It made no difference to me initially whether GABA was a neurotransmitter. My goal was the elucidation of its function in the nervous system, whatever it might be. However, I did sense that the “transmitter question” seemed to agitate a number of physiologists.
It would be difficult to discuss anything about the brain today without referring to chemical neurotransmitters. According to some estimates there may be as many as one hundred different chemical substances secreted by brain neurons, yet in 1950 even the possibility that neurons in the brain communicated chemically was rarely mentioned. Acetylcholine and norepinephrine (noradrenaline) had finally been accepted as neurotransmitters, but they were thought to act only at the autonomic nervous system synapses controlling the slow-responding visceral organs. How all of this changed is the subject of this chapter.
Henry Dale had briefly raised the possibility of the existence of brain neurotransmitters in his 1936 Nobel Lecture when he mentioned that acetylcholine had been found in the brain. He was alluding to work in his own laboratory by John Gaddum, who had recently reported finding it there.2 Dale remarked that acetylcholine must be there for some purpose, and he referred to Wilhelm Feldberg’s preliminary observations of physiological responses evoked by placing drugs that affect acetylcholine in the brain. However, Dale was characteristically cautious and concluded these remarks by saying, “We need a much larger array of well-authenticated facts, before we begin to theorize.”
Zénon Bacq recalled that Dale did express an interest in investigating whether there are neurotransmitters in the brain, but did not know how to proceed:
I remember that Dale often showed his interest in this huge problem during informal discussions. He was inhibited because of the anatomy of the CNS, it was impossible to utilize the methods used with the heart, smooth and striped muscles, viz. isolation of an organ, complete control of its blood vessels and nerves and bioassay or chemical analysis of perfusion liquids. He asked his friends, experts in the anatomy of the CNS, but the unique suggestion has been to use a rare species of marsupial in which the isolation of a well defined part of the brain could have been possible. One had to wait for the rapid postwar development of the biochemistry of the CNS; biochemically minded pharmacologists could build progressively a convincing picture which is now universally accepted. New sophisticated techniques were introduced.3
In general, little consideration was given in the early 1950s to any role neurotransmitters might play in the brain despite several reports of finding acetylcholine and later noradrenaline (norepinephrine) in animal brains. In 1948 Marthe Vogt and Wilhelm Feldberg, for example, had described the concentration of acetylcholine in different brain regions.4 By 1950 there were even reports that acetylcholine concentrations in some brain regions were increased following electrical stimulation, but this was not accepted as evidence that brain neurons normally secrete this substance.5
In 1954 Marthe Vogt, who had by that time left Dale’s laboratory at the National Institute of Medical Research for a position at the University of Edinburgh, described the concentration of noradrenaline and adrenaline in different brain regions of cats and dogs.6 Using Cannon’s term “sympathin” to refer collectively to both noradrenaline and adrenaline, she reported that “sympathin” is most highly concentrated in the hypothalamus and other brain areas from which it had been shown that electrical stimulation can evoke sympathetic visceral responses. Nevertheless, she did not conclude that “sympathin” is a neurotransmitter in the brain, and she ended her paper with the comment that the available evidence “did not answer the question of the function of sympathin in the central nervous system.”7 It is clear from the literature of the time that even those who had played major roles in proving that noradrenaline and acetylcholine are neurotransmitters in the peripheral nervous system hesitated to conclude that they played that role in the brain.
Wilhelm Feldberg had started to investigate what acetylcholine was doing in the brain, but when his Rockefeller Foundation support ended in 1936 he had to leave Henry Dale’s laboratory for a position in Australia, where he worked on other problems. Feldberg was able to return to England in 1938, but when World War II began the following year the investigation of possible neurotransmitters in the brain was not a high-priority area of research. After the war Feldberg resumed his investigations, using a cannula system that he developed for injecting drugs into the ventricles of the brain in awake and unrestrained animals.
In his 1963 book, A Pharmacological Approach to the Brain, Feldberg listed more than fifty different physiological and behavioral responses he had observed following injection of various drugs into the ventricles of the brain.8 Some of these drugs, like physostigmine, were known to affect acetylcholine. Among other physiological effects, he observed changes in blood pressure, pupillary size, salivation, and respiration. He also reported behavioral responses such as increased playfulness, rage, fear, and excessive eating (hyperphagia) when he injected the anesthetic (hypnotic) chloralose into the ventricles of the brain. The hyperphagia was similar to what Walter Hess, a 1949 recipient of the Nobel Prize, had observed in cats following electrical stimulation of the hypothalamus.9 Feldberg, however, did not mention neurotransmitters in his book, and he acknowledged that injecting drugs into the brain ventricles left open many questions about where and how the drugs were working: “To locate the site where a drug acts when penetrating the brain from the cerebral ventricles … is the main task anyone is faced with when applying drugs to the inner or outer surface of the brain.”10
Even as late as 1963 Feldberg seems to have taken pains to avoid using the word “neurotransmitter” to explain any of the effects that followed the administration of drugs into the brain. Writing that same year, John Gaddum commented that although such monoamines as adrenaline, nor-adrenaline, serotonin, and dopamine had attracted much attention because they were found in the brain, “there is no direct evidence that any of these amines is released by nerves in the central nervous system; but there is much indirect evidence from experiments on tissue extracts suggesting that they play an important part in controlling the brain.”11
Feldberg, who had been elected to the Royal Society in 1947, attracted a number of young scientists to work with him, and they helped to stimulate interest in developing techniques for injecting drugs into specific brain regions rather than into the ventricles of the brain. Although these younger associates were sometimes intimidated by the long hours Feldberg spent in the laboratory, they had the highest respect and affection for him.12 One young scientist—the only psychologist working in the laboratory at the time—related how a casual remark by Feldberg forever changed the way he did research. He had proudly shown Feldberg some experimental results that had borderline statistical significance. Feldberg replied with his “distinctive German accent and a kind smile,” “What are all these statistics? A good scientist doesn’t rely on statistics, he gets control of his variable. You should get control over your variables.”13
This is similar to the advice Dale gave Feldberg about fifty years earlier when he told him to keep perfecting his experiment “until your method is working to perfection.” Feldberg continued to be active in the laboratory, publishing his last experimental paper in 1987 at the age of eighty-seven. He retired in 1990 following an accusation by an antivivisection group that he was mistreating animals by not properly anesthetizing them during surgery.14 An investigation found that the accusation was exaggerated, but it may have had some merit. The affair was troublesome and painful for Feldberg, who died, not long after, in 1993.
