Selective δ-opioid receptor ligands: Potential PET ligands based on naltrindole

Article abstract of DOI:10.1016/S0960-894X(01)00112-3

Two series of δ-selective ligands related to the prototypic δ-antagonist naltrindole have been prepared and evaluated in opioid binding assays with the aim of developing new PET ligands for the δ-opioid receptor. One compound (5d) had significantly higher selectivity than naltrindole, but with substantially reduced binding affinity. For those compounds retaining similar affinity to naltrindole, those having ethyl and fluoroethyl substituents afforded the highest levels of selectivity. However, none of the compounds combined the high level of affinity and selectivity ideally suited to the development of an imaging agent.

Full text of DOI:10.1016/S0960-894X(01)00112-3

Bioorganic & Medicinal Chemistry Letters 11 (2001) 939–943
Selective ꢀ-Opioid Receptor Ligands: Potential PET Ligands
Based on Naltrindole
Joanne Clayson,a Andrew Jales,b Robin J. Tyacke,b Alan L. Hudson,b David J. Nutt,b
John W. Lewisa and Stephen M. Husbandsc,*
aSchool of Chemistry, University of Bristol, Bristol, UK
bPsychopharmacology Department, University of Bristol, Bristol, UK
cDepartment of Pharmacy and Pharmacology, University of Bath, Bath, UK
Received 18 December 2000; revised 26 January 2001; accepted 15 February 2001
Abstract—Two series of d-selective ligands related to the prototypic d-antagonist naltrindole have been prepared and evaluated in
opioid binding assays with the aim of developing new PET ligands for the d-opioid receptor. One compound (5d) had significantly
higher selectivity than naltrindole, but with substantially reduced binding affinity. For those compounds retaining similar affinity to
naltrindole, those having ethyl and fluoroethyl substituents afforded the highest levels of selectivity. However, none of the com-
pounds combined the high level of affinity and selectivity ideally suited to the development of an imaging agent. # 2001 Elsevier
Science Ltd. All rights reserved.
Introduction
for that receptor. Methods such as positron emission
tomography (PET) offer a unique means of studying
biochemical and pharmacological processes within liv-
ing beings. The most commonly utilised radioisotopes
for PET imaging are C11 and F18. Both radioisotopes
can often be introduced into a suitable ligand without
significant impact on the binding affinity or selectivity.
To be successful a radioligand needs not just excellent
affinity and selectivity for its target site, but must be
able to cross the blood–brain barrier, exhibit a high
level of specific binding and should not be metabolised
to a large number of labelled metabolites.
Selective, non-peptide, competitive antagonists for the
d-opioid receptor have been reported as having possible
therapeutic benefits in a number of clinical applications.
Naltrindole (NTI: 1), the standard d-antagonist has
been found to be an immunosuppressant and thus of
potential utility in the transplant of organs.1 It has also
been reported to suppress the development of morphine
tolerance and physical dependence2 and to block some
of the behavioural effects of cocaine.3 For these reasons
there is continued interest in the development of alter-
natives to naltrindole, preferably having still higher
selectivity for the d-receptor.
Naltrindole remains the standard d-opioid antagonist
and possesses a structure that offers a number of
opportunities for the design of new PET ligands. The
most detailed SAR studies to date relating to nal-
trindole have focused on substitutions on the benzenoid
ring of the indole moiety. An extensive series of aryl
substituted naltrindole analogues have been synthe-
sised,4,5 with perhaps the major findings being that
d-selectivity is retained by substitution at the C-70 car-
bon while introduction of an appropriate basic, or
0
cationic groupat C-5 leads to k-selectivity.6
Important to the further study of d-opioid receptor
function is the development of imaging agents selective
Schmidhammer and co-workers have studied the effect
of substitution at the 14-hydroxy group, synthesising a
number of O-alkyl derivatives. From this work, it was
apparent that a 14-ethoxy group retained d-selectivity,
and was superior to other ether groups such as methoxy
*Corresponding author. Tel.: +44-1225-826826; fax: +44-1225-
826114; e-mail: s.m.husbands@bath.ac.uk
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00112-3

