306 547.435-292:6I2.8I7 THE LIBERATION OF ACETYLCHOLINE BY POTASSIUM. BY W. FELDBERG1 AND J. A. GUIMARAIS1,2. (From the National Institute for Medical Research, London, N.W. 3.) (Received November 22, 1935.) RECENTLY Brown and Feldberg [1935-6] have shown that KCl liberates acetylcholine (ACh.) from the superior cervical ganglion of the cat. The theoretical implication, as regards the mechanism of liberation of acetylcholine by nerve impulses, made it desirable to examine this effect on other organs. Bezn ak [1934] has published a few experiments showing that potassium ions liberate acetylcholine from the frog's isolated heart. The organs used in the present research were the submaxillary gland, the tongue and the sweat glands of the cat. The submaxillary gland has a cholinergic nerve supply of postganglionic parasympathetic axons from the chorda tympani. The tongue has a double cholinergic innervation, from postganglionic parasympathetic fibres to the blood vessels in the chorda tympani, and from the motor fibres in the hypoglossal to voluntary fibres. The innervation of the sweat glands in cats provides an instance of a postganglionic cholinergic neve supply from the true sympathetic system. Cats. METHODS. (1) Salivary glands. Babkin, Gibbs and Wolff [1932] have shown that stimulation of the chorda fibres to the submaxillary glands, normally without effect on the arterial blood-pressure, causes a fall of pressure after eserine. This fall is abolished by atropine [Feldbferg, 1933]; it is due to the acetylcholine set free by the stimulation, which, being prevented from destruction by eserine, gets into the general circulatlon. Similarly any acetylcholine liberated by KCI from the gland might be expected to produce a fall of pressure in the eserinized cat. We have therefore 1 Rockefeller Foundation Fellow. 2 Fellow of Junta de Educagao Nacional, Portugal.
LIBERATION OF ACETYLCHOLINE BY POTASSIUM. 307 examined the effect on the arterial blood-pressure of small doses of KCl injected into the artery of the submaxillary gland before and after eserine, and the liability of this effect to paralysis by atropine. The cats were anesthetized with chloralose, or the brain was destroyed under preliminary ether according to the method described by Burn and Dale [1914]. Both vagi and cervical sympathetic nerves were cut in the neck, and the arterial blood-pressure was recorded from the femoral artery. The cats were eviscerated by removal of the stomach, small and large intestine and spleen. On one or both sides the branches from the upper part of the common carotid artery were tied, except those supplying the submaxillary gland. A short glass cannula was tied into the central stump of the lingual artery and closed with a rubber tube. The KCI was injected through this tube with a syringe; during the injection the carotid artery was clamped below the origin of the lingual artery, the clamp being removed as soon as the injection was completed. The chorda-lingual nerves were cut, and glass cannu]ae were tied into the Wharton's ducts to observe the effect of the KCI on the secretion from both submaxillary glands. Dogs. KCI was injected into the submaxillary artery, the venous effluent from the gland being collected and tested for acetylcholine. The dogs were anaesthetized with chloralose; the vagi and chordalingual nerves were cut. The preparation for the intra-arterial injections of KCI was the same as that described for the cat. The venous outflow was collected by means of a glass cannula introduced into the superficial jugular vein, all branches not originating in the submaxillary gland being tied. The blood was made incoagulable by intravenous injection of chlorazol-fast-pink, 1 c.c. per kg. of an 8 p.c. solution of the dye being injected. 2 mg. of eserine and 0-2 mg. of atropine per kg. were injected 20 min. before the KCI injections were made. One effect of the KCI was to cause such intense constriction of the vessels of the gland that the venous outflow stopped almost completely. To overcome this constriction the KCI injections were followed by an injection of a few c.c. of Locke solution. 7-10 mg. KCI per c.c. were injected in a solution containing a low NatI concentration, or no NaCl, in order to render the solution isotonic. (2) Perfusion of the cat's tongue and foot. The perfusions were carried out with Locke solution containing eserine in a concentration of 1: 500,000, using a Dale-Schuster pump. The preparation and perfusion arrangements were essentially the same
