a comparison of the physiological actions (from the

A COMPARISON OF THE PHYSIOLOGICAL ACTIONSAND CHEMICAL CONSTITUTION OF PIPERIDINE,CONIINE AND NICOTINE. BY B. MOORE AND R. ROW.(Five Figures in Text.)

(From the Physiological Laboratory of University College, London.)

ALTHOUGH the constitution of the nicotine molecule has not yet beencompletely established, it is. certain that it contains two well-markedorganic groups. It is generally agreed among organic chemists thatone of these groups is a pyridene ring; regarding the other group thereis still a considerable amount of controversy. The presence of a pyri-dene ring is shown, by the formation of nicotinic acid (i.e. /8-pyridine-carbonic acid) when nicotine is treated with oxidizing agents, such asnitric or chromic acid1: and by the fact that it is an unsaturatedcompound capable of forming addition products with bromine orhydrogen2.

One view as to the character of the second group is that it alsocontains a pyridene ring which is wholly or partially reduced: so thatnicotine might be regarded as a dipyridyl derivative. This view issupported by the researches of Cahours and Etard', who have shownthat nicotine on cautious oxidation with potassium ferricyanide inalkaline solution loses 4 atoms of hydrogen and yields a body with theformula C1oH10N2 which they have named isodipyridene. If this com-pound is formed by direct oxidation from the nicotine molecule,unaccompanied by a molecular rearrangement, it points to the presenceof two pyridene rings (one reduced) in the nicotine molecule since itprobably has the formula C5H4N-CQH8N.

1 Huber, Annalen der Chemie u. Pharmacie, CXLI. S. 271. 1867; Berichte der deut.Chem. Gesellsch. Jahirg. iII. 8. 849. 1868.

2 Cah ours and Etard, Bull. de la Soc. Chim. de Paris, xxxIv. p. 449; Liebrecht,Ber. der deut. Chem. Gesellsch. 1885, S. 2969; ibid. 1886, S. 2587.

Comptes rendus Acad. LXXXVIII. p. 999; xc. pp. 275, 1315; xci. p. 1079; xCviLp. 1218.

B. MOOIE AND R. ROW.

Since then Etardi has shown that nicotine forms acetyl andbenzoyl derivatives, and concludes that it hence contains one nitrogenatom in a secondary and one in a tertiary form, and most probably hasthe formula

aCH CHK25HC a aHE,HOU, \ OCH2

N NH

More recently this formula has been disputed by Pinner2, who isof the opinion that no second pyridene ring is present, but instead areduced pyrrol ring in which the nitrogen atom is united to a metbylgroup.

According to Pinner, nicotine is ,8-pyridyl-methyl-pyrrolidine withthe formula

/CHHC a-CH-CH2\

11 OH2.HC CH N -OH7

N OCHThis formula is deduced from the behaviour of certain bromo-

derivatives obtained by Pin n er, which on decomposition by weakalkalies yield, among other products, methyl-amine. Since one pyridenering is certainly known to be present in the nicotine molecule, if amethyl group is present in the remlainder of the molecule, the empiricformula prevents the possibility of a second pyridene ring.

Pinner assumes that the methyl group is united to nitrogen, andhence that nicotine has the formula given above.

Other formulae have also recently been advanced3, but these possesslittle probability from the chemical standpoint, and are further notsupported by the comparative physiological behaviour of nicotine asstated in this paper, and hence need not here be discussed.

1 Comptes rendus Acad. cxvii. pp. 170, 278. 1893.2 Berichte der deut. Chem. Gesellsch. Jahrg. xxvi. S. 292, 765, 2135. 1893; ibid. xxvii.

S. 1053; ibid. xxviii. S. 456, 1932.3 See F. Blau, Monatsheft f. Chemie, xiii. S. 330. 1892; Berichte der deut. Chem.

Gesell8ch. Jahrg. xxvi. 8. 628, 1029. 1893; ibid. xxvii. S. 2535: Olive ri, Gazzetta, xxv.1, 59. 1895.

274

PIPERIDIArE, CONIINE, NVICOTINE.

Whichever formula be accepted it is certain that the nicotinemolecule contains a pyridene ring united to a reduced ring containinganother nitrogen atom, and this reduced ring, is very probably either apiperidine or pyrrolidine ring.

Now piperidine and pyrrolidine are two closely allied bases onlydiffering in composition by the presence of one extra CH2 group in thecase of piperidine, as is shown in the following formula:

NH NH

H C CH12 H2C CH2II , IIHC\ CH 2 H2()- -CH2

OH2

Pyrrolidine is obtained by the reduction of pyrrol in the samefashion as piperidine is obtained by the reduction of pyridine, andfurther the physical properties of the two bases bear a close resem-blance.

Again, the nitrogen atom on which the physiological propertiesdepend has the same grouping in the two rings being united to two CH2groups in the same manner.

It is hence fairly probable that pyrrolidine and piperidine are alsoclosely allied in their physiological action'.

The presence of the two constituent groups in the nicotine moleculesuggests the problem-what share has each group in producing thephysiological actions of the whole ?

This question we have attempted to solve to some extent by testingthe separate physiological actions of pyridene and piperidine andcomparing them with those of nicotine, using the piperidine to repre-sent the reduced portion of the nicotine molecule, with which it isclosely allied in chemical nature.

Our attention was first attracted to the subject by the enormous risein arterial blood-pressure caused by intravenous injection of nicotine.Taking this rise in blood-pressure as a criterion, we tested the effects ofintravenous injection of both pyridene and piperidine and found thatwhile pyridene caused a marked fall, piperidine caused a much moreextensive rise.