940
J. Clayson et al. / Bioorg. Med. Chem. Lett. 11 (2001) 939–943
or propyloxy.7 An alkyl groupside chain was of partic-
ular interest to us as it allows, with appropriate modi-
fication of the chemistry, to introduce a radiolabel into
the molecule. However Schmidhammer showed that the
14-hydroxyl groupcould only easily be alkylated with
dimethyl- or diethyl-sulfate, or by an allylic or benzylic
alkylating group.7,8 This limited chemistry clearly
reduced the range of options for introduction of a
radiolabel. We were therefore interested in determining
if alkyl groups such as ethyl and propyl could also be
utilised in the more readily alkylated 14-amino series.
Additionally, it seemed that the indolic nitrogen would
also provide a means of attaching an appropriate side
chainy as this modification does not appear to adversely
affect selectivity.9 An example of this latter approach is
the d PET ligand [11C]MeNTI (2) studied by the group
at Johns Hopkins.10ꢀ14
Synthesis
The 14-amino morphindoles (6) were prepared from
14-amino dihydrocodeinone (3) which is synthesised
from N-cyclopropylmethylnorthebaine by a previously
reported method (Scheme 1).15 Fischer indole synthesis
utilising phenylhydrazine in a 1:1 mixture of HCl and
glacial AcOH gave 4 in reproducible yield (55%). In the
absence of HCl acetylation of the 14-amino group
occurred. Presumably the presence of HCl results in the
hydrolysis of any acetylated product to leave only the
primary amine (4). Alkylation could be achieved in high
yield (80–96%) utilising the appropriate alkylhalide at
130 ꢁC in a sealed tube. The alkyl halides used were:
Br(CH2)2F/NaI, Br(CH2)3F/NaI, CH3CH2I, CH3(CH2)2I.
BBr3 mediated 3-O-demethylation16 suffered from diffi-
cult isolation and poor yields of the desired products
ꢁ
.
Scheme 1. (i) PhNHNH2 HCl, AcOH/HCl 1:1, heat, 1.5 h, 51%; (ii) RX, MeCN, NaHCO3, 130 C, sealed tube, 80–96%; (iii) PrSNa, HMPA,
110 ꢁC; (iv) HCO2H, CH3CO2COCH3, 91%; (v) Ph3CCl, TEA, DMAP, DCM; (vi) NaH, DMF, Br(CH2)nF; (vii) 33% aq HCO2H, DCM.
yA preliminary account of this work was made to the European Opioid Conference, Guildford, UK, April 1997.

J. Clayson et al. / Bioorg. Med. Chem. Lett. 11 (2001) 939–943
941
Scheme 2. (i) Ph3CCl, TEA, DMAP, DCM; (ii) NaH, DMF, R0-X (X=Br or I); (iii) 50% aq AcOH, DCM.
and for this reason thiolate17 mediated demethylation
was utilised. This cleanly gave the desired products
when no fluorine was present in the side chain, but for
the fluoroethyl (5c) and fluoropropyl (5d) analogues it
resulted in partial or complete substitution of fluoride
by propane thiolate. It was not possible to isolate the
products of demethylation for 5c or 5d in sufficient
quantity for pharmacological testing.
of the side chain by a single methylene unit resulted in a
3-fold further reduction in d-affinity with no effect at k
or m, thereby reducing selectivity. The thioether (6c) was
assayed to determine the effect of further extending the
side chain. Whilst d-affinity remained constant, affinities
at k and m increased significantly, resulting in very
limited d-selectivity. As would be expected from SAR
studies in other series of opioid ligands, the 3-O-methyl
ethers (4 and 5a–d; Table 1) were of lower affinity at
each receptor compared to their 3-hydroxyl analogues.
The 14-amino compound (4) was of very low affinity
at each receptor, but alkylation provided a dramatic
increase in affinity, particularly at d and m receptors.
Interestingly, in the 3-methoxyl series the propylamino
analogues (5b and 5d) were of higher selectivity than
their ethylamino counterparts (5a and 5c), reversing the
selectivity observed with the 3-hydroxy series and pre-
viously reported for naltrindole and its ethers.7 5d
proved to be the most selective ligand in this study, but
was of too low affinity for consideration as a PET ligand.
14-Aminocodeindole (4) was also utilised in the pre-
paration of 14-formylamido analogues 11a and 11b
(Scheme 1). Initially demethylation with propane thio-
late gave the morphindole (7) in good yield (79%) fol-
lowed by formylation using formic acetic anhydride.18
Protection of the phenol as the triphenylmethyl ether (9)
was followed by indolic-N alkylation with NaH and the
appropriate alkyl halide (Br(CH2)nF). Deprotection to
leave the phenols was achieved in 33% aq formic acid/
dichloromethane.
Indolic N-ethyl-, propyl-, fluoroethyl- and fluoro-
propylnaltrindole analogues (15a–d) were prepared by
alkylation of 3-O-triphenylmethylnaltrindole (13) using
Br(CH2)nF or I(CH2)CH3 followed by deprotection
with aqueous acetic acid (Scheme 2). 13 was prepared
by tritylation of naltrindole using trityl chloride/tri-
ethylamine/DMAP. Analogous derivatives (15e and 15f)
of the d-partial agonist oxymorphindole, where NMe
replaces NCPM, were prepared by an identical route.
Formylation of the 14-amino groupled to a compound
(8) with similar d-affinity and selectivity as the 14-alkyl-
amino analogues discussed above, indicating that the
amino groupneed not be basic. Subsequent fluoro-
ethylation and fluoropropylation at the indolic nitrogen
also had little or no effect on affinity or selectivity of
these 14-formylamino analogues.
In the analogues directly related to naltrindole (i.e.,
having a 14-hydroxyl group) the effects of alkylation at
the indolic-N were again relatively minor (Table 2).
Ethylation to 15a gave a compound with a similar bind-
ing profile to naltrindole. d-Selectivity was slightly reduced
due to lower d-affinity, while addition of an extra
methylene unit, to 15b, again caused a further dropin d-
affinity and selectivity. The fluoroalkyl analogues (15c and
15d) behaved similarly with the ethyl analogue having a
slightly more favourable profile than the propyl.
Results and Discussion
The new ligands were assayed for their ability to dis-
place selective tritiated ligands from guinea pig brain
membranes.19 The radioligands used were [3H]DAMGO
(m), [3H]DPDPE (d) and [3H]U69593 (k). Of the 14-
aminomorphindole analogues (6a–c), only 6a with an
ethylamino side chain bound with d-affinity approach-
ing that of naltrindole (0.77 vs 0.15 nM; Table 1). How-
ever, k and m affinity was also reduced for 6a giving a
broadly similar selectivity profile to naltrindole. Extension
Thus, none of the newly synthesised compounds, or
MeNTI (2), had an improved binding profile compared

