemori, A. E.; Larson, D. L.; Portoghese, P. S. The Irreversible
Narcotic Antagonistic and Reversible Agonist Properties of the
Fumaramide Methyl Ester Derivative of Naltrexone. Eur. J .
Pharmacol. 1981, 70, 445-451.
Refer en ces
(1) Takemori, A. E.; Portoghese, P. S. Selective Naltrexone-Derived
Antagonists. Annu. Rev. Pharmacol. Toxicol. 1992, 32, 239-269.
(2) Portoghese, P. S.; Sultana, M.; Takemori, A. E. Design of
Peptidomimetic δ Opioid Antagonists Using the Message-Ad-
dress Concept. J . Med. Chem. 1990, 33, 1714-1720.
(3) Le Bourdonnec, B.; El Kouhen, R.; Lunzer, M. M.; Law, P. Y.;
Loh, H. H.; Portoghese, P. S. Reporter Affinity Labels: An
o-Phthalaldehyde Derivative of â-Natrexamine as a Fluorogenic
Ligand for Opioid Receptors. J . Med. Chem. 2000, 43, 2489-
2492.
(4) (a) Wong, O. S.; Sternson, L. A.; Schowen, R. L. Reaction of
o-Phthalaldehyde with Alanine and Thiols: Kinetics and Mech-
anism. J . Am. Chem. Soc. 1985, 107, 6421-6422. (b) Simmons,
S. S., J r.; J ohnson, D. F. The Structure of the Fluorescent Adduct
Formed in the Reaction of o-Phthalaldehyde and Thiols with
Amines. J . Am. Chem. Soc. 1976, 98, 7098-7099. (c) Simmons,
S. S., J r.; J ohnson, D. F. Reaction of o-Phthalaldehyde and Thiols
with Primary Amines: Fluorescence Properties of 1-Alkyl (and
Aryl)thio-2-Alkylisoindoles. Anal. Biochem. 1978, 90, 705-725.
(d) Garcia Alvarez-Coque, M. C.; Medina Hernandez, M. J .;
Villanueva Camanas, R. M.; Mongay Fernandez, C. Formation
and Instability of o-Phthalaldehyde Derivatives of Amino Acids.
Anal. Biochem. 1989, 178, 1-7.
(5) Ananthan, A.; J ohnson, C. A.; Carter, R. L.; Clayton, S. D.; Rice,
K. C.; Xu, H.; Davis, P.; Porreca, F.; Rothman, R. B. Synthesis,
Opioid Receptor Binding, and Bioassay of Naltrindole Analogues
Substituted in the Indolic Benzene Moiety. J . Med. Chem. 1998,
41, 2872-2881.
(6) Kshirsagar, T.; Nakano, A. H.; Law, P. Y.; Elde, R.; Portoghese,
P. S. NTI4F: A non-peptide fluorescent probe selective for
functional delta opioid receptors. Neurosci. Lett. 1998, 249, 83-
86.
(7) Portoghese, P. S.; Sultana, M.; Nelson, W. L.; Klein, P.; Takemori,
A. E. δ Opioid Antagonist Activity and Binding Studies of
Regioisomeric Isothiocyanate Derivatives of Naltrindole: Evi-
dence for δ Receptor Subtypes. J . Med. Chem. 1992, 35, 4086-
4091.
.
(18) Portoghese, P. S.; Lipkowski, A. W.; Takemori, A. E. Binaltor-
phimine and Nor-Binaltorphimine, Potent and Selective κ-Opioid
Receptor Antagonists. Life Sci. 1987, 40, 1287-1292.
(19) The ED50 ratio represents the ED50 of 2 in the presence of the
antagonist divided by the control ED50
.
(20) (a) Tsao, P. I.; Von Zastrow, M. Type-specific Sorting of G
Protein-Coupled Receptors after Endocytosis. J . Biol. Chem.
