J . Med. Chem. 1995,38,1337-1343
1337
Electrophilic N-Benzylnaltrindolesas 6 Opioid Receptor-SelectiveAntagonists Vijaya L. Korlipara, Akira E. Takemori,? and Philip S. Portoghese* Department of Medicinal Chemistry, College of Pharmacy, and Department of Pharmacology, Medical School, University of Minnesota, Minneapolis, Minnesota 55455 Received January 3, 1995@
The N-benzyl group of N-benzylnaltrindole (1, BNTI), a potent and selective 82 opioid receptor antagonist, was employed as a scaffold to hold electrophilic moieties (isothiocyanate and haloacetamide) in an effort to obtain selective affinity labels (2-4 and 8-11>.The corresponding acetamide derivatives (5-7) also were synthesized to serve as nonelectrophilic controls. The 0-andp-isothiocyanates (2 and 4) and the haloamides (8-11) were selective 6 opioid receptor antagonists in the mouse vas deferens (MVD) preparations, while the rneta isomer 3 was a d-selective full agonist (IC5,, = 5 nM). The fact that the effect of 2 and 4 was found to increase as a function of time in MVD suggests a covalent mechanism for the wash resistant component. The rn-isothiocyanate 3 was found to be a d-selective and irreversible agonist in the MVD, and it is suggested that it may be covalently binding to a n agonist recognition site. In the mouse abdominal stretch antinociceptive assay, compounds 2-4 and 9 were d-selective antagonists but exhibited &J& selectivity ratios lower than that of BNTI.
Introduction With the recent cloning and sequencing of opioid receptors, it is now well established that there are a t least three types of G protein-coupled receptors (D, 6, and K ) that possess high sequence homo1ogy.l Electrophilic affinity labels have played a prominent role in the pharmacologic characterization of opioid receptom2 Among the affinity labels having nonequilibrium 6 antagonist activity, naltrindole 5’-isothiocyanate (5’NTII) and [~-Ala~,Leu~]enkephalin-Cys~ (DALCE)have been used to investigate the putative 6 2 and 61 receptor subtype^.^^^ In an effort to expand the armamentarium of subtypeselective 6 antagonists, and on the basis of the high 8 2 antagonist selectivity of N-ben~ylnaltrindole~ (BNTI, l), we have synthesized a series of analogues that contain isothiocyanate (2-4) and haloacetamide (8-11) substituents on the ortho, meta, or para position of the benzyl group. Here we report on the pharmacologic selectivity of compounds in this series and compare the activities with the nonelectrophilic analogues 5-7.
BNTI as a lead compound in order to develop affinity labels with 6 opioid receptor selectivity. In order to cover a range of reactivity and chemical selectivity, the isothiocyanate and haloacetamide groups were considered for attachment to the benzyl group at the ortho, meta, or para position. The rationale for preparing regioisomers was based on the hypothesis that different receptor subtypes may possess different reactivity with the electrophilicgroup of the affinity labels.6 This could occur if only one of the subtypes contains a receptorbased nucleophile in close proximity to the electrophilic group of a reversibly bound regioisomer. Such differentiation may be the basis for the 62 selectivity of 5’-NTII.7 The choice of electrophilic groups was based on their chemical reactivity. Isothiocyanates react readily with the amino and thiol groups of amino acid residues, while bromoacetamides and iodoacetamides react with a broader spectrum of nucleophiles that include amino, thiol, hydroxyl, and carboxylate. The acetamides 6-7 were synthesized t o serve as the nonelectrophilic analogues of the haloacetamides 8-11.
Chemistry
1
Design Rationale and Chemistry BNTI (1)was recently reported to be a potent and selective 6 opioid receptor antagonist which displays 62 opioid receptor antagonism in viva5 We have employed + Department of Pharmacology, Medical School. @Abstractpublished in Advance ACS Abstracts, March 15, 1995.
Target compounds 2- 11 were synthesized from the previously reported5 0-,m-, and p-aminobenzyl derivatives 12-14 (Scheme 1). Isothiocyanates 2-4 were prepared from compounds 12- 14 using thiophosgene in the presence of sodium bicarbonate in a chloroformwater mixture a t room temperature. Purification of the reaction mixture using dry column chromatography worked well on a small scale (0.25 mmol) but led to the formation of impurities when done on a larger scale because of the longer time required to conduct the chromatographic purification (4-5 h). The acetamides 5-7 were prepared by reacting the amines 12-14 with acetic anhydride. Any diacetylated products, 15,which were formed were converted to the desired acetamides by methanolysis. Bromoacetamides 8-10 were obtained by acylation of 12-14 with bromoacetyl bromide. An earlier attempt to synthesize the
0022-2623/95/1838-1337$09.00/00 1995 American Chemical Society
1338 Journal of Medicinal Chemistry, 1995, Vol. 38, No. 8
Korlipara et al.