The fall due to pyridene had already been described by Lauder

1 We have not yet been able to obtain a supply of pyrrolidine, but intend subsequentlyto study its physiological action.

PH. XXII. 19

29w75

B. MOORE AND R. ROW.

Brtinton and Tunnicliffe' some years previously, and quite recentlyTunnicliffe2 working with Brunton has briefly described the immenserise due to piperidine which he finds is still obtained after section ofthe cord at the level of the occipito-atlantoid ligament or after sectionof the vagi, and is not due to increased heart's action but to in-creased peripheral resistance.

Our first experiments with piperidine were made a few days afterthe publication of Tuninicliffe's results and hence possess no claims topriority, but, at the time they were made, we were unacquainted withthese results, and were guided into trying the effects of piperidine byits probable presence in the nicotine molecule.

These preliminary experiments having made it somewhat probablethat the effect of nicotine upon arterial blood-pressure was connectedwith the presence of a reduced pyridine or pyrrol group in its molecule,we proceeded to test other piperidine derivatives with the view ofascertaining whether this was a common property of the reducedpyridine ring. Injection of piperin, the alkaloid of pepper, whichcontains piperidine united to piperic acid, was found to cause a similar,though less extensive rise; while coniine (i.e. a-propyl-piperidine,

CH2CH2 CH1CH2 CH-C2,H7) produced a still greater rise than piperidine itself

NHFurther investigation has shown us, not only that the rise of blood-

pressure is due to very similar actions in all these cases, but that manyof the physiological properties of the three bodies, piperidine, coniineand nicotine are closely allied, and due, we believe, to a common cause-the presence of a reduced pyridene or pyrrol group-although intensifiedin different degree by the presence of different side groups introducedinto the molecule.

For the purpose of more clearly correlating the actions of the threecompounds, our experiments are described under various headings. Ineach section, the known facts which have been recorded separately foreach of the compounds by different observers are stated as concisely aspossible, because these when looked at collectively greatly strengthenthe case for a closely allied action due to a closely allied chemicalconstitution. Only those points in which our results differ from those

1 This Journal, xvii. p. 275. 1894-5.2 Centralblatt.ffir Physiologie, x. S. 777. 1897.

276

PIPERIDINE, CONIINE, NICOTINE.

of others, or those which, so far as we have been able to ascertain, arenew, are treated of in detail.

General Effects.The effects of subcutaneous injection of nicotine in frogs have been

described by various observers', and those of piperidine and coniine byothers2. The accounts given by these various writers for the differentdrugs are closely alike and clearly show that the action is a similar onein all three cases.

We have found that the effects obtained vary somewhat accordingto whether the free base or a salt obtained by its neutralization isadministered. In the first place, equal results are obtained with muchsmaller quantities of the free bases; and, in the second, there is muchless of that increased nmuscular activity which has been described bymany observers as an initial stage in nicotine poisoning. The probablecause of these differences in intensity of action may, we think, be thatit is in both cases the free base which is really active and that whenthe salt is given it is only gradually liberated by the action of thealkalies of the blood and lymph. The increased muscular activity isalso probably due in great part to the irritation caused by the freealkali.

Whatever be the reason, 2 milligrams of free nicotine or ten of freepiperidine usually produce as great an effect in a frog as ten milligramsof nicotine or 50 of piperidine administered as hydrochlorides. Theeffect is also produced more quickly, and in case recovery takes place itdoes so more rapidly, when the smaller dose of free base is given.

Piperidine is less strongly active than either coniine or nicotine,this we have found to hold throughout all our experiments. Theresult of subcutaneous injection of 10-20 mnilligrams of nicotine8 orof coniine, or of 50-100 milligrams of piperidine in each case as ahydrochloride in medium sized frogs (weighing about 25 grams), is theonset of a muscular paralysis which gradually becomes complete. Theappearance of the animals, and the characteristics of the paralysis arein all three cases identical. The fore-limbs become cataleptic, remaining

I For literature see Langley and Dickinson, This Journal, xi. p. 304. 1890.2 See Cushny, Journ. of Exper. Med. New York, i. p. 202. 1896.3 Exact doses cannot be stated for two reasons; in the first place, all animals are not

equally susceptible; in the second, the solutions of the drugs, especially those of piperidineand coniine, rapidly undergo oxidation and so lose activity. Hence the doses statedthroughout the paper must only be taken as rough approximations.

19-2

277


in whatever position they are placed. Both fore and hind-limbs becomeset in a peculiar position much more at right angles to the axis of thebody than normal. The frog when placed on its back can only turnover with difficulty, and later does not turn at all. The hind-limbsbecome completely paralysed, remain extended, and finally fail torespond to the strongest stimulus. Respiratory movements cease,although the heart still continues beating in a normnal manner fora long time afterwards.

Peripheral and central effects on the nervous system, and onskeletal muscles.

Motor paralysis. If the frogs be pithed after the final conditionabove described has established itself, and the sciatic nerves be exposedand stimulated; it is found, in all three cases, that even thestrongest electrical stimulation calls forth no response in the cor-responding muscles, these are however excitable and give a normalcontraction when directly stimulated. The effect obtained is usuallyas perfect as if the animal had been completelv curarized, but ofcourse the doses given above are much greater than that required ofcurare.

The peripheral action of coniine on motor nerve fibres has long beenknown. The peripheral motor paralysis caused by nicotine was firstdescribed by Rosenthal' and has been confirmed by other observers.In the case of piperidine our results differ slightly from those ofCushn y2, who was unable to obtain complete motor paralvsis; but,did however obtain it with a-methyl- and a-ethyl-piperidine andwith coniine in smaller doses.

