RELATION BETWEEN CHEMICAL CONSTITUTION EFC. 1103 The Relation Bet ween Ch.emical Constitution and Physiological A c t io n. A Lecture delivered before the Chemical Society 011 December 6th, 1917. By FRANK LEE PYMAN. THE study of the relation between chemical constitutioii and physiological action is a branch of research which has a definite place in the investigation of medicinal substances. Chemical research on a drug begins with the attempt t o isolate the priii-ciple to which its physiological action is due and when this has proved successful the next step is the determination of the con-stitution of the active principle by analytic and synthetic methods. The knowledge is thus gained that some compound of known chemical structure has a particular physiological effect and the way is then clear for the study of the relation between chemical constitution and physiological action by the preparation of a number of substances related to the parent compound and com-parison of their actions on the living organism.The history of quinine affords an illustration of this sequence. Cinchona bark was employed as a remedy for malarial fevers in the fifteenth century. Later on the alkaloid quinine was isolated and recog-nised as the chief active principle of the drug. Chemical investi-gation eventually established the structure of the alkaloid and attempts have since been made to improve or vary its medicinal properties by slight alterations of the molecular structure such as reduction of the vinyl group and replacement of the methoxyl by higher alkyloxyl groups.Moreover a number of compounds with the following general formula have been synthesised which, in common with quinine are distinguished as powerful febrifuges combining a low toxicity to man and a high toxicity to infusoria and paramoecia. The study of the relation between chemical constitution and physiological action may have various objects philosophic o 1104 PYMAN THE RELATION RETWEEN CHEMICAL practical. From the purely scientific point of view it is of great interest to determine the change in physiological action resulting from modification of the chemical structure of an active compound and to elucidate the groups within its molecule to which its pre-dominant physiological action is due ; whilst from the practical standpoint the work may be directed to the physiological, chemical or physical improvement of a drug-for instance it may be desired to eliminate some undesirable secondary effect while maintaining the chief physiological action of the drug or to pre-pare a derivative more stable or more soluble than the parent compound.The relatioil betweeii chemical constitutioii and physiological action has a significance in the discovery of new drugs similar to the relation between chemical constitution and colour in the dis-covery of new dyes. I n the latter case however a single physical property the absorption of light of different wave-lengths is studied ; whereas the term physiological action has no simple meaning but covers any action on the living organism. The bactericidal action of phenol the hypnotic properties of diethyl-barbituric acid and the local anzesthetic action of cocaine are ex-amples of physiological action which are no more comparable with one another than are the chemical structures of the three drugs.Moreover it should be borne in mind that the same superficial signs of physiological action may be due to different causes. Purga-tion for instance is caused by saline cathartics such as magnesium sulphate which act by increasing the bulk of fluid in the intes-tines; and by vegetable purgatives such as derivatives of anthra-quinone which act by irritating the epithelium of the intestines, thus promoting peristalsis. The difficulty of generalisation in the relation under discussion may be instanced by the effect of intro ducing a methyl group into the ortho-position of a phenol, where in the case of the pareiit compuiid the resulting 0-cresol is a more powerful germicide than phenol whilst a similar substi-tution in phydroxy-B-phenylethylamine leads to a substance, 4-hydroxy-/3-m-tolylethylamine which has only one-half of the llressor properties of the parent compound.The subject of this paper must therefore be subdivided eventually into a number of fragments on the relation between chemical constitution and a particular physiological effect ; but before proceeding with these, some general remarks on the action of drugs may be made. An example of physiological action which everyone can appreciate without special knowledge is the effect of certain volatile compounds on the terminations of the olfactory nerves producing the sense of smell.Many compounds of similar constitutioll hav CONSTITUTION AND PHYSIOLOOICAT~ ACTION. 1 105 the same type of smell-for inst'ance the lower fatty acids whilst each member may have a specific odour-which in this particular case serves to distinguish the individual members from formic to valeric acids. Sense of taste also provides an occasional means of discrimination not only between side chains of different lengths -p-ethoxyphenylcarbamide (dulcine) being sweet whilst the higher alkylphenylcarbamides are not-but also in certain cases between stereoisomerides-cl-histidine for example tasting sweet whilst I-histidine is tasteless. It is noteworthy that stereo-chemical influences often have profound effects on the physio-logical actions of quite different classes of compouiids particularly in actions on nerve-endings as Cushiiy has poiiitetl out ( L a n c e / , September 9th 1916 459) ; thus 7-hyoscyamine has about 100 times the niydriatic action of the d-variety and [-adrenaline has many times the pressor eflect of d-adrenaline.In the case of pilocarpine which contains Cwo asyinrrietric carbon atoms a chaiige of sign of one of these results in the formation of the stereoisomeric isopilocarpine which has only a fraction of the activity of pilo-carpine itself. The asymmetry of a nitrogen atom may also con-dition a difference in pllysiological action ; when Icanadine is methylated a mixture of the a- and P-methochlorides is obtained, the isomerism of which is due t o the asymmetry of the nitrogen atom; these produce a typical curare effect (paralysis of nerve endings in voluntary muscle) in the frog the &salt being how-ever twelve times as powerful as the a-salt.Stereoisomerides, however do not always show large diff ereiices in physiological action even in actions on nerve-endings ; cl- and I-homatropine differ little from each other and dZ-homatropine in mydriatic action, whilst &cocaine a stereoisomeride of natural lzvorotatory cocaine, has a local anasthetic action which although quicker and more intense is also more evanescent than that of cocaine. Very little is known about the cause of the variatioii in the physiological actions of stereoisomerides but recently an explana-tion has been suggested by Windaus (Ntrcl~r.K. Ges. Wiss. Cfittingen 1916 301) for the different physiological behaviour of the stereoisomerides 8- and 6:-eholestanol. These compounds differ greatly in their power of inhibiting the hzmolytic action of saponins such as digitonin the former having this property in a high degree whilst the latter has oiily slight preventive proper-ties. Now the &compound has been found to combine with digitonin t o give an