A highly selective κ-opioid receptor agonist with low addictive potential and dependence liability

Article abstract of DOI:10.1016/j.bmcl.2006.02.017

Buprenorphine analogs have been synthesized. In the studies of analgesic and addictive effects in mice and [³⁵S]GTPγS binding assay in human brain tissue, an analog of buprenorphine where the tert-butyl is replaced by a cyclobutyl moiety (16) has been identified as a selective κ-partial agonist which gives antinociceptive effects, but has low abuse potential. The results may lead to lower degrees of dysphoria than full κ-agonists.

Full text of DOI:10.1016/j.bmcl.2006.02.017

Bioorganic & Medicinal Chemistry Letters 16 (2006) 3609–3613
A highly selective j-opioid receptor agonist with low
addictive potential and dependence liability
a,
b
Hee Sock Park,a,b Hee Yoon Lee,a,* Yong Hae Kim, Jin Kyu Park,
*
Edwin E. Zvartauc and Heeseung Leed
aDepartment of Chemistry, KAIST, Daejeon 305-701, Republic of Korea
bDong Kook Pharmaceutical Co. Ltd, Seoul 135-283, Republic of Korea
cDepartment of Psychopharmacology, Pavlov Medical University, St. Petersburg, Russia
dDepartment of Anesthesiology and Pain Medicine, Ewha Womans University, College of Medicine, Seoul, Republic of Korea
Received 21 November 2005; revised 6 February 2006; accepted 8 February 2006
Available online 2 May 2006
Abstract—Buprenorphine analogs have been synthesized. In the studies of analgesic and addictive effects in mice and [35S]GTPcS
binding assay in human brain tissue, an analog of buprenorphine where the tert-butyl is replaced by a cyclobutyl moiety (16) has
been identified as a selective j-partial agonist which gives antinociceptive effects, but has low abuse potential. The results may lead to
lower degrees of dysphoria than full j-agonists.
Ó 2006 Elsevier Ltd. All rights reserved.
Opioid analgesics are very important in human life.
For example, unrelieved pain destroys a person’s qual-
ity of life and is one of the driving forces behind inter-
ests in physician-assisted suicide and euthanasia.1
However, opioids have several undesirable effects. Phy-
sician reluctance to prescribe opioids in the past has
related to the associated risk of addiction and respira-
tory arrest.2
HO
HO
HO
O
O
O
N
NMe
N
MeO
HO
Buprenorphine
μ Partial agonist
Agonist
A potent analgesics
MeO
MeO
HO
HO
Etorphine
Diprenorphine
μ/κ
Agonist
Pure antagonist
The importance of subtle structural differences be-
tween agonists and antagonists is evidenced far more
exten-sively with opioids than any other class of
drugs.3 The effect by altering substituents can also
be seen in the structurally related family of buprenor-
phine (BUP), etorphine, and diprenorphine having
two-carbon bridge substituent not in morphine
(Fig. 1).4
δ
Tranquilizer for
large animals
κ Partial agonist
To reverse the effects
of etorphine
Figure 1. Buprenorphine and its analogs.
The present study aims to provide BUP analogs having
better analgesia, but with lower side effects. Analgesia
and self-administration (SA) assays6 in mice are the pri-
mary bioassays used in this study and provide an initial
model to assess the relative benefit/risk ratio of therapy
and dependence.
Although BUP may not be entirely devoid of abuse
potential, its propensity to produce addiction appears
to be much lower than that of other potent opioids.5
However, it is still required that non-addictive opioids
should be developed.
BUP analogs, 4, 4d, 5, 14, 15, 16, and 17, were prepared
by modification of the known procedures (Scheme 1).7,8
The compound 19 was reacted with cyclopropyl magne-
sium bromide to give the mixture of 2 and 2d in the ratio
of about 7:3 (HPLC data). Compounds 2 and 2d were
separated by the silica gel column chromatography to
Keywords: j-partial agonist; Opioid; Addiction; Buprenorphine.
*
Corresponding authors. Tel.: +82 42 869 2818; fax: +82 42 869
5818; e-mail: kimyh@kaist.ac.kr
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.02.017

3610
H. S. Park et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3609–3613
HO
free bases 4, 4d, 5, and 14–17 were converted to the cor-
MeO
MeO
responding HCl salts in 94–97% yields. HCl salts of 4,
4d, 5, and 14–17 were prepared for the biological
evaluations.
i
ii
O
O
O
NMe
NMe
NMe
MeO
HO
MeO
HO
MeO
MeO
Compounds, 4, 4d, 5, 14, and BUP (1st set: Tables 1–5),
were screened for analgesic and addictive activity6 in
male DBA mice after intravenous injection.