This result as pointed out by Cushny clearly shows the effect ofintroducing a side group in increasing the toxicity. Our experimentsshow throughout a similar effect. Cushny attributes the failure toinduce complete motor paralysis to stoppage of the heart before asufficient dose could be given. We have never found a direct stoppingof the heart due to injection of neutralized piperidine even in verylarge doses (100-200 miligrams); the heart has always gone onbeating for one or two hours after complete motor paralysis andstopping of respiration. Cushny is here dealing with an interestingstage of partial paralysis, which is also described in the case of

l Comptes rendus de la Societe de Biologie, p. 91. 1867.2 Journ. of Exper. Med. p. 204. 1896.

278


mammals after administration of pituril by Langley and Dickinson'as occurring before complete paralysis comes on. In this stage, asinale induction stimulus causes an apparently normal effect; but ifsingle stimuli are repeated at intervals of half a second the contractionsrapidly dwindle away and soon no effect whatever is obtained. Onapplying a tetanizing current the effect obtained is a somewhatprolonged twitch and not a tetanus. If the tetanizing current beapplied at short intervals, a jerk is obtained each time; but if theinterval is made too short, the jerks diminish and disappear. Inno case is a tetanus obtained on prolonging the period of tetanicstimulation, but the muscle relaxes again and remains relaxed aslong as the tetanizing current is applied. Cushny supposes thatthis partial paralysis may be due to piperidine rendering the nerveplates in muscle more liable to fatigue. According to this observer,the irritability of the nerve plate does not seem to be affected bypiperidine, because the minimal current which on application to thenerve was capable of causing a muscular contraction coutinued to doso after poisoning by piperidine. This is interesting as showing thatthe injury of the nervous mechanism within the muscle is one affectingits power to carry impulses and not its excitability. But, in ouropinion, there is nothing in the experiments to show that the seatof injury is the nerve plate. As in the case of curare, we onlyknow from such experiments that the seat of injury lies in theintramuscular nervous mechanism, which does not necessarily implyin the nerve plates.

Seat ofparalysis. The difficulty of the problem increases when onegoes on to inquire as to the seat of the general paralysis abovedescribed; as to whether it is entirely due to the peripheral paralysisof motor fibres which undoubtedly takes place, or whether it is duechiefly or in part to a paralysis, of the centres in the cord preceding oraccompanying the peripheral motor paralysis.

Accordingly one finds most diverse opinions expressed in theliterature on the subject, not only as to the action of the threedifferent drugs but also as to the action of each.

Some observers have assumed, merely because reflex action ceases,that the central nervous system is affected8; others have exposed the

1 An Australian alkaloid, the physiological action of which has been compared withthat of nicotine by Langley and Dickinson.

2 Loc. cit. p. 281.8 Langley and Dickinson, This Journal, xE p. 272. 1890,

279


sciatic nerve and stimulated after dividing it; on finding that reflexaction ceased simultaneously with or even before motor action on thesame side', they have concluded that the spinal cord is involved in theparalysis. A little consideration will show that both conclusions areillogical. Failure to obtain a reflex may be due to break down of thereflex arc at any point, or to partial break down at two or more differentpoints. Suppose one is testing through the whole arc from sensorynerve-endings at one end, via cord, to motor nerve-endings at the other.If paralysis is obtained, it may be due (a) to injury of sensory endingsor of the connections of these with sensory nerve fibres, (b) to injury ofnerve cells or connections of these in the cord or spinal ganglia, (c) toinjury of motor nerve-endings or connections of these with motor nervefibres, (d) to injury of nerve trunks, (e) to injury of two or more up toall portions of this nervous mechanism. It is evident therefore that ifthe positive conclusion that the cord is affected is to be drawn in cases,such as those we are discussing, where unldouLbted peripheral paralysiscoexists; at least the peripheral nerve-endings of the reflex arc which isbeing tested must be completely protected from the action of the poison.The fact, that after administration of any of the three substances weare considering, the reflexes gradually become weaker and disappear,and that it is only after the disappearance of all reflex action thatcomplete peripheral motor paralysis can be demonstrated, does not showthat the disappearance of reflexes is due to paralysis of the cord; thedisappearance of reflexes may have been equally well the signal ofparalysis of sensory or motor nerve-endings. When the central end ofthe sciatic is stimulated in such a condition, the failure to obtain a reflexmay be due to the poisoning of the peripheral motor endings and not toany effect on the cord:

In the case of coniine, it is usually stated that the motor nerves areperipherally paralysed, and afterwards the motor centres in the brainand cord2. We have not been able to obtain any clear evidencehowever that the cord and brain are really involved. The argument

I v. Anrep, Arch. f. Anat. u. Physiol. 1879, Physiol. Abth. Supp. S. 182.2 Kolliker first showed conclusively that the main paralysing action of coniine was a

peripheral one on the motor nerves (Virchow's Arch. x. S. 235. 1856), he found that afrog's leg left attached by the sciatic nerve only, could be reflexly stimulated at a timewhen the other leg through which the poison had circulated could not be made to reacteither to peripheral stimulation through its nerve or reflexly. Similar results wereobtained by Guttmann (Berl. klin. Wochensch. 1866, S. 58), who denies any actionexcept a peripheral one on motor nerve fibres. See also Cushny, Journ. of Exper. Med.I. p. 208. 1896.