Scheme 1
BrCHpCOBr c--
NH-pyBr
8, ortho 9. meta 16, para
0
"c=s
"2
12, ortho 13, meta 14, para
2, ortho 3, mta 4, Pam
NH-$--Cky 0
11
16
ortho
meta Pam
p-bromoacetamide 10 from the amine 14 and bromoacetic acid in the presence of dicyclohexylcarbodiimideand 1-hydroxy-W-benzotriazole(HOBt) resulted in the formation of 16 and the HOBt conjugate 17. Compound 16 may have been formed by the reaction of 17 with the parent amine 14. Alternatively, 16 could have been formed from nucleophilic attack by amine 14 on bromoacetamide 10. The p-iodoacetamide 11 was synthesized from 10 using excess sodium iodide in THF.
Pharmacological Results
Smooth Muscle Experiments. The opioid agonist and antagonist potencies of the target compounds and their precursors were determined on the electrically stimulated guinea pig ileum8 (GPI) and mouse vas deferensg(MVD) preparations (Table 1). The antagonist potencies were determined by employing morphine (MI, ethylketazocine (EK), and [D-Ala2,D-Leu5]enkephalin'0 (DADLE) as p-, K- and 6-selective agonists, respectively. Opioid antagonism is expressed as an IC50 ratio, which is the IC50 of agonist in presence of antagonist divided by the control Ic50.Agonist IC50 values were obtained either directly after incubation with target compound without washing or after five washes with buffer. The incubation times of the target compounds ranged from 10 to 60 min.
All target compounds (100 nM), with the exception of the isothiocyanate 3, were inactive as agonists in these tissue preparations. However, at 1pM some of the compounds displayed weak partial agonist activity, with up to 50%agonism in the MVD. Compounds 2 and 4 exhibited relatively good 6 opioid antagonism, with little or no activity at p and K receptors (Table 1). Time-dependent studies involving incubation of the ligands with the MVD were carried out in an effort to distinguish between possible entrapment and covalent binding to the receptor. If indeed the ligand was involved in a covalent bond formation with the receptor, a time-dependent increase in IC50 ratio would be expected. However, if the ligand were to bind reversibly, there should be little change in the IC50 ratio after 15 min, as equilibrium is attained within that period. It was found that the IC50 ratios of both the 0 - and p-isothiocyanates 2 and 4 exhibited a time-dependent increase in the postwash IC50 ratio, suggesting that the isothiocyanate moiety may have reacted covalently with a receptor nucleophile. The postwash IC50 ratio of 2 in the MVD is shown in Figure 1. The m-isothiocyanate 3 was unique in that it was found t o be a full agaonist at 6 opioid receptors with a potency one-tenth that of DADLE (Figure 2). The agonism of 3 was not reversed by washing with either
BNTIs as 6 Opioid Receptor-Selective Antagonists
Journal of Medicinal Chemistry, 1995, Vol. 38,No. 8 1339
250
/ I N-u
,
T
200
.Q
150
c)
2
0 v)
0
16 / I N-4
100
50
0 10
30
60
Time (min) HO
f"
Figure 1. Time course of antagonism of DADLE by isothiocyanate 2 (100 nM) in the MVD.
