A κ opioid pharmacophore becomes a spinally selective κ-δ Agonist when modified with a basic extender arm

Article abstract of DOI:10.1021/ml1001294

We have explored the concept of a molecular extender arm attached to a κ opioid agonist pharmacophore 3 (ICI-199,441) in an effort to potentially interact with a complementary group on a neighboring opioid receptor. The molecular arm containing a terminal

Full text of DOI:10.1021/ml1001294

pubs.acs.org/acsmedchemlett
A K Opioid Pharmacophore Becomes a Spinally Selective
K-δ Agonist When Modified with a Basic Extender Arm
Ye Tang, Jie Yang, Mary M. Lunzer, Michael D. Powers, and Philip S. Portoghese*
Department of Medicinal Chemistry, College of Pharmacy, University of Minnesota, Minneapolis,
Minnesota 55455, United States
ABSTRACT We have explored the concept of a molecular extender arm attached
to a κ opioid agonist pharmacophore 3 (ICI-199,441) in an effort to potentially
interact with a complementary group on a neighboring opioid receptor. The
molecular arm containing a terminal amine group was lengthened incrementally
from 11 up to 18 atoms. Increasing the number of atoms in the arm produced
virtually no change in the mouse intracerebroventricular (i.c.v.) antinociceptive
potency. In contrast, the intrathecal (i.t.) potency of 6 (KDA-16) with a 16-atom arm
was dramatically increased, as reflected by its antinociceptive i.c.v./i.t. ED50 ratio of
∼130. Further lengthening led to a decreased ED50 ratio. In vivo selective antagonist
studies of KDA-16 revealed that κ and δ opioid receptors were responsible for the
greatly enhanced i.t. potency. Calcium release experiments in HEK-293 cells suggested
that KDA-16 selectively activate κ-δ heteromers. These data are consistent with the
reported possible presence of heteromeric κ-δ opioid receptors in mouse spinal
cord but not in the brain. The use of a molecular extender arm may be useful for
developing spinally selective analgesics.
KEYWORDS Opioid receptors, heteromers, molecular extender arm
As opioid agonists that selectively activate spinal cord
opioid receptors could be a viable approach to reducing side
effects elicited by analgesics at supraspinal sites, we have con-
sidered the possibility that an opioid pharmacophore with
a molecular arm of specific length could potentially interact
with a residue on a neighboring receptor, thereby modulating
opioid selectivity. Considering the existence of some acidic
amino acid residues on the extracellular loops of opioid recep-
tors, we envisaged that a protonated amine group on a molecu-
lar arm might associate with a carboxylate group on a
neighboring receptor to modify opioid activity. Such ligands
might be capable of bridging protomers in homomers or
heteromers.
vidence for homomeric and heteromeric opioid recep-
tors1-4 has led to the design of bivalent ligands that
have the potential for bridging opioid receptors with
E
two pharmacophores linked to one another through a spacer.5,6
Such ligands whose spacer is in the range of 18-21 atoms
have pharmacological properties that are consistent with the
activation or antagonism of heteromeric opioid receptors in
vivo and in vitro.7 Thus, bivalent ligands have been reported
for targeting of heteromeric μ-κ, κ-δ, and μ-δ opioid recep-
tors. Significantly, the spacer length for the presumed bridg-
ing of receptors is in the range of 16-21 atoms for these
bivalent ligands, which is of sufficient length for cross-linking
the receptor protomers in an oligomeric array.5-10
Given the colocalization of δ and κ receptors in spinal cord
axons of rodents13 and reports on the possible association
of these receptors as heteromers,12 we have explored such
an approach through use of a molecular arm attached to a
κ opioid receptor agonist pharmacophore 3 (ICI-199,441),14
in an effort to design spinally selective ligands (Table 1). Thecon-
stitution of the molecular arm attached to the pharmaco-
phore was designed to maintain a hydrophobic-hydrophilic
balance and allowed us to readily adjust the arm length. It is
composed of a number of glycine units (n = 0-3) and a
diglycolic alkyldiamine moiety in which the methylene unit
Different pharmacological selectivity profiles upon intra-
thecal (i.t.) versus intracerebroventricular (i.c.v.) administra-
tion have been observed during the study with the bivalent
ligand KDN-21 (1),6 which has κ and δ opioid receptor anta-
gonist pharmacophores tethered with a spacer. KDN-21 was
characterized by substantially higher i.t. receptor selectivity
and potency with little or no i.c.v. selectivity. Similarly, a
single pharmacophore κ-δ opioid receptor agonist 60-GNTI
(2)11 was reported to be active i.t. but only weakly active i.c.v.