280


that smiall doses cause increased muscular activity like faint strychninepoisoning has not much weight, as there is no ground for the assumptionthat this takes place by a direct action on the cord. If the resistanceof any part of the reflex arc is diminished increased muscular actionmust be the result; if cceteris paribus the peripheral motor ends bestimulated the muscular contractions will be increased beyond thenormal. There is hence as much ground for believing that theincreased muscular activity is due to a peripheral stimulus whichprecedes a peripheral paralysis, as for believing that it is due to acentral stimulus preceding a central paralysis.

Most observers are also agreed that nicotine causes paralysis chieflyby acting on the central nervous system, but on no better evidencethan in the case of coniine.

According to v. Anrep1 the cataleptic condition of the fore-legspreviously described is independent of the central nervous system, forit still takes place after brain, medulla, and spinal cord have beendestroyed. The same observer states that the paralysis of the motornerve-endings is preceded by a stage in which no reflex action can beobtained, which is shown by an absence of effect upon the muscles onstimulating the central end of the sciatic, occurring some time before noeffect is obtained on stimulating the peripheral. Since no precautionswere taken to prevent peripheral action from destroying the reflex, thisproves nothing.

Langley and Dickinson2 give six successive stages of nicotinepoisoning, which are as follows: (1) Stage of excitation. (2) Stage ofspasms. (3) Stage of quiescence. (4) Stage of flaccidity. (5) Stageof paralysis of the central nervous system; no reflex can be obtained.

1 Arch. f. Anat. u. Physiol. 1879, Physiol. Abth. Sutpp. S. 167. This observer is alsoof opinion that the clonic spasms seen in nicotine poisoning are due to a central actionon the medulla and cord. Such clonic spasms occur with all three drugs, and probablyhave a common origin. von Anrep gives two proofs that the spasms have a centralorigin, viz. (1) that they cease after cutting the nerves to the part, (2) that they continuewhen the blood is prevented from reaching the part so that there can be no peripheralexcitation. The first proof is insufficient, for before cutting the nerves a peripheralexcitation of nerve-endings would cause exaggerated effects to follow normal stimuli fromthe cord, and these would cease on cutting the nerves; the second proof would besufficient if experimentally true, but we have only obtained the opposite result. In a legwhich has all been ligatured save the sciatic nerve we have never seen any clonic spasmswith any of the three drugs, and in the case of piperidine and coniine after ligature of theabdominal aorta Cushny (loc. cit.) has obtained a similar result, there being no cloniccontractions of the legs observed.

2 Loc. cit. p. 272.

281


(6) Stage of paralysis of motor nerve-endings; the muscles are stillirritable and the heart still beating.

We agree that all these six conditions may apparently be recognizedwhen nicotine is administered and the animal observed without anyfurther precaution; but we deny that stages 5 and 6 show that centralparalysis takes place first, followed later by peripheral paralysis, orindeed that there is necessarily any central action whatever. Thegradual disappearance of reflex action may be due to an increasingblock at the periphery, and the final stage of complete motor paralysismay only mean the completion of this peripheral blocking. In fact,we have found, when means are taken to prevent such peripheralblocking in one limb, that a reflex can here be obtained after motorparalysis is complete in the other limb to which the poison has hadaccess'.

In the case of piperidine and coniine the problem of the seat ofparalysis has been recently more carefully studied by Cushn y 2, whofound that ligature of the abdominal aorta in frogs prevented the usualconvulsive movements of the hind-limbs after injection of piperidine.In other frogs in which one iliac artery was tied, and then the reflexexcitability determined by Ttirck's method after injection of piperidine,it was found that no departure from the normal reflex time was ob-tained in either leg, although the movements in the unligatured limbwere reduced to weak jerks. Cushny therefore concludes that thereflex arc is unaffected except by action at the ends of the motornerves.

We have tried to compare the share of peripheral and centraleffects in producing paralysis by ligaturing in the frog, the whole of onehind-limb except the sciatic nerve, and then injecting in differentexperiments each of the three drugs. The doses injected were 20milligrams in each case of coniine or nicotine or 100 milligrams ofpiperidine. The results obtained with all three drugs were as far as wecould judge identical.

In each case, after destroying the brain, both sciatic nerves wereexposed at the thighs; a ligature was passed round one leg at thethigh and tightly tied excluding the sciatic; the dose of the drug to betried was then injected into the dorsal lymph sac or peritoneal cavity.At short intervals the effects of stimulating by a tetanizing current(a) the web of each foot, and (b) each sciatic nerve were observed. It

2 Journ. of Exper. Med. I. p. 202. 1896.

282

I Vide infra.


was soon found that (with all three drugs) the reflexes to stimulationof the foot on the side which had been ligatured were much strongerthan on the side which bad not been ligatured; this occutrred at a stagewhen contraction was still obtained on each side on stimulating thesciatic of that side. The minimal strength of stimulus which gave anobvious effect on the ligatured side was taken, this gave no effect onthe unligatured side. This difference in effect increased as the experi-ment progressed, and finally a stage was reached at which, neitherstimulation of the web of either foot, nor of either sciatic produced anycontraction on the unligatured side, although the muscles of this sidestill remained excitable to direct stimulation. At this stage, however,it was found that strong stimulation of the web of the unligatured footcaused an obvious reflex contraction of the gastrocnemius on theligatured side.

These experiments clearly show, in our opinion, that there is astrong peripheral action on motor nerve terminations from the begin-ning and not merely at a stage when reflexes are abolished; that thisperipheral action plays a large share in the abolition of the reflexes;and that peripheral paralysis may be complete at a stage when there iscertainly not complete central paralysis.

