Berbanes - American Chemical Society

Mar 6, 1986 - yohimbine, phentolamine, and prazosin. Their effects on presynaptic a2-adrenoceptors were studied on chemical neurotransmission of isola...
1 downloads 0 Views 658KB Size
J. Med. Chem. 1987,30,1355-1359

1355

Berbanes: A New Class of Selective a,-Adrenoceptor Antagonists E. Sylvester Vizi,*+Istvan Toth,* George T. Somogyi,f Lajos Szabo,o Laszlo G . Harsing, Jr.,? and Csaba Szantays Institute of Experimental Medicine, Hungarian Academy of Sciences, H-2450 Budapest, Hungary, Central Research Institute for Chemistry, Hungarian Academy of Sciences, H-1525 Budapest, Hungary, a n d Organic Chemical Department, Technical University of Budapest, H-1521 Budapest, Hungary. Received March 6, 1986 The a-adrenoceptor blocking properties of some berbanes have been studied and compared with those of idazoxan, yohimbine, phentolamine, and prazosin. Their effects on presynaptic a,-adrenoceptors were studied on chemical neurotransmission of isolated rat vas deferens and longitudinal muscle strip of guinea pig ileum; xylazine or norepinephrine was employed as agonist. The al-adrenoceptor blocking activities of the berbanes were tested on isolated rat vas deferens and rabbit pulmonary artery by using phenylephrine or norepinephrine as agonist. The antagonistic activity of the berbanes and the reference compounds on a2-and a,-adrenoceptors was characterized by t h e apparent pA2 values. Of t h e compounds studied, [8aS*-(8aa,12aa,13aa)]-lla-hydroxy5,6,8a,9,10,11,12,12a,l3,13a-decahydro-8H-benzo[g]-1,3-benzodioxolo[5,6-a]quinolizine, 6d, proved to be the most selective antagonist at the presynaptic a2-adrenoceptors (a1:a2ratio = 1659). Since blockade of a,-adrenoceptors located on noradrenergic axon terminals leads to a n increase of norepinephrine release, this compound could have potential as a n antidepressant agent.

There is now considerable evidence that there are a t least two distinct subclasses of a-adrenoceptors: al- and a,-adreno~eptors.~-~ This pharmacological classification is independent of anatomical arrangemente6 a,-Adrenoceptors can be differentiated from a,-adrenoceptors by the specificity of a series of agonists and antagonists acting on these receptors. Such a subclassification of a-adrenoceptors opens possibilities for the development of new drugs with a high selectivity for each receptor subtype. Several structures have been reported that show significant selectivity as antagonists at a1 (e.g., prazosin) or a,-adrenoceptors (e.g., b e n z o d i ~ x a n s , ~ ~benzo',~ quinolizines,8J0 pyridinylpiperazines,ll benzazepines,12 benzofuroquinolizines128).Combination of 1,4-benzodioxan and imidazoline moieties1J3led to the discovery of idazoxan (RX 781094,2-(1,4-benzodioxan-2-yl)-2-imidazoline), which possesses significant a,-adrenoceptor selectivity. Convincing experimental evidence is available that norepinephrine (NE), via stimulation of a,-adrenoceptors located a t the varicose axon terminals of noradrenergic neurons, controls the further release of Recently, it has been suggested that a common underlying mechanism for the action of antidepressants may be the ability to alter a,-adrenoceptor sensitivity,lsJgthus increasing NE re1ea~e.l~ In addition, the a,-adrenoceptors on noradrenergic neurons in the cerebral cortex are thought to be supersensitive, which would tend to reduce the release of NE. Therefore, it can be suggested that treatment with a compound that acts as an antagonist at presynaptic a,-adrenoceptors could provide an effective and novel treatment for depression. We report here the synthesis and pharmacological testing of new berbane derivatives, which can be regarded as depyrrolo analogues of the biologically potent yohimbine and reserpine alkaloid^.^^^^ The berbanes may therefore possess similar a-adrenoceptor activity. Chemistry Preparation of the berbane skeletonz1has been reported earlier.22-27The tetracyclic ketones 2-4 were synthesized from 12292"7 (Schemes I and 11). The derivatives 2,3, and 4 were reduced with NaBH,, yielding alcohols 5 , 6, and 7, respectively. The stereochemistry of the alcohols has been verified 'Institute of Experimental Medicine, Hungarian Academy of Sciences. *Central Research Institute for Chemistry, Hungarian Academy of Sciences. $Technical L'niversity of Budapest.