308 W. FELDBERG AND J. A. GUIMARAIS. as those described by Dale, Feldberg and Vogt [1936] for the tongue, and by Dale and Feldberg [1934] for the foot. The arterial cannulae were short and were connected with the perfusion pump by a rubber tube, through which the KlC injections were made with a syringe. During the actual injection the tube was clamped on the pump side of the point of injection, and the pump was stopped. The potassium was injected in an isotonic solution containing a low concentration of NaCl, or none. (3) Acetylcholine estimation in the venous outflow. The acetylcholine was estimated on the eserinized leech muscle and the arterial blood-pressure of the cat. The identification was further substantiated by showing that the depressor action was abolished by atropine, that the active principle was destroyed by alkali, and that the acetylcholine equivalents obtained from the same sample on both tests corresponded. This detailed pharmacological analysis was not completed with the venous fluid from every organ, but in each case more than one of the characteristics of acetylcholine were demonstrated. The active principle behaved in all details like acetylcholine, and we shall therefore refer to it as acetylcholine (=ACh.). The high potassium content of the venous effluent complicated its assay for ACh. Control experiments with KCI were therefore always carried out. In the experiments on the leech the potassium excess was eliminated by diluting the venous fluid with KlC-free Locke solution. EXPERIMENTAL. (1) Salivary glands. Cats. Arterial injections of 4-10 mg. of KCI (in 0-5 c.c. of NaCl-free Locke solution) had usualjy no appreciable effect on the arterial bloodpressure of a cat, apart from an occasional small pressor action. If the administration of KCl was repeated 20-30 min. after the intravenous injection of 0-1-0*2 mg. of eserine per kg., it produced, after a latent period of a few seconds, a fall of pressure, which was abolished by atropine. This is illustrated by' the experiment given in Figs. 1 and 2, taken from cats under chloralose. Intravenous injections of the same amounts of KCI had no depressor action after eserine. Control injections of NaCl were made into the submaxillary artery without causing a fall of pressure. Brown and Feldberg [1935] have shown that caesium has only a very weak effect in liberating ACh. from the superior cervical ganglion. In agreement with this finding we obtained, even after eserine, no
LIBERATION OF ACETYLCHOLINE BY POTASSIUM. 309 depressor effect with Cs arterially injected into the gland under conditions in which KCI produced a pronounced fall of pressure. Fig. 2, Fig. 1. Cat, 3.3 kg., chloralose, eviscerated, blood-pressure; arterial injections of 10mg. KCl into submaxillary gland. A before and B after 0-66 mg. of eserine; C after 2 mg.' of atropine. Fig. 2. Cat, 3'6 kg., chloralose, eviscerated, blood-pressure; 0-7 mg. of eserine intravenously. At A arterial injection of 10 mg. KCI; at B of 16 mg. CsCJ. for instance, shows the effect of arterial injections of 10 mg. KCI (A) and 16 mg. CsCl (B) on the arterial blood-pressure of a 3-6 kg. cat under chloralose to which 0-7 mg. of eserine has been given. PHT. LXXXVT. 20
310 W. FELDBERG AND J. A. GUIMARAIS. In spinal cats the depressor effect of arterial KCI injections after eserine was not regularly obtained, the blood-pressure being usually less sensitive to ACh. than that of the cat under chloralose. If the bloodpressure responded well to ACh., the depressor action of arterial KCI injections after eserine could always be demonstrated even with low arterial pressure. In Fig. 3 B, for instance, which is taken from a spinal cat, the blood-pressure was only 30 mm. Hg. Nevertheless, 4 mg. of KCI, injected arterially, produced a slight but definite fall of pressure (B), which could be obtained repeatedly. Before the eserine was injected, the KCI injections had no depressor action (A). The arterial KCI injections had only a slight secretory effect if eserine had not been injected; as a rule less than a drop of saliva was secreted, E a _~~~~~~~~~~~ _~~~~~~~~~~~...... _, Fig. 3. Cat, 3-1 kg., spinal, eviscerated, blood-pressure and salivary secretion. Arterial injection of 4 mg. KCI before (A) and 20 min. after (B) 0-6 mg. of eserine. and sometimes no secretion occurred at all. After the intravenous injection of eserine the secretory effect of the KCI increased gradually, being at its height after 20-30 min., 4-7 drops of saliva being then evoked by the KC1 injection (see Figs. 2 A and 3 B). Dogs. The venous effluent from the salivary gland of a dog, after the administration of eserine, but without KCI injection, had no action on the eserinized leech muscle or the cat's blood-pressure; but samples collected during and shortly after arterial injection of KCI contained recognizable amounts of ACh. (Figs. 4 and 5). 20 mg. of KCI, in 2 c.c., had been injected arterially and the venous outflow had been collected. Part of the venous effluent was tested in Fig. 4 A in a 10 p.c. dilution on the leech muscle, causing a contraction slightly greater than that produced by ACh. 1 in- 5 x 108. The effect of a potassium solution similarly diluted, and prepared on the assumption that the injected KCI
LIBERATION OF ACETYLCHOLINE BY POTASSIUM. 311 was completely recovered in the venous outflow, was tested at C. The actual potassium concentration of the venous effluent was undoubtedly weaker, since there is no reason to believe that all the injected KCI had been collected. The venous effluent collected during a control injection Fig. 4. Eserinized leech, 10 p.c. venous outflow from dog's salivary gland collected during injection of KCI (A), and Locke's solution (B). C=KCI control; D=ACh. 1 in 5 x 108. Fig. 5. Cat, 2-5 kg., chloralose, blood-pressure; 1 c.c. venous outflow from dog's salivary gland collected during KCI (C, D, H and J), and Locke's solution (A and G) injection. B = 10 mg. KCI; E = 1 c.c. ACh. 1 in 108; F = I c.c. ACh. 1 in 5 x 107. Between H and J 2 mg. of atropine. of Locke solution was tested at B. In Fig. 5 the same samples were assayed on the arterial blood-pressure of a cat. The absence of ACh. from the fluid collected after a control injection of NaCl into the salivary gland, and the lack of effect of KCI alone on the cat's blood-pressure, are shown in Fig. 5 A and B. The sample collected during the KCI 20-2
312 W. FELDBERG AND J. A. GUIMARAIS. injection into the gland had a strong depressor action (C, D), the effect being slightly less than that of ACh. 1 in 5 x 107, and being, moreover, abolished by atropine (H and J). Under the conditions of the comparison, with unknown constituents of the effluent affecting one or the other reaction to an unknown degree, the estimates obtained on the leech with a tenfold dilution and on the blood-pressure with the undiluted fluid show a remarkable conformity. (2) Tongue perfusion. After the injection of KCI into the lingual arteries the venous effluent from the tongue contained ACh. 8 mg. of KCl in 2 c.c. already just sufficed to produce a distinct effect. If higher amounts of KCl were injected, the ACh. concentration of the venous fluid became greater. In Fig. 6, 40 mg. of KCI had been injected in 5 c.c. of Locke's solution Fig. 6. Cat, 3 kg., chloralose, eviscerated, blood-pressure; A, B, D and F = 2 c.c. venous outflow from cat's tongue collected during KCl injection. A made alkaline, B made acid; C collected before KCI; E and G =2 c.c. ACh. 1 in 108. Between E and F 2 mg. of atropine. into the lingual arteries and about 6 c.c. of venous fluid were collected. 2 c.c. of this sample were injected at D into the cat, causing a fall of pressure slightly greater than that produced by 0-02 of ACh. (E). The depressor effects of the venous fluid (F) and of ACh. (G) were completely abolished by 2 mg. atropine. At C a control injection of 2 c.c. of venous outflow, collected before the KCI was injected, was tested; it had no depressor effect. A sample collected during another injection of KCI into the tongue was divided into two parts, one part being made alkaline by the addition of 0-2 c.c. N/10 NaOH per c.c., the other being made acid by the addition of 0-2 c.c. N/10 HCI per c.c.; both samples were