942
J. Clayson et al. / Bioorg. Med. Chem. Lett. 11 (2001) 939–943
Table 1. Opioid receptor binding affinities of 14-amino analogues of naltrindole
Ki (nM)
Compound
R
R0
R00
d
m
k
m/d
k/d
4
5a
5b
5c
5d
6a
6b
6c
H
H
H
H
H
H
H
H
H
H
CH2CH3
(CH2)2CH3
CH2CH2F
(CH2)3F
CH2CH3
(CH2)2CH3
(CH2)3S(CH2)2CH3
CHO
Me
Me
Me
Me
Me
H
H
H
H
H
381ꢂ296
27.4ꢂ8.8
19.0ꢂ7.6
10.1ꢂ2.3
12.9ꢂ2.8
0.77ꢂ0.08
2.41ꢂ0.76
1.78ꢂ0.48
0.89ꢂ0.32
0.55ꢂ0.15
0.82ꢂ0.15
0.15ꢂ0.01
14700ꢂ4510
671.2ꢂ446.7
5191ꢂ776
213.7ꢂ69.6
5127ꢂ2353
90.3ꢂ11.8
124.9ꢂ3.9
8.11ꢂ3.20
36.2ꢂ3.7
7325ꢂ3810
2679ꢂ1935
1784ꢂ1339
515ꢂ62
39
25
94
21
397
117
52
5
41
19
74
273
51
240
74
23
8
3105ꢂ36
57.3ꢂ15.3
54.8ꢂ22.6
14.5ꢂ6.0
26.7ꢂ3.4
18.0ꢂ1.3
18.3ꢂ2.0
14.3ꢂ1.1
8
30
32
9
11a
11b
1: NTI
(CH2)2F
(CH2)3F
—
CHO
CHO
—
32.0ꢂ2.0
58
26
183
H
—
20.9ꢂ3.5
27.5ꢂ7.7
95
Table 2. Opioid receptor binding affinities of naltrindole analogues substituted on the indolic-N
Ki (nM)
Compound
R
R0
d
m
k
m/d
k/d
15a
15b
15c
15d
15e
15f
CPM
CPM
CPM
CPM
Me
Me
CPM
CPM
CH2CH3
(CH2)2CH3
(CH2)2F
(CH2)3F
(CH2)2F
(CH2)3F
H
0.26ꢂ0.05
0.44ꢂ0.04
0.24ꢂ0.11
0.40ꢂ0.10
6.44ꢂ4.34
2.66ꢂ1.05
0.15ꢂ0.01
0.49ꢂ0.23
22.0ꢂ8.4
16.4ꢂ1.1
16.7ꢂ1.3
18.7ꢂ0.8
30.0ꢂ13.5
70.7ꢂ12.9
27.5ꢂ7.7
39.2ꢂ1.9
14.0ꢂ2.3
11.9ꢂ0.3
85
38
70
47
5
27
183
81
54
27
35
25
18
75
95
17
8.30ꢂ1.40
9.84ꢂ0.39
116.7ꢂ11.2
200.4ꢂ51.1
14.3ꢂ1.1
1: NTI
2
Me
8.33ꢂ0.86
to NTI. Only 5d displayed higher selectivity and this
was at the expense of substantial loss in affinity. Of
those compounds with comparable affinity to NTI,
those possessing an ethyl or fluoroethyl group at the
14-position or at the indolic nitrogen displayed greatest
selectivity, consistently higher than their propyl ana-
logues. Although the differences were relatively small,
the observation that this holds true at both the 14-posi-
tion (as also reported by Schmidhammer)7 and at the
indolic-nitrogen is intriguing.
binding profile. However, these results would suggest
that, in fact, neither the newly synthesised compounds
nor the current PET ligands has the level of selectivity
ideally suited for imaging studies.
Acknowledgements
This work was funded by MRC ROPA No. G9513279
and US Public Health Service Grant DA 00254.
Subsequent to our preliminary communication, Mathews
et al.20 disclosed their work with [F-18]fluoroethylnal-
trindole (=15c). High binding affinity (0.09 nM) and
high selectivity (>100-fold over m and k) in rat brain
homogenates was reported. Although we did not find
quite the same level of affinity or selectivity for 15c in
our study, together with 6a, 15c had the most promising
References
1. Arakawa, K.; Akami, T.; Okomoto, M.; Akioka, K.;
Nakai, L.; Oka, T.; Nagase, H. Transplant. Proc. 1993, 25,
738.