By 1960, serotonin and GABA (γ-aminobutyric acid) had also been found in the brain, although they too were not accepted as neurotransmitters. Eugene Roberts, who would play a leading role in establishing GABA as a major inhibitory neurotransmitter, was well aware, as the quotation that heads this chapter indicates, that in the 1950s most neurophysiologists remained adamant in rejecting theories of chemical transmission in the brain. They correctly pointed out that finding a chemical substance in the brain, or even proving that it can produce various physiological and behavioral effects when injected into the brain, does not prove that it is secreted by neurons during normal synaptic transmission. Similar observations could be made, they argued, if the chemical substances were important for some metabolic function within neurons.
During the 1950s so little was known about brain chemistry, let alone brain neurotransmitters, that it was difficult even to propose any hypotheses that might explain how chlorpromazine and the other early psychotropic drugs worked. These drugs, which were being marketed successfully in Europe and North America during the second half of the 1950s, had all been discovered accidentally.15 It could not have happened any other way, as there was no basis for predicting the effects of these drugs. Brain neurotransmitters were not implicated in any speculation about how the drugs worked. Nor were any abnormalities in neurotransmitter activity hypothesized to be the cause of mental illness as is commonly done today. The successful marketing of the initial psychotropic drugs did, however, provide a strong incentive and a source of support for investigating how these drugs worked, and this eventually led to speculation about their action on brain neurotransmitters.
One hypothesis proposed to explain psychosis, while in the end not correct, did serve to stimulate ideas about possible biochemical causes of mental illness. This hypothesis involved both serotonin and the hallucinogen lysergic acid diethylamide (LSD). In 1953 John Gaddum (soon to be Sir John), who had been a major contributor to Henry Dale’s research program from 1927 to 1933, was professor of pharmacology at the University of Edinburgh. Gaddum had been studying the effects of LSD. He knew, of course, that this drug was extremely potent in producing hallucinations and delusions, major symptoms of schizophrenia. He had actually tried LSD on himself—once in the laboratory with his colleague Marthe Vogt monitoring him and another time at home while being observed by his wife and daughter. In both cases, the reaction was so severe that it was feared he was going mad.16
When it was learned that the chemical structure of LSD and serotonin were quite similar, Gaddum began investigating how the two interacted. At the time, serotonin was known only as a substance extracted from the intestinal tract that made blood vessels and other smooth muscles contract. Because of this action on blood vessels, it was first called vasotonin, but later the name serotonin was adopted. Gaddum found that LSD blocked serotonin’s effect on smooth muscles. Shortly afterward, serotonin was found in the brain and Gaddum hypothesized that LSD might produce insanity because it blocks the action of serotonin. He proposed that serotonin might “play an essential part in keeping us sane.”17 This was in 1954, and that same year others also speculated that a serotonin deficiency might cause schizophrenia.18 However, serotonin was not considered a neurotransmitter; rather, it was just another interesting chemical substance found in the brain. Nevertheless, the hypothesis had raised the possibility that mental illness might be caused by an abnormality in brain chemistry. This idea was the origin of the expression attributed to the neurophysiologist Ralph Gerard that: “There is no twisted thought without a twisted molecule.”
In 1955 Bernard “Steve” Brodie and his collaborators at the National Institute of Health in Bethesda used a specific fluorometric assay to study serotonin in the brain. He found that LSD suppresses the action of brain serotonin while reserpine, a drug that was being used to treat schizophrenia, released serotonin from a bound state. These results led Brodie to consider that at least some kinds of mental illness might be caused by a serotonin deficiency in the brain, but he hesitated to call serotonin a neurotransmitter. Brodie concluded only that he had provided evidence that “serotonin has a role in brain function.”19 Although today serotonin is considered by many to be implicated in various mental illnesses, at the time the hypothesis of serotonin deficiency as the cause of schizophrenia was weakened when it was found that other drugs that produced hallucinations and delusions do not block serotonin.