R
R
O
1
2
3
4
5
R
R
iii
i
MeO
HO
Rank order of full agonists in their analgesic potencies
was as follows: 5 > 4 > 4d (Table 1). The potency of 5
was 100–544 times higher than that of BUP. The full
agonists, 4, 4d, and 5, showed higher analgesic effects
than those of 14 and BUP (Table 2).
ii
O
O
O
NCN
NMe
NMe
MeO
MeO
HO
MeO
R
R
HO
R
HO
Comparisons of additive properties of the compounds
are summarized in Table 3. Analysis showed that
there are different SA potencies in the following or-
der: 5 > 4 > 4d P 14 P BUP. Comparisons of efficacy
criteria (Table 4) showed no difference between
samples.
6
7
2d R
(diastereomer of 2)
4d R
R
iv
MeO
MeO
HO
v
ii
O
O
O
NH
N
N
The important results have been obtained from the
analysis of benefit/risk ratio. In comparing the analge-
sic and reinforcing effect (Table 5), the highest safety
index was found in 4. Addictive safety of the com-
pounds when assessed together with analgesic poten-
R1
R1
MeO
MeO
HO
MeO
HO
HO
R
9
R
R
R1
R
R1
8
R
R
10
11
12
13
14
15
16
17
tial
decreased
in
the
following
order:
4 > 5 > BUP = 4d P 14. Addiction liability of BUP is
demonstrated in the clinical study.5
Scheme 1. Synthetic pathway for 1–17. Reagents and conditions: (i)
RMgBr, THF, reflux, 1.5 h; (ii) KOH, diethyleneglycol, 220 °C, 2 h;
(iii) CNBr, CH2Cl2, reflux, 4 h; (iv) KOH, diethyleneglycol, 175 °C,
1.5 h; (v) R1CH2Br, Mg(OAc)2, DMF, 100 °C, 2 h.
Compounds, 15, 16, 17, and BUP (2nd set: Tables 6–10),
were also screened for analgesic and addictive activity6
in male SHR mice.
On the basis of analgesic effect analysis, 15 and 17
demonstrated lower potencies when compared to
BUP (Tables 6, 7, and 10). Their efficacies were either
a bit lower (17) or similar (15) to that of BUP. While
16 was 33–299% more potent than BUP depending on
the test (Table 6). It was more efficient in terms of
maximum analgesic action across analgesic tests
(Table 7).
give 62% and 26% yields, respectively. The compound 3
was separated by the fractional crystallization in 60%
yields. These compounds, 2, 2d, and 3, were identified
1
by comparing their H NMR spectra with those of the
known tetrahydrothebaine analogs.10 O-Demethylated
4, 4d, and 5, were obtained by hydrolysis of methoxy
moiety of 2, 2d, and 3.
Compounds 14–17 were prepared in several steps from 2
and 3. Compounds 2 and 3 were converted to 6 and 7
with CNBr in 95% yields. Compounds 6 and 7 were
hydrolyzed to free amines 8 and 9 in 89–90% yields.
Amines 8 and 9 were alkylated to give 10–13 in 78–
80% yields. Hydrolysis of 10–13 afforded compounds
14–17 in 68–79% yields. Finally, diastereomerically pure
Table 2. Efficacies (max analgesic effects, %) of 4, 4d, 5, 14, and BUP
Test
4
4d
5
14
BUP
Tail clip
Tail flick
Hot plate
Writhing
100
100
100
100
100
100
100
100
100
100
100
100
No effect
18.6
10.4
46.7
65.3
58.5
100
100
Table 1. Potencies of 4, 4d, 5, 14, and BUP in analgesic tests
a
Test
ED50 (lg/kg)
4
4d
5
14
BUP
ND
Tail clip
Tail flick
Hot plate
Writhing
27.5 (14.0–55.0)
19.7 (12.5–31.0)
11.0 (7.8–15.0)
0.59 (0.4–0.9)
122.0 (62.0–241)
197.0 (106–365)
77.0 (47.0–126)
13.0 (7.0–24.0)
3.7 (1.8–7.8)
2.4 (1.6–3.8)
1.5 (0.8–3.1)
0.055 (0.04–0.08)
ND
ND
ND
312.0 (200–600)
195.0 (120–310)
29.5 (16.0–53.0)
235.9 (131–424)
ND, not determined.