Scheme I

R3

R4

0 2

4 2-4

a b C

d e

A2

R1

R3

R4

COOCH3 H H COOCH3 H H CzH5 C2H5 H H C H2 H CN CH2

CH2 CH2

by IR and NMR spectroscopy (Tables 1-111). The intensive Bohlmann-band system in the IR spectra (1) Timmermans, P. B. M. W. M.; van Zwieten, P. A. J . Med. Chem. 1982,25,1389. (2) Vizi, E. S. Adrenoceptors and Catecholamine Action; Kunos, G., Ed.; Wiley: Chichester, 1984;pp 65-109. (3) Timmermans, P. B. M. W. M.; van Zwieten, P. A. Handbook of Hypertension; van Zwieten, P. A., Ed.; Elsevier: Amsterdam, 1984;Vol. 3, pp 239-248. (4) Timmermans, P. B. M. W. M.; de Jonge, A.; Thoolen, M. J. M. C.; Willfert, B.; Batink, H.; van Zwieten, P. A. J . Med. Chem. 1984,27, 495. (5) Berthelsen, S.; Pettinger, W. A. Life Sci. 1977,21, 595. (6) Mouille, P.; Dabire, H.; Fournier, B.; Schmitt, H. Eur. J. Pharrnacol. 1981,73, 367. (7) Chapleo, C. B.; Myers, P. L.; Butler, R. C. M.; Doxey, J. C.; Roach, A. G.; Smith, C. F. C. J. Med. Chem. 1983,26, 823. (8) Lattimer, N.; Rhodes, K. F.; Ward, T. J.; Waterfall, J. F.; White, J. F. Br. J . Pharmacol. 1982, 75, 154P. (9) Doxey, J. C.; Roach, A. G.; Strachan, D. A,; Virdee, N. Br. J . Pharrnacol. 1983, 79, 311P. (10) Lattimer, N.; McAdams, R. P.; Rhodes, K. F.; Sharma, S.; Turner, S. J.; Waterfall, J. F. Naunyn-Schrniedeberg's Arch. Pharrnacol. 1984,327, 312.

0022-2623/87/1830-1355$01.50/00 1987 American Chemical Society

1356 Journal of Medicinal Chemistry, 1987, Vol. 30, No. 8

Vizi et al.

Table I. Chemical and IR Data

starting compd 3a 3b 3c

product yield, % 6a 71 6b 69 6c" 59.8 6d" 16.2 2a 5a 58 2b 5fQ 21 5b" 38 2c 5c 60 4a 7e" 37 4c 7c" 49 3d 6g 61 3e 6k 87 61 6m 90 6c 6n 92 6d 60 89 3a 6h 47 3b 6i 41.5 2a 5h 42 6c 61' 47 6a 6~ 59 Separation by flash chromatography. HCl salt.

Table 11. 'H NMR Data" 1 H, s compd C-6 H 6a 6.50 6.50 6b 6.50 6c 6.50 6d 6.55 5a 6.50 5f 6.52 5b 6.55 5C 6.55 7e 6.50 7c 6.60 6g 6.50 6k 6.55 6m 6.53 6n 6.52 60 " 6 (CDC13).

1 H, s C-9 H 6.62 6.72 6.72 6.70 6.65 6.70 6.70 6.75 6.60 6.65 6.70 6.72 6.64 6.69 6.66

mp, "C formula 159.5-162 C20H25N05 148-150 C2OH26N05 173-175 C18H23N03 214-217 C18H23N03 166 C20H25N05 C20H25N05 185-187 C20H25N06 205 C18H23N03 217-219 202-203 C20H25N05 212-216' Ci8H2,NO3Cl C21H31N03 135-139 C19H22N203 225-226 165-167 C21H29N04 153 C20H25N04 155 C20H25N04 C19H26N04C1 243-245' C19H25N04 152-155 C19H26N04C1 245-250' C17H24N03Br 2029 C21H27N05 142-146 C02CH3. C02C2H5.e CN. f AcO. 2 H, s OCH7O 5.85 5.84 5.85 5.85 5.85 5.84 5.85 5.85 5.85 5.85 5.85 5.86 5.84

