The First Hint at the Existence of Chemical Neurotransmitters
Adrenaline might then be the chemical stimulant liberated on each occasion when the impulse arrives at the periphery.
Thomas Renton Elliott (1877–1961) is commonly given credit for being the first to suggest the existence of chemical neurotransmitters.1 That Elliott was a remarkably astute scientist and scholar there can be no doubt, but the question of whether Elliott actually concluded that nerves secrete chemical substances is more ambiguous than usually reported.
After completing the academic requirements for the medical degree in 1900 at the age of twenty-three, he decided to get involved in research before starting the clinical training. He soon demonstrated his potential for research and was awarded the prestigious Coutts-Trotter fellowship, which enabled to work with John Langley in the physiology department at Cambridge. Elliott left Cambridge in 1906 in order to complete his medical education. It was during the period 1900 to 1906 that Elliot completed the major portion of the work that led him to suggest that adrenaline was released at sympathetic nervous system synapses. What he meant by that, however, has generally been misunderstood.
When Elliott started at Cambridge, Walter Gaskell had recently stepped down as professor and head of the physiology department and John Langley had succeeded him. Langley assigned Elliott the task of extending his own observations on the similarity of the effects produced by adrenaline and by sympathetic nerve stimulation. This became the focus of Elliott’s work. Not wanting to be quick to publish, he systematically explored the action of adrenaline on different visceral organs and glands in different animal species and compared the responses to those obtained by stimulating the innervating sympathetic nerves. In doing this, Elliott went well beyond confirming his mentor’ s earlier observations.
Elliott presented his results at a meeting of the Physiological Society in May 1904. He first reviewed Langley’s evidence that adrenaline acts directly on the smooth muscles and not on the innervating nerves. He also noted that the effects of adrenaline on the muscle actually increase after cutting the nerve. This important observation, which was later confirmed and extended by others, is now called “denervation supersensitivity,” a phenomenon that will be discussed further in the chapter on Walter Cannon, who used it to detect the release of adrenaline. Elliott reported that at virtually all visceral organs the effect of adrenaline is identical to that produced by stimulating the innervating sympathetic nerve, and he concluded his presentation with the often-quoted statement that “Adrenaline might then be the chemical stimulant liberated on each occasion when the [sympathetic nerve] impulse arrives at the periphery.”2
Elliott had not only extended Langley’s observations by providing many more examples of the similarities in the effects of adrenaline and sympathetic nerve stimulation, but he also suggested that adrenaline might normally be released at the synapse. However, a careful reading of what Elliott actually wrote does not support the assumption that he had proposed that it is the sympathetic nerves that secrete the adrenaline. Although he hypothesized that adrenaline might be released whenever a sympathetic nerve impulse reached the muscle he seems to have concluded, in agreement with Langley, that it was the organ stimulated that did the secreting, not the nerve.
Elliott followed up his presentation at the Physiological Society with a 68-page publication entitled “On the Action of Adrenalin.” The paper contains an impressive amount of data demonstrating that adrenaline is effective only on smooth muscles and glands innervated by sympathetic nerves. He also speculated that synapses at the terminals of parasympathetic nerves and between all pre- and postganglionic nerves, as well as between spinal motor nerves and skeletal muscles, might share a biochemical similarity, although the substance is not adrenaline.
The cranial or sacral [divisions of the parasympathetic branch of the autonomic nervous system] on the other hand together with all the preganglionic “synapses” are rather related biochemically to the junctions of the skeletal [spinal] nerves with striped [skeletal] muscles.3
Elliott had based this speculation on the fact that muscarine, nicotine, and some other drugs, but not adrenaline, either blocked or facilitated transmission at all of these synapses. This was a most prescient speculation, as it is now known that one substance, acetylcholine, is secreted at all these sites. At the time, however, there was little interest in acetylcholine, and none of the drugs effective at these sites was thought to be natural substances in the body, as adrenaline was.