Although a spate of possible biochemical explanations of mental illness began to be proposed, none of these early proposals mentioned brain neurotransmitters. In 1959, for example, the psychiatrist Seymour Kety, director of the Laboratory of Clinical Science at the National Institute of Mental Health, wrote a major review of the extant biochemical hypotheses of schizophrenia. The review was published in two parts as lead articles in successive issues of the influential journal Science.20 All the theories Kety reviewed were based on the idea that some abnormal metabolic processes produced a toxic substance that acted on the brain. Kety concluded that all these theories of schizophrenia had serious shortcomings, but what is particularly revealing is that neurotransmitters were not referred to in any way.21
In a second article published in 1967, Kety and Joseph Schildkraut proposed that depression is caused by a norepinephrine deficiency. Even at this late date, however, Kety and Schildkraut hesitated to call norepineprine a neurotransmitter in the brain, although they implied that acetylcholine might be:
While norepinephrine functions as a chemical transmitter substance at the terminals of the peripheral sympathetic nervous system, the role of the other amines [epinephrine, serotonin, and dopamine] in the central nervous system is far from clear…. None of the biogenic amines has yet been definitely established as a chemical neuro-transmitter in the brain, however, and some investigators have suggested that one or more of these amines may act instead as modulators or regulators of synaptic transmission mediated by some other chemical transmitter—for example, acetylcholine.22
Arvid Carlsson, who later shared the 2001 Nobel Prize in Physiology or Medicine for his work on the neurotransmitter dopamine, has described the resistance in the 1950s to the idea of brain neurotransmitters:
Our experiments in the late 1950’s provided the first direct evidence for a role of an endogenous agonist [here, meaning neurotransmitter], present in brain tissue, in animal behavior, thus foreshadowing the paradigm shift from electrical to chemical signaling between nerve cells in the brain. As might be expected, these observations and interpretations at first met with considerable skepticism by some of the most prominent representatives of this field.23
Carlsson noted that this skepticism existed even among some of the very investigators, such as Henry Dale and his collaborators, who had done so much to prove the existence of neurotransmitters in the peripheral nervous system. This skepticism was apparent at the Ciba Foundation Symposium on Adrenergic Mechanisms held in London in March 1960.24 Carlsson noted:
At this meeting practically all prominent workers and pioneers in the catecholamine field were present. It was dominated by the strong group of British pharmacologists, headed by Sir Henry Dale [who was eighty-five at the time]. I was impressed to see how the British pharmacologists, as well as many other former Dale associates, behaved toward Sir Henry like schoolchildren to their teacher, although some of them had indeed reached a mature age.25
Carlsson presented evidence that dopamine and other catecholamines acted like neurotransmitters in the brain. In his paper, which was entitled “On the Biochemistry and Possible Functions of Dopamine and Nor-adrenaline in the Brain,” and in his Nobel Prize address in 2000, Carlsson noted that the evidence he presented in 1960 was met with “a profound and nearly unanimous skepticism.” This was partly because, Carlsson remarked, dopamine had not yet been accepted as anything more than a precursor of norepinephrine (noradrenaline) and also because brain dopamine seemed to have little effect on muscles. In his Nobel Lecture, Carlsson recalled that:
Dale expressed the view that L-DOPA is a poison…. Martha Vogt concluded that the views expressed … regarding a function of serotonin and catecholamines, respectively, in the brain would not have a long life. W. D. M. Paton referred to some unpublished experiments indicating that the catecholamines are located in glia. In his concluding remarks John Gaddum stated that at this meeting nobody had ventured to speculate on the relation between catecholamines and the function of the brain. But this was what I insisted upon throughout the meeting, so the clear message to me was that I was nobody.26
If even pharmacologists were still skeptical about the existence of brain neurotransmitters in 1960, it is not surprising that there was strenuous resistance to the idea among neurophysiologists. Hugh McLennan, for example, wrote in his authoritative 1963 monograph Synaptic Transmission that: “In the vertebrate central nervous system there is only one synapse identified whose operation can, with assurance, be ascribed to acetylcholine.” He was referring to Eccles’ 1952 experiment demonstrating that acetylcholine is used as a neurotransmitter by spinal motor neurons. McLennan concluded that he had not found a single instance that he was willing to accept as evidence of a chemical neurotransmitter at a brain synapse.27
Throughout much of the 1960s it was still not clear that the various chemical substances found in the brain were neurotransmitters. This uncertainty was reflected in the common usage of the expression “putative neurotransmitter.” In 1965 the psychologist Neal Miller reviewed research demonstrating that satiated animals could be made to eat or drink by inserting adrenergic or cholinergic drugs, respectively, into specific brain regions. Miller noted that because acetylcholine and norepineprine are considered neurotransmitters in the peripheral nervous system, we had a right to ask, “Does similar coding of transmission occur in the brain, and if it does, is it related to specific forms of behavior? Biochemists have shown that the hypothalamus is especially rich in both acetylcholine and norepinephrine. What are they doing there?”28
In retrospect, this reluctance to speak of neurotransmitters in the brain is even more surprising considering that during the 1950s and 1960s a number of new techniques were actually providing anatomical, chemical, and physiological evidence that brain neurons secrete neurotransmitters. Gas chromatography and mass spectrometry had made it possible to detect minute quantities of chemical substances released when nerves were stimulated electrically and even when they were normally active. Thus, for example, it was found that the acetylcholine concentration in the visual cortex was seven times greater when the eyes were open than when closed.29
Another important line of evidence for brain neurotransmitters came from the development of a new staining technique that made it possible to visualize biogenic amines within neurons. The technique was developed in the early 1960s by Swedish histochemists headed by Nils-Åke Hillarp in Götteborg and Bengt Falck in Lund. With appropriate pretreatment, norepinephrine, epinephrine, serotonin, and later dopamine could be made to fluoresce different colors. When appropriately treated, norepinephrine, for example, became a bright green, while serotonin took on a yellow fluorescence. It became possible to see norepinephrine, epinephrine, serotonin, and dopamine in the cell body (soma) of neurons and throughout their axons and terminals. By 1965 the neuronal circuits in the brain that carried epinephrine, norepinephrine, and serotonin had been mapped, and by 1971 dopamine pathways were also mapped. Because the dopamine pathway differed from the norepinephrine pathway, this supported the suspicion that dopamine was a separate neurotransmitter, not simply, as previously thought, a precursor of norepinephrine.30 Although it was not sufficient evidence, the fact that these substances were located in specific neural pathways and not found in all neurons added to the argument that they might be acting as neurotransmitters in these brain regions.