of all these compounds supports the trans quinolizidine structure.28a Saari, W. S.; Halczenko, W.; King, S. W.; Huff, J. R.; Guare, Jr., J. P.; Hunt, C. A.; Randall, W. C.; Anderson, P. S.; Lotti, V. J.; Taylor, D. A.; Clineschmidt, B. V. J . Med. Chem. 1983, 26, 1213. De Marinis, R. M.; Hieble, J. P.; Matthews, W. D. J . Med. Chem. 1983, 26, 1213. (a) Huff, J. R.; Anderson, P. S.; Baldwin, J. J.; Clineschmidt, B. V.; Guare, J. P.; Lotti, V. J.; Pettibone, D. J.; Randall, W. C.; Vacca, J. P. J . Med. Chem. 1985, 28, 1756. Doxey, J. C.; Roach, A. G.; Smith, C. F. C. Br. J . Pharmacol. 1983, 78, 489. Starke, K. Annu. Rev. Pharmacol. Toxicol. 1981, 21, 7. Langer, S. Z. Pharmacol. Rev. 1980, 32, 337. Vizi, E. S. Bog. Neurobiol. (Oxford) 1979, 12, 181. Vizi, E. S. Non-Synaptic Interaction Between Neurons; Wiley: Chichester, 1984. Crews. F. T.: Smith, C. B. Science (Washington, - D.C.) 1978, 202, 322. ' Crews, F. T.; Smith, C. B. J . Pharmacol. Exp. Ther. 1980,215, 143. Jirkovsky, J.; Protiva, M. Collect. Czech. Chem. Commun. 1963, 28, 2577. Normal berbane, [8aR*-(8a@,12aa,13aa)]5,6,8a,9,10,11,12,12a,l3,13a-decahydro-8~-dibenzo[a,g]quinolizine; alloberbane, [8aS*-(8aa,12aa,13aa)]5,6,8a,9,10,11,12,12a,l3,13a-decahydro-8~-dibenzo[a,g]qninolizine; epialloberbane, [8aS*-(8aa,12aa,13ap)]5,6,8a,9,10,11,12,12a,l3,13a-decahydro-8~-dibenzo[a,g]-

quinolizine.

1 H, m C-11 H 4.25 eq 4.25 eq 4.05 eq 3.75 ax 3.80 ax 4.35 eq 3.85 ax 3.65 ax 4.20 eq 3.75 ax 4.25 eq 4.20 eq 4.98 eq 5.03 eq 4.86 ax

OH 3500 3460 3460 3450 3480 3350 3400 3350 3450 3350 3390 3450

3350 3400 3400 3400 3480 g HBr salt.

half-width, Hz 5 6

IR, cm-' Bohlmann bands 2750-2800 2750-2800 2750-2800 2750-2850 2750-2850 2750-2800 2750-2800 2750-2800 2750-2800 2750-2850 2750-2800 2700-2750 2750-2850 2750-2850 2750-2850 2750-2800 2750-2850 2750-2850 2750-2800 2750-2800

3 H, s COOCH, 3.82 3.68

17OOc 17OOc 1720' 1710' 1700' 173OC 2230' 17051

17201 1720f

1710d

3 H, s AcO

10

30 20 6 24 30 11

3.70 3.71 3.80

30 8 8

10 10 30

1.92 1.90 1.96

Table 111. 13C NMR Data of 6a-c" 6a 6b 6c 105.29 c-1 105.02 105.37 145.88* 145.72* 145.84* c-2 145.62* 145.56* c-3 145.55* 108.29 108.27 c-4 108.35 127.79 127.80 C-4a 128.08 29.53 29.89 29.87 c-5 52.84 52.59 52.75 C-6 62.48 62.29 62.32 C-8 36.47 37.11' 36.70 C-8a 20.83 24.09 20.48 c-9 33.00 47.19 31.66" (2-10 66.32 c-11 65.57 65.87 37.28 36.46 c-12 49.69 34.44 33.86 C-12a 37.40' 34.95 31.25" 34.86 C-13 64.29 64.26 63.96 C-13a 131.79 131.72 131.40 C-13b COOCH, 51.83 51.75 OCH,O 100.57 100.56 100.56 COOkH, 175.69 176.24 u 6 (CDClJ. Assignments of signals (*), (+), or (") may be interchanged.

Generally it has been observed2sbthat the chemical shift data and the coupling constants of the carbinol proton (22) Szabo, L.; Toth, I.; Honty, K.; Toke, L.; Tamas, J.; Szantay, Cs. Chem. Ber. 1976, 109, 1724.