Among the conclusions Elliott reached in his 1905 paper was that
In all vertebrates the reaction of any plain [smooth] muscle to adrenalin is of a similar character to that following excitation of the sympathetic (thoraco-lumbar) visceral nerves supplying that muscle. The change may be either to contraction or relaxation…. A positive reaction to adrenalin is a trustworthy proof of the existence of sympathetic nerves in any organ…. Sympathetic nerve cells with their fibres, and the contractile muscle fibres are irritated by adrenalin. The stimulation takes place at the junction of muscle and nerve.4
Elliott did not conclude in any of his publications that sympathetic nerves secrete adrenaline. It is possible that John Langley may have discouraged him from further speculation in print as, according to Henry Dale, Langley was himself “impatient of speculative theory” and advised his students to “make accurate observations and get the facts. If you do that the theory ought to make itself.”5
Elliott’s speculation had little effect on others at the time, and Langley did not even mention the possibility that sympathetic nerves might secrete adrenaline in his 1921 textbook on the autonomic nervous system.6 This, ironically, was the same year that Otto Loewi reported the results of an experiment demonstrating that nerves actually secrete chemical substances.
Henry Dale, whose role in this story will be discussed in the following chapter, was a close friend of Elliott, and he may have been partly responsible for discouraging Elliott from pursuing his ideas at the time. Dale told Elliott about several of his own observations where the effects of adrenaline and sympathetic nerve stimulation were not identical. And well they might not have been, as we now know that it is not adrenaline but noradrenaline (norepinephrine), a closely related substance, that is actually secreted by sympathetic nerves in most mammals. Despite having demonstrated a number of similarities between the effects of adrenaline and nerve stimulation, Langley reported that in no case is “there complete correspondence between the action of a drug and the effects of nerve stimulation.”7
Elliott left Langley’s laboratory in 1906 and begin his clinical training at University College Hospital in London. He completed the medical degree in 1908, with multiple honors, and was appointed to the clinical staff there.8 Elliott also taught in the medical school, and, with Thomas Lewis, he introduced changes in medical education that encouraged the staff as well as the students to become involved in clinical research based on the experimental biological sciences.
Although he was heavily involved in clinical work and teaching, Elliott found time for additional research on the sympathetic nervous system. Invited to give the Sidney Ringer Memorial Lecture in 1914, he spoke about the adrenal gland’s relation to the sympathetic nervous system. Leaning heavily on the work of Walter Gaskell, Elliott reviewed the embryological evidence demonstrating that the secretory cells in the adrenal medulla, the chromaffin cells, arise from the same primordial cells (the ectodermal sympathoblasts) from which postganglionic sympathetic neurons originate.
We have seen how the ganglion cell [postganglionic sympathetic nerve] and the adrenalin cell are both derived from what is almost a common cell with power to transmit a nervous impulse or to excrete adrenalin.9
Elliott speculated that sympathetic nerves at one time had the capacity to produce and secrete adrenaline as do the chromaffin cells of the adrenal gland, but during the course of evolution they lost this ability.
Their present anatomical separation may be an index of a differentiation of functions which was once held [emphasis added] by the two in common, when the adrenalin liberation was a part of the nervous impulse.10
Elliott speculated that the terminals of the postganglionic sympathetic nerves might have acquired the capacity to store the adrenaline secreted by the adrenal gland.
It is conceivable that as the nervous cell developed its peculiar outgrowths for the purpose of transmitting and localizing the nervous impulse, it might lose its power of producing adrenalin and come to depend on what could be picked up from the circulating blood and stored in its nerve endings. Removal of the glands would cut off this source of supply, and paralysis of the nerves would result sooner or later in that territory where the nerves had been functionally most active and had consumed their stores.11
Elliott pursued the idea that adrenaline secreted by the gland was stored at the nerve-muscle synapse. He found, however, that a week after adrenalectomy some sympathetic nerves still had the capacity to evoke a response.12 In the end, Elliott concluded that circulating adrenaline from the adrenals is stored at most sympathetic synapses, but he was noncommittal about whether it is stored in the muscle or the nerve. He considered this “a question of secondary interest” because he regarded the place where the two meet—the myo-neural junction—as a separate unit.13 Although Elliott considered it possible that adrenaline might be involved in neural transmission even if a membrane separated the nerve from the muscle, Langley concluded that this would make chemical action at the synapse less likely, “for it would involve the secretion of a substance from the nerve endings.”14
Elliott continued to follow the later developments that eventually proved the existence of chemical neurotransmitters, but he did no further work on the problem. He wrote nothing more about the adrenal glands or the sympathetic nervous system after 1914. He preferred to guide and advise others doing research without getting directly involved in the work. Nevertheless, Elliott’s early research on the sympathetic nervous system and his contributions to medical education were so highly regarded that in 1913, although not yet thirty-six, he was elected to the Royal Society.