Other techniques provided additional evidence for the existence of brain neurotransmitters. One involved the use of a device consisting of several fine glass tubes (cannulae) glued to a recording electrode. After this device was inserted into the brain, it was possible to use electrophoresis to move minute quantities of the ions of a “putative” neurotransmitter across a small group of neurons.31 The electrode recorded changes in the firing rate of the neurons exposed to the neurotransmitter. It thus became possible to demonstrate that neurons exposed to the “putative” neurotransmitters responded by increasing or decreasing their spontaneous firing rates. One of the first results with this technique was reported by Rosamond Eccles (John Eccles’ daughter) and D. R. Curtis, who demonstrated in 1958 that the Renshaw cells in the spinal cord were excited by acetylcholine.32
In 1961 John Gaddum described a “push-pull cannula” that could be inserted into different brain regions.33 This device consisted of cannulae bound together and connected to precision pumps. The pumps made it possible to “push” a saline solution across a group of neurons and then withdraw (“pull”) the fluid out. The withdrawn fluid could then be analyzed to determine if any “putative” neurotransmitter had been secreted by the neurons.34 In describing the advantages of this apparatus, Gaddum wrote that,
(1) It detects the substances liberated under fairly normal conditions.
(2) It gives an indication of the turnover rather than the store, and can be used to study the effect of factors such as nervous stimulation.
(3) It should be possible to localize the site of liberation of substances in such tissues as the central nervous system more precisely than by other means.35
Thus, evidence began to accumulate that at least some of the chemical substances found in the brain are located in specific neuronal pathways, that they are released when the neurons are active, and that the release of these chemicals increased or decreased the firing rate of neurons exposed to them in ways that duplicated the effects of electrical stimulation.
Probably the most convincing evidence for the existence of brain neurotransmitters came from the electron microscope. This instrument was first developed by Siemens in Germany and RCA in the United States in the late 1930s, but because of World War II its potential was not realized until after the war. By the 1950s the optics of the electron microscope had been improved and techniques for fixing, staining, embedding, and sectioning neuronal tissue for examination in this microscope had been developed. Its great power of magnification made it possible to see the “ultrastructure” of neurons. In 1954 George Palade and Sanford Palay, working at the Rockefeller Institute for Medical Research, reported seeing small vesicles in neurons that they estimated to be between 300–500 angstroms in size. It was also possible to finally see the gap at the synapse, which was estimated to be approximately 200 angstroms.36
Two years later Sanford Palay, then in the department of anatomy of Yale University, used the electron microscope to describe more fully the anatomy of brain synapses in the rat. He noted that the clear gap, which could be seen with the electron microscope, “is impressive confirmation of the neuron doctrine enunciated and defended by Ramón y Cajal during the early part of this [the nineteenth] century.” Palay commented that the dense patches seen on postsynaptic membranes where the two neurons were opposed may be “considered as the ultimate points for transmission of the nervous impulse.” He also described the synaptic vesicles as being present only on the presynaptic side of the synapse, making the synapse, as Cajal had concluded “clearly polarized.” Palay speculated that the vesicles may have a “direct role in the transmission of nervous impulses across the synapse.”
If an analogy may be drawn between these synaptic terminals [those in the brain] and the myoneural junction which has essentially the same structure, the small vesicles may be considered as containing small units of a chemical transmitter, like acetylcholine, or precursor of this transmitter, which are discharged into the intrasynaptic space.37
In 1959 V. P. Whittaker in Cambridge, England, commented that “chemical transmission at nerve endings is now being increasingly discussed in terms of the ‘synaptic vesicle’ theory.” After using a technique that made it possible to separate different components of the presynaptic terminals based on their relative density, Whittaker reported that the component containing the synaptic vesicles had the greatest concentration of acetylcholine or serotonin.38 The conclusion that the synaptic vesicles held the chemical neurotransmitter until it was released was confirmed in 1968 when Tomas Hökfelt in Sweden combined the electron microscope and the fluorescent staining technique to demonstrate that the hypothesized neurotransmitters were indeed located in the synaptic vesicles of presynaptic terminals.39
John Eccles, who had led the opposition to chemical transmission, had converted completely by 1965 to the view that virtually all synapses, in the brain as well as elsewhere in the nervous system, are chemical. In his Charles Sherrington Lecture Eccles stated that both excitation and inhibition were transmitted chemically.40 By this time several chemical substances had gained acceptance as neurotransmitters, and Eccles stated that glycine appeared to be the major inhibiting neurotransmitter in the spinal cord, while GABA probably played that role in the brain. Eccles’ statement was based in part on the work of Aprison and Werman in Indianapolis, who had found that glycine is concentrated in the region of the spinal cord where inhibitory interneurons were known to be located.41 GABA had been located in the brain, and its role as a major neurotransmitter inhibitor was supported by the finding that drugs such as bicuculine and picrotoxin, which block GABA, produce a generalized excitement of neural activity and, depending on dose, a convulsion. Strychnine, which blocks glycine inhibition at spinal motor neurons, does not block GABA-induced inhibition.