(Figure 1). Studies ofstandard κ and δ agonists and antagonists
that exhibit i.c.v./i.t. selectivity ratios, which are dependent
on the route of administration, also suggest different phenotypic
receptors.12 When taken together, these studies have suggested
that δ-κ opioid receptor heteromers are possibly localized in
the spinal cord but not in the brain.
Received Date: June 6, 2010
Accepted Date: October 6, 2010
Published on Web Date: October 27, 2010
r
2010 American Chemical Society
7
DOI: 10.1021/ml1001294 ACS Med. Chem. Lett. 2011, 2, 7–10
|

pubs.acs.org/acsmedchemlett
Figure 1. Structures of KDN-21, 60-GNTI, and ICI-199,441.
Table 1. Opioid Agonist Selectivity of Ligands (4-8) in the Mouse Tail-Flick Assaya upon Intracerebroventricular or Intrathecal Administration
ED50 (nmol/mouse)
compd
compd nameb
n
m
R
i.c.v.
i.t.
i.c.v./i.t. ratio
3
4
5
6
7
8
ICI-199,441
KDA-11
KDA-14
KDA-16
KDA-18
N/A
N/A
0
N/A
4
N/A
0.58 (0.41-0.83)
0.52 (0.37-0.73)
0.48 (0.34-0.66)
0.58 (0.24-1.43)
0.40 (0.28-0.59)
0.47 (0.32-0.69)
0.55 (0.47-0.64)
1.05
3.45
CH3
CH3
CH3
CH3
CHO
0.15 (0.11-0.21)
1
4
0.17 (0.12-0.24)
2.81
2
3
0.0045 (0.003-0.08)
0.035 (0.022-0.042)
0.020 (0.011-0.038)
130.91
11.43
23.44
3
2
2
3
a Ref 12. b Ref 15.
is varied (m=2-4). The molecular arm was extended from
the meta-amino position of pharmacophore 3, since this
attaching point has been proved not to radically change the
selectivity or potency of the pharmacophore.5 The series
consisted of four ligands 4-7 (KDA-11 to KDA-18)15 having
an arm length from 11 to 18 atoms between the pharmaco-
phore and the terminal secondary methylamino group.
Ligands 4-7 were first evaluated for agonist activity in the
mouse tail flick assay (Table 1) by either intracerebroven-
tricular (i.c.v.) or intrathecal (i.t.) administration. All ligands
(4-7) containing a basic group at the terminus of the extender
arm exhibited antinociceptive activity characteristic of full
agonists. When given i.c.v., these ligands displayed poten-
cies similar to the parent ligand 3 (ICI-199,441). In contrast,
when administered intrathecally, a potency increase that
ranged from 3 to 122 times higher than the parent com-
pound 3 was observed. It is noteworthy that the highest
i.t. potency was achieved at a chain length of 16 atoms
(6, KDA-16), and further extension of the arm led to reduce
the activity. In this regard, 6 was exceptionally potent by the
i.t. route (ED50 = 4.5 pmol) relative to i.c.v. administration
(ED50 = 580 pmol). Ligand 6 afforded the highest i.c.v./i.t.
ED50 ratio of ∼130, as compared to the ratio of 1.05 for the
parent compound 3 (Table 1). Depending on the arm length,
other ligands in this series also showed 3-11-fold increase of
their i.c.v./i.t. ratio. Thus, the in vivo data were consistent
with our original design concept that attachment of a molecular
extender arm to the pharmacophore might increase efficacy.
