J . Med. Chem. 1994,37, 2138-2144
2138
4-[ (Alkylamino)methyl]furo[3,2-c]pyridines: A New Series of Selective K-Receptor
Agonists Alan Naylor,'ls Duncan B. Judd,s David I. C. Scopes,g Ann G. Hayes: and Philip J. Birch+ Departments of Medicinal Chemistry and Neuropharmacology, Glaxo Group Research Ltd., Ware, Hertfordshire SG12 ODP,England Received January 13, 19948
The synthesis of 5-(arylacetyl)-4-[(alkylamino)methyllfuro[3,2-clpyridines (16-23, 26, 27) and their activities as K-opioid receptor agonists are described. K- Agonist potency was particularly sensitive to the nature of the basic moiety. In particular, in the rabbit vas deferens (K-specific tissue), the 3-pyrrolidinol analogue 17 (ICs0 = 2.7 nM) was found to be approximately 5-fold more potent than the corresponding pyrrolidine analogue 16 (ICs0 = 15 nM). In the rat and hamster vasa deferentia (p-specific and &specific tissues, respectively), 17 showed only weak antagonist activity ( ~ K > B 5.5), underlining its selectivity for the K-opioid receptor. The major activity for 17 is resident in the 4S,3'S-isomer 26 (rabbit vas deferens ICs0 = 1.1nM). Compound 26 displays excellent antinociceptive activity, as determined in the mouse acetylcholine-induced abdominal constriction test (EDSO= 0.001 mg/kg, sc). n The high level of interest in the medicinal chemistry of opioids during the past decade has stemmed, in large part, from the identification of three distinct opioid receptor subtypes: p, K, and 6.1 It is well established that activation of each of these receptors can produce antinociceptive effects in a variety of animal modeh2 In particular, CI considerable effort has been expended in obtaining GR45809 selective K-agonists with the expectation that such compounds should be strong analgesics3devoid of many of the undesirable side effects associated with morphine-like p-agonists (e.g., respiratory depression, constipation, physical dependen~e).~ In addition to playing a role in antinociception, activation of the K-opioid receptor ,can , o produce neuroprotective effects in certain animal models CI Ya of cerebral ischaemia,s Le., clincal use of K-agonists may provide a novel approach for the treatment of stroke. (3) (16-23,26,27) As part of a program of research aimed at identifying selective K-receptor agonists as potential analgesic6 and Chemistry neuroprotective' agents, our interest in this area has The general route shown in Scheme 1 was used to centered on the 2-[ (alky1amino)methyllpiperidineclass synthesize the majority of the compounds described in of K-agonist 1. It has previously been shown that certain this paper. The novel methyl 4,5,6,7-tetrahydrofuro[3,2compounds within this group [e.g., 2 (GR45809)l are both potent and selective agonists at the K-OpiOid r e c e p t ~ r . ~ * ~clpyridine-4-carboxylateintermediates 6 and 7 were prepared from the known furanethylamines 412 and 5,13 As an extension of these discoveries,we and others6J0have respectively,and converted into the amides 8-11 by heating studied the effect of fusion of benzenoid (e.g., 3) and with the appropriate neat amine. Subsequent reduction heteroaromatic ring systems onto the piperidine nucleus with lithium aluminum hydride followed by acylation of of 1. The finding that these modificationsfurnished potent the resulting diamines 12-15 with 3,4-dichlorophenylacetic and selective K-agonists encouraged us to evaluate the acid, using 1,l'-carbonyldiimidaole activation, provided profiles of structures incorporating an electron-rich furan the desired analogues 16-19, 22, and 23. In the case of nucleus. We now report on the synthesis and K-agonist analogues possessing the 3-pyrrolidinolgroup, some 0,Nactivity of a novel series of 4-[(alkylamino)methyllfurodiacylation was observed at this latter stage. However, [3,2-clpyridines(16-23,26,27).Structure-activity studselective O-deacylation could be readily effected upon iesQhave shown that within the 2- [(alkylamino)methyllhydrolysis with lithium hydroxide. is one piperidine series (l), (3,4-di~hlorophenyl)acetyl~~ The 3-pyrrolidinone20was obtained via Swern oxidation of the optimum N-acyl groups for K-agonist activity. of the alcohols 17 (Scheme 2). Preparation of the 1,2,3,6Accordingly, in the present work, this feature has been tetrahydropyridinyl derivative 21 necessitated an alternaretained and we have focused on selected modifications tive approach via the alcohol 28 which was obtained by of the basic moiety which, based upon our previous studies: reduction of the ester 6 (Scheme 3). 0,N-Diacylation of can result in high K-agonist potency. the alcohol 28 with 3,4-dichlorophenylaceticacid followed by O-deacylation gave the amido alcohol 29. Swern Department of Medicinal Chemistry. oxidation of 29 and immediate reductive amination of the + Department of Neuropharmacology. Abstract published in Advance ACS Abstracts, June 1, 1994. resulting crude aldehyde with 1,2,3,6-tetrahydropyridine
0022-262319411837-2138$04.50/0 0 1994 American Chemical Society
4- [(AIkylamino)methyl]furo[3,2-~]pyridines
Journal of Medicinal Chemistry, 1994, Vol. 37, No. 14 2139
Scheme 1.