almost inactive additive product whilst ocholestanol does not combine with digitonin. This case is of special importance because of the close relation of the cholestanols to cholesterol a constituent of the livin 1106 PYMAN THE RELATION BETWEEN CHEMICAL organism; it may be that a similar difference in the ability of stereoisonierides t o combine with constituents of the nerve cells is the cause of their different action in other cases also.A point to be considered in connexion with the relation between chemical constitution and physiological action is the effect of the physical and chemical properties of the substance on its distribu-tion in the organism. The influence of physical properties such as aolubility in different media may be of great importance as in the case of hypnotics where Meyer and Overton found that the narcotic effect of a series of aliphatic compounds on tadpoles was proportional to the partition coefficients of their solubilitaies in oil and water. An indication of the effect of chemical properties on the distribution of drugs in the organism was afforded by the work of Ehrlich (compare ‘‘ v.Leyden-.Festschrift,” 1898). He showed that basic dyes such as methylene-blue stained tha grey nerve substance whereas their sulphonic acids did not and this difference suggested that bases which are liberated in the blood-stream by the alkali are extracted by the nerve substances whilst their sulphonic acids remain in solution as alkali salts. Similarly, the facility with which an alkaloid is extracted from aqueous alkaline solutions by immiscible solvents may reasonably be sup-posed to affect its distribution in the organism and Ehrlich gave examples of the change of action when certain basic drugs are converted into derivatives containing a free acid grouping or into quaternary salts. I n the case of alkaloids it is a general rule that the introduc-tion of a free carboxyl group into the molecule profoundly modifies the physiological action of the parent compound.Benzoylecgonine, of which cocaine is the methyl ester has no local anzesthetic action; quitenine the acid obtained by oxidising the vinyl group of quinine to a carboxyl group is non-toxic but regains its toxicity on ethylation ; the lactone pilocarpine becomes inactive on the addi-tion of a molecule of alkali hydroxide which forms the alkali salt of the corresponding hydroxy-acid whilst a similar loss of physio-logical activity is shown by a series of tropeines containing a lactone group which lose their atropine-like action on the addition of a molecule of alkali hydroxide. The formation of quaternary salts likewise very largely affects the physiological properties of alkaloids and was the subject of study many years ago by Crum-Brown and Fraser.To give an example from more recent work Laidlaw (Biochem. J. 1910 5, 243) found that 6 7-dimethoxy-3 4-dihydroisoquinoline (I) had a strychnine-like effect whilst its methochloride (11) was devoid o CONSTITUTION AND PHYSIOLOGICAL ACTION. 1 107 this property which however reappeared in its reduction product G 7-climethoxy-2-methyltetrahydroisoquinoline (111). CH CH CH2 /\/\N /\/\NMeCl MeO/\/)Nl\le MeO\/\/CH2 nzeol I I 'CH2 CH2 (111. ) CH2 3feo\/\/ (11. j CH2 ( 1 9 1 A similar relation was observed with papaverine (IV) its metho-chloride (V) and the reductioii product of the latter laudanosine $JH2*C6H3(0Mf1).2 f H,*C,H,(OBle), C C M ~ o / \ / \ N w.1 p32*CGH,(OM42 CH M eO/\/\NMe M"o\/\/ I ICH2 WI.1 CH2 (VI) ; here also the tertiary bases were characterised by strychnine-like action which was not obtained with papaverine niethochloride.Having directed attention to the complication introduced into the relation under discussion by the effect of physical and chemical properties on the distribution of drugs we may now consider certain difficulties of generalisation. We have seen that certain compounds closely allied in chemical constitution differ remarkably in their action and we find on the other hand groups of substances which are almost indistinguishable physiologically but have little in common from the point of view of chemical constitution; one such group is formed by the alkaloids nicotine lobeline and cytisine, another by muscarine arecoline and pilocarpine.Experience has shown however that the members of a group of chemical com-pounds of similar constitution often resemble one another in physiological action and in such cases it is of interest to observe the effect of slight alterations in chemical structure. Many such investigations have been carried out and will be known to you. To-night I shall confine my attention to a few lines of work i 1108 PYMAN THE EELATION BETWEEN CHEMICAL this field which have been carried out or materially advanced wit.hin the last ten years selecting especially those with which the Wellcome Research -Laboratories have Trope ines. The compounds knowii as tropeiiies the amino-alcohol tropine.Atropine, group is the dl-tropyl ester of tropine, CH2-CH-CH2 I I I been a,ssociated. are the and the acyl derivatives of parent inember of the readily yields this sub-I NMe CH*O*CO*CHYh*CH,*Of1. CH2-CH-CH, I t 1 stance on hydrolysis. By esterifyiiig tropine with other acids, tropeines containing different acyl groups may be prepared. A number of these have been examined physiologically the best known being homatropine; the mandelyl ester of tropine which was described by Ladenburg in 1883. I n this paper I propose to give an account of the work on the relation between chemical constitution and physiological action in the tropeines carried out some years ago by Dr. H. A. D. Jowett and myself with the co-operation of Dr.H. H. Dale F.R.S. (Srventh Interlfat. C'ongr. Appl. Chenz. 1909 IVA 1 335) in coiitinuatioii of an investi-gation coinmeiiced by Dr. Jowett in collaboration with Dr. C. R. Marshall. The tropeines appeared to us to be specially suitable for a study of the relation between chemical constitution and physiological action since they are easily prepared give neutral salts readily soluble in water and can be tested physiologically under uniform conditions. Their salts were dissolved in distilled water to give solutions equivalent in tropine content to a 1 per cent. solution of homatropine hydrobromide and the niydriatic effects of these solutions were then compared. By means of two pipettes delivering drops of equal size a drop of one of the two solutions to be compared was allowed to fall into the right eye of a cat and a drop of the other exactly a t the same moment into the left eye the head being held until all was absorbed so that none escaped by overflow of tears.I n the case of the less active tropeines the times required to produce the maximum mydriatic effect were much the same in all cases so that the more active of two was easily recognised. I n the case of the highly active tropeines the rapidity of actio'n as well as the maximum mydriatic effect had to be considerecl. It should be iioted that the mydriasis caused by the mor CONSTITUTION AND PHYSIOLOGICAL ACTION. 1 109 powerful tends to produce consensual niyosis in the other eye so that a small difference of activity is exaggerated and easily de twted.The effect of concentrated solutions has not been tested but tropeines which produce no perceptible effect in dilute solutions may give evidence of mydriatic effect when applied in concentrated form; thus Gottlieb ( . ~ T c ? z . e s p . Pnth. I’hnrnl. 