Berbanes: Selective a,-Adrenoceptor Antagonists Scheme I1

5

6

7 5-7 a b

CH;! CH2 C H2 C H2 CH2 CH2

C

d

e f 9 h i i k I

m n 0

P

Re R4 R5 OH COOCH3 H H H OH H COOCH3 H OH H H H H OH H COOCH3 H OH H H C O O W OH H C2H5 H H H OH CH2OH H H OH H CH2OH H OH H H H H OH H CN H OH CH3 H H H OH CH3 H H H OCOCH3 H H H OCOCH3 H H OCOCH3 H C03GH5 H H OH R2

R1

C2H5 CH2 CH2 H CH2 CH3 CH3 CH2 CH2 CH;!

R3

reflect the steric position of the OH group. The signals at higher chemical shift values (4.05-4.35 ppm) possessing smaller half-widths (5-10 Hz) are characteristic of equatorial protons (axial OH group), while the signals of axial protons appear at lower values (3.65-3.85 ppm) with wider half-widths (20-30 Hz). The steric position of C-10 and C-12 substituents in most cases can also be deduced from and Jll,12 or half-width of C-11 the coupling constants Jlo,ll Ha28c,d

Reduction of 3a-e yielded alcohols Ga-e,g,k, respectively, as the main products. The proton NMR data indicate the p-axial position for the C-11 OH. The equatorial orientation of the (2-10 and C-12 COOMe groups (compounds 6b and 6a, respectively) follows from the 13C NMR data (see Table 111). An axial COOMe substituent in either the C-10 or the (2-12 position would exert a shielding(23) Szabo, L.; Toth, I.; Szantay, Cs.; Radics, L.; Virag, S.; Kanyo, E. Acta Chim. Acad. Sei. Hung. 1979,100, 19. (24) Szabo, L.; Toth, I.; Toke, L.; Kolonits, P.; Szantay, Cs. Chem. Ber. 1976, 109, 3390. (25) Toth, I.; Szabo, L.; Bozsar, G.; Szantay, Cs.; Szekeres, L.; Papp, J. Gy. J.Med. Chem. 1984,27, 1411. (26) Szabo, L.; Honty, K.; Toke, L.; Szantay, Cs. Chem. Ber. 1972, 105, 3231. (27) Szabo, L.; Honty, K.; Toke, L.; Toth, I.; Szantay, Cs. Chem. Ber. 1972, 105, 3215. (28) (a) Bohlmann, F. Chem. Ber. 1985, 91, 2157. (b) Bhacca, N. S.; Williams, D. H. Applications of NMR Spectroscopy in Organic Chemistry; Holden-Day: San Francisco, 1964. (c) Pierre, L.; Vinceus, M.; Vidal, M. Bull. SOC.Chirn. Fr. 1971, 1755. (d) Casu, B.; Reggiani, M.; Gallo, G. G.; Vigevani, A. Tetrahedron 1966,22, 3061.