Our experiments do not show that central action is completelyabsent, nor that there is no effect on sensory terminations. Suchactions or their absence could only be demonstrated by quantitativeexperiment on extent of reflex action before and after administration ofthe drugs, under such conditions that the centre only became affectedand not either set of peripheral nerve terminations. Such quantitativeexperiments are exceedingly difficult to carry out, on account of thechange in excitability of the peripheral terminations due to cutting offthe blood supply, and on this ground it is impossible to certainly provethe presence or absence of a small amount of central effect.

In our experiments it was necessary, in order to obtain a reflexcontraction of the gastrocnemius on the ligatured side, by stimulation ofthe web of the foot on the unligatured side, to use a much strongerstimulus than was necessary to obtain such an effect in a normal frogin which the brain only had been destroyed. But on making a controlexperiment, in which one leg was ligatured and no drug given, but thestrength of stimulus necessary to obtain such a reflex determined afterthe lapse of an equal time; it was found that here also a much strongerstimulus was necessary. Hence we are unable to state whether thesedrugs possess a slight amount of central paralysing action or not, but

283


we are certain that they have a marked peripheral paralysing action,and are inclined to believe that this may be sufficient to account for allthe paralysis observed.

Effects on peripheral gantglion cell connections. It is here that wemeet the only important exception to the otherwise very similarphysiological actions of the three drugs. Langley and Dickinson' asis well known, have shown that nicotine in moderate doses paralysesthe nerve cells of various sympathetic ganglia-or rather the endingsof pre-ganglionic fibres in sympathetic nerve cells-before paralysingthe peripheral (post-ganglionic) endings of the nerve fibres, and haveturned the effect to good account in determining the position of nervecells in the track of sympathetic nerve fibres.

The same observers2 have further tested a large number of otherdrugs including coniine for a similar action. None of these drugs werefound to possess an action in this respect at all approaching in intensitythat of nicotine, although coniine is included in a group with threeothers (out of sixteen tested) as having a slight effect. The experimentsconsisted in stimulating in the rabbit the cervical sympathetic aboveand below the superior garnglion, before and after administering thedrug either by painting on the ganglion, or by intravenous injection,and observing the effects on the ear and on the pupil of the eye. Itwas found that while 7 to 10 milligrams of nicotine completely paralysed,50 nmilligrams of coniine were required to obtain a partial paralysis:painting the superior cervical ganglion also produced a slight, but onlya slight paralysis.

This restult we can completely confirm. At Dr Langley's suggestion,we have tested the effects of the three drugs in this manner on thecells of the superior cervical ganglion in the rabbit, and found that5 to 10 milligrams of nicotine caused complete paralysis to stimulationbelow the ganglion, while 50-60 milligrams of coniine were requiredto produce even a partial paralysis which rapidly passed off. In thecase of piperidine, even with a dose of 100 milligrams, we were unableto detect any paralysing action whatever. It might be thought thatthis difference in action was due to the unreduced (pyridene) groupwhich the nicotine molecule contains, and which is not present in theother two drugs. On testing this experimentally, however, by injectionof pyridene, we found that such an explanation would not hold, for

1 Proc. Roy. Soc. XLVI. p. 423. 1889; This Journal, xi. p. 283. 1890; Langley, ibid.p. 131.

' This Journal, xi. p. 509. 1890.

284


pyridene alone has no paralysing action. It has already been statedthat the presence of a side group and the character of this group havean important action in increasing by a variable amount the potencyof the drug. It may be that this is the true explanation of thedifference in action here, and the character of the group added innicotine may cause it to react chemically with more ease on theperipheral ganglion cells or their connections'. Whatever be thecause, the experimental fact is that nicotine powerfully stimulatesand then paralyses the cells of the superior cervical ganglion or theirconnections, while both pyridene and piperidine do not appreciablyaffect them.

Effects on the respiration.

All three drugs cause at first a slight quickening and markeddeepening of the respirations in mammals2, so causing an appearanceof dyspncea. In the case of coniine and nicotine the rate becomesslower again and the individual inspirations shallower, uintil finallywith a sufficient dose they altogether cease, and the animal succumbsunless artificial respiration be applied. The heart is little affectedby a dose sufficient to cause death by respiratory failure. In the caseof piperidine, after the period of excitation there is a slight fall beneaththe normal, but recovery takes place rapidly, and even large doses arenot fatal. Thus, in a rabbit weighing about 1l5 kilogram, underchloroform, we have injected subcutaneously 250 milligrams withoutcausing death.

The excitatory effect on the respiration is shown on the accomn-panying blood-pressure tracings'. Here the animal was under artificialrespiration, and before each injection there was no voluntary action ofthe muscles of respiration. But on injection there was on each occasionviolent respiratory efforts, producing the effect shown on the blood-pressure tracing.

We have not attempted to decide whether this preliminary exci-tation and subsequent paralysis of the respiration is due to excitation

1 If the nicotine molecule contains pyrrolidine instead of piperidine (see Introduction),it may be that this action may be due to a difference between these two bases. This weintend to test as soon as we can obtain a supply of pyrrolidine.

2 For the effects in the frog see " General effects."3 See section on "Effects on arterial blood-pressure." In a prolonged blood-pressure

experiment where coniine or nicotine are injected, it is always necessary to employ arti-ficial respiration to prevent death from respiratory failure.

28a

B. MOORE AND R. RO W.

and paralysis of the respiratory centre or of the motor nerve termi-uations of the respiratory muscles. Langley and Dickinson' found,in the case of pituri, that previous section of the vagi lessened thequickeninig effect, but that the respirations still became much deeper;while it is stated that coniine kills by paralysing the respiratorymuscles. Analogy drawn from the action of the drugs on the nervousmechanism of skeletal muscle would indicate that the paralysis is toa large extent peripheral2.