When World War I started in 1914, Elliott was stationed in France with the Medical Corp. He rose to the rank of colonel and was awarded the distinguished service order and made a commander of the order of the British Empire for his research on the treatment of war-related injuries. Elliott was very much involved in coordinating research on the causes and treatment of wound shock, which at the time was responsible for the loss of many lives. In this role, Elliott met the Harvard physiologist Walter Cannon, who was serving with the American forces in France in 1917. The two became good friends and remained in contact, mostly through correspondence, throughout their lives. Later, as will be described, Cannon became aware of Elliott’s early work on adrenaline and the sympathetic nervous system, and this influenced his own work on the same problem.
In 1953, seventeen years after Henry Dale had shared the Nobel Prize with Otto Loewi for proving the existence of neurohumoral secretions, he expressed his debt to Elliott in a dedication that read:
To T. R. Elliott,
Who had so much to do with the
beginning of these adventures
and, long after they have ended,
is still my counselor and friend.15
Elliot’s ill health in later years eventually forced him to retire, but he continued to provide thoughtful and helpful advice to all who sought him out. He died in 1961 at the age of eighty-three.16
A year after Elliott’s 1905 publication demonstrating a strong relationship between adrenaline and the sympathetic nervous system, Walter Dixon (1871–1931), a leading figure in British pharmacology, described a study that suggested that a humoral factor was secreted by the vagus nerve, the parasympathetic nerve that innervates the heart. Dixon was one of the first in Great Britain to have the title of professor of pharmacology, a position he had held at King’s College, London. When he transferred to Cambridge University, Dixon assumed responsibility for building a school of pharmacology committed to research. He was well aware of what was going on in John Langley’s physiology laboratory, and he certainly was familiar with Elliott’ s research on adrenaline and the sympathetic nervous system, as he was thanked by Elliott in a footnote for providing help with the manuscript.
Dixon was highly respected as a researcher and teacher. The young Henry Dale published one of his first papers with Dixon, and he always regarded him as a friend and a respected elder colleague. Dixon’s interests varied widely, and he tended to initiate the investigation of a chemical substance and then switch his interest to another substance almost immediately. He was one of the first to investigate the properties of mescal and other hallucinogens, and he tried several of these drugs on himself. After trying mescal he reported:
When sitting with closed eyes, balls of red fire pass slowly across the field of vision. Later these changed to kaleidoscopic displays, with ever-increasing colours, or revolving wheels of colour being arranged in a definite pattern, which constantly changes. Only seen with closed eyes. After-images are prolonged.17
Dixon was also interested in addiction, and he studied the action of such drugs as alcohol, morphine, cocaine, cannabis, and tobacco. Later he served on the League of Nations Committee on Drugs of Addiction and also on the league’s Committee on the Standardization of Drugs, which was headed by Henry Dale. From his varied experience with drugs, Dixon had formulated a general theory of drug action. He proposed that all drugs interact with endogenous substances to form a new chemical substance that is responsible for the effects produced. Dixon wrote that the reason “few drugs exert a similar effect upon all tissues” is that the drugs interact with different endogenous substances. This idea, if not identical to, was clearly very close to Langley’s theory of “receptor substances,” described in the previous chapter.
Much of Dixon’s research was guided by his belief in the importance of endogenous substances for understanding not only the action of drugs but also the action of nerve impulses. In 1906 he reported the results of an experiment that appeared to demonstrate that chemical mediation was involved at the synapse between the vagus nerve and the heart. It had been known since 1845 that stimulating the vagus nerve slowed heart rate, and by 1900 frogs were commonly being used in physiology classes to demonstrate the action of the vagus nerve and to illustrate how various drugs affected that nerve’s capacity to slow heart rate. It was also well known that while adrenaline accelerated heart rate, pilocarpine, muscarine, and some similar alkaloid drugs slow heart rate in the same manner that stimulating the vagus nerve did.