a According to Litchfield and Wilcoxon procedure.11

H. S. Park et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3609–3613
Table 3. Potencies of 4, 4d, 5, 14, and BUP in SA test
3611
Parameter
4
4d
5
14
BUP
a
Unit ED50
Optimal unit doseb
Dosec
0.62 (0.37–1.04)
1.28 (0.72–2.28)
2.56d
65.3d
0.028 (0.021–0.036)
1.35 (0.62–2.9)
3.2
70.9
2.04 (1.0–3.9)
7.68
143.54
1.54
28.5
0.04
1.25
a lg/kg/injection, calculated by the aid of N+ values, taken from the ascending part of dose–response curve. A unit dose supporting an initiation of
SA in 50% of subjects.
b lg/kg/injection, unit dose corresponding to the maximum values of reinforcing criteria (delta and R). A unit dose supporting an initiation of SA in
maximum number of subjects.
c Obtained lg/kg, dose consumed during SA session at the concentration corresponding to the highest R value.
d Corresponding to maximum N+ criterion.
Table 4. Efficacies of 4, 4d, 5, 14, and BUP in SA test
Parameter (max)
4
4d
2.0 18.76
5
14
BUP
delta
R
13.75 3.15
0.23 0.06
87.5
25.38 4.68
0.30 0.04
100
13.00 4.80
0.19 0.06
75
13.00 2.36
0.24 0.06
87.5
0.21 0.13
75
N+
Table 5. Safety indices of 4, 4d, 5, 14, and BUP
Compound
Writhing test ED50
(lg/kg)
Safety index for
unit ED50 dose
Safety index for optimal
unit dose
Ratio of dose consumed for the SA
session to analgesic ED50
BUP
4
29.5
0.59
0.07
1.05
0.1
0.26
2.6
4.9
48.3
5.0
4d
5
13.0
0.055
235.9
0.2
0.51
0.006
0.73
0.01
22.7
0.3
14
Table 6. Potencies of 15, 16, 17, and BUP in analgesic tests
Test
ED50 (lg/kg)
15
16
17
BUP
Tail clip
Tail flick
Hot plate
Writhing
633.0 (357–1122)
358.7 (208.7–617.7)
186.7 (117.8–295.8)
22.4 (11.5–43.6)
28.4 (15.4–52.0)
10.1 (5.3–19.3)
17.6 (9.4–33.0)
11.2 (5.0–26.8)
ND
46.6 (22.5–96.7)
80.3 (53.5–120.5)
190.3 (89.7–404.0)
47.4 (24.1–93.3)
40.4 (24.9–65.5)
23.3 (14.5–37.5)
22.3 (10.5–47.3)
Table 7. Efficacies (max analgesic effects, %) of 15, 16, 17, and BUP
Compound 16 appears to be one of candidates in
the safe drugs (Table 10). Compound 16 exhibits
very interesting pharmacological properties: it is
more potent and effective as an analgesic when com-
pared to BUP and at the same time it is less potent
as a reinforcing agent. Compound 16 may not exhib-
it drug dependence liability at the doses tested in
cancer and non-cancer pain compared with BUP.
The comparatively long analgesic effect of 16 must
be noted.12
Test
15
88.9
84.5
100
100
16
17
44.4
81.0
100
100
BUP
Tail clip
Tail flick
Hot plate
Writhing
100
100
100
100
100
100
100
100
Table 8. Potencies of 15, 16, 17, and BUP in SA test
Parameter
Unit ED50
15
16
17
BUP
1.15
In addition, the [35S]GTPcS assay of 16 was carried
out using human membranes which was initially doc-
umented in human neuroblastoma SH-SY5Y cells13 as
a valuable approach to demonstrate and quantify
affinity of different opioid receptor agonists and par-
tial agonists.14
3.6
4.1
3.78
(1.9–6.9) (2.6–6.3) (2.5–5.6) (0.6–0.21)
Optimal unit dose 4.29
Dose 180
8.7
1,240
4.34
200
1.92
80
In experiments with initiation of iv self-administration
in drug-naive mice, 16 demonstrated much lower poten-
cy ratios (0.06–0.28) when compared to BUP (Table 8).