LIBERATION OF ACETYLCHOLINE BY POTASSIUM. 313 kept at room temperature for 20 min., then neutralized and injected at A and B. The depressor action of the sample kept in alkali (A) was abolished, whereas the acidified portion (B) had retained its depressor action. It is possible that the rich cholinergic parasympathetic nerve supply from the chorda tympani to the arteries and mucous glands of the tongue accounts for most of the effects demonstrated in this organ; it is also probable, however, that the motor innervation to the voluntary muscle fibres is responsible for some contribution to the total amount of ACh. liberated in response to KCI. (3) Sweat glands. The difficulty we encountered in these experiments was that the concentration of ACh. in the venous outflow was too low to allow dilution for testing, and the potassium concentration of the outflow was Fig. 7. Cat, 3-8 kg., chloralose, eviscerated, 0-4 mg. eserine intravenously, blood-pressure; 1 c.c. venous outflow from cat's foot collected during (B and D) and before (A) KCI injection. C and E, 10 mg. KCl; F and G, 20 mg. KCl. Between C and D, 1-5 mg. of atropine; between F and G suprarenals excised. itself often high enough to upset the test. Nevertheless, we have been able in several experiments to show that the venous effluent, collected during and shortly after potassium injections, had a depressor action on the eviscerated eserinized cat, which did not react, or reacted differently, to corresponding doses of KCI. The depressor action corresponded to that of a solution of ACh. 1 in 2 to 4 x 108, and was abolished by atropine (see Fig. 7). We might mention in this connection that intravenous injections of only 10-15 mg. of KCI often cause a rise of the blood-pressure of the eviscerated, eserinized cat, due entirely to an output of adrenaline from the suprarenals. This is illustrated by Fig. 7. This output of adrenaline in response to KCI is due to a direct action on the suprarenal medulla. In unpublished experiments of one of us (W. F.), in collaboration with G. L. Brown, it could
314 W. FELDBERG AND J. A. GUIMARAIS. be shown that by injections into the central stump of the cceliac artery of an eviscerated cat, the vena cava and the abdominal aorta being tied below the suprarenals, even 15-2 mg. of KCI caused an output of adrenaline, as shown by the rise of blood-pressure, whereas intravenous injections of these amounts of KCl were ineffective. The output of adrenaline was not abolished by intravenous injection of large paralysing doses of nicotine (20-30 mg.), but only slightly reduced. We have not made control experiments in which the sweat glands were excluded from the perfusion by ligation of the pads, as we were unable to demonstrate the liberation of ACh. with regularity. It is therefore not certain whether the acetylcholine found in the venous outflow of the foot arises entirely or only in part from the sweat glands, and in part from other tissues, such as striated muscles DISCUSSION. Brown and Feldberg have shown that KCI causes liberation of acetylcholine at ganglionic synapses. In our experiments the action has been on organs containing peripheral endings of postganglionic fibres. In some cases the motor endings in voluntary muscle may also have been concerned. This apparently general action of KCI, in tissues possessing a cholinergic innervation, lends support to the suggestion, made by Brown and Feldberg, that the discharge of acetylcholine may be effected by the K ions mobilized in the passage of the nerve impulse. No experiments have yet been made to test whether the chemical transmitter of adrenergic impulses is also liberated by potassium. SUMMARY. It has been shown that KCI injected into the arterial blood supply liberates acetylcholine from various organs (salivary glands, tongue, sweat glands). We wish to make grateful acknowledgment of our debt to the Medical Research Council for hospitality and to Sir Henry Dale for help and advice. REFERENCES. Babkin, B. P., Gibbs, 0. S. and Wolff, H. G. (1932). Arch. exp. Path. Pharmaak. 168, 32. BeznAk, A. B. L. (1934). J. Physiol. 82, 129. Brown, G. L. and Feldberg, W. (1935). Ibid. 84, 12P. Brown, G. L. and Feldberg, W. (1936). Ibid. 86, 2. Burn, T. H. and Dale, H. H. (1914). J. Pharmacol., Baltimore, 6, 417. Dale, H. H. and Feldberg, W. (1934). J. Phy8iol. 82, 121. Dale, H. H., Feldberg, W. and Vogt, M. (1936). Ibid. (in the Press). Feldberg, W. (1933). Arch. exp. Path. Pharmak. 170, 560.