At the time there were many new drugs available, most of which were extracts from plants known to the natives of the various countries then being explored by Europeans. Muscarine, for example, was extracted from the poisonous mushroom Amanita muscaria, while pilocarpine, which causes salivation when chewed, was obtained from South American shrubs of the genus Pilocarpus. The various effects of muscarine in particular were studied, and this led to the realization that it mimicked not only vagal stimulation but most of the effects produced by stimulating other parasympathetic nerves. It was also known that the drug atropine, likewise obtained from a plant, blocked the effects of both muscarine and parasympathetic nerve stimulation.18 While it was not known how muscarine exerted its effect, Langley had demonstrated, as he had done with adrenaline, that it did not act on the axon or nerve fiber itself. These observations led to the speculation that some natural alkaloid substance might play a role in mediating parasympathetic nerve effects.
At a meeting of the Therapeutic Society in London held in December 1906, Dixon reported the results of an experiment on the vagus nerve. A brief report of the study was published the following year in a relatively obscure magazine. The title of both the presentation and the published report, “On the Mode of Action of Drugs,” reflected Dixon’s broader interest in how drugs interact with endogenous chemical substances. He described the reason for undertaking the investigation:
When a muscle contracts, when a gland secretes, or a nerve ending is excited, the cause in each case may be due to the liberation of some chemical substance, not necessarily set free in the circulation, as in the case of secretin, but more likely liberated at the spot upon which it is required to act. In order to test the validity of this reasoning, I investigated the action of the vagus nerve upon the heart.19
What Dixon did was to extract a substance from the heart of a dog, the donor animal, before and after stimulating the vagus nerve. It was a crude experiment even at the time.
Animals were killed by pithing; they were bled and the vagus nerves were then placed on the electrodes and excited for half an hour. The heart was next extirpated, placed in boiling water for ten seconds and extracted with alcohol.20
The fluid extracted from the boiled heart underwent further processing. As Dixon described the process, the extracted fluid was first
evaporated until dryness and then taken up again with 100 per cent alcohol. This was again evaporated off again on the water bath, and a few drops of normal saline solution added. The solution so obtained was found to have the power of inhibiting the frog’s heart.
Dixon wrote that fluid extracted from a nonstimulated heart also slowed heart rate, but to a lesser degree.
[The non-stimulated heart] also gave a supply of this inhibitory substance, but in a smaller degree than in the excited heart. I interpret these experiments to mean that some inhibitory substance is stored up in that portion of the heart [emphasis added] to which we refer as a “nerve ending,” that when the vagus is excited this inhibitory substance is set free, and by combining with a body in the cardiac muscle, brings about this inhibition.
It seems clear that Dixon believed the inhibitory substance was stored in the area of the heart where the vagus nerve joins it. Although a neural impulse is required to release that substance, Dixon never suggested that it was secreted by the nerve. He also concluded that drugs like muscarine that inhibit heart rate must also “act by liberating the inhibitory hormone,” whereas drugs like atropine, which block the action of both muscarine and vagal stimulation, “either prevent the liberation of the hormone, or saturate the substance in the end organ upon which it acts.” What Dixon concluded from his experiment was made even clearer by the recording secretary who summarized his 1906 presentation:
Professor W. E. Dixon gave an account of his experiments on vagus inhibition. He was of the opinion that the heart contains a substance—“pro-inhibitin,” which as a result of vagus excitation is converted into a chemical body—“inhibitin.” This substance, combining with the heart muscle, results in cardiac standstill.21
Dixon concluded that drugs and neural impulses both liberate endogenous substances:
Drugs act, as far as we can judge, by influencing the tissues in exactly the same way as the physiologist affects the tissues when he stimulates a nerve. In any case the analogy between the activity of a tissue, produced on the one hand by exciting a nerve electrically and on the other by the administration of some drug, is so close as to warrant the conclusion that the ultimate effect is produced by the same mechanism in both cases.
Dixon’s observation of the action of drugs like muscarine on structures innervated by parasympathetic nerves was analogous to Thomas Elliott’s much more extensive investigation of the action of adrenaline on structures innervated by sympathetic nerves. Because Dixon knew Elliott and his work very well, this tends to support the conclusion that Elliott never meant to imply that sympathetic nerves secrete adrenaline. At the very least, Elliott never clearly committed himself to that position at the time.