GABA and glycine are amino acids and represent a different chemical class of neurotransmitters than either acetylcholine or the biogenic amines norepinephrine, dopamine, and serotonin. In time it was learned that the amino acids, which are small molecules, are the most common neurotransmitters. We now know that in addition to the inhibitory neurotransmitters, there are many excitatory amino acids, such as glutamate and aspartate, that act as neurotransmitters in the brain. Glutamate is now considered to be the most common mammalian excitatory neurotransmitter.42
Several novel peptides were also shown to act as neurotransmitters. The discovery of one of these peptides, Substance P, has an interesting history. It was originally given its name by John Gaddum, who did much of the early work on this substance with Ulf von Euler when they were both working in Dale’s laboratory. Gaddum suggested the name Substance P because it was a fine powder when extracted from the intestines. It was initially thought to act like acetylcholine because it dilated blood vessels and produced a contraction of the smooth muscles of the intestines. However, unlike acetylcholine, its action is not blocked by atropine. It was later found that Substance P is a peptide and that it is a major neurotransmitter for conveying sensory information about pain.43 So by happenstance, the name Substance P became especially apt in that it could stand equally for a powder, peptide, and pain.
Still another substance found to be a neurotransmitter in the brain is histamine. It will be recalled that Henry Dale had studied histamine even before it was identified as such, but this research was done in the context of its action as a substance found in certain tissue that can produce a profound vasodilation and lowering of blood pressure. Histamine is a biogenic amine synthesized from amino acid precursors. In the 1980s histamine was found in neurons, and the distribution of this substance and its receptors in the brain have subsequently been described. The histamine neurotransmitter has been shown to have a wide range of behavioral and physiological effects involving sleep and arousal, eating and drinking, motor activity, learning, sexual behavior, aggression, and pain perception.44
During the 1950s and 1960s a great amount was learned about the way neurotransmitters are synthesized, stored, and released, and also how they induce changes in postsynaptic neurons. In 1970 Ulf von Euler, Julius Axelrod, and Bernard Katz shared the Nobel Prize in Physiology or Medicine. In the presentation of the award it was noted that Henry Dale and Otto Loewi and their collaborators had “showed that impulse transmission takes place by chemical means”:
As with all fundamental discoveries, the discovery of a chemical mediator in nervous transmission led to revolutionary new thinking. Neurochemistry and neuropharmacology developed into rapidly expanding branches of science. A host of new questions arose. How were such highly active transmitter substances synthesized, stored, and released? How could they appear, produce their effects, and disappear within a fraction of a second, which must happen if chemical mediation was to explain the very fast chain of events taking place in nervous processes? What kind of substances were involved? Each of today’s prize-winners had made his own special contribution toward solving problems in this field.
Von Euler’s work on the synthesis of norepinephrine was described above. Julius Axelrod also worked primarily on norepinephrine, and he demonstrated that the action of this neurotransmitter is terminated primarily by being taken back up into the presynaptic neuron from which it had originally been released. Axelrod was a member of the Laboratory of Clinical Sciences at NIMH, which was headed by Seymour Kety. It will be recalled that Kety was interested in the possibility that schizophrenia was caused by an error in the metabolism of norepinephrine. There had been reports that adrenochrome, an abnormal oxidation product of epinephrine, was present in schizophrenic patients. It had been shown that adrenochrome could be formed from norepinephrine, but this work was not done on animals or patients. Kety was able to obtain some radioactive norepinephrine containing tritiated hydrogen (3H), which had just then become available, and he used it to trace the fate of norepinephrine in animals. This work was discontinued when it did not confirm the “adrenochrome hypothesis.” As there was some radioactive norepinephrine left over, Axelrod was able to use it for his more basic research. He administered it to rats and was able to show that it was accumulated in sympathetic nerves in the periphery.
It was then found that drugs like cocaine, which potentiate sympathetic effects, inhibit the uptake of norepinephrine by sympathetic nerves. This work was pursued by Axelrod and others, who eventually proved that the reuptake mechanism is the main way that norepinephrine effects are terminated in the brain as well as the periphery.45 The reuptake mechanism was also shown to be the major way of terminating the effects of dopamine, serotonin, and GABA as well as norepinephrine. Prior to Axelrod’s work, termination of all neurotransmitter action was thought to follow the model of acetylcholine, whose action is terminated by the action of an enzyme.
Bernard Katz, who shared the Nobel Prize with Axelrod and von Euler, studied acetylcholine. He was able to show that this neurotransmitter is released in small packets, or quanta, each of which produces a small electrical response at the junction between nerves and skeletal muscles. When a nerve fires more rapidly, more of these quanta are released. Katz also showed that even in a resting state neurons continually release a small number of quanta of acetylcholine; this, while insufficient to provoke a response from a muscle, is necessary to maintain the synaptic connection. Bernard Katz was still another of the many established or promising German scientists who were able to get out of Nazi Germany.46
By the end of the 1960s, more and more chemical substances were being considered possible neurotransmitters in the brain. Scientists recognized that it was necessary to establish some criteria that had to be met before a chemical substance could be accepted as a neurotransmitter, for its mere presence in the brain was not sufficient qualification, even if it could be shown to have some physiological or behavioral effects. At the very least, the candidate neurotransmitter had to be shown to be secreted by neurons at the synapse and to be capable of producing the same effect on postsynaptic neurons as stimulation of the presynaptic neuron. The candidate neurotransmitter had to be demonstrably present in the synaptic vesicles. There had to be evidence that the enzymes necessary for the synthesis of the candidate neurotransmitter were present in neurons and that there existed some mechanism to terminate its action, such as enzyme degradation or reuptake.
When it became possible to detect receptors for different neurotransmitters, the demonstration of their presence became an additional criterion. The concept of receptors has had a long history. As described in chapter 2, before 1900 John Langley hypothesized the existence of “receptor substances” in order to explain how a drug could produce opposite effects, either excitation or inhibition, at different sites. Over time, Langley’s “receptor substance” was shortened to “receptor.”47 Although the concept of receptors was criticized as vague, the need for such a concept persisted. In the 1920s, several investigators, but particularly the pharmacologist Alfred Clark, described the properties of receptors more fully. Clark determined that drugs are generally effective only when applied to the cell surface (membrane) and not when placed inside the cell itself. He estimated the size of different drug molecules and the length of the cell membrane, and he also determined the minimum amount of different drugs that could produce a maximum response. From these data, Clark was able to argue that drugs must act on discrete receptor units on the membrane that do not occupy its entire surface.