0-
17
naltrexone (500 nM) or Naltrindolell (NTI) (500 nM). However, when the agonist effect curve of 3 was determined in the presence of NTI (20 nM), the agonist dose-response curve was shifted by a factor of 40 to higher concentration. Consequently, it is likely that 3 acts at the same receptor as NTI. In the GPI compound, 3 showed no agonism or antagonism at 1pM. Among the haloacetamide derivatives 8-11, the mbromoacetamide 9 was the most potent 6 antagonist, with an IC50 ratio (511) a t least 10-fold greater than other members in this group. However, it was found that a substantial fraction of the antagonism of the bromoacetamides 8- 10 either could be removed upon washing or exhibited no apparent time dependence. This suggests that the antagonist effect did not involve a significant amount of covalent association with the receptor within the time frame of the experiment. The o-iodoacetamide 11 exhibited wash resistant antagonism, but it did not appear to be time-dependent. This is consistent +ith a noncovalent localization of 11 in the membrane, as no time-dependent increase of the antagonist effect was observed. Nonelectrophilic analogues 5-7 corresponding to the haloacetamides 8- 11 were evaluated for comparison purposes. Like the bromoacetamides, washing the treated preparation afforded lower IC50 ratios and no substantial wash resistant antagonism. In this connection, it appeared that the antagonism produced by m-acetamide 6 was not removed by washing as easily as 5 or 7. It appears likely that residual 6 may give rise to this persistent antagonism. In Vivo Studies. The isothiocyanate regioisomers 2-4 and the m-bromoacetamide 9 were evaluated for 6 antagonist activity in mice using the abdominal stretch
0.01
I 0.1
I I
I
I
10
100
I
1000 10000
Concentration (nM) Figure 2. Dose-response curves of DADLE control (01, m-isothiocyanate 3 (H), and 3 in presence of NTI (A)in the MVD. The preparation was incubated with NTI (20 nM) for 15 min prior to testing of 3.
assay as described previously.ll The antinociception produced by the 61 agonist [~-Pen~,~-Pen~lenkephalin~~J~ (DPDPE) and the 6 2 agonist [D-Ser2,LeU6]enkephalinThr6J0J4(DSLET) was determined 1.5 and 24 h after the mice were pretreated (5 nmol icv) with the test compounds. The results were expressed as ED50 ratios, which represent the ED50 of the pretreated divided by the ED50 of the control (Table 2). All of the compounds were more potent as antagonists 24 h after pretreatment compared t o 1.5 h. This was particularly evident for the antagonism of DSLET which generally exhibited substantially higher ED50 ratios relative t o those of DPDPE. The o-thiocyanate 2 possessed the highest DSLET ED50 ratio (8.8at 24 h) which was comparable t o that of BNTI (1)(11.1). However, unlike BNTI, 2 appeared not t o differentiate between the 6 receptor subtypes as reflected by the low ddd1
Korlipara et al.
1340 Journal of Medicinal Chemistry, 1995, Vol. 38, No. 8
Table 1. Opioid Antagonist Potencies of N-Benzyl-NTI Analogues in the MVD and GPI Preparations
compd 1 (BNTI)
R
2
H 0-NCS
3 4
m-NCS p-NCS
5
O-NHCOCH3
6
m-NHCOCH3
7
p-NHCOCH3
8
o-NHCOCHZBr
9
m-NHCOCHZBr
10
p-NHCOCHZBr
11
DADLE (6) 208 f 28 57 f 16 23 f 5.9 39 f 10 216 f 20 agoniste 18 f 3.7 6.7 f 1.0 20 f 5.6 54 f 14.2 10 f 1.9d 33 f 8.1 16 f 6.7 18 f 6.4 31 f 6.1 1.6 f 0.92 3.1 f 0.94 7.1 f 2.25 35 f 9.4 13 f 2.4 511 f 115 44 f 13.4 51 f 19.0 30 f 6.8 3.6 f 0.57 12 f 5.1 1.4 f 0.16 50.7 f 9.0 51.2 f 12.6 25.5 f 6.6d
IC50 ratioa incubation time (min) 3oc 3oc 1od 3od 6od 3oc
1.6 f 0.76 0.6 f 0.13
selectivity ratiob EK ( K ) SfP 131 0.4 f 0.08 57 1.3 f 0.3
0.8 f 0.2 1.3 f 0.4
1.2 f 0.6 0.6 f 0.2
25
18
1.9 f 0.6
1.3 f 0.3
29
42
5.3 f 1.5
0.9 f 0.2
0.9 f 0.2
0.2 f 0.05
31
31
1.7 f 0.3
1.1 f 0.3
21
33
3.8 f 0.9
0.9 f 0.4
135
511
0.7 f 0.1
0.9 h 0.1
30
30
0.7 f 0.3
0.5 f 0.2
51
51
M 01)
6fK
208 45
1od 36 3oc 3od 3oc
6.2
33
1od 3od 3oc 1od 3od 6od 3oe 30d 3oc 10d 30d 3oc 1od 3od 6od 3oc 1od 3od
a IC50 ratios were determined in the presence of 100 nM antagonist after a 30 min incubation. Selectivity ratios were calculated using the value 1when IC50 ratios were