Efficacy values of reinforcing actions of BUP and 17
were close to each other (Table 9).
The assays of specific receptor binding, defined with
naloxone, were conducted, as described previously.14a
The concentration of protein in human brain cortical
membranes (n = 4) was 1.50 mg/mL 0.03 (SEM),

3612
H. S. Park et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3609–3613
Table 9. Efficacies of 15, 16, 17, and BUP in SA test
Parameter (max)
15
16
17
BUP
delta
R
11.56 10.14
0.21 0.14
56
103.5 41.65
0.45 0.15
83
21.56 8.35
0.24 0.07
56
16.22 4.78
0.28 0.05
67
N+
Table 10. Safety Indices of 15, 16, 17, and BUP
Compound
Writhing test
ED50 (lg /kg)
Safety index for unit
ED50 dose
Safety index for optimal
unit dose
Ratio of dose consumed for the SA
session to analgesic ED50
BUP
15
22.31
22.4
11.2
47.4
0.05
0.16
0.37
0.08
0.086
0.19
0.78
0.09
3.59
8.04
16
17
110.7
4.2
determined by the method of Bradford15 with bovine
serum albumin as the standard. The values of Kd
and Bmax are 0.83 (nM) and 64.4 (fmol/mg protein)
for [3H]DAMGO, 0.70 (nM) and 50.2 (fmol/mg
protein) for [3H]U69593, and 4.98 (nM) and 70.3
(fmol/mg protein) for [3H]DPDPE, respectively.
stimulation of 7.2% over baseline. U69593 was defined
in this assay as a full j-agonist and increased stimula-
tion of [35S]GTPcS by ca. 13.4% over baseline. EC50
values of U69593 and 16 were 64.3 and 1.84 nM,
respectively.
The effects of agonist stimulation of 16 and U69593
were antagonized by norBNI, a selective j-opioid
receptor antagonist16 as shown in Figure 2. The shift
in EC50 caused by 3 nM norBNI for 16 and U69593
Second, the agonist stimulation of [35S]GTPcS to hu-
man brain cortical membranes was studied with 16
and selective reference agonists at opioid receptors.
Compound 16 increased the in vitro binding of
[35S]GTPcS to human brain cortical membranes
dependent upon concentration but with a maximum
16 + 20nM NLX
16
120
110
100
U69593 + 3nM nor BNI
U69593
120
110
100
-11
-10
-9
-8
-7
-6
Log[16]
-10
-9
-8
-7
-6
-5
-4
Log[U69593]
16 + 3nM NTI
16
120
110
100
16 + 3nM nor BNI
16
120
110
100
-11
-10
-9
-8
-7
-6
-11
-10
-9
-8
-7
-6
-5
Log[16]
Log[16]
Figure 3. Stimulation of [35S] GTPcS binding by 16 in the presence of
specific l-antagonist (naloxone, upper panel) and d-antagonist (nal-
trindole, lower panel) in human brain cortical membranes. Shown are
mean values SEM from two independent experiments. Each exper-
iment was carried out in duplicate.
Figure 2. Stimulation of [35S]GTPcS binding by U69593 and 16 in the
presence of specific j-antagonist (norBNI) in human brain cortical
membranes. Shown are mean values SEM from 3 to 4 independent
experiments. Each experiment was carried out in duplicate.

H. S. Park et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3609–3613
3613
Supplementary data
100
75
50
25
0
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2006.02.017.
References and notes
1. Joranson, D. E. Symptom Control in Cancer Patients
(Japan) 1995, 6, 105.
2. Sapir, R.; Catane, R.; Strauss-Liviatan, N.; Cherny, N. I.
J. Pain Symptom Manage. 1999, 17, 266.
-11
-10
-9
-8
-7
-6
3. Snyder, S. H.; Pasternak, G. W. Trends Pharmacol. Sci.
2003, 24, 198.
4. Deschamps, J. R.; Flippen-Anderson, J. L.; George, C.
Biopolymers 2002, 66, 287.
5. Tzschentke, T. M. Psychopharmacology 2002, 161, 1.
6. (a) Semenova, S.; Kuzmin, A.; Zvartau, E. E. Pharmacol.