Pharmacologists needed the concept of receptors not only to explain why a drug has different effects at different sites, but also why normally only one of the mirror images of the same molecule—the so-called left and right isomers—is effective. This was explained by analogy of the “key and lock,” where the key is the neurotransmitter and the lock is the receptor.48 Even though the arguments for the existence of receptors may now seem persuasive, pharmacologists of the time were reluctant to accept the concept. Dale, for example, remained critical of the concept for quite a while, mainly because there was no way of detecting its physical presence.49 Before 1970 the concept of receptors was rarely discussed even at pharmacology meetings, while most neurophysiologists regarded the concept as overly imaginative theorizing and the multiplication of hypothetical receptors as an unnecessary complication.
The first receptor to be identified was the nicotinic–acetylcholine receptor. This work was initiated by Carlos Chagos in Rio de Janeiro, who studied the role of acetylcholine in activating the electric organ of the electric eel (Torpedo marmorata). Later studies used alpha-bungarotoxin, an active component of cobra toxin, which was found to combine with the nicotinic receptor with a high degree of specificity.50 This receptor had been shown to control the electric organ of the Torpedo, but what made it possible to use bungarotoxin to identify the nicotinic receptors was the exceptionally high concentration of these receptors in the electric organ. It is estimated that the nicotinic cholinergic receptors may constitute as much as 20 percent of the weight of the electric organ in the Torpedo fish.
Identifying receptors for neurotransmitters in the brains of mammals, where the concentration of receptors is much lower than in the electric organ, proved to be considerably more difficult. Solomon Snyder, who was a major contributor to this work, has described how this was accomplished.51 The identification and location of receptors for the enkephalins in the brain, which is described below, served as the model for identifying the receptors for most neurotransmitters. Basically, the technique involved “labeling” receptors by attaching radioactive substances to chemicals that bind to them. The radioactive substance might be the neurotransmitter itself or some ligand—an agonist or antagonist—that binds to the same receptor with a high degree of specificity.52 Techniques for recovering and measuring the bound radioactive ligand were developed, and from this it was possible to determine the relative concentration of receptors in different brain areas and the relative strength of the binding affinity of different drugs competing for the same receptor.53 During the 1970s, in addition to the enkephalin receptor, acetylcholine, norepinephrine, glycine, GABA, and histamine receptors were found in the brain.
Although most neurotransmitters were first discovered in the brain and accepted after they were shown to act on the nervous system and to have met the other established criteria, this was not always the way it happened. In the case of the peptides now known as enkephalins, for example, their existence was first inferred from the action of opium and related opioid drugs. Opium is obtained from the poppy plant, and morphine and codeine are extracts of opium. By the mid-twentieth century several synthetic opioids were also available. It was inferred that all these different opioids must be acting on a specialized brain receptor because analgesia and euphoria were obtained only from drugs with very similar chemical structures.
The search for the opiate receptor in the brain began around 1970. Although there were some suggestive earlier results, it is commonly acknowledged that Solomon Snyder and Candace Pert, then a Ph.D. student working in Snyder’s laboratory at Johns Hopkins University, were the first to obtain convincing evidence of opiate receptors in the brain. This was in 1973.54 The race was close—researchers at New York University and in Uppsala, Sweden, identified opiate receptors in brain tissue later that same year.55
The discovery of an opiate receptor in the brain triggered a feverish search for an endogenous opioid in the brain on the rationale that, as one investigator put it, why else “would God have made opiate receptors unless he had also made an endogenous morphine-like substance.”56 It did not require much imagination to appreciate the potential importance of finding a natural substance in the brain that could induce pleasure and alleviate pain. Several teams of researchers entered the race, but the first to succeed were Hans Kosterlitz and his young associate John Hughes, who reported in 1975 that they had extracted an endogenous opiate-like substance from brain tissue.
Hans Kosterlitz left Nazi Germany in 1934 while still relatively young. He was able to continue his education at the University of Aberdeen, where he earned a Ph.D. and a medical degree. Kosterlitz eventually became professor and head of the department of pharmacology in Aberdeen. In 1973, when Kosterlitz had to give up his academic appointment at the compulsory retirement age of seventy, he established the Unit for Research on Addictive Drugs and continued his research on opiates. When one of his graduate students found that the mouse vas deferens is extremely sensitive to morphine and related opioids, Kosterlitz and Hughes used this discovery to develop a bioassay useful in the search for an opioid substance that might be in the brain.57 They knew they would need a lot of fresh brains for this search, and decided that the pig brain might be most readily obtained from the local slaughterhouse.
John Hughes, because he was much younger than Hans Kosterlitz, had the task of going to the slaughterhouse early in the morning to collect pig brains. He later reported that a bottle of Scotch whiskey was much more helpful in getting the abattoir workers to cooperate than any story he could tell them about the importance of the science.58 Hughes brought back carts filled with pig brains, from which were extracted various substances to test on the vas deferens. In 1975 Hughes and Kosterlitz reported that they had successfully isolated an opioid substance in the pig brain.59 The substance produced the appropriate response from the mouse vas deferens, and the response was blocked by the specific opioid antagonist naloxone.60 Initially, this brain extract was called an “endorphin,” a contraction of endogenous morphine-like substance, but Kosterlitz and Hughes proposed the name enkephalin (from the Greek meaning “in the brain”). The more neutral term “enkephalin” was preferred because it did not restrict its effects to the known euphoric and analgesic effects of opiates.