1896 37 218) has stated that lactyltropeine and hippuryltropeine produce no mydriasis when introduced into the conjunctival sac in 2 per cent. solution but that 10 t o 20 per cent. solutions produce mydriasis commencing in half an hour. Further it must be pointed out that only the effect of lortd application has been tested; tropine itself although it has no local action on the eye pro’duces a striking mydriasis .in cats when given internally in large doxs and certain tropeines which have no local action for example the lactone of o-carboxyphenylglyceryltropeine produce mydriasis on injection.Briefly the problem investigated was the relation between the chemical constitution of the acyl group of a tropeine and the mydriatic effect produced by the instillation of a neutral solution equivalent in tropine content to a 1 per cent. solution of hom-atropine hydrobromide into the conjunctival sac of a cat. NO attempt was made to determine the cause of the mydriatic effect, which may have been due to action of the atropine type that is, paralysis of the motor nerve-endings of the sphincter (contractor) muscle of the pupil or t o action of the cocaine type that is, stimulation of the nerve-endings in the dilator muscle. Thirty tropeines were prepared and exanlined comparatively by this method.The mydriatic action of many of these had been recorded previously and references to the earlier results are given below. The mydriatic action of a further fifteen tropeines which we ourselves did not examine is also taken into consideration. For the purpose of discussion the forty-five tropeines inay be divided conveniently into six groups. T. Tropeines of aliphatic acids ................................. 8 11. Tropeiiies of substituted benzoic acids .................. (i 111. Tropeines of substituted hydratropic acids ............... 1 1 1V. Tropeines of substituted phenylacetic acids ............ 1 3 V. Tropeincs of sizbstituted phenylpropionic acids ......... 5 VC. Tropeines of acids in which the phenyl and cnrhosyl gronps are separated by an imino-group ............2 The tropeines of each group have been tabulated in order to show the results obtained a t a glance 1110 PYMAN THE RELATION BETWEEN CHEMICAL I . Tropeines of Aliphatic Acids. Mydriatic action A f Present Previous results corn--*- parison. Tropeine. Pormnls. Action. Observer. Actioii. 1. Acetyl- - Gottlieb ............... CH3*C0,T 2. Glycollyl- ............ CH2( OH)'CO,T - Marshall' 3. Lactyl- ............... CH;CH(OH)'CO,T + lo'& Gottlieb 4. Succinyl- ............ ('CH,'CO,T) - Gottlieb 5. Tartryl- ............... [*CH(OH)'CO,T] -6. Fumsroyl- ............ (:CH*CO,T) -7. Methylparaconyl- ... I I - Marshall1 -............... -1- Marshall' -l - y / CHMe*CHCO,T 0 'CO 'CH, CMe,-CH.CO,T O'CO 'CH, 8.Terebyl- I I Jowett and Hann T. 1906 89 357. Gottlieb stated that acetyl- and succinyl-tropeines can be brought in the solid state on to the conjunctival sac of a cat without. per-ceptible mydriatic effect but that lactyltropeine produces mydriasis commencing in half an hour under these conditions although it is inactive when applied as a 2 per cent. solution. The above table shows that previous observers had only reported mydriatic activity of dilute solutions in one instance that of terebyltropeine. This compound was again examined in the course of the present work and found t o be inactive. Tartryl-and fumaroyl-tropeines were also inactive so that no aliphatic tropeine that has yet been tested possesses mydriatic properties when applied as a dilute solution t o the eyes of a cat.11. Tropeines of Substituted Benzoic Acids. Mydriatic action. Previous results. comparison. && - Present Order of Tropeine. Formula. Action. Observer. Action. a.ctivity. 9. Benzoyl- ............ C,H;CO,T + Schmiede- 4- 3 10. Phthaloyl- ......... C6H,(COaT) -11. o-Hydroxybenzoyl- HO'C,H;CO,T - FaIck 4 1 12. m-Hydroxybenzoyl- HO*C,,H,'CO,T + Volkers _I_ 2 13. p-Hydroxybenzoyl- HO'C,H,*CO,T -14. Protocatechoyl- ... (HO)2C6H3'C02T - Marshall -berg. R'. Buchheim Arch. exp. Path. Pharm. 1876 5 463. Ladenburg AnnuZen 1883 217 82 CONSTITUTION AND PHYSIOLOGICAL ACTION. 11 11 Our examination of the above tropeines confirmed the state-ments of previous observers except in the case of o-hydroxy-benzoyltropeine.So far from being inactive this proved to be the most active of the tropeines of substituted benzoic acids, m-hydroxybenzoyltropeine being the next in order of activity. The tropeines of phydroxy- and 3 4-dihydroxy-benzoic acids both containing a para-hydroxyl group were inactive. 111. Tropeines of Substituted Hydratropic Acids. Mgdriatic action. Previous Present I- - results. comparison. i 2 2 . ' .r( g $ & F 4 Tropeine. Formula 0 4 8 Cushny' 4- 1 ;] CH,*OH + Laidlaw2 + Erbe3 + Lewinand 16. I-Tropyl-(hyoscyamine) I 17. d-Tropyl-(d-hyoscyamine) CHPh'C0,T 15. dl-Tropyl-(atropine) 18. Atropine methonitrate. I CH,-OH CHPh*CO,T,MeNO Grube4 CH,*O~CO*CH, CHPh*C02T CH,C1 CHPh'C0,T CH,Br Guillery5 CIIPh'C0,T CII;O *SO,H 22.Atropinesulphuric acid I - Trendel-CIIPh'C0,T onhurg CII,*OH CPh( OH)-CO,T i ............... Guillery5 Guillery5 19. Acetyltropyl- 1 20. Chlorohydratropyl- I 2 1. Bromohydratropyl- 1 ...... + Lewiii and ...... + Lewin and 23. Atroglyceryl- ............... 1 -b 2 26. Atropyl- ..................... CPh(:CH,)*CO,T - Lewin and 24. Atrolactyl- .................. CI'hMe( OH)'CO,T -1- Volkers Guillerys J . physioE. 1904,30 176. Barrowcliff and Tutin T. 1909 95 1906. Inaug. Diss. Miinchen 1903. Inaug. Diss. Gottingen 1905. " Die Wirkungen von Arzneiinitteln und Giften auf das Auge," Berlin, Arch. exp. Path. Pharm. 1913,73 118. 1905 p. 209. The tropeines of substituted hydratropic acids present several points of interest the most striking being the difference in activity between I- and d-hyoscyamine.Cushny working with the partly racemised substances found that Ehyoscyamine was about fourtee 11 12 PYMAN THE RELATION BETWEEN CHEMICAL times as active a inydriatic as d-hyoscyarniae. Later Laidlaw showed that the ratio of activity between the pure salts was much greater the mydriatic action of the Z-compound being about one hundred times that of the dextro-compound. The mydriatic action of atropine is therefore mainly due to the 7-constituent. Methylation of the nitrogen atom of atropine decreases the mydriatic action atropine methonitrate being apparently inter-mediate in action between atropine and homatropine. Acetylatropine is stsated by Lewin and Guillery to cause mydriasis and paralysis of the accommodation when applied in 1 per cent.solution whilst the same authors report that chloro-and bromo-hydratropyltropeines in 2 per cent. solution cause greater irritation to the eyes than atropine. The chloro-com-pound although less active than atropine gives sufficient mydriasis for ophthalmic purposes whilst t'he action of the bromo-compound is even slower and less intense than that of the chloro-compound. They found that atropyltropeine caused 110 mydriasis in 2 per cent. solution. Atropinesulphuric acid the acid sulphuric ester of atropine and a t the same time an internal salt' has no mydriatic action in 1 per cent. solution. Atroglyceryltropeine is of particular interest since it contains two hydroxyl groups in the positions of those of atropine and hornatropine respectively.CH,*OH CH,*OH H I C,II,*C CO,T C,H~&CO,T C,B ,*d*c4 qr. k J tz O H I Atsro p i 11 c. A t roglycerylt ropcine. HornatPopine. When examined on cats by the comparative method i t proved to be intermediate in activity between atropine and homatropine, but was less active than homatropine for the human eye. Atrolactyltropine is described as a powerful mydriatic strikingly similar in this respect t o homatropine. With the exceptioii of homatropiiiesulpliuric acid which is in-active like atropiiiesulphuric acid in the previous s,ection all the tropeines of substituted phenylacetic acids have mydriatic proper-ties. The effect of stereoisomerism on the activity is much less marked than in the previous section the enantiomorphous forms of homatropine differing only slightly in action the lmo-form being again the more active.Homatropine methobromide dilates the pupils of cats' eyes more completely and more quickly than a solution of hornatropin CONSTITUTION AND PHYSIOLOGICAL ACTION. 