Biochem. Behav. 1995, 50, 17; (b) Kuzmin, A. V.; Patkina,
N. A.; Zvartau, E. E. Brain Res. 1994, 652, 1.
7. Compounds 2, 4, 6, 8, 10 ,and 14 were reported (Kim, Y.
H.; Park, J. K.; Kim, K.; Park, H. S. U.S. Patent
5,849,915, 1998.
8. Bentley, K. W.; Hardy, D. G. J. Am. Chem. Soc. 1967, 89,
3281.
9. Bentley, K. W.; Hardy, D. G.; Meek, B. J. Am. Chem. Soc.
1967, 89, 3273.
10. (a) Fulmor, W.; Lancaster, J. E.; Morton, G. O.; Brown, J.
J.; Howell, C. F.; Nora, C. T.; Hardy, R. A., Jr. J. Am.
Chem. Soc. 1967, 89, 3322; (b) Soars, M. G.; Mattiuz, E. L.;
Jackson, D. A.; Kulanthaivel, P.; Ehlhardt, W. J.; Wrigh-
ton, S. A. J. Pharmacol. Toxicol. Methods 2002, 47, 161.
11. (a) Litchfield, J. T.; Wilcoxon, F. J. Pharmacol. Exp. Ther.
1949, 96, 99; (b) Schuster, C. R.; Johanson, C. E. Neurosci.
Biobehav. Rev. 1981, 5, 315.
Log[16]
Figure 4. Relative efficacy of 16 [35S]GTPcS stimulation compared
with U69593. Shown are mean values SEM from three independent
experiments. Each experiment was carried out in duplicate.
was 22.5-fold and 15.9-fold, respectively. The calculat-
ed Ke values of 16 and U69593 for norBNI inhibition
were 0.14 nM (lower panel, Fig. 2) and 0.20 nM
(upper panel, Fig. 2), respectively. Similar values for
inhibition of two agonists indicate that these agonists
activated the same receptor. Stimulation of
[35S]GTPcS by 16 was not antagonized by naloxone
or naltrindole, a preferential antagonist of l or d
receptors, respectively (Fig. 3). The results also
showed that 16 behaved as a partial agonist at the
j-opioid receptor in human brain cortical membranes
and with a relative efficacy of 35% when maximum
stimulation of U69593, a known selective full j-ago-
nist, was set as 100% (Fig. 4).
12. Compound 16 and BUP have a long duration of action
compared to other compounds in analgesic tests; see
Supporting information.
In summary, a variety of BUP analogs have been syn-
thesized. In the studies of analgesic and addictive ef-
fects in mice and [35S]GTPcS binding assay in
human brain tissue, the new compound 16 has been
identified as a selective j-partial agonist which gives
antinociceptive effects, but has low abuse potential.
It is reported that the activation of j-receptors leads
to the suppression of unpleasant l and d mediated
side effects such as the rewarding effect.17 The fact
that this compound has a profile (j-partial agonist)
may lead to lower degrees of dysphoria than full
j-agonists.18 The compound may be valuable for the
development of long-acting analgesic, as a requirement
for protracted use in neuropathic pain and drug abuse
medication.
13. Traynor, J. R.; Nahorski, S. R. Mol. Pharmacol. 1995, 47,
848.
14. (a) Emmerson, P. J.; Liu, M.-R.; Woods, J. H.; Medzih-
radsky, F. J. Pharmacol. Exp. Ther. 1994, 271, 1630; (b)
Ko, M. C. H.; Lee, H.; Harrison, C.; Clark, M. J.; Song,
H. F.; Naughton, N. N.; Woods, J. H.; Traynor, J. R. J.
Pharmacol. Exp. Ther. 2003, 306, 179; (c) Remmers, A. E.;
Clark, M. J.; Mansour, A.; Akil, H.; Woods, J. H.;
Medzihradsky, F. J. Pharmacol. Exp. Ther. 1999, 288, 827.
15. Bradford, M. M. Anal. Biochem. 1976, 72, 248.
16. Portoghese, P. S.; Lipkowski, A. W.; Takemori, A. E. Life
Sci. 1987, 4, 1287.
17. Narita, M.; Funada, M.; Suzuki, T. Pharmacol. Ther.
2001, 89, 1.
18. Pfeiffer, A.; Brantl, V.; Herz, A.; Emrich, H. M. Science
1986, 233, 774.