Effects on the circulation.

Perfusion through the excised frog's heart. These experiments wereperformed on hearts enclosed in the usual fashion in a Schifer's heart-plethysmograph. The druas were weighed out, slightly diluted withRinger's circulating fluid, neutralized as exactly as possible with dilutehydrochloric acid, and made up with Ringer's fluid to one per cent.solution (reckoned as base). These stock solutions were diluted downto 1 per mille with Ringer's fluid when required for use. By thisprocedure, the final reaction of the dilute solutions on litmus papercould not be distinguished from that of the Ringer's fluid.

The perfusions were carried ouit under a pressure of 20-25 centi-metres of saline, and in each case a normal tracing was taken withRinger's fluid before perfusing the drug. We have obtained recordsof the action of all three drugs on the same heart, the normal beatbeing re-obtained after each drug by perfusing Ringer's fluid, and theheart continuing to beat spontaneously throughout the experiment.Portions of the record of one of these experiments are reproduced inFig. 1. It will be seen from these that the action of all tbree drugsis the same; the heart is slightly slowed and the duration of the systoleis much prolonged3. Solutions of this strength may be perfused fora long time without producing any other effect; the heart also doesnot becomne irregular or cease. Irregularity is usually produced withi per cent. solutions, but often passes off; with one per cent. solutions,

1 This Journal, xi. p. 282. 1890.2 See arguments in preceding section.8 Movement up means contraction, the plateau at the top shows the duration of

systole, movement down shows dilatation. After each of the tracings 2, 3 and 4, Ringer'sfluid alone was perfused and the rate and form shown in tracing 1 were recurred to, Theslowing and prolongation of systole are apparent to the eye, on bathing the frog's heartin situ with a strong neutralized solution 2 per Jent. of piperidine or nicotine.

286

PIPERIDINE, CONIINE, ATICOTINE.

the heart becomes irregular and ceases to beat; but this is probablymore due to alteration in tonicity than to any direct action on the heartfibres.

Tracing 1. Riniger's Fluid. Tracing 3. Coniine.

Tracing 2. Piperidine. Tracing 4. Nicotine.

Fig. 1. Tracings showing the slowing and prolongation of systole produced byperfusion through the excised frog's heart.

Perfusion through blood vessels. These experiments were carried outon toads and frogs in which the brain and cord were thoroughlydestroyed by pithing. A cannula was inserted either into the bulbusaortce, or one of the aortae (the other being included in the ligature)and the sinus venosus opened by a free incision. Normal saline wasthen perfused under a pressure of about 20 centimetres until the fluiddropping away was no longer coloured and the flow had become constantas estimated by counting. The neutralized drug in A per cent. solution'was then perfused at a like pressure and the effect on the calibre of thevessels observed by counting the number of drops from minute tominute.

The results obtained showed a considerable variation in differentexperiments, but such variations are found with all three drugs, andin all respects their action is alike. Thus in some experiments weobtained an almost complete closure of the arterioles; while in othercases the flow only went down temporarily to half its amount and thenreturned to normal value or slightly beyond it'. In no case have weobtained an increased flow first, so that we believe we are justifiedin concluding, that neutralized solutions of all three drugs constrict

1 Prepared in normal saline solution, as described in the previous section for Ringer'sfluid, so as to ensure an identical reaction with the saline used for washing out.

2 In such cases of partial failure, increasing the strength of the perfused drug even upto 1 per cent. does not cause increased constriction.

287


arterioles after destruction of brain and spinal cord, in spite of theseapparently spontaneous variations which we are unable to explain.

Direct application to the mesenteric vessels. In order to test whetherthese drugs acted directly on the involuntary muscle fibres of the smallarteries, we have followed the method of applying a drop directly to amesenteric artery and vein, under observation through a microscopeprovided with a micrometer eyepiece. The frog was pithed and a loopof mesentery pinned out for observation in the usual manner, and thena drop of the drug under examination in 1 per cent. neutralized solutionwas run over the mesenteric vessels. We were met by the astonishingresult that these drugs so applied cause not constriction but dilatation.All three cause dilatation, but the effect is most marked in the case ofconiine, which may cause the arteriole to increase threefold in diameter.This effect is not due to the drugs being used in too strong solution, sothat an initial stage of excitation passes rapidly off into paralysis beforebeing observed, for no contraction is observed with weaker solutions'.

We attempted to discover whether the effect might be ascribed tolocal irritation, by making observations on the miesentery after injectionof each of the three drugs in 1 per cent. solution into the dorsal lymphsac, but no decided effect upon the mesentery in either direction couldbe observed We are hence unable to give any explanation of thisunexpected result, and can only record it as an experimental obser-vation 8.

Some effects of intravenous injection in mammals. References havealready been made to the action of piperidine and nicotine in raisingblood-pressure after intravenous injection in mammals, to these may beadded that the rise after nicotine has been described by Traube,

1 Neither is such a fall of blood-pressure obtained in mammals even after injection oflarge doses as might be expected if such paralysis of involuntary muscle-fibre took place.

2 We have also failed to obtain constriction of mesenteric arterioles after injection of1 c.c. of 1 per cent. solution of dried suprarenal into the dorsal lymph sac in the frog, thisresult is probably due to absorption of the injected material by other tissues before it canreach the blood vessels; for we can confirm Oliver's observation that a particle of drysuprarenal or a drop of 1 per cent. solution applied locally to a mesenteric arteriolealmost causes obliteration at the spot where it is applied.