11 13 2 6. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. IV. Tropeines of Substituted Pheiiylacetic Acids. Mydriatic action. - Present Previous results. com-parisoil. Tropeine. Formula. Phenylacetyl- ... CH,Ph*CO,T (homatropine) *. CHPh(OH) *CO,T 1 ...... dl-Mandelyl-d-Mandelyl-Z-Mandelyl- ......... dl-MandeIyI-dl-Mandelyl-etho -Homatropine-sul-methobromide C'HPh(OH)'W2T,1L1eSr bromide ......... CHPh(OH)'CO,T,Et.B1. nhurir acid ......CHPh(O'SO.H)'CO,T " I I m-Methylmandelyl- C ,H,Mo 'CH( OH) *CO,T o -Methylmandelyl -p-Methylmrtndelyl -Phenvlchloro -acetyl- ......... CHPhC:1'C02T acetyl- ......... CHPh( NH2)'C0,T Phenylamino-Pht halidecarbosyl- C ,H;CH 'CO ,T I I co-0 Jowett and Pyman T. 1907, Action. Observer. Actioii. -I- + VOlkers + + i-4- Symoris . j -+ Symons -I-- Trendel-enbnrg 4-4- + 4 t--k Marshall 3-91 92. hydrobromide of the same strength but is less active for human eyes. Homatropine ethobromide is less active than the metho-bromide. Of the three methylmandelyltropeines the ortho- and meta-compounds equal each other and are more powerful than hornatropine in mydriatic power when tested on cats but the para-compound is slightly less active.Phenylacetyl- phenylchloroacetyl- phenylaminoacetyl- and phthalidecarboxyl-tropeines which contain 110 free alcoholic hydroxyl group are all active but much less so than honiatropine. Cinnamoyltropeine reported by Ladenburg as " hardly inydriatic," was found to have no mydriatic properties under the conditions of our investigation. 8-Phenyl-a-hydroxypropionyl-t r opeine is isomeric with atropine (u-p hen y l-P-h y d r oxy pr opionyl-tropeine) and also with atrolactyltropeine (a-phenyl-a-hydroxy-propionyltropeine) ; it only differs from honiatropine in that the phenyl and secondary alcohol groups are separated by a methylene group. It is a powerful mydriatic for cats' eyes and begins to dilate the pupil considerably earlier than atropine but the atropine dilation when it once begins quickly overtakes the othe 11 14 PYMAN THE RELATION BETWEEN CHEMICAL V.Tropei?i.es of Substit7ited Phenylpropionic Acids. Mydriatic action. Tropoine. Formula. 40. Lactone of o-carb- C,H4'CH(OH)'CH'C0,T 41. isocoumarincarb- ~6H,*CH:C.C02'I! 43. b-Phenvl-a-hvdr-39. Cinnamoyl- ...... CHPh:CH*CO,T I 0 oxyphenyl- I glyceryl- co oxyl- I I co- 0 > Present Previous com-results. parison. - 7-- Ladenburg -- Symons -- Symons -+ 1 oxypiopioiyl- CH,Ph'CH(OH)'CO,T 43. B-2-Pyridyl-a-hydroxfpro-pionyl- ......... C,H4N'CH,'CH(OH)'C0,T aiid becomes maximal a little earlier. When tested on human eyes it was found to be inferior to homatropine hydrobromide. &2-Pyridyl-a-hydroxypropionyltropeine has been included in the above group.It differs from the preceding member by the substitution of pyridine for benzene and although active is con-siderably weaker. VI. Tropeines of Acids in which the Groups are Separated b y an Pheityl and Cccrboxyl Imin 0-group. Mydriatic action. Present com-Previous results. parison. Troneine. Formula. Action. Observer. A4ction. 7-44. Hiipuryl- C,H,'CO'NH'CH,'CO,T { T:$ to 207; Gottlieb 45. Phenvlcarb-amo- ...... C,H,*NH*CO,T -I-Gottlieb states that hippuryltropeine behaves similarly t o lactyltropeine that is to say it only exercises a mydriatic effect when the solid substance is introduced into the conjunctival sac t o give a concentrated solution and is inactive in dilute solution. Phenylcarbamot,ropeine proved to have a slight mydriatic action.I n the foregoing tables the tropeines are classified according t CONSTITUTIOX AND PHYSIOLOGICAL ACTION. 11 15 their chemical constitution. The thirty members which we com-pared by the method given above may be grouped also in order of their 1. 11. 111, IV. v. VI. rnydriatic properties. Atropine B-Pheiiyl-a-hydroxypropionyl- I- Most active. t ropeine I A troglyceryl tropeiii e Intermediate i l l activity be-tween atropine and hom-atropine . dl-Homat ropiiie d - and Z-Homatropine Quaternary salts of hoinatropiiie 0- m- and p-Rlethylhomatropine J 8- 2-Pyridyl -a-hydroxypropionyl-Phthalidecarboxyltropeine Phenylchloroacetyltropeine Phenylaminoacet~yltropeine Phenylacetyltropeine Benzoyltropeine o -Hydroxybenzoyltropeine nz-Hydroxybcnzoyltropeine Phenylcarbamotropeine Tar tryltropeine Fumaroyltropeine rile t hylparaconyltropeinc Terebyltropeine Lactone of o-carbosyphenylglyceryl-is2Coumarincarboxyltropeine P hthalo ylt ropeine.p-Hydroxybenzoyltropeine Protocatechoyltropeine Cinnamoylt ropeine 1 Of a similar order of activity. All active but less so than i homstropine. t ropei ne Faintly active. i Inactive. I tropeine Of the fifteen tropeines which we did not examine by the com-parative method natural hyoscyaniine which is the lzvo-variety, is nearly twice as active as atropine d-hyoscyamine much less so. Acetylatropine chlorohydratropyltropeine bromohydratropyl-tropeiiie atrolactyltropeine and atropine metrhonitrate appear to be equivalent in niydriatic power t o the members of groups I1 and I11 of the above table whilst the following are stated to be inactive in dilute solutioii : Acetyltropeine.Atropyltropeine. Glycollyltropeine. Hippuryltropeine. Lactyltropeine. Atropinesulphuric acid. Succin yltropeine. Homatropinesulphuric acid. Before considering whether any general conclusions can be drawn froin these results attention inay be directed to the general-isatioii which has found its way into the literature sometimes i 11 16 PYMAN THE RELATION BETWEEN CHEMICAL association with Ladenburg’s name. This generalisation states that a tropeine to have mydriatic properties must contain (1) a benzene nucleus and (2) an alcoholic hydroxyl group in the side chain containing the carboxyl group.Now Ladenburg stated in 1883 that rri-hydroxybenzoyltropine had mydriatic properties so it seemed unlikely that the geiieral-isation was due to him. Accordingly after a careful but unsuccessful search of the literature for such a generalisation under Professor Ladeliburg’s name we coniiiiuiiicated with him, arid learned that he was unable to recollect framing it. In the light of the evidence afforded above i t would appear that the first postulate of this generalisation is approximately correct ; 110 tropeine of an aliphatic acid has yet been found to possess mydriatic properties in dilute solution but on the other hand, the closed chain need not necessarily be