To evaluate the importance of a basic group on the terminus
of the molecular arm, analogue 8 with a formamide moiety
to reduce the basicity of the amino group was synthesized as
a control ligand for KDA-16 (6). In vivo evaluation showed
that 8 possessed an i.c.v./i.t. ED50 ratio of ∼23, exhibiting an
i.c.v. potency equivalent to that of the ligand 6 but with 4.4-
fold less i.t. potency relative to 6. The reduced i.c.v./i.t.
potency ratio suggested that the terminal-protonated amino
group in KDA-16 might be involved in association with a
counterion on a neighboring receptor or the same receptor.
r
2010 American Chemical Society
8
DOI: 10.1021/ml1001294 ACS Med. Chem. Lett. 2011, 2, 7–10
|

pubs.acs.org/acsmedchemlett
Table 2. Antagonism Effects on Ligand 6 with Selective Antagonists
The pharmacological selectivity of 6 was characterized
using the selective antagonists, norBNI16 (κ), NTI17 (δ), and
CTOP18 (μ) (Table 2). When administered i.t., compound 6
was potently antagonized at κ and δ receptors. Using the
same antagonist dose i.c.v. did not cause a significant shift,
possibly due to differential sensitivity of different phenotypic
receptors to these antagonists in the brain and cord. Given
the evidence for the coexistence of κ and δ receptors in spinal
neuronons6,7,13 and reported studies with ligands selective for
heteromeric δ-κ receptors, these data suggest that KDA-16 (6)
was able to activate both κ and δ opioid receptors in the mouse
spinal cord but not in the brain. In this regard, the finding that
norBNI and NTI synergistically antagonized 6 upon i.t. admin-
istration is of possible significance (Figure 2), given its high
i.c.v./i.t. ED50 selectivity ratio and κ/δ selectivity. One possibility
is that this arises as a consequence of allosterism between
heteromeric κ-δ opioid receptors.6,7,12
upon Intrathecal Administration
antagonist
ED80 ratioa
norBNI (κ)
NTI (δ)
CTOP (μ)
28.50b
9.97b
no changec
a ED80 ratio = ED80 of ligand in the presence of antagonist divided by
the ED80 of the ligand alone. b This is a one point experiment. Dose of
norBNI, 2.5 nmol/mouse; 20 min peak time; NTI, 5.0 nmol/mouse; 20 min
peak time; and CTOP, 6.0 pmol/mouse; 20 min peak time. c No change
means that with the antagonist, the ligand ED80 was not significantly
different from the control value and no shift was expected.
To investigate this further, ligand 6 was tested using the
calcium release assay19 on HEK-293 cells stably expressing
μ, κ, δ, κ-δ, κ-μ, and μ-δ opioid receptors. These six cell lines
also contained a transiently expressed chimeric Giiq protein,
which stimulates the release of calcium upon activation. This
study revealed that 6 exhibited potent agonist activity in the
subnanomolar range on cells coexpressing κ/δ opioid recep-
tors. In contrast, other coexpressed and singly expressed
receptors were activated 66-335 times less potently (Figure 3).
Given that κ and δ opioid receptors are known to organize as
heteromers in cultured cells,1,11 the results of the cell-based
study are consistent with our in vivo observations and suggest
that KDA-16 (6) might selectively activate κ-δ receptor
heteromers rather than κ or δ homomers.
Figure 2. Synergism of norBNI and NTI against ligand 6 (20 pmol,
i.t.) in HSD:ICR(CD1) mice. norBNI in the presence of NTI (() was 9.20
times more potent than norBNI alone (2). NTI in the presence of
norBNI (1) was 9.28 more potent than NTI alone (9). Antinociception
is quantified as the percent maximal possible effect (%MPE), which
is calculated as %MPE = (test - control/10 - control) ꢀ 100.