3 v. Basch and Oser (Mediz. Jahrbiucher. Wien 1872, S. 367) found that injection ofnicotine in mammals caused primary pallor followed by flushing, which result was stillobtained after section of the cervical cord, and both statements are confirmed by L.angleyand Dickinson (loc. cit. p. 298). The latter observers found in some regions dilatationfollowed by constriction, in others constriction followed by dilation; a variation in theamount of these effects might possibly explain some of our anomalous experimentalresults.

288


v. Basch and Oser, and Langley and Dickinson', we have not beenable to find any reference to such an action in the case of coniine.

The vaso-motor effects have been tested previously in the case ofnicotine by observing the appearance of pallor or flushing of thedifferent regions after injection. A detailed account of these vaso-motor changes is given by Langley and Dickinson '. These observersfound in the rabbit great primary pallor followed by flushing in thewhole of the abdominal viscera and spleen, ear, and skin generally:and in the region of distribution of the fifth nerve primary fluishingfollowed by great pallor, which lasted for some minutes. After repeatedinjections these vasomotor effects became lessened and disappeared.

WTe have taken tracings of certain effects of the three drugs afterintravenous injection in the cat or dog, and here reproduce a record ofeach of these; the similarity of the three tracings is obvious.

The animals were in all cases aneesthetized with chloroform andkept under chloroform and morphia throughout the experirnent. Asufficient dose of atropine was usually given at the beginning of theexperiment to paralyse the vaguis action of the heart. In preliminaryexperiments it was found that each of the three drugs at first caused avagus action on the heart, slowing it and preventing any excessive riseof blood-pressure, and that after two or three small doses this effectdisappeared. In order to eliminate this effect in animals in whichcomparative effects of the three drugs were to be obtained, atropinewas therefore administered.

The uppermost tracing in each figure is of the auricular beat3', theforce of this is somewhat increased but the rhythm is little altered.An attenmpt was made to obtain a tracing of the ventricular beat also,but this was prevented by the violent respiratory efforts which followedeach injection of either of the drugs. The separate heart-beats shownon the blood-pressure tracing indicate sufficiently however that therhythm of the ventricle is not appreciably affected with the doses given,while the force of each beat is usually slightly increased.

The second tracing from above is that of kidney volume taken bymeans of a kidney box constructed on the same principle as that usedby Schafer and Moore to record changes in the volume of the

1 This Journal, xi. p. 292. 1892, where the other references are also given (p. 304).2 Loc. cit. p. 298.3 Taken by the method described by Oliver and Schaifer, This Journal, xviii. p. 256.

1895.

289

4

4

P4 *

.m .rD

~4.

000a

04 0

. 0

-

a)

o

0

.4.

014a

C)

0

to

o0

*0 0* 4.4

Oc)

a)

PH. XXII.

aS

0

0_a

C)Cs

5.

4-4

04a)

0

P4

P4

N

a)

ho

Ca)

0

O~ .

OC4)

0

.0*a

a)_4

0

Ca

ro

.4 tt

4.

00 .

40

C4

C)

.na

-FzP

44.O

Wa:

OI

20


spleen'. The recording lever of the tambour was arranged to givesmall excursions, so that the whole of each change in volume might berecorded.

The third tracing is of arterial blood-pressure taken from thecarotid.

The fourth tracing is a signal showing the time at which intravenousinjection (usually into the crural vein) of the specified amount of eachdrug took place.

The fifth tracing gives the time in seconds.On comparing the three figures it is seen that the form of the curve

of blood-pressure is very much alike in all cases, that there is very littleaction on the heart, and- that simultaneously with therise in blood-pressure a decrease in volume of the kidney commences and accompaniesthe rise of blood-pressure; as the blood-pressure slowly falls again thekidney slowly fills and the two effects pass off together. Such an effectcan of course only be due to an active constriction of the kidney vessels.

These results taken in conjunction with those given on perfusion inother sections make it evident that these drugs produce their effect onblood-pressure by affecting some mechanism peripheral to the centralnervous system. Only two views hence suggest themselves, viz. thatthe action is an excitatory one on the cells of peripheral ganglia, orthat it is a direct excitation to contraction of the muscle cells of thearterioles.

The former view is that taken by Langley and Dickinson2 in thecase of nicotinie, and is that with which we are most inclined to agree;

1 This Journal, xx. p. 5. 1896. Such a kidney box may be constructed by mouldinggutta percha made plastic in warm water over an inverted porcelain capsule of the propersize laid on a glass plate and moistened with water. Some of the gutta percha is pressed downall round to make a flange, and a sufficient amount must be taken to make the hollow cupso formed about ; inch thick, otherwise it will not have sufficient rigidity. Two such cupsare made, and at one place on each flange the rubber while still plastic is depressed, sothat when the two cups are apposed there is a rounded opening for the passage of therenal vessels. Into one half an open glass tube is inserted, which is bent inside so as tolie alongside the rim of the cup. This tube gives connection by a rubber tube with arecording Marey's tambour, and air is used as the communicating medium. The whole ismade tight with vaseline and kept together by a rubber band passed round the two halves.

2 This Journal, xi. p. 297. 1890. There is however no experimental evidence for thestatement there made that the cardio-inhibitory centre in the medulla is first stimulatedand then paralysed by nicotine. AU the experimental results given may be equallyexplained on the assumption that the ganglion cells in the heart are first stimulated andthen paralysed by the drug. And such an action would be more analogous to the othereffects of the drug than an action on the central nervous system.