that of benzene since /3-2-pyridyl-a-hydroxypropionyltropeiiie which contains a pyridine instead of a benzene residue is active.The second postulate that a tropeine to be inydriatic must have ail alcoholic hydroxyl group in the side chain containing the carboxyl group is incorrect for mydriatic substances are obtained when the hydroxyl group of atropine is exchanged for acetoxyl, chlorine or bromine and when the hydroxyl group of hornatropine is exchanged for hydrogen chlorine o r an amino-group or when i t is closed by the formation of a lactone; moreover benzoyl- and 0- and nz-hydroxybenzoyl-tropeines are mydriatic. The loss of mydriatic properties on the replacement of the hydroxyl group in atropine and honiatropine by the sulphuric acid residue is possibly due to the same cause which operates in the case of substances containing a free carboxyl group such as benzoylecgonine and quitenine.Whilst however the second postulate is incorrect as regards the qualitative mydriatic action of tropeines i t must be remem-bered that those tropeiiies which we found more active than or equal to homatropine in mydriatic properties contained an alcoholic hydroxyl group. Of the tropeines of hydroxybenzoic acids the o- and m-sub-stituted compounds were active whilst the p and also the 3 4-di-hydroxy-compounds were inactive. Substitution in the pposition in this case causes the mydriatic action to vanish and in the case of the methylmandelyltropeines also the para-compound is less active than the ortho- and meta-isomerides. The tropeines of substituted hydratropic phenylacetic and phenylpropionic acids were all active with the exception of atropinesulphuric acid homatropinesulphuric acid the lactone o CONSTITUTION AND PHYSIOLOGICAL ACTION.1 117 o-carboxyphenylglyceryltropeine and those containing an un-saturated linking in the side-chain containing the carboxyl group. Consideration of the above material led us to conclude that no generalisation as to the relation between niydriatic action and chemical constitution could be made which would offer a strict explanation of the) results obtained. Before leaving the subject of the tropeines attention may be directed to some points of' interest in connexim with allied mydriatics. Norhyoscyamine which differs from hyoscyamine in containing an imino- in the place of an N-methyl group has only one-eighth of the mydriatic effect of hyoscyamine and the racemic form noratropine is again about one-eighth as active as atropine (Carr and .Reynolds T.1912 101 946. Physiological tests by Laidlaw). I t has already been pointed out that the steric structure of the acyl radicle of a tropeine influences its niydriatic properties. The steric structure of the basic portion of the molecule is also important for the tropyl and mandelyl esters of +-tropine have 110 mydriatic properties (Liebermann and Limpach Her. 1892 25, 933. Physiological tests by Liebreich). A iriinoctlkyl Esters. The question as to what portions of the cocaine molecule (I) are essential to the local anaesthetic action of the alkaloid has long been the subject of investigation and the collated results of numerous workers have shown by a series of eliminations that the anaesthetic properties of cocaine are a,ssociated with its func-tion as an aminoalkyl ester.It has been found that the carboxy-methyl (C0,Me) group is not an essential factor since tropa-cocaine (11) which contains no such group produces the same effect; and further the presence of a bridged or simple ring containing nitrogen is unnecessary since eucaine (111) which possesses only a simple not a bridged ring and stovaine (IV), alypine (V) and novocaine (VI) which contain no such ring have well-marked local anathetic properties. CH2*CH-CH*C0,&le CH,*CH-CH I &Me CI1*O*COPh 1 kMe bH-O*COPh I i i I C H,*CH-CH, (11.1 CH,*CM6-CH2 CH,*Nhfe, AH bH-O*COPh C,H ,*C*O*COPh I I I I C H ;CH-CH C*3 (111.) W. 1118 PYMAN THE RELATION BETWEEN CHEMICAL CH2*NMe2 C,H,*i<O*COPh ( C2H5)2N*CH2*CH2*O*CO*C,H,*NH2 (21) From the above considerations it follows that local aimsthetic action is associated with the aininoalkyl ester structure and we may now inquire what complexes in such eaters are necessary for the possession of local anzesthetic properties. Aminoalkyl esters have the general formula R*CO-O-(CR,R2),-NR,R4 ; they are formed by the esterification of an acid with an alcohol containing an umino-pap and may be dealt with conveniently from this point 0f view. The ucyl group of aminoalkyl esters possessing local anathetic properties is in most cases aromatic and in the majority of sub-stances of practical application is the benzoyl group as instanced by the compounds numbered I to V.Fourneau (J. Pharm. C'kim., 1910 [vii] 2 337 397) has however recorded that the valeryl, bromovaleryl and bromoheptoyl esters of dimethylaminodi-methylethylcarbinol (the benzoate of which is stovaine) have anzsthetic properties so that the presence of a ring complex in the acid does not appear to be essential. I n the case of cocaine replacement of the benzoyl by substi-tuted benzoyl or other acid radicles leads to substances with much weaker local anzesthetic properties. Thus the phenylacetyl deriv-ative is much less powerful the o-chlorobenzoyl and m-niixobenzoyl derivatives have only a slight local anmthetic action and the m-hydroxybenzoyl compound still less whilst the substances in which the benzoyl is replaced by the valeryl rrL-aminobenzoyl, phthaloyl cinnamoyl or isatropyl radicles are inactive (Ehrlich, Liebreich and Poulsson.Compare Ehrlich and Einhorn BeT., 1894 27 1870). Substitution i n the benzoic acid nucleus of aminoalkyl benzoates is not however necessarily associated with weak local anzesthetic action for the p-aminobenzoyl esters of many amino-alcohols are strong local anaesthetics novocaine being diethylaminoethyl p-aminobenzoate whilst the dialkylaminoalkyl 3 4-diaminobenzoates have also considerable local anzesthetic properties (Einhorn D.R.-P. 194365). I n the case of cocaine, the mbstitution of phthaloyl f o r benzoyl gave an inactive com-pound and similarly whilst diethylaminoethyl benzoate, NEh=CH,*CH,*O*COPh is stated to have local .anzesthetic proper-ties (E.Schering D.R.-P. 175080) diethylaminoethyl phthalate proved to be inactive (Pyman T. 1908 93 1793. Physiologica CONSTITUTION AND PHYSIOLOGICAL ACTION. 1 1 19 tests by Dale and Symons). Passing now to aromatic acids in which the carboxyl of the acyl group is not directly attached to the benzene nucleus w0 have seen that the replacement’ of the benzoyl by the phenylacetyl group in cocaine gives a substance having local anaesthetic properties ; with a-eucaine also the phenyl-acetyl compound has well-marked local anaesthetic properties (Vinci Virch. -4rch. 1898 154 549); i t was found however that replacement of the paminobenzoyl group by the paminophenyl-acetyl group in novocaine and anEdhesine (ethyl p-aminobenzoate) gave inactive compounds diethylaminoethyl and ethyl p-amino-phenylacetates (Pyman this vol.p. 167. Physiological tests hy Dale and Symons). Cinnamic acid usually but not invariably confers local anaes-thetic properties on aminoalkyl esters. As we have seen cinnamoyl-cocaine is inactive but the a-eucaine derivative has local anaesthetic properties (Vinci loc. cit .). Tetramethyldiamino-dimethylethylcarbinyl cinnamate that is an ‘( alypine ” in which cinnamic acid takes the place of benzoic acid produces an anzesthetic effect lasting twice as long as that brought about by the same quantity of cocaine (Farbenfabriken vorm. F. Bayer & Co. D.R.