J. Clayson et al. / Bioorg. Med. Chem. Lett. 11 (2001) 939–943
943
2. Abdelhamid, E. E.; Sultana, M.; Portoghese, P. S.; Take-
mori, A. E. J. Pharmacol. Exp. Ther. 1991, 258, 299.
3. Reid, L. D.; Hubbell, C. L.; Glaccum, M. B.; Bilsky, E. J.;
Portoghese, P. S.; Porreca, F. Life Sci. 1993, 52, PL67.
4. Portoghese, P. S.; Farouz-Grant, F.; Sultana, M.; Take-
mori, A. E. J. Med. Chem. 1995, 38, 402.
12. Madar, I.; Lever, J. R.; Kinter, C. M.; Scheffel, U.;
Ravert, H. T.; Musachio, J. L.; Mathews, W. B.; Dannals,
R. F.; Frost, J. J. Synapse 1996, 24, 19.
13. Madar, I.; Lesser, R. P.; Krauss, G.; Zubieta, J. K.; Lever,
J. R.; Kinter, C. M.; Ravert, H. T.; Mussachio, J. L. Ann.
Neurology 1997, 41, 358.
5. Ananthan, S.; Johnson, C. A.; Carter, R. L.; Clayton, S. D.;
Rice, K. C.; Xu, H.; Davis, P.; Porreca, F.; Rothman, R. B. J.
Med. Chem. 1998, 41, 2872.
6. Olmsted, S.; Takemori, A. E.; Portoghese, P. S. J. Med.
Chem. 1993, 36, 179.
14. Smith, J. S.; Zubieta, J. K.; Price, J. C.; Flesher, J. E.;
Madar, I.; Lever, J. R.; Kinter, C. M.; Dannals, R. F.; Frost,
J. J. J. Cereb. Blood Flow Metab. 1999, 42, 43.
15. Kirby, G.; McLean, C. J. Chem. Soc. Perkin Trans. 1
1985, 1443.
7. Schmidhammer, H.; Krassnig, R.; Greiner, E.; Schutz, J.;
White, A.; Berzetei-Gurske, I. P. Helv. Chim. Acta 1998, 81, 1064.
8. Schmidhammer, H.; Schwarz, P.; Wei, Z. Y. Helv. Chim.
Acta 1998, 81, 1215.
9. Portoghese, P. S.; Sultana, M.; Takemori, A. E. J. Med.
Chem. 1990, 33, 1714.
10. Lever, J. R.; Kinter, C. M.; Ravert, H. T.; Musachio, J. L.;
Mathews, W. B.; Dannals, R. F. J. Labelled Comp. Radio-
pharm. 1995, 36, 137.
11. Lever, J. R.; Scheffel, U.; Kinter, C. M.; Ravert, H. T.;
Dannals, R. F. Eur. J. Pharmacol. 1992, 216, 459.
16. Hutchins, C. W.; Cooper, G. K.; Purro, S.; Rapoport, H.
J. Med. Chem. 1981, 24, 773.
17. Michne, W. F. J. Med. Chem. 1978, 21, 1322.
18. Sheehan, J. C.; Yang, D.-D. H. J. Am. Chem. Soc. 1958,
80, 1154.
19. Smith, J. A. M.; Hunter, J. C.; Hill, R. G.; Hughes, J. J.
Neurochem. 1989, 53, 27.
20. Mathews, W. B.; Kinter, C. M.; Palma, J.; Daniels, R. V.;
Ravert, H. T.; Dannals, R. F.; Lever, J. R. J. Labelled Comp.
Radiopharm. 1999, 42, 43.