In a second paper published the same year, Kosterlitz and Hughes reported that they had identified and synthesized two enkephalins, met- and leu-enkephalin, in the brain. The two were chemically identical, except that the former contained methionine and the latter, leucine. They wrote, “The discovery in the brain of two endogenous pentapeptides with potent opiate activity raises a number of questions that cannot be adequately dealt with in this paper. It will now be possible, however, to test the hypothesis that these peptides act as neurotransmitters or neuromodulators at synaptic junctions.”61 Among the questions Kosterlitz and Hughes mentioned as needing to be investigated was whether tolerance to and dependence on endogenous enkephalins could develop, as they did with synthetic opioids and natural opium derivatives. Many wondered whether a natural “high” could be obtained from these brain enkephalins.
Only a few weeks after Hughes and Kosterlitz published their report, Solomon Snyder, Gavril Pasternak, and Rabi Simantov reported finding the same two enkephalins in rat brains. The Snyder group also mapped the distribution of the met-enkephalin receptor in different brain areas. The high concentrations of enkephalins in the locus coeruleus and the amygdala seemed to explain the euphoric and analgesic effects of opioid drugs, as these structures are thought to play a significant role in mediating emotional responses. Similarly, the high concentration of enkephalins in the medial hypothalamus and certain regions of the spinal cord was consistent with what was known about the regulation of the affective component of pain and the transmission of pain signals.62
The new radiolabeling techniques eventually made it possible to identify receptors for virtually all neurotransmitters. In the 1970s both Solomon Snyder at Johns Hopkins and Philip Seeman in Toronto reported labeling two different dopamine-receptor families, now called D1 and D2. It was subsequently found that most neurotransmitters have more than one receptor type. In some instances numerous receptors respond to the same neurotransmitter.
In 1957, even before dopamine receptors had been identified, Arvid Carlsson and Margit Lindquist in Göteborg hypothesized about the consequences of blocking what were at the time hypothetical dopamine receptors. They were trying to explain the antipsychotic action of reserpine and drugs like chlorpromazine by determining what they had in common. Starting with evidence suggesting that both blocked dopamine transmission, Carlsson and Lindquist found that while reserpine reduced dopamine in the brain, other drugs with antipsychotic properties did not. They were especially surprised to find that drugs like chlorpromazine actually increased the amount of dopamine metabolites in the brain. Carlsson and Lindquist then theorized that the chlorpromazine-like drugs might block dopamine receptors and that this might activate a feedback loop calling for more dopamine to be released. Such a mechanism could explain the increase in dopamine metabolites despite there being no decrease in brain dopamine levels. This brilliant insight led to the discovery of the first of many mechanisms that compensate for changes induced in neurotransmitter systems. The feedback loop that Carlsson and Lindquist hypothesized was later identified, and it was shown to act on dopamine autoreceptors located on dopamine neurons. Activation of these dopamine autoreceptors regulates the amount of dopamine released. Later, the ability to tag receptors with radioactive substances made it possible to show that the number and sensitivity of receptors increase or decrease (up and down regulation) in response to blockade or excessive stimulation, respectively. Many other mechanisms that compensate for changes induced in one part of a neurotransmitter system have since been found.
It will be recalled that many neurophysiologists had rejected chemical transmission because it was assumed to be too slow to mediate the fast responses observed at many synapses. This position could no longer be defended after Paul Greengard, who shared the year 2000 Nobel Prize in Physiology or Medicine, began to describe the molecular mechanisms involved in chemical transmission of the nerve impulse. Greengard identified two different mechanisms, which he categorized as either fast or slow chemical transmission.63 The fast chemical transmission (both excitatory and inhibitory) takes place in less than one millisecond, while even the slow transmission takes only several milliseconds—fast enough to do everything the brain is known to do.64 Walter Cannon was certainly correct when he wrote in 1939 that not enough was known about chemical transmission to allow many of its critics to assume that it was too slow a process to do the job required.
Just when it began to appear that all synapses were mediated by chemical transmission, a few electrical synapses—estimated to be about one percent of the total—were found. Use of the electron microscope showed that the synaptic gap in what proved to be purely electrical synapses is so narrow that it is virtually obliterated. For that reason, they are called “tight junction,” or “gap junction,” synapses. There are no presynaptic vesicles or any other evidence of chemical transmission at electrical synapses. Transmission across this extremely narrow gap is faster than at chemical synapses, and researchers believe it can be bidirectional as well as unidirectional.
Perhaps because transmission across electrical synapses is so fast, they are found in neural circuits controlling escape behavior in invertebrates such as the squid and the crayfish. However, electrical synapses cannot be considered esoteric and unimportant, because they have been found in retinal cells of the eye, the locus coeruleus, the corpus striatum, and other areas of the mammalian brain. In the hippocampus, a region important for the formation of memories, GABA-secreting cells are linked by gap junctions in a way that facilitates a synchronous firing that maximizes their influence.65 It is also interesting that in some animals with relatively simple nervous systems, such as the round worm (C. elegans), which has only 302 neurons and an estimated 10,000 synapses, half the synapses are chemical and the other half are electrical gap junction synapses.
It is ironic that although chemical transmission was once considered the exception, now it is the electrical synapse that is the exception. Moreover, our ideas about chemical transmission have broadened considerably over the years. Initially, chemical transmission was viewed as essentially doing what electrical transmission was thought to do, but in another way. That is, the action of neurotransmitters was thought to be restricted to exciting or inhibiting a postsynaptic cell. However, we now realize that neurotransmitters produce different types of changes. Rapid changes in the excitability of a postsynaptic cell occur when a neurotransmitter induces changes in the concentration of sodium, calcium, or potassium ions within that cell. Neurotransmitters acting on membrane receptors produce these changes by causing specific ion channels (“transmitter-gated ion channels”) in the membrane to open or close. These changes can be completed in milliseconds.