-P. 173631) whiIst y-diethylaminopropyl cinnamate (apothesine) also has well-marked local anaesthetic properties (E. A. Wildman and L. Thorp U.S.Pat. 1193649). It may be noted that the aminoalkyl esters of aminocinnamic acid are stated to have several times the local anzsthetic power of those of aminobenzoic acid (Meister Lucius & Bruning D.R.-P., 187593). With regard to the nature of the substituted amino-group required in an alkamine ester having local anmthetic properties, there is little available information. Most of the best known local anzesthetics contain a tertiary amino-group but Peucaine which has powerful local anaesthetic properties contains a secondary amino-group. Norcocaine in which the N-methyl group is re-placed by the imino-group has greater local anaesthetic properties than cocaine (compare Ehrlich and Einhorn Zoc. c i t . ) but the primary amine corresponding with novocaine namely &amino-ethyl paminobenzoate (VII) (Forster T.1908 93 1865. Physiological tests by Dale) is devoid of local anasthetic action. The nature of the alkyl groups replacing the hydrogen atoms of the amino-group appears t o affect the local anzesthetic properties in some degree; thus piperidylethyl benzoate (VIII) is only slightly active whilst s-di-/3-benzoyloxy-l 4-diethylpiperazine (IX) has very distinct action and pB-dibenzoyloxytriethylamine (X) i 1120 PYMAN THE RELATION BETWEEN CHEMICAL slightly active whilst BP-dibenzoyloxymethyldiethylamine (XI) (Pyman loc. c i t . ) is inactive. NH2* C H,*CH,*O*CO* CGH,*N H C,H,oN*CH3*CH,*0.C'OPh (VII.) (vzrr.) PhCO.O*CH,.CH,*N<~~~:~~~>N.CH,*CHB*O*COPh (IX) NEt(CH,*CH,*O*COPh) N Me( CH,*CH,-O*COPh), (X.) (XI.1 The ct7colmZ residues of these esters are very varied in character; they may be primary secondary or tertiary and may separate the acyl residue and the substituted amino-group by chains of a varying number of carbon atoms. As inst'ances of active aminoalkyl esters derived from primary alcohols we have novocaine and the dialkylaminoethyl benzoat es ; in these oiily two carbon at'oms separate the acyl- and amino-groups. y-Diethylaminopropyl cinnamate is an example of an ester of a primary alcohol in which the two groups referred t o are separated by three carbon atoms. Cocaine tropacocaine and the eucaines are derived from secondary alcohols containing a chain of three carbon atoms between the acyl- and amino-groups whilst instances of local anaesthetic property in secondary alcohols in which these groups are separated by only two carbon atoms are furnished by P-benzoyloxy-/3-3 4-methylenedioxyphenylethyldi-methylamine (I) and By-dibenzoyloxydimethylpropylamine (11) (Pymaii loc.cit .). 0-C H +' 1 1 C H N Me, dlH*O-COPb . dH,*@*COPh Finally typical examples of local anaesthetics derived from tertiary alcohols are stovaine and alypine in which the acyl and amino-groups are also separated by a chain of two carbon atoms. The general conclusions to be drawn from the above summary are that in aminoalkyl esters having local anmthetic properties (1) the acyl group is usually aromatic (2) the amino-group may be secondary or tertiary and may be associated with simple or bridged ring complexes and (3) the alcohol group may be primary CONSTITUTfON AND PIIYSIOLOaICAL ACTION.112 1 secondary or tertiary and may separate the acyl and amino-groups by a chain of either two or three carbon atoms. Adrenaline and the Amines. Adrenaline the active principle of the suprarenal gland is a substance of powerful physiological action. Its action simulates the effects of exciting sympathetic nerves and in consequence has been termed I ‘ sympathomimetic.” Therapeutically it is chiefly used to prevent bleeding by its vasoconstrictor action when applied locally ; when injected intra-venously it causes amongsi; other symptoms a large rise of blood pressure also partly due to vasoconstriction and the measure of this pressor effect when accompanied by other symptoms of sympathomimetic action serves for the comparison of adrenaline with allied compounds.Adrenaline is of comparatively simple constitution being 8-3 4-trihydroxy-&phenylethylmethylamine (I) and the question as to the relative influences of the different portions of its mole-cular structure has been the subject of much investigation. It was a t one time suggested that the catechol nucleus (11) was the essential active group for catechol causes a rise of blood pressure on intravenous injection whilst the other half of the molecule, 8-hydroxyethylmethylamiiie (111) has no such action. OH O H ()OH I ’ ’\OH CH2*OH \/ \/ CH,*NHMe I CH*OH 6H2*NHMe (1.) (11.) (111.) Barger and Dale ( J . Physiol. 1910 41 19) however showed that the rise of blood pressure produced by catechol was not due to sympathomimetic action hut t o an action of an entirely different type whilst on the other hand many aliphatic and aromatic amines had an action very similar t o that of adrenaline.They studied the relation between chemical constitution and sympatho-mimetic action in a large number of amines gradually appro,aching adrenaline in constitution and as a quantitative index of the activity of the compounds they adopted the effect on arterial blood pressure. The aliphatic amines were first examined then those containing a phenyl group and finally phenylalkylamines in which one two or three hydroxyl groups were introduced as substituents into the benzene nucleus. I n the aliphatic series the following VOL. UXI. Y 1122 PYMAN THE RELATION BETWEEN CHEMICAL results were obtained with primary amines under comparable con-ditions : Substance.Pressor effect. 'i J (1) Methylamine (2) Ethylamine (3) iaoPropylamine - insignificant. (4) n-Propylamine ( 5 ) isoButylamine (6) n-Butylamine positive. (7) isoamylamine, (8) n-Amylamine, (9) n-Hexylamine, (10) n-Heptylamine, (1 1) n-Octylamine, several times that of No. 6. distinctly greater than that of No. 7. greater than that of No. 8. less than that of No. 9. less than that of No. 10. With still higher members of the series comparison was difficult, since they became increasingly toxic. Of mcolndary amines diethylamine was found to be inactive, methylisoamylamine C,H,,*NHMe was considerably weaker than isoamylamine whilst diisoamylamine had very little of the action.In the aliphatic series therefore the most active member proved t o be n-hexylamine. The next group examined consisted of aromatic amines in which the benzene nucleus was otherwise unsubstituted. (1) Aniline Ph'NH, did not show the specifio action. (2) Benzylamine Ph*CH;NH, had a mere trace of the aotion. (3) a-Phenylethylamine Ph'CHMe'NH, was very feebly active. (4) B-Phenylethylamine Ph*CH,'CH,'NH, was more active than Nos. (3) and (a) and its activity was distinctly greater than that of n-hexylamine the most active of the aliphatic amines. (5) y-Phenylpropylamine Ph*[CH,];NH, was No. much (4). less active than B-Phenylethylamine which proved to be the most active of this series contains the skeleton of adrenaline but differs from it in lacking (1) the 3 4dihydroxyl substituents of the benzene nucleus, (2) the hydroxyl substituent of the &carbon atom and (3) the methyl group attached to the nitrogen atom.The effect of the two last substitutions singly or together on P-phenylethylamine was tested by the examination of 8-hydroxy-0-phenylethylamine, Ph-CH(OH)*CH,*NH, P-phenylethylmethylamine, P h* CH,*CH,*NH Me, and fi-hydroxy-8-phenylethylmethylamine, P h* CH (OH) CH,* NHMe, none of which differed noticeably in activity from P-phenylethyl-amine. Further work was directed to determining the influence of phenolic hydroxyl groups on the action of these phenylalkyl CONSTITUTION AND PHYSIOLOGICAL ACTION. 