Considering the flexibility of the 16-atom molecular arm,
we attribute the substantially greater i.t. potency of KDA-16
(6) to the possible association of its protonated amine with
acidic amino residues (Glu or Asp) on an extracelluar loop of
its neighboring δ opioid receptor. The κ opioid pharmacophore
might first associate with a κ opioid receptor, followed by
bridging of the arm to a carboxylate residue on a neighboring
δ receptor (Figure 4). It may be more than a coincidence that
the extender molecular arm of 6 is close to the range of
spacers in bivalent ligands that target heteromeric receptors,
given its potent activity and selectivity in the spinal cord. The
molecular extender arm approach could be applied to design
spinally selective ligands as potential analgesics that have
minimal supraspinal side effects.
Figure 3. Intracellular calcium release assay mediated by increas-
ing concentrations of the ligand 6 was conducted in HEK-293 cells
stably expressing opioid receptors. The response was measured as
relative fluorescence units (RFU) in the Y-axis (no. of replications
g 3). ΔRFU is the change in RFU, which is calculated by subtract-
ing the lowest baseline value after time = 32 s from the maximal
RFU measured during that interval.
Figure 4. Conceptual model of the interaction of 6 (KDA-16) with heteromeric κ-δ opioid receptors. The κ pharmacophore (blue) binds to
the κ receptor, placing its cationic protonated amine moiety in proximity to an anionic moiety on an outer loop of the associated δ receptor,
resulting in bridging to both receptors. Arm lengths shorter or longer than that in ligand 6 are not optimal for association of counterions.
r
2010 American Chemical Society
9
DOI: 10.1021/ml1001294 ACS Med. Chem. Lett. 2011, 2, 7–10
|

pubs.acs.org/acsmedchemlett
SUPPORTING INFORMATION AVAILABLE Experimental
procedures and analytical data for compounds 4-8. This material
is available free of charge via the Internet at http://pubs.acs.org.
(12) Lunzer, M. M.; Portoghese, P. S. Selectivity of δ and κ opioid
ligands depends on the route of central administration in
mice. J. Pharmacol. Exp. Ther. 2007, 322, 166–171.
(13) Wessendorf, M. W; Dooyema, J. Coexistence of kappa- and
delat-opioid receptors in rat spinal cord axons. Neurosci. Lett.
2001, 298, 151–154.
(14) Barlow, J. J.; Blackburn, R. P.; Costello, G. F.; James, R.;
Le Count, D. J.; Main, B. G.; Pearce, R. J.; Russell, K.; Shaw,
J. S. Structure/Activity Studies Related to 2-(3,4-dichloro-
phenyl)-N-methyl-N-[2-(1-pyrrolidinyl)-1-substituted-ethyl]-
acetamides: A Novel Series of Potent and Selective κ-Opioid
Agonists. J. Med. Chem. 1991, 34, 3149–3158.
AUTHOR INFORMATION
Corresponding Author: *Tel: 612-624-9174. Fax: 612-626-6891.
E-mail: porto001@umn.edu.
Funding Sources: This research was supported by Grant DA01533
from the National Institute on Drug Abuse.
(15) Nomenclature employed in this series is as follows: K,
κ receptor; D, δ receptor; and A, agonist; the digits refer to
the numbers of atoms between the pharmacophore and the
terminal amino group.
ACKNOWLEDGMENT We thank Ajay S. Yekkirala for capable
technical assistance and helpful discussion.
(16) Portoghese, P. S.; Lipkowski, A. W.; Takemori, A. E. Binaltor-
phimine and nor binaltorphimine, potent and selective kappao-
pioid receptor antagonists. Life Sci. 1987, 40, 1287–1292.
(17) Portoghese, P. S.; Sultana, M.; Takemori, A. E. Design of pepti-
domimetic δ opioid receptor antagonists using the message-
address concept. J. Med. Chem. 1990, 33, 1714–1720.
(18) Gulya, K.; Krivan, M.; Nyolczas, N.; Sarnyai, Z.; Kovacs, G. L.
Central effects of the potent and highly selective mu opioid
antagonist D-Phe-Cys-Tyr-D-Trp-Orn-Thr-Pen-Thr-NH2 (CTOP)
in mice. Eur. J. Pharmacol. 1988, 150, 355–360.