From the perspective of what is known today, it is difficult to account for the results Dixon reported. While we now know that the vagus nerve secretes acetylcholine to inhibit heart rate, it is highly unlikely that Dixon extracted this substance from the heart he was working with. Acetylcholine is very unstable, and the elaborate processing Dixon used to extract the substance from the stimulated heart makes it highly unlikely that any acetylcholine present would have remained active. The synaptic physiologist Hugh McLennan later wrote about Dixon’s methodology that it is: “inconceivable that this substance [acetylcholine] was present in Dixon’s extract.”22 There may well have been some substance present in the heart after it was stimulated for thirty minutes that could explain the slowing of heart rate Dixon observed, but it is unlikely to have been acetylcholine. It is not useful to speculate further about Dixon’s results as neither he or anyone else ever attempted to replicate the experiment. Dixon never published any actual data. He presented no numbers indicating how much the heart rate decreased or how reliably he could reproduce his results. He simply stated that the extracted fluid inhibited the frog’s heart.23
Dixon, commenting on the lack of influence of his earlier work, later noted that “the birth of the chemical transmission era was perhaps premature and the theory on which our discussion was centered, aroused little comment at the time, and its youth and adolescence were equally discouraging.”24
After Dixon died in 1931, Henry Dale wrote that “the late W. E. Dixon was at the time almost alone in proposing that the vagus nerve, secreted a chemical substance responsible for slowing the heart.” Dale later stated that “it was beyond doubt that Dixon, following Elliott’s suggestion concerning adrenaline, had at that early date a conception of the general nature of the mechanism which later evidence has completely justified.”25 It is true that Dixon had “a conception of the general nature of the mechanism”: he concluded that the chemical substance mediated the effect of the nerve impulse; but he did not conclude that this substance was secreted by the nerve.
Dale’s statement about Dixon’s contribution was made in 1934 when he and Otto Loewi were well on the way to proving that parasympathetic nerves secrete acetylcholine. Dale appears to have been overly generous in crediting Dixon and Elliott with the origin of the idea that autonomic nerves secrete chemical substances. While both Elliott and Dixon proposed that chemical substances are involved in sympathetic and parasympathetic nerve action, neither concluded that such a substance is secreted by nerves. Dixon was particularly clear in his belief that the substance liberated came from the stimulated heart muscle, not the nerve.
As noted above, Dale was originally skeptical, if not opposed, to Elliott’s idea that adrenaline is released (irrespective of its origin) at sympathetic nerve terminals. Dale’s skepticism was based partly on his observation (with George Barger) that several other amines were more potent than adrenaline in reproducing sympathetic effects and also on his observations that the effects of adrenaline and sympathetic nerve stimulation were not always identical. These observations led Dale and Barger to conclude in 1910 that:
To suppose that such bases [amines] and sympathetic nerve impulses alike owe their action to the liberation of adrenine seems to us to create additional difficulties for the conception.26
The problem of how the vagus nerve causes slowing of heart rate was not picked up again until 1920, when Otto Loewi began his seminal studies that eventually proved that the nerves slowing heart rate secrete acetylcholine. Ironically, at the same meeting of the British Medical Association where Dixon presented his results on cardiac slowing, Reid Hunt, an American pharmacologist working at the time with Paul Ehrlich in Frankfurt, reported that acetycholine is the most powerful substance known for lowering blood pressure, a well-established parasympathetic response.27 Hunt concluded that
[Acetycholine] is a substance of extraordinary activity. In fact, I think it safe to state that, as regards its effect upon the circulation, it is the most powerful substance known. It is one hundred times more active than choline, and hundreds of times more active than nitroglycerine; it is a hundred times more active in causing a fall in blood-pressure than is adrenaline in causing a rise.28
Because acetylcholine was not known to be a natural substance found in animals, it did not occur to either Dixon or Hunt that their reports might be related. However, about a decade later, Hunt’s report did influence Henry Dale, who at the time was demonstrating that acetylcholine was much more potent than muscarine or any other known substance in mimicking all the effects of parasympathetic nerve stimulation. The story of how Dale came to investigate acetylcholine is the subject of the next chapter.