1 123 amines and in the first place the effect of introducing a single hydroxyl group was ascertained with the following result8 : ( 1) p-Hydroxy-B-phenylethylamine, HO*C6H;CH;CH2~NH *.( 2) m-Hydroxy -B-phenylethylamine, (3) o-Hydroxy-8-phenylethylarnine, (4) 4-Hydroxy -B -m-t olylethylamine, Me HO(->CH;CH,*NH,. 8-p-Dihydr&y-B-phenylethylamine, HO'C,H;CH( OH)*CH;NH,. p -Hydroxy - i-aminoacetophenone, HO~C,H;CO*CH,'NH,. p -Hy droxy - 8-phenylethylm et hylaminc HO*C6H;CH,'CH,'NHMe. p -Hydroxy-b -phenylet h ylet h ylamine , HO*C6H;CH;CH;NH Et. p-Hydroxy-8-phenylethyldimethylami HO*CIH4*CH,'CH;NMe2. p-Hydroxy-B-phenylethyltrimethyl . . _.-ammonium iodide, HO'C6H4'CH,'CH,'NMe,~. dl-p -Hydroxy -a -p henylet hylamine, HO'C,H;CHMe*NH,. I-p-Hydroxy-a-phenylethylemine, ( 13) p-Hydroxyphenylethylacetarnide, (14) Tyrosine ethyl ester, HO *C ,H4*CH2*CH2*NHA~.HO *C 6H4*CH2'CH( CO ,El t) *NH,. 9 .ne, 3 to 5 times as active as 8-phenylethylamine. Had about 1-20th of the activity of adrenaline. Equal to (1). No more active than 8-phenyl-ethylamine. Half as active as (1). Less active than (1). Feebly about 1-10th as About the same as (1). Considerably less active than (1) and (7). Very much less active than (1) and (7). Action entirely different from that of adrenaline resemb-ling that of nicotine. Very slightly active. Very slightly active; not Inactive. active as (1). different from (1 1). Inactive. The foregoing results show that the introduction of a hydroxyl group into phenylethylamirie is accompanied by an increase in activity in the case of the p- and m-compounds but not in that of the o-substituted compound.Here again as in the unsubstituted phenyl series neither the introduction of a hydroxyl group in the &position (5) nor methyl-ation of the nitrogen (7) increases the activity of the parent com-pound whilst the introduction of a larger alkyl group (8) or second methyl group (9) on the nitrogen atom seriously diminishes the activity of the compound. The next group of compounds examined contained two phenolic hydroxyl groups and included (a) derivatives of acetocatechol, ( 6 ) derivatives of ethylcatechol and ( c ) derivatives of hydroxy-ethylcatechol. By determining the doses which produced rises of blood pressure to equal submaximal heights the approximate average activity values were found t o be ae follows 1124 PYMAN THE RELATION BETWEEN CHEMICAL (a) ( 1) 3 4-Dihydroxy-o-aminoacetophenone, (OH),C,H,*CO'CH;NH ..........................................1.5 (OH),C,H,'CO*CH;NHMe Weaker (1); greater i t h ~ ~ ~ o * No. (4) (2) 3 4-Dihydroxy-o-methylaminoacetophenone, (3) 3 4-Dihydroxy-w-ethylaminoacetophenone, (4) 3 4-Dihydroxy-o-propylaminoacetophenone, (OH)2C,H,'CO'CH,'NHEt ....................................... 2.25 (OH),C,H;CO'CH,'NHPr ....................................... 0.25 (OH),C6H,*CH;CH,*NH ....................................... 1 (OH),C,H,'CH;CH,'NHMe .................................... 5 ( 13) ( 5 ) 3 4-Dihydroxy-B-phenylethylamine, (6) 3 4 -Dihydroxy - 8-phenylethylmethylamine, (7) 3 4-Dihydroxy-8-phenylethylethylamine, ( 8 ) 3 4-Dihydroxy-8-phenylethylpropylaminey (OH)2C,H3.CH,*CH,*NHEt ....................................1.5 (OH),C,H,*CH(OH)*CH;NH ................................. 50 (OH),C,H3*CH2~CH;NHPr .................................... 0.25 (c) (9) d1-8-3 4-Trihydroxy-j3-phenylethylamine, ( 10) dE-8-3 4-Trihydrosy-~-phenylethylmethylamine (dl-adren-(1 1) 1-8-3 4-Trihydroxy-~-phenylethylmethylarnine (Z-adren-aline) (OH),C,H;CH( 0H)'CH;NHMe.. ...................... 35 aline) .................................................................. 50 In these series the N-propyl derivatives were much less active than the N-methyl and N-ethyl derivatives but there is no con-sistency in the relative values of the amino- N-methyl and N-ethyl derivatives; in the (a) series the N-ethyl in the ( b ) series the N-methyl and in the ( c ) series the amino-compound was the most active of those examined.Two amines containing three phenolic hydroxyl groups were also examined namely 2 3 4-trihydroxy-o-aminoacetophenone, and 2 3 4-trihydroxy-P-phenylethylamine, (OH),C,H,*CO*CH,-NH , (OH),C,H,*CH,= CH,-NH,. I n each case the pressor action was somewhat weaker than that of the corresponding catechol base. Consideration of the above results led Barger and Dale t o the following conclusions '' The optimum carbon-skeleton for sympathomimetic activity consists of a benzene ring with a side-chain of two carbon atoms the terminal one bearing the amino-group. Another optimum condition is the presence of two phenolic hydroxyls in the 3 4-position relative t o the side-chain ; when these are present an alcoholic hydroxyl still further intensifies the activity.A phenolic hydroxyl in the ortho-position does not increase the activity. Many physiologically active amines occur in nature as the result of decarboxylation of amino-acids by bacteria. Of those men CONSTITUTION ,4ND PHYSIOLOGICAL ACTION. 11 25 tioned above for instance isoamylamine and p-hydroxy-fl-phenyl-ethylamine are derived from leucine and tyrosine respectively. HO*C6H4*CH,*CH(NH,)*C0&l + HO*C,H4*CH2*CH2*NH,. Tyrosine. p-Hydroxy -B-phenylethylamine. Derivatives of ethylamine containing heterocyclic nuclei are formed similarly thus indole-ethylamine from tryptophan and aminoethylglyoxaline from histidine.Aminoethylglyoxaline occurs naturally in ergot and is an intense stimulant of plain muscle; several of its derivatives and allied compounds listed below have been prepared and compared with it physiologically (Ewins T., 1911 99 2052; Pyman ibid. 2172; 1916 109 186. Physio-logical tests by Laidlaw and Dale): ( 1) 4-Aminomethylglyoxaline.. ...................... C3H3N2'CH,'NH2. (2) 5-Methyl-4-aminomethylglyoxaline.. .......... Me'C3H2N2'CH2*NH,. (3) 5-Methyl-4-methylaminomethylglyoxaline Me'C3H2N2'CH2 'NHMe, (4) 4-B-Aminoethylglyoxaline ..................... C,H3N,'CH2'CH,*NH2. i t C,H,N,'CH'CH,*NH,. ( 5 ) By-bis( 4-Glyoxaline) -propylamine (6) B-Hydroxy-B-glyoxaline-4-ethylamine ...... C,H3N,'CH(OH)'CH,'NH2.(7) 5-Methyl-4-aminoethylglyoxaline ............ Me'C3H2N2*CH,'CH,*NH,. ( 8 ) l-Methyl-4-aminoethylglyoxaline ............ MeC,H,N:CH,'CH,*NH,. (9) l-Methyl-5-aminoethylglyoxaline ............ Me'C,H N;CH,*CH,'NH,. C3H3N2'CH2. ............ ( 10) 4 -7 . Aminobutylglyoxaline .................... .C,H3NS*CH;CH,'CHMe"H, Of these compounds No. 6 was found to be less active than aminoethylglyoxaline (No. 4)) and No. 7 had only about a 1/2OOth of the characteristic stimulant action of No. 4 whilst the remain-ing members of the series only showed this action to a slight extent. Here also the optimum side-chain has two carbon atoms between the cyclic system and the amino-group the compounds in which one (No. 1) and three (No. 10) carbon atoms separated these groups being much less active.The introduction of an alcoholic hydroxyl group (No. 6) o r of methyl substituents into the nucleus (No. 7) or into the imino-group (Nor:. 