Evidence has accumulated that acetylcholine (ACh) is capable of acting on neurones at sites other than the subsynaptic membrane. It has been shown for instance that it can excite mammalian sensory nerve fibres in the skin and mesentery (Brown & Gray, 1948;Douglas & Gray, 1953), carotid sinus pressure receptors (Diamond, 1955) and C fibres carrying activity from cutaneous touch receptors (Douglas & Ritchie, 1960). Among the first to suggest that acetylcholine might excite post-ganglionic sympathetic fibres were Coon & Rothman (1940), who injected acetylcholine into the skin and obtained sympathetic effects which were prevented by ergotamine and by degeneration of the sympathetic supply to the area. Recently, Brandon & Rand (1961) and Daly & Scott (1961) have reinvestigated the sympathomimetic effect of acetylcholine on the spleen, previously described by Farber (1936). They showed that close arterial injection of acetylcholine into the spleen produced contraction of that organ and this effect could be prevented by reserpine, by adrenergic blocking agents, by hexamethonium and by degeneration of the sympathetic supply. It seemed possible that this sympathomimetic effect of acetylcholine might be due to its exciting the sympathetic post-ganglionic adrenergic C fibres, and experiments were carried out to test this hypothesis. Some of the results of these experiments have already been communicated to the Physiological Society (Ferry, 1962). METHODSCats were anaesthetized with ethyl chloride and ether and then with intravenous chloralose, 80 mg/kg. The abdominal cavity was opened by a mid-line incision and the spleen and its arterial supply were separated from the stomach and omentum by tying and cutting the hepatic, left gastric and gastro-epiploic arteries. In some experiments the stomach was removed to gain better access to the arterial supply of the spleen.Acetylcholine solution was injected into the splenic artery through a polythene cannula placed in a suitable sidebranch and pointing toward the heart. The arteriesused for retrograde cannulation were the hepatic artery, the smaller of the two divisions of the splenic artery, or a small branch of the splenic artery just before it enters the spleen. In one experiment one of the gastro-epiploic arteries was used. The arrangement of the cannula was such that injections could be made without interrupting the natural blood supply during the injection.

(1960) and by Brandon & Rand (1961). These authors used the contractile response of the effector organ as an index ofthe amount of transmitter liberated. Interpretation of their results is, however, difficult because not only are we ignorant of the relation between transmitter liberation and effector response, but, as Vane (1962) has pointed out, many drugs alter non-specifically the tissue response to catechol amines. We have therefore done our experiment by collecting venous blood from the spleen of the cat and estimating the noradrenaline liberated by stimulation of the nerves.Cocaine, anticholinesterases and hexamethonium do not appear to affect the liberation of noradrenaline by the nerves or its uptake by the tissues of the spleen. METHODSCats were anaesthetized with ethyl chloride, ether and intravenous chloralose 80 mg/kg. The spleen, its nerve supply, arteries and venous drainage were prepared as described by Brown & Gillespie (1957), except that instead of tying the splenic nerve early in the preparation, we left ligation as late as possible in order to prevent those effects of neuronal rest that may appear after 1-2 hr (Brown, Davies & Ferry, 1961). In some experiments the spleen was perfused with the cat's own blood through a constant-output perfusion pump. This consisted of a tube of silicone rubber compressed by three plates. The two outer plates were input and output valves, the larger, middle or ventricular plate compressed most of the tubing between the valves. The valves and ventricles were operated by pulses of air produced in three coupled compressors which ran at 90 c/min. The output of the pump could be varied by altering the stroke volume. In most experiments the stomach and duodenum were removed in order to gain better access to the arterial supply of the spleen. Blood was taken from the cat through a polythene cannula placed in the stump of the superior mesenteric artery and, after passing through the pump, was ejected into the coeliac axis through a cannula placed in the hepatic artery. The coeliac axis was tied between the aorta and the hepatic artery after the pump hadbeenconnected and had begun to pump blood into the coeliac axis; there was thus no interruption of the blood supply to the spleen. Experiments showed that, if the splenic blood supply was interrupted, the resistance to * M.R.C. Scholar.

Brown & Gillespie (1957) studied the output of sympathetic transmitter in the venous blood from the spleen of the cat resulting from stimulation of the splenic nerves. They showed that the output depended upon the frequency of stimulation, being maximal at 30/sec, and falling away at higher and lower frequencies. No noradrenaline could be detected in the venous blood at frequencies lower than 10/sec. The administration of the adrenergic blocking agents dibenamine and dibenyline increased the output at all frequencies below 30/sec. The explanation suggested for these results was that combination with tissue receptors precedes the metabolic removal of liberated noradrenaline; at frequencies below 10/sec the unpoisoned receptor mechanism can remove all the liberated noradrenaline, and so none can be detected in the venous effluent. At frequencies between 10 and 30/sec increasing amounts overflow into the circulation because the receptor mechanism is swamped. The administration of a blocking agent prevents the uptake of transmitter by the tissue, and the noradrenaline then appearing in the venous blood gives a measure of the amount liberated by the nerve endings. It is therefore possible by measuring the amount of transmitter in the venous blood, before and after a blocking agent is given, to determine, for any frequency of stimulation, the amount of transmitter liberated, the amount taken up by the tissue, and the amount normally overflowing.Some experiments made for another purpose suggested that' previous activity might modify the amount of the transmitter overflowing when the nerves were stimulated. It had been found, for instance, that a first group of stimuli at 30/sec given to the nerve after a rest of 1-11 hr yielded 450 pg/stimulus, whereas a second, given 10 min later, produced an overflow of 1025 pg/stimulus (mean of 5 observations). These experiments led us to think that interruption of the normal constant centrifugal discharge of the sympathetic neurones might alter the peripheral processes of liberation

No abstract

1 The purpose of this investigation was to determine the long-term effects of a single dose of persistent anticholinesterases on muscle action potentials evoked by nerve stimulation. 2 Action potentials (APs), elicited by stimulation of the phrenic nerve, were recorded intracellularly in muscle fibres of mouse diaphragm. The time between stimulus and AP was measured and the variability of this latency was calculated during trains of APs. At the beginning of trains of APs there was an increase in latency, and this delay was also measured. 3 Within 3h of subcutaneous injection, a single dose (500nmolkg-1) of the anticholinesterase, ecothiopate produced about 90% reduction in the acetylcholinesterase activity of homogenates of mouse diaphragm muscle, but five days after injection, this activity was not different from values in untreated animals. The initial delay of APs and the variability of latencies were increased four fold and two fold respectively, remained at these maxima from the 1st to the 5th day after ecothiopate, and returned to the values in untreated animals between 15 and 27 days after ecothiopate. 4 These effects of ecothiopate on AP latency were dose-dependent and were also seen in extensor digitorum longus and soleus muscles. 5 Other anticholinesterases used were BOS (pinacolyl S-(2-trimethylaminoethyl)methylphosphonothioate), a quaternary compound, and diisopropyl fluorophosphate, a tertiary compound, which had effects similar to those of ecothiopate; the greater duration of the effects of this compound may be related to the greater duration of reduction in cholinesterase activity. 6 Ecothiopate had no effect on the delay or variability of latencies of endplate potentials which were recorded in cut-fibre preparations 5 days later. 7 It is concluded that the effects of ecothiopate on the latencies of indirectly-evoked muscle APs are postjunctional, may not be related to the degree of reduction in cholinesterase activity at the time of recording, and are not directly linked to necrosis.