2000, 275, 11130-11140. (b) Hasbi, A.; Allouche, S.; Sichel, F.;
Stanasila, L.; Massotte, D.; Landemore, G.; Polastron, J .; J auzac,
P. Internalization and Recycling of δ-Opioid Receptor Are
Dependent on
a Phosphorylation-Dephosphorylation Mecha-
nism. J . Pharmacol. Exp. Ther. 2000, 293, 237-247. (c) Ko, J .
L.; Arvidsson, U.; Williams, F. G.; Law, P. Y.; Elde, R.; Loh, H.
H. Visualization of Time-dependent Redistribution of δ-Opioid
Receptors in Neuronal Cells during Prolonged Agonist Exposure.
Mol. Brain Res. 1999, 69, 171-185.
(21) Mosberg, H. I.; Hurst, R.; Hruby, V. J .; Gee, K.; Yamamura, H.
I.; Galligan, J . J .; Burks, T. S. Bis-penicillamine Enkephalins
Possess Highly Improved Specificity Toward δ Opioid Receptor.
Proc. Natl. Acad. Sci. U.S.A. 1983, 80, 5871-5874.
(22) (a) Farrens, D. L.; Altenbach, C.; Yang, K.; Hubbell, W. L.;
Khorana, H. G. Requirement of rigid-body motion of transmem-
brane helices for light activation of rhodopsin. Science 1996, 274,
768-770. (b) Dunham, T. D.; Farrens, D. L. Conformational
changes in rhodopsin. Movement of helix F detected by site-
specific chemical labeling and fluorescence spectroscopy. J . Biol.
Chem. 1999, 274, 1683-1690.
(23) (a) Konopka, J . B.; Margarit, S. M.; Dube, P. Mutation of Pro-
258 in transmembrane domain 6 constitutively activates the G
protein-coupled R-factor receptor. Proc. Natl. Acad. Sci. U.S.A.
1996, 93, 6764-6769. (b) J avitch, J . A.; Fu, D.; Liapakis, G.;
Chen, J . Constitutive activation of the â2 adrenergic receptor
alters the orientation of its sixth membrane-spanning segment.
J . Biol. Chem. 1997, 272, 18546-18549. (c) Rasmussen, S. G.;
J ensen, A. D.; Liapakis, G.; Ghanouni, P.; J avitch, J . A.; Gether,
U. Mutation of a highly conserved aspartic acid in the â2
adrenergic receptor: constitutive activation, structural instabil-
ity, and conformational rearrangement of transmembrane seg-
ment 6. Mol. Pharmacol. 1999, 56, 175-184.
(24) Marjamaki, A.; Frang, H.; Pihlavisto, M.; Hoffren, A. M.;
Salminen, T.; J ohnson, M. S.; Kallio, J .; J avitch, J . A.; Scheinin,
M. Chloroethylclonidine and 2-aminoethyl methanethiosulfonate
recognize two different conformations of the human R2A-adren-
ergic receptor. J . Biol. Chem. 1999, 274, 21867-21872.
(8) The apparent Ki reflects both the reversible and irreversible
binding components.
(9) Compound 4 inhibited [3H]diprenorphine binding to δ receptors
with a Ki value of 1.7 ( 0.4 nM. Like many morphinans and
NTI analogues in particular, compound 4 is a “sticky” ligand
and could not be washed out completely from the membrane
preparation.
(10) Henderson, G.; Hughes, J .; Kosterlitz, H. W. A New Example of
a Morphine-Sensitive Neuro-Effector Junction: Adrenergic Trans-
mission in the Mouse Vas Deferens. Br. J . Pharmacol. 1972, 46,
764-766.
(11) Rang, H. P. Stimulant Actions of Volatile Anaesthetics on
Smooth Muscle. Br. J . Pharmacol. 1964, 22, 356-365.
(12) Portoghese, P. S.; Takemori, A. E. TENA, A Selective Kappa
Opioid Receptor Antagonist. Life Sci. 1985, 36, 801-805.
(13) Fournie-Zaluski, M.-C.; Gacel, G.; Maigret, B.; Premilat, S.;
Roques, B. P. Structural Requirements for Specific Recognition
of Mu or Delta Opiate Receptors. Mol. Pharmacol. 1981, 20,
484-491.
J M010004U