8 and 9) also gave less active compounds. Protozoacidal Drzigs. The fourth and last exainple of the relation which I desire t o discuss to-night concerns the action of certain alkaloids in proto-zoal diseases. Malaria is a condition in which the blood is infested with plasmodia and is treated by means of quinine which has a specific action on the parasites. Amebic dysentery is similarly due to infection with the E n t a m a b a histolytica and responds best to the. action of emeiine. Experiments have recently been conducted in both fields to determine whether some derivative of the alkaloid8 mentioned.or one of the alkaloids associated wit 1126 PYMAN THE RELATION BETWEEN CHEMICAL them in cinchona bark or ipecacuanha root respectively have any advantages therapeutically. The line of attack has been some-what similar in the two cases; the toxicities of the drug and a number of its derivatives to protozoa and mammals were first deter-mined in the laboratory and the more promising derivatives were then tested clinically. I n the case of the inquiry into the value of certain cinchona derivatives (A. C. MacGilchrist Znd. J . Med. Res. 1914-1915, 2 315 336 516; 1915-1916 3 l) the relative lethality of each derivative to different species of infusoria (as representing protozoa) and to guinea-pigs (as representing mammds) was determined with the object of finding some indication as to which derivative would be most useful for the treatment of malaria that is would kill the parasite and yet cause least inconvenience or harm to the host.The results obtained are tabulated below the substances being given in the order of their lethality t o infusoria ethylhydro-cupreine hydrochloride being the most toxic. The minimum lethal dose to guinea-pigs is also given together with formulz designed to show the structural differences a t a glance; the two1 pairs of stereoisomerides quinine and quinidine cinchonine and cinchon-idine differ in the sign of the carbon atom bearing the alcoholic hydroxyl group. M.L.D. for guinea-pigs mg per Substances in order of lethality to infusoria. kil og. Formula. 1. Ethylhydrocupreine hydrochloride 0.65 EtO'C1,H18ON2*CH,*CH,.2. Cinchonine sulphate .................. 0.425 H'C17HI8ON2'CH:CHa 3. Quinine sulphate ..................... 0.525 Me0 'C ,H,,ON,*CH:CH,. 4. Hydroquinine hydrochloride ......... 0.6 MeO*C ,Hl8ON,*CH,'CH,. 5 . Quinidine sulphate ..................... 0.4 MeO'C ,H,,ON,'CH:CH,. 6. Cinchonidine sulphate.. ................ 0.6 R~C,,H,,ON,~CH:CH,. From these results it appeared that ethylhydrocupreine in which the vinyl group of quinine is reduced whilst ethoxyl takes the place of methoxyl was a promising subject for clinical trial in malaria. On a clinical comparison however the order of their value was found to be as follows: 1. Hydroquinine hydrochloride. 2. Quinine sulphate. { Quinidine sulphate. 5. Ethylhydrocupreine hydrochloride.6. Cinchonidine sulphate. Cinchonine sulphate. I n the case of the alkaloids of ipecacuanha determinations of the relative toxicity of a number of compounds t o free living amcebz gave the following results (Pyman and Wenyon J . Pharmacol. 1917 in the press). Emetine cephaeline #-methyl CONSTITUTION AND PHYSIOLO~ICAL AOTION. 1127 emetine and N-methylcephaeline were approximately equally amcebacidal ; N-methylemetine methochloride rubremetino hydro-chloride noremetine and the hydrochloride B, C,,H,70,NC1,,HC1 ,5H20, obtained by the oxidation of cephaeline were inferior to these, whilst psychotrine su1phat.e was much inferior. These results indicate that the full amcebacidal action characteristic of emetine is only exhibited when the nucleus is intact.The exact constitu-tion of the nucleus of these alkaloids is a t present unknown but it is certainly present intact and fully reduced in emetine, cephaeline N-methylemetine N-methylcephaeline and noremetine, for these substances are interconvertible in a simple manner differ-ing only in the number of methyl groups attached to the oxygen and nit,rogen atoms of the :molecule. Noremetine. Cephieline. Emetine. Psychotrine. N-Methylcephaeline. N-Methylemetine. It is interwting t o note that four of these compounds were very active. The inferiority of noremetine may conceivably be due to the fact that this compound contains four hydroxyl groups in place of the four methoxyl groups of emetine for Laidlaw (Zoc. cit.) has shown that amongst other isoquinoline derivatives a similar change of constitution produces an alteration in physio-logical action.N-Methylemetine methochloride is a biquaternary salt and as such a difference in its action from that of the parent tertiary base is not surprising. The fact that the hydro-chloride B still retains some amcebacidal properties is interesting, because of the comparatively simple constitution of this substance, which is devoid of the guaiacol residue and one of the nitrogen atoms of cephaeline. The hwo remaining substances rubremetine and psychotrine are not fully saturated rubremetine hydrochloride containing eight hydrogen atoms fewer than emetine hydrochloride, whilst psychotrine contains two atoms of hydrogen fewer than cephaeline. Some of the above compounds and certain others have also been tested on Entamaba histolytica in vitro (Dale and Dobell J .Pharmacol. 1917 in the press) and here again emetine cephaeline 1128 RELATION BETWEEN CHEMICAL CONSTITUTION ETC. and N-methylemetine proved to be active as was also the 0-methyl ether of psychotrine (which had not been tested on free living amoebae) whilst psychotrine again proved to have only a slight action. It is curious to contrast the similarity of action between cephae-line and its methyl ether emetine with the difference between psychot>rine and its methyl ether. The toxicity of many of the above compounds was determined and the laboratory results indicated that N-me t h yleme t ine and 0-me t h y 1 p syc hot ri n e were 1 ess toxic than emetine to mammals. Since t,hey were at the same time equal to emetine in amoebacidal properties it was thought that they might prove to be superior to this alkaloid in the treatment of amoebic dysent'ery but unfortunately clinical trials have shown that this is not the case (G. C. Low Brit. Med. J. November 13th 1915; C. M. Wenyon and F. W. O'Connor J . Roy. Army Med. Corps 1917 28 473; M. W. Jepps and J. C. Meakins Brit. Med. J. November 17th 1917 648). A review of the subjects discussed b n i g h t leads to the con-clusion that i t is very difficult to improve upon naturally occurring active principles the use of which in medicine is due to accumu-lated experience. I n point of maximum effect none of the natural compounds discussed to-night-hyoscyamine cocaine adrenaline, quinine emetine-is surpassed by its derivatives; but on the other hand it has been possible in some of the cases to prepare derivatives o r synthetic analogues which have proved to be of service in medicine. In conclusion I should like t o thank Dr. €3. H. Dale F.R.S., of the staff of the Medical Research Committee for help on the physiological side and Dr. H. A. D. Jowett for his collaboration in the previous paper which forms the basis of much of the work recorded above