Scheme 42b
cym (S)-(-)-malic (6) R1
(4) R’ = H (5) R1 =
H
(8) R’
iH ~ 2 , ~ 3( ’c =H ~ ) ~ (9) R’ = H’ R2 = Rb = Me (10) R1 = H R2,R3’= CH2CH(OH)(CHz)z (11) R1 =Me, R2,R3 = CHzCH(OH)(CH2)2
i
(7)R1 = Me
Me
acid
(J‘
OH
cym (9-3 1
(S)-30
Q*
OH
H
~
3
2
a (i) MeCOCl; (ii) PhCHZNH2; (iii) MeCOC1; (iv) LiAlH4; (v) Hz, Pd-C.6 Structures depict absolute configuration for the synthesis of (S)-32; an identical set of conditions was used to obtain the corresponding optical antipode.
(i) CHO-COZMe, HC1; (ii) HNR2R3; (iii) LiAlH,; (iv) (a) 3,4C~Z-C~H~CHZCO~H, 1,l’-carbonyldiimidaole,(b) LiOH.
Scheme 2.
R0-m(i)
&
D
- 0
- 0
(17) a
o
(20)
(i) (COC1)2, DMSO, EtsN.
8 -& -
Scheme 3.
(i)
H
(6)
(28)
?-I
-
n
(iii) (iv)
I
I
(21) a
Biological Results and Discussion
OH
c4Na
H
(ti)
Having confirmed that this is indeed the case, we elected to modify the procedure by using much milder conditions. Following initial activation using acetyl chloride, reaction with benzylamine was conducted at 20-25 “C, and then, further treatment with acetyl chloride gave (S)-30. Using this method, epimerization of the 3-hydroxyl group could be avoided; subsequent reduction with lithium aluminium hydride gave the enantiomer (S)-31 in >98% ee (as determined by lH NMR using (R)-(-)-2,2,2-trifluoro-l(9-anthryllethanol as the chiral solvating agent), which upon hydrogenolysis provided (S)-32. A similar sequence starting from (R)-(+)-malicacid gave (R)-32 in 95% ee. Since the K-receptor agonist activity was found to predominate in 23 (mixture of R- and S-isomers at C-4), derived from (S)-3-pyrrolidinol,the diastereomeric acetate derivatives 24 and 25 were prepared, separated by HPLC, and converted to the 4S,3’S- and 4R,3’S-isomers 26 and 27, respectively (Scheme 5). The absolute stereochemistry at C-4 of the piperidine nucleus was determined by X-ray analysis of compound 27 (Figure 1).
(i) LiAlH4; (ii) (a) 3,4-Cl&~H&H&OzH, 1,l’-carbonyldiimi-
dazole, (b) LiOH; (iii) (COC1)2, DMSO, EtSN; (iv) 1,3,5,6-tetrahydropyridine, NaCNBHs.
and sodium cyanoborohydride afforded the required analogue 21. The component diastereoisomers of 17 (i.e., 22 and 23) were synthesized by employing 3(R)-and 3(S)-pyrrolidinol in the general route illustrated in Scheme 1. The preparation of 3(R)-and 3(S)-pyrrolidinolwas based upon the method described by Joull6 et al.14 (Scheme 4). However, this method, which involves direct reaction of @)-(-)-malic acid with benzylamine followed by lithium aluminium hydride reduction of the resulting hydroxysuccinimide, was reported to afford N-benzyl-3(S)pyrrolidinol [(S)-311 in only 84 7% enantiomeric purity.
The K-agonist activity of the 4-[(alkylamino)methyl]furo[3,2-clpyridines was determined in vitro using the isolated rabbit vas deferens (LVD) preparation, which is rich in K-opioid re~ept0rs.l~ The receptor selectivities for 16 and 17 were assessed by comparing the activity determined in the LVD preparation with that in the rat vas deferens (RVD)16and hamster vas deferens (HVD)17 tissues which are rich in p - and &opioid receptors, respectively. Antinociceptive activity was determined using the mouse acetylcholine-induced abdominal constriction test1*following subcutaneous administration of the agonist, ED50 values being determined in each case. All the compounds in this series were found to behave as high efficacy agonists in the LVD. Modification of the basic moiety was found to produce a pronounced effect on K-agonist potency (Table 1). In particular, the 8-pyrrolidinol analogue 17 (GR91272) was found to be approximately 5-fold more potent than the corresponding pyrrolidine analogue 16. Interestingly, this potencyenhancing effect of the 3-pyrrolidinol moiety is not necessarily observed in other aromatic-fused systems (cf., the 5-hydroxytetrahydroisoquinolineseries 33 and 34, Table 1). In the rat and hamster vasadeferentia, 16 showed no agonist nor antagonist activity up to 10-5 M and 17 showed only weak antagonist activity ( ~ K