Journal of Medicinal Chemistry, 1987, Vol. 30, No. 8 1357

gauche effect on carbon C-8a in comparison with the respective value of 6c (6c, 36.47 ppm; 6b, 36.70 ppm; 6a, 37.11 ppm). Upon reduction of 3c, 6d was isolated as the minor product; the lH NMR spectrum indicates an a-equatorial OH group. The major product of reduction of ketone 2c was 5c, having a (2-11@-equatorial hydroxyl group. Similarly, reduction of keto esters 2a and 2b gave rise to alcohols 5a and 5b with C-11 @-equatorialhydroxyl groups and aequatorial COzMe groups. Upon reduction of keto ester 2b, in addition to alcohol 5b as the main product, a minor product 5f was separated by flash chromatography. Alcohol 5f, as expected, contains the C-11 OH group in an a-axial position. Reduction of compounds 4a and 4c gave different stereochemical results. Reduction of the keto ester 4a produced alcohol 7e as the major product, containing the C-11 OH in an a-axial position. Ketone 4c gave alcohol 7c with the C-11 0-equatorial hydroxyl group. In order to prepare dialcohols 6h,i and 5h for pharmacological studies, ketones 3a,b and 2a were reduced with a large excess of NaBH4. For the same purpose, alcohols 61,266c, and 6d were acetylated to give 6m and 6n with a @-axialand 60 with an a-equatorial acetoxy group. Conversion of methyl ester 6a to the corresponding ethyl ester 6p was carried out in the presence of 4-MeC6H4S03H catalyst in boiling EtOH. BBr, dealkylation of 6c gave 6j. Biological Results and Discussion Antagonist activities of these compounds on presynaptic (a2)and postsynaptic (a1)adrenoceptors were determined in the rat isolated field stimulated vas deferens.29 In this preparation, presynaptic a,-adrenoceptor agonists such as xylazine characteristically inhibit stimulation-evoked contractions with an EDb0of 8 X lo-* M. Phenylephrine, a relatively pure a,-adrenoceptor agonist, produced dosedependent contractions. These pre- and postsynaptic effects were preferentially blocked by selective inhibitors of al- and a,-adrenoceptors. Quantitative analysis of the antagonism of both selective agonists by the compounds studied resulted in a calculation of the ratio of pA2(a1): pA2(a2)as a measure of the selectivity for either a-adrenoceptor site. Of the reference drugs reported in Table IV, idazoxan was the most potent and selective a,-antagonist in the rat vas deferens, while yohimbine was found to be less potent and selective. However, several members of the berbane series (Table IV) showed a highly selective a,-adrenoceptor antagonist activity, with very low potency on postsynaptic a,-adrenoceptors in the vas deferens. There was a significant difference between the alloberbane (6) and epialloberbane (7) skeletons as far as their potency on a,-adrenoceptors is concerned: alloberbanes proved to be about 10 times more potent. Of the compounds tested, [8aS*-(8aa,12a(u,13aa)]-lla-hydroxy-5,6,8a,9,10,11,12,12a,l3,13a-decahydro-8H-benzo[g] -1,3-benzodioxolo[ 5,6a]quinolizine (6d) proved to be the most selective a2adrenoceptor antagonist. Compound 6c also proved to be very active as an a,-adrenoceptor antagonist (a1:a2ratio = 602). The selectivity ratios for 6d, idazoxan, yohimbine, and phentolamine were 1659, 117.5, 4.7, and 0.57, respectively. Prazosin possessed by far the highest affinity for the postsynaptic a,-adrenoceptors which were previously stimulated with phenylephrine, followed in decreasing order of potency by prazosin > phentolamine > yohimbine > idazoxan > 6d, respectively. (29) Vizi, E. S.; Somogyi, G. T.; Hadhazy, P.; Knoll, J. NuunynSchmiedeberg's Arch. Pharmacol. 1973, 280, 79.

1358 Journal of Medicinal Chemistry, 1987, Vol. 30, No. 8

Vizi et al.

Table IV. Presynaptic az-Adrenoceptor and Postsynaptic al-Adrenoceptor pA2 Values in Rat Vas Deferens and Presynaptic a2-Adrenoceptor pAz Values in the Longitudinal Muscle Strip of Guinea Pig Ileum vas deferens0 pA2 f SD long. muscle strip pAz f SD' compd a,-adrenoceptorb al-adrenoceptorc selectivity:d a l a 2ratio xylazine 1-NE 5a 7.41 f 0.10 4.86 f 0.83 354.8 6.40 f 0.51 6.17 f 0.81 5b 5.64 f 0.30 5.98 f 0.98 0.46 4.82 f 0.44 6.55 f 0.88 5c 7.12 f 0.02 5.02 f 0.12 125.9 7.02 f 0.13 5.95 f 0.37 5.25 f 0.30 9.1 6.72 f 0.89 6.31 f 0.81 5h 6.21 f 0.11 6a 7.18 f 0.36 6.18 f 0.44 10 7.86 f 0.19 7.80 f 0.46 4.80 f 0.22 10 6.61 f 0.29 7.82 f 0.15 6b 5.80 f 0.16 4.85 f 0.15 602 0.13 f 0.08 7.62 f 0.34 6c 7.63 f 0.44 4.95 i 0.11 1659 8.07 f 0.20 8.26 f 0.46 6d 8.17 f 0.01 6g 6.41 f 0.08 257 6.32 f 0.22 6.69 f 0.13 6h 5.60 f 0.63 6i 5.91 f 0.20 81.2 6.51 f 0.41 6.30 f 0.31 6j 5.74 f 0.11 54.9 125.8 6.67 f 0.78 6k 6.10 f 0.07 204.1