Structure-activity relationship of strychnine derivatives modified in the non-aromatic part. Iskander, G. M., Bohlin, L. 1978. Acta pharmaceutica Suecica, 15(6), 431–8. PMID 749546
Structure-activity relationship of strychnine derivatives modified in the non-aromatic part
GEORG M. ISKANDER and LARS BOHLIN
Department of Pharmacognosy, Faculty of Pharmacy, Biomedical Center, Box 579, S-751 23 Uppsala, Sweden
ABSTRACT The CD₅₀ (convulsant dose) and LD₅₀ of fourteen synthesized strychnine derivatives were determined after subcutaneous injection in mice. Various changes in the non-aromatic part of the strychnine molecule generally led to decreased activity. The quaternary Nₙ-alkyl salts of these derivatives were almost purely muscle relaxants.
In a previous paper from this department [1], the convulsant and muscle-relaxant effects of 2- and 3-substituted strychnine derivatives were studied. Previously, some studies have been reported which compared the structure of strychnine derivatives and the pharmacological activity.
Thus, Joachimoglou [2] showed that hydrogenation of strychnine decreased the activity. Many authors [3-8] reported the curarizing effect of quaternary Nₙ-alkyl salts of strychnine. Several studies [9-12] described the preparation and the pharmacological activity of various betaines of strychnic acid. Karrer et al. [13] studied the curare activity and toxicity of some alkyl chloride derivatives of strychnidine and dihydrostrychnidine in frogs, rabbits and mice. Szabó et al. [7, 14] ran two structure-activity studies with several strychnine derivatives. They showed that the lactam group was responsible for the characteristic strychnine effect. By transformation of the alkaloid into quaternary compounds, the strychnine-like actions were suspended. However, other modifications of the molecule influenced the strychnine effect only to a lesser degree.
To obtain a more detailed picture of the active sites of the strychnine molecule and to establish the structure-activity relationship of these derivatives, we have now correlated the changes in convulsant activity, and other effects such as muscle-relaxation, with skeletal changes in the non-aromatic part of the strychnine molecule.
Experimental
Chemistry
Melting points were determined with a Leitz Mikroskopheiztisch 350. IR spectra were recorded with a Perkin-Elmer 157G spectrophotometer (KBr-discs). The NMR spectra were recorded on a Varian HA 100 D. The relevant chemical shifts are given in ppm from tetramethylsilane (internal standard). Mass spectra were obtained with an LKB 9000 instrument at 70eV with direct inlet. The liquid chromatograms, for purity check, were carried out for the free bases on an instrument having a Waters model 6000 pump, a U6k injector and a Varian UV detector (254 nm) with the solvent system chloroform-methanol (containing 2 % conc. ammonia) 93:7 on a μ-Porasil column (30 cm x 4.5 mm ID) and a flow rate of 3 ml/min.
Strychnine derivatives synthesized and used for pharmacological studies
Strychnidine (II). The substance is prepared by the reduction of strychnine with excess of lithium aluminium hydride (5 mol equiv.) in refluxing tetrahydrofuran. Needles from methanol; m.p. 254°C (lit. 256°C) [13]. IR: 750, 1230, 1470, 1597, 1610, 2840 and 2930 cm⁻¹. MS: m/e 320 (M⁺; 100 %), 290 (13), 289 (12), 196 (11), 194 (10), 180 (24), 170 (12), 156 (16), 144 (27), 143 (18), 130 (12). NMR (CDCl₃): 5.84 (1H, m, C₂₂ olefinic proton); 6.38—6.87 (2H, m, aromatic protons); 6.95—7.22 (2H, m, aromatic protons).
21,22-Dihydrostrychnine (III). The substance is prepared by the catalytic hydrogenation of strychnine, as reported earlier. Needles from aqueous methanol (50 %); m.p. 218—220°C (lit. 220—222°C) [17]. IR: 755, 1100, 1400, 1460, 1475, 1595, 1670 and 2900 cm⁻¹. MS: m/e 336 (M⁺; 100 %), 165 (20), 164 (39), 144 (10), 143 (20), 142 (CDCl₃): 7.17—7.37 (3H, m, C₁₁,₁₂,₁₃ aromatic protons); 7.97—8.20 (1H, m, C₄ aromatic proton).