292


for, although the strong peripheral action on skeletal muscle might leadone to expect a similar effect on involuntary muscle, we bave beenuinable to obtain such an effect experimentally and have on the contraryalways seen its opposite.

We have not observed any appreciable fall below normal after therise has passed away in the case of any of the three drugs; except nearthe end of an experiment. Usually, there is a rapid rise followed by aslow fall which becomes asymptotic to the normal level before theinjection, and remains a few millimetres above it for a long time.

If it is desired to obtain a full effect from any given injection, aconsiderable pause must be made after the previous injection. When abig effect is produced by injection of any one of the drugs and imme-diately after this has passed off, a similar amount of the same or oneof the others is injected, the effect obtained is always comparativelyfeeble. But in a few minutes' time a maximal effect is again obtain-able.

When the animal (cat or dog) which is being experimented upon isnot put under artificial respiration, after injection of a certain amountof coniine or nicotine (20-70 milligrams) the respiration ceases to bestimulated by the injections, becomes feebler and finally stops, so thatthe animal dies of asphyxia probably due to peripheral paralysis of therespiratory muscles.

Such an effect is never obtained with any reasonable dose ofpiperidine; thus we found that a rabbit of two kilos. could stand injectionof as much as 200 milligrams of piperidine without injury.

Hence in the case of an animal under natural respiration no changeis observed in the effects upon blood-pressure following injection ofconiine or nicotine, for the aninmal dies before the mecbanism isparalysed by whicb the blood-pressure is affected. But when theanimal is kept under artificial respiration and small doses of eitherconiine or nicotine (say 5 milligrams) are injected at short intervals(say of 5 minutes), we have found that after a certain amount of eitherdrug has been administered (30-50 milligrams in a dog of about5 kilos), no further effect on blood-pressure can be obtained. Thisposition cannot be reached by injection of piperidine alone except thedrug be given in large doses1; but when produced by either coniine

1 In one experiment on a dog of 2-85 kilos., 70 milligrams of piperidine prevented theeffects on blood-pressure of subsequent injection of either piperidine, coniine, or nicotine;but usually it is difficult to obtain such an abolition with piperidine alone.

20-2

293

B. MOORE AAND R. RO W.

alone or nicotine alone, no one of the three drugs is then capable ofproducing the slightest effect on blood-pressure.

The animal does not die in this condition provided the artificialrespiration is kept up; the heart continues to beat strongly andregularly and the blood-pressure remains moderately high. After thelapse of an hour or more the vasomotor effects of injection begin toreturn slightly.

~.._ ... ..... ...... .. ..- .._.. .. .......

Fig. 5. Absence of effect on injection of 1 milligram of neutralized nicotine, in the sameanimal as in figs. 2, 3 and 4, after previous admilnistration of large doses.

As an example of this aboliti'on of effect a tracinag (Fig. 5) is given,showing the negative result of inijection of 1 milligram of nicotine; thiswas taken from the same animal as had previously given a prodigiousrise of blood-pressure with half the dose (see fig. 4). At this stage theanimal (dog of 7.5 kilos) had received 10 milligrams of nicotine,30 milligrams of coniine and 60 milligrams of piperidine. After thisnegative result was obtained, and another similar with an additionalmiilligram of nicotine, 10 milligrams of nicotine were injected at onedose: this produced a slight effect, very much less than J milligram inthe beginning of the experiment, but afterwards no effect co'uld beobtained even with 20 milligrams of either coniine or nicotine. It isremarkable that there is no very serious fall in blood-pressure evenwith these excessive doses, such as might be expected if there wascomplete vasomotor paralysis at that stage when the drugs cease toraise blood-pressure. Thus in Fig. 5 the blood-pressure is about I ofits original value at the beginning of the experiment, about 3 hourspreviously, and even after excessive doses of coniine and- nicotine afterthis period there was iio appreciable fall in blood-pressure.

294


In another experiment on a dog of 6-5 kilos. complete absence of allreaction to any of the three drugs was observed after injection of only15 milligram of coniine as acetate; but usually more than this amountis requiired.

CONCLUSIONS.

The three alkaloids piperidine, coniine, and nicotine are very similarin physiological action although the intensity of action varies; and webelieve that this similarity is due to the presence of a reduced pyridine(or pyrrol) ring in each molecule, and further that the intensification ofthe action is due to the introduction of an organic radicle as a sidegroup into this ring.

The points of similarity in physiological actioni may be summarizedas follows

1. The subcutaneous injection of each drug (in frogs) in sufficientdose causes complete motor paralysis.

2. This motor paralysis is mainly due to paralysis of the intra-muscular part of motor nerves. This is shown by the fact that muscleremains irritable directly but not through its nerve, when the poisonhas been allowed to circulate through the muscle, and that anunpoisoned muscle may be caused to contract reflexly through a spinalcord which has been subjected to the poison. We have not been ableto demonstrate that there is no central action, btut this, if present,must be slight.

3. The excised frog's heart is somewhat slowed, and the durationof the systole prolonged by each of the three drugs.

4. The heart in situ (in mammals) is at first slowed but afterwardsunaffected.

5. The arterial blood-pressure is enormously raised, and the rise isdue to constriction of arterioles and not to increased heart's action.This constriction of arterioles is independent of connection with thecentral nervous system and is probably due to vasomotor excitationin peripheral ganglia.

6. At a certain stage this vasomotor mechanism probably becomesparalysed, for further administration of any of the three drugs nolonger affects the arterial blood-pressure.

295

a comparison of the physiological actions (from the

Documents