Neurotransmitters (and hormones), called “first messengers,” can also induce more enduring changes by activating so-called second messengers located in postsynaptic neurons. Many of these changes in neuronal function take place over hours, days, and weeks, not milliseconds. Through a cascading series of molecular steps, the “second messengers” can induce changes in the expression of specific genes. The changes in gene expression can alter virtually everything about a postsynaptic cell, including, among many others changes, the number of receptors, the production of neurotransmitters, and the amount of protein carrier used in the reuptake of neurotransmitters or in the production of the enzymes used to terminate the action of some neurotransmitters. Moreover, anatomical changes can be induced in, for example, the branching of dendrites and the number of spines they contain. (Synapses on dendritic spines are thought to be primarily excitatory.) These more enduring changes are thought to underlie the neural plasticity that supports learning and memory, while the faster changes are more likely to support the transfer of information through neural pathways. A single neurotransmitter may induce both the fast and the slower, more enduring changes depending on the receptor activated.
It is also known now that in addition to producing more enduring changes, chemical transmission can exert its influence at greater distances than electrical stimulation could. Whereas electrical stimulation could influence neurons located directly across the synaptic gap, some neurotransmitters travel in the blood or cerebral spinal fluid to reach receptors located a considerable distance from where they are released. In this respect, neurotransmitters are similar to hormones and are reminiscent of the common origin of sympathetic neurons and the cells of the adrenal medulla that release the hormone adrenaline into the bloodstream.
Although the distinction can be blurred, sometimes neuromodulators are distinguished from neurotransmitters. The primary function of neuromodulators is not thought to be the transmission of nerve impulses across the synapse, but rather the modification of the effectiveness of postsynaptic neurons. This may be accomplished, for example, by altering the synthesis, release, reuptake, or metabolism of neurotransmitters. Neuromodulators generally have a slower onset and longer duration of action than neurotransmitters.
More recently, the definition of neurotransmitters has had to be broadened to include some gasses, such as nitric oxide and carbon monoxide. Study of these gasses, which are called “novel neurotransmitters,” has necessitated a rethinking of what constitutes a neurotransmitter. The physiological role of nitric oxide (NO) was discovered in the 1980s when Robert Furchgott found that acetylcholine does not directly cause blood vessels to relax, but does so by causing nitric oxide to be released from the endothelium cells that line blood vessels. Nitric oxide also causes the blood vessel relaxation that supports penile erection. It has been found that nitric oxide can be released in the body by many different agents. Nitroglycerin, for example, which relieves ischemic cardiac (angina) pain, causes nitric oxide to be released. In 1998 Robert Furthgott shared the Nobel Prize with Louis Ignarro and Ferid Murad for their work demonstrating that a gas can act as a “messenger molecule.”
It has subsequently been shown that nitric oxide is present in central nervous system neurons and is released following stimulation of brain cells.66 Among the brain areas most extensively studied in this regard are the cerebellum, striatum, cerebral cortex, hippocampus, hypothalamus, and the ascending reticular formation. Nitric oxide is not stored. Its synthesis is triggered on demand by an influx of calcium into the neuron. Because it is not stored, the location of nitric oxide is facilitated by tracing the enzyme nitric oxide synthase (NOS).
Studies of the possible function of brain nitric oxide are underway. There is evidence that it may play a role in the long-term facilitation of the synaptic transmission that underlies learning and memory. Eric Kandel of Columbia University and his colleagues have demonstrated that the long-lasting facilitation of synaptic transmission (long-term potentiation) may be dependent on the enzyme NOS.67 It has been suggested that nitric oxide acts as a “retrograde messenger,” potentiating synaptic transmission by either increasing the postsynaptic response or increasing the amount of presynaptic neurotransmitter that is released. There is also some evidence, obtained from “knockout mice,” that nitric oxide is involved in regulating sexual and aggressive behavior.68
Nitric oxide seems to break every rule of nerve transmission. It is not synthesized in advance and stored in protective synaptic vesicles. Instead, it is synthesized on demand and then immediately released, normally together with other neurotransmitters. Nitric oxide does not bind to receptors on the membrane of postsynaptic cells; rather, it diffuses right through the postsynaptic membrane into the cell, where it binds to certain enzymes. It has a short half-life, measured in seconds, so there is no need for its action to be terminated by enzymatic degradation or by reuptake.
In consideration of all these differences, Solomon Snyder, who has studied the central nervous system properties of nitric oxide and other gasses, has suggested that we need a more “liberal conceptualization” of neurotransmitters. Snyder has proposed defining neurotransmitters as simply any “molecule, released by neurons or glia, that physiologically influences the electrochemical state of adjacent cells.”69 From the outset of his research he has been more interested in molecules that convey instructional messages by any means and less concerned with the criteria for calling a substance a neurotransmitter.70 In any case, it may not be surprising that when there was a great amount of skepticism about brain neurotransmitters, a long list of criteria had to be established before a candidate neurotransmitter would be accepted. Now that chemical transmission is considered the rule, a more liberal definition of neurotransmitters seems reasonable.
We have come a long way in the past one hundred years. Around 1900 the neuron doctrine had just been accepted, although there still remained a few who continued to argue that nerve cells fibers were connected, with no gap between them. It was, however, generally accepted that there was at least a functional, if not a physical, gap that separated neurons and neurons from the muscles and glands