21,22-Dihydrostrychnidine (IV). Dihydrostrychnine (0.672 g; 0.002 mol) is extracted in a Soxhlet apparatus with dried tetrahydrofuran and the hot solution dropped into a stirred refluxing suspension of LiAlH₄ (0.284 g; 15 mol equiv.) in tetrahydrofuran (200 ml). After 5 h reflux, the reaction mixture is cooled; water (1.5 ml), NaOH solution (1.5 ml, 10 %) and finally water again (4.5 ml) are added successively in order to destroy the excess of LiAlH₄. The hydroxides of lithium and aluminium are filtered off and the filtrate evaporated
in vacuo. The residue (0.58 g) is purified by column chromatography on alumina (Woelm, basic, activity III). Elution with chloroform-benzene (19:1) afforded the product (0.4 g; 59 %) as colourless plates from methanol; m.p. 221—224°C (lit. 212—214°C) [17]. IR: 750, 1060, 1095, 1305, 1460, 1595, 1615 and 2900 cm⁻¹. MS: m/e 322 (M⁺; 100 %), 292 (15), 180 (15), 156 (14), 144 (15), 143 (17).
21,22-Dihydroxy-21,22-dihydrostrychnine (V). This compound is prepared by the method of Kogure and Kotake [18]. Colourless plates from methanol: m.p. 228—230°C (after drying in high vacuum) (lit. 228—230°C) [18]; IR: 750, 1095, 1380, 1480, 1595, 1665 (C=O = 0), 2900, 3070 and 3300 cm⁻¹ (C₂₁,₂₂-OH). MS: 368 (M⁺; 100 %), 312 (26), 311 (19), 197 (22), 196 (22), 168 (20), 144 (26), 143 (28), 130 (21). NMR (DMSO-D₆): 4.20—5.20 (2H, broad peak, disappeared on shaking with D₂O, C₂₁,₂₂-OH); 7.08—7.54 (3H, m, C₁₁,₁₂,₁₃ aromatic protons); 7.84—8.05 (1H, m, C₄ aromatic proton).
21-Hydroxy-22-oxo-21,22-dihydrostrychnine (VI). The compound is prepared by the method of Prelog and Kathriner [19]. Colourless needles from ethanol; m.p. 238—246°C (lit. 239—241°C) [19]. IR: 760, 1130, 1410, 1475, 1600, 1665 (C=O = 0), 1710 (C₂₁ = 0), 2910 and 3400 cm⁻¹ (C₂₂-OH). MS: 366 (M⁺; 100 %), 338 (61), 295 (41), 280 (30). NMR (CDCl₃): 4.65 (1H, broad peak, C₂₂-OH); 7.07—7.43 (3H, m, C₁₁,₁₂,₁₃ aromatic protons); 7.92—8.28 (1H, m, C₄ aromatic proton).
16-Alkoxystrychnines. These substances are prepared by crystallizing pseudostrychnine from the appropriate alcohol [20, 21].
16-Methoxystrychnine (VIII). Needles from methanol; m.p. 195°C (decomp.) (lit. 196—198°C) [21]. IR: 760, 1395, 1460, 1475, 1590, 1680 and 2900 cm⁻¹. MS: m/e 364 (M⁺; 70 %), 348 (13), 334 (100), 333 (72), 332 (74), 193 (13), 192 (13), 191 (15), 144 (30), 143 (17), 130 (18). NMR (CDCl₃): 3.22 (3H, s, C₁₆-OCH₃); 5.90 (1H, m, C₂ olefinic proton); 6.95—7.30 (2H, m, C₁₁,₁₂ aromatic protons); 7.80 (1H, q, C₁ aromatic proton, J₁,₂ = 2 Hz and J₁,₃ = 8 Hz); 8.12 (1H, q, C₄ aromatic proton, J₄,₃ = 2 Hz and J₄,₂ = 8 Hz).
16-Ethoxystrychnine (IX). Needles from ethanol; m.p. 215—217°C (lit. 217—219°C, decomp.) [21]. IR: similar to 16-methoxystrychnine. MS: m/e 378 (M⁺; 14 %), 348 (15), 334 (75), 333 (42), 332 (100), 303 (25), 144 (25), 143 (15), 130 (22). NMR (CDCl₃): 1.28 (3H, t, C₁₆-OCH₂-CH₃, in the ethyl group); 3.72 (2H, q, C₁₆-OCH₂-CH₃, J = 7 Hz); 5.88 (1H, m, C₂ olefinic proton); 6.95—7.30 (2H, m, C₁₁,₁₂ aromatic protons); 7.86 (1H, q, C₁ aromatic proton, J₁,₂ = 2 Hz and J₁,₃ = 8 Hz); 8.12 (1H, q, C₄ aromatic proton, J₄,₃ = 2 Hz and J₄,₂ = 8 Hz).
16-Isopropoxystrychnine (X). Needles from 2-propanol; m.p. 157—160°C. IR: Similar to 16-methoxystrychnine. MS: m/e 392 (M⁺; 43 %), 349 (68), 334 (100), 333 (36), 332 (54), 210 (45), 185 (29), 144 (27), 143 (29), 130 (25). NMR (CDCl₃): 1.22 (6H, m, C₁₆-OCH(CH₃)₂ in isopropyl group); 3.78 (1H, m, C₁₆-OCH(CH₃)₂); 5.90 (1H, m, C₂ olefinic proton); 7.00—7.45 (2H, m, C₁₁,₁₂ aromatic protons); 7.92 (1H, q, C₁ aromatic proton, J₁,₂ = 2 Hz and J₁,₃ = 8 Hz); 8.10 (1H, q, C₄ aromatic proton, J₄,₃ = 2 Hz and J₄,₂ = 8 Hz).
Quaternisation of the free bases
The following alkaloids are quaternized: strychnine, strychnidine and 21,22-dihydrostrychnine. The alkaloid (0.1—0.2 g) is dissolved in chloroform (2 ml, A. R. grade) that was previously saturated with methyl chloride gas at room temperature. The mixture is then left to stand overnight and the crystalline product obtained is filtered, washed with dry chloroform and dried (yield 80—90 %). The IR, MS and NMR spectra of these salts are similar to those of the parent alkaloids.
19-Methylstrychnin-19-ium chloride (XII). Colourless leaflets from ethanol-ethyl acetate; m.p. 290°C (lit. 288.5—291°C) [13]. MS: m/e 334 (M⁺-15; 100 %).
19-Methylstrychnidin-19-ium chloride (XIII). Colourless needles from ethanol-ethyl acetate; m.p. 300°C (decomp.) (lit. 310°C) [13]. MS: m/e 320 (M⁺-15; 100 %).
19-Methyl-21,22-dihydrostrychnin-19-ium chloride (XIV). Needles from ethanol-ethyl acetate; m.p. 322—325°C. MS: m/e 336 (M⁺-15; 100 %).
Pharmacology
The methods used for determination of CD₅₀, LD₅₀ and
in vitro determination of muscle-relaxant effect have been described in a previous paper [1].
Results and discussion
The formulae of the derivatives tested are given in Figs 1 and 2 and the results are summarized in Table 1.
The removal of the C₁₉ carbonyl function as in strychnidine (II) decreased the observed convulsive and lethal effects. This is in accord with the findings of Szabó and Weimann [7] who showed that the lactam group is responsible for the characteristic strychnine effect. Also, several authors [9, 11, 12] reported that the convulsive action of strychnine is due chiefly to the ketopiperidine nucleus. Sandberg and Kristiansson [15] reported that the alkaloid diabolene, with an open amide ring, had a very low convulsant activity and toxicity compared to strychnine. Also, Szabó
et al. [14] showed that an acid amide link in the dihydroindole nitrogen is necessary for strychnine to produce cerebrospinal convulsion.
It seems, therefore, that the presence of an amide group is not only necessary but it should be part of a ring, as in the strychnine molecule, in order to give the optimal activity.
The double bond in 21,22-position may be another site of action in the strychnine molecule. We found that 21,22-dihydrostrychnine (III) had about 15 times less activity than strychnine. This confirms the result from the first study in this series [15]. Verpoorte
et al. [16] have determined 13C-NMR spectra of strychnine and strychnine derivatives and shown that the loss of the double bond in icajine caused a more flexible 7-membered ring and a less rigid nature of the Nₐ-bridge. Presumably, the stereochemical change in the hydrogenated compound (III) may be one reason for its decreased activity.
21,22-Dihydrostrychnidine (IV) is about twice as active and four times as toxic as strychnidine itself. This result is very remarkable because strychnine, brucine, icajine and vomicine, all show a marked decreased activity and toxicity upon hydrogenation [15]. However, the two structural changes in (IV) did not seem to reinforce each other as expected.
When hydroxyl groups are introduced at the 21,22-positions as in 21,22-dihydroxy-21,22-dihydrostrychnine (V) and 21-hydroxy-22-oxo-21,22-dihydrostrychnine (VI), a combined convulsant and muscle-relaxant effect appeared. In no case did (V) and (VI) have the typical extension component of the tonic convulsion. In the study by Sandberg and Kristiansson [15], an introduction of an epoxy function in the 21,22-positions decreased the convulsant and toxic activity but no muscle-relaxant effect appeared. The decreased activity supports the previous suggestion that the 21,22-double bond might be an active site for the receptor.
The new muscle-relaxant effect observed, as described here, may be due to an increased water solubility of the parent alkaloid and thus diminishing passage through the blood-brain barrier, or alternatively a formation of pseudosalts (viz. the hydroxyl group relative to the nitrogen atom) which may induce some sort of hydrogen bonding.
The 16-alkoxystrychnines (VIII, IX, X) produce both clonic and tonic convulsions. The convulsive activity and lethal effects decreased with increasing size of the alkyl substituent (Table 1). The unalkylated 16-hydroxy compound (VII) is the most active in this series.
The salts of strychnine and its derivatives (XII, XIII), in which the tertiary nitrogen is quaternized, showed only a muscle-relaxant effect. 19-Methyl-21,22-dihydrostrychnin-19-ium chloride (XIV) also showed convulsive effect but without the typical extension component of the tonic convulsion. An explanation for this may be that quaternization of strychnine prevents passage through the blood-brain barrier and thus stops the compound reaching the active site of action in the central nervous system.
Many studies [3—6, 9, 12] have been carried out on the muscle-relaxant effect or quaternary Nₐ-alkylsalts of strychnine. Busquet and Vishniac [4] reported that strychnine iodomethylate did not produce convulsions even in lethal doses, which confirms our own results with compounds (XII) and (XIII). Hunter [5] also studied the curarisation effect of "strychnine dimethylsulphate" in mice, dogs, and cats, and found that it had a very short duration and was accompanied by a significant blood pressure depression; the curarization effect was however antagonized by neostigmine.
In our study, 19-methylstrychnin-19-ium chloride (XII) had an LD₅₀ value of 30 mg/kg when injected intraperitoneally in mice, and again the effect was antagonized by neostigmine.
The reversed order and great difference in LD₅₀ doses between 19-methylstrychnin-19-ium chloride (XII) and 19-methylstrychnidin-19-ium chloride (XIII) is difficult to explain but Karrer
et al. [13] also obtained the same results for (XII) and (XIII) in mice, when injected intraperitoneally.
The strychnine derivatives (V, VI, XII, XIII, XIV)¹ which showed muscle-relaxant activity (positive screen-grip test) [23] were tested on the rat diaphragm preparation [1]. The 50 % inhibition values include: (VI) 64 µg/ml, (XII) 85 µg/ml and (XIV) 96 µg/ml, and the effect was not antagonized by neostigmine, thus confirming previous findings [1].
Walther [24] reported the same non-antagonistic effect of neostigmine for strychnine on the rat diaphragm muscle. No explanation can be offered for this apparent difference in effect by neostigmine
in vitro and the above observed activity
in vivo.
In conclusion, it is worth observing that many factors can be responsible for the differences in effects between strychnine and its derivatives so far studied, such as the lipid solubility and stereochemical changes of the molecule; but because of the complex effects induced by the strychnine molecule, both central and peripheral, it is not possible, within the scope of this investigation, to give more detailed explanations of the observed changes in activity.
Acknowledgement
We want to thank Professor Finn Sandberg for valuable discussions during the work. We are also grateful to Mr. Jan Strömbon and Mrs. Cilly Stolt for their help with the technical and animal work.
References
1. L. Bohlin, Y. Ali and G. M. Iskander,
Acta Pharm. Suec. 12, 461 (1975)
2. G. Joachimoglou,
Z. Angew. Chem. 36, 93 (1923)
3. C. Bartorelli,
Arteno Parmense 6, 318 (1934)
4. H. Busquet and C. Vishniac,
C. R. Soc. Biol. (Paris) 127, 664 (1938)
5. A. R. Hunter,
Anaesthesia 7, 206 (1952)
6. R. Hazard, J. Cheymol, P. Chabrier and H. Drouin,
J. Physiol. (Paris) 49, 198 (1957)
7. L. Szabó and L. Weimann,
Acta Pharm. Hung. 35, 26 (1965)
8. S. L. Cowan and H. R. Ing,
J. Physiol. 82, 432 (1934)
9. G. Pappalardo,
Farm. Sci. Tec. 2, 401 (1947)
10. S. P'Abbate,
Boll. Soc. Ital. Biol. Sper. 26, 340 (1950)
11. G. Ottaviano,
Boll. Soc. Ital. Biol. Sper. 26, 338 (1950)
12. G. Ottaviano,
Boll. Soc. Ital. Biol. Sper. 25, 1495 (1949)
13. P. Karrer, C. H. Eugster and P. Waser,
Helv. Chim. Acta 32, 2381 (1949)
14. L. Szabó, L. Weimann and O. Clauder,
Acta Pharm. Hung. 38, 84 (1968)
15. F. Sandberg and K. Kristiansson,
Acta Pharm. Suec. 7, 329 (1970)
16. R. Verpoorte, P. J. Hylands and N. G. Bisset,
Org. Magn. Reson. 9, 567 (1977)
17. A. E. Oxford, W. H. Perkin and R. Robinson,
J. Chem. Soc. 2389 (1927)
18. A. Kogure and M. Kotake,
J. Inst. Polytech. Osaka City Univ. Ser. C. 2, 39 (1951)
19. V. Prelog and A. Kathriner,
Helv. Chim. Acta 31, 505 (1948)
20. K. Warnat,
Helv. Chim. Acta 14, 999 (1931)
21. A. S. Bailey and R. Robinson,
J. Chem. Soc. 704 (1948)
22. N. G. Bisset, C. G. Casinovi, C. Galeffi and G. B. Marini Bettolo,
Ric. Sci. Parte 2. Sez. B 35, 6 (3), 273 (1965)
23. F. Sandberg, R. Verpoorte and A. Cronlund,
Acta Pharm. Suec. 8, 341 (1971)
24. H. Walther,
Acta Biol. Med. Ger. 11, 93 (1963)
Received September 18, 1978.
¹ (V) and (XIII), too small amounts to be tested.
Table 1
Pharmacological results of strychnine derivatives tested
| No. | Strychnine derivative | Muscle relaxant effect | Clonic convulsion | Tonic convulsion flexion | Tonic convulsion extension | CD₅₀ mg/kg | LD₅₀ mg/kg | No. of animals |
|---|
| I | Strychnine* | - | + | + | + | 0.43 | 0.47 | 30 |
| II | Strychnidine | - | + | + | + | 47.9 | 80.2 | 56 |
| III | 21,22-dihydrostrychnine | - | + | + | + | 6.0 | 7.28 | 34 |
| IV | 21,22-dihydrostrychnidine | - | + | + | + | 18.2 | 20.0 | 39 |
| V | 21,22-dihydroxy-21,22-dihydrostrychnine | + | (+)* | + | - | 206 | 230 | 40 |
| VI | 21-hydroxy-22-oxo-21,22-dihydrostrychnine | + | (+)* | + | - | 175 | 204 | 37 |
| VII | 16-hydroxystrychnine* | - | + | + | + | 1.10 | 1.21 | 36 |
| VIII | 16-methoxystrychnine | - | + | + | + | 3.12 | 3.48 | 32 |
| IX | 16-ethoxystrychnine* | - | + | + | + | 4.57 | 4.84 | 60 |
| X | 16-isopropoxystrychnine | - | + | + | + | 6.61 | 7.24 | 32 |
| XI | Strychnine-N-oxide* | - | + | + | + | 78.7 | 80.0 | 33 |
| XII | 19-methylstrychnin-19-ium chloride | + | - | - | - | > 250 | - | 28 |
| XIII | 19-methylstrychnidin-19-ium chloride | + | - | - | - | 79.4 | - | 32 |
| XIV | 19-methyl-21,22-dihydrostrychnin-19-ium chloride | + | + | + | - | 110 | 120 | 30 |
a Data from ref. [15].
b Short clonic convulsions (< 5 sec.) upon touch.