(19) Yekkirala, A. S.; Kalyuzhny, A. E.; Portoghese, P. S. Standard
Opioid Agonists Activate Heteromeric Opioid Receptors:
Evidence for Morphine and [D-Ala2-MePhe4-Glyol5]Enkephalin
as Selective μ-δ Agonists. ACS Chem. Neurosci. 2010, 1 (2),
146–154.
REFERENCES
(1)
Jordan, B. A.; Devi, L. A. G-protein-coupled receptor hetero-
dimerization modulates receptor function. Nature 1999, 399,
697–700.
(2)
George, S. R.; Fan, T.; Xie, Z.; Tse, R.; Tam, V.; Varghese, G.;
O'Dowd, B. F. Oligomerization of mu- and delta-opioid recep-
tors. Generation of novel functional properties. J. Biol. Chem.
2000, 275, 26128–26135.
(3)
(4)
Gomes, I.; Jordan, B. A.; Gupta, A.; Trapaidze, N.; Nagy, V.;
Devi, L. A. Heterodimerization of mu and delta opioid recep-
tors: A role in opiate synergy. J. Neurosci. 2000, 20, RC110.
Wang, D.; Sun, X.; Bohn, L. M.; Sadee, W. Opioid receptor
homo- and heterodimerization in living cells by quantitative
bioluminescence resonance energy transfer. Mol. Pharmacol.
2005, 67, 2173–2184.
(5)
(6)
Daniels, D. J.; Kulkarni, A.; Xie, Z.; Bhushan, R. G.; Portoghese,
P. S. A bivalent ligand (KDAN-18) containing δ-antagonist and
κ-agonist pharmacophores bridges δ2 and κ1 receptor pheno-
types. J. Med. Chem. 2005, 48, 1713–1716.
Bhushan, R. G.; Sharma, S. K.; Xie, Z.; Daniels, D. J.; Portogh-
ese, P. S. A bivalent ligand (KDN-21)reveals spinal δ and κ
opioid receptors are organized as heterodimers that give rise
to δ1 and κ2 phenotypes. Selective targeting of δ-κ hetero-
dimers. J. Med. Chem. 2004, 47, 2969–2972.
(7)
(8)
Xie, Z.; Bhushan, R. G.; Daniels, D. J.; Portoghese, P. S.
Interaction of bivalent ligand KDN21 with heterodimeric
δ-κ opioid receptors in human embryonic kidney 293 cells.
Mol. Pharmacol. 2005, 68, 1079–1086.
Daniels, D. J.; Lenard, N. R.; Etienne, C. L.; Law, P.-Y.; Roerig,
S. C.; Portoghese, P. S. Opioid-induced tolerance and depen-
dence in mice is modulated by the distance between phar-
macophores in a bivalent ligand series. Proc. Natl. Acad. Sci.
U.S.A. 2005, 102, 19208–19213.
(9)
Zhang, S.; Yekkirala, A.; Tang, Y.; Portoghese, P. S. A bivalent
ligand (KMN-21) antagonist for μ-κ heterodimeric opioid
receptors. Bioorg. Med. Chem. Lett. 2009, 19, 6978–6980.
(10) Yaguo Zheng, E. A.; Harikumar, K. G.; Hopson, J.; Powers,
M. D.; Lunzer, M. M.; Miller, L. J.; Portoghese., P. S. Induced
association of μ opioid (MOP) and type 2 cholecystokinin
(CCK2) receptors by novel bivalent ligands. J. Med. Chem.
2009, 52, 247–258.
(11) Waldhoer, M.; Fong, J.; Jones, R. M.; Lunzer, M. M.; Sharma,
S. K.; Kostenis, E.; Portoghese, P. S.; Whistler, J. L. A heterodimer-
selective agonist shows in vivo relevance of G protein-coupled
receptor dimers. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 9050–
9055.
r
2010 American Chemical Society
10
DOI: 10.1021/ml1001294 ACS Med. Chem. Lett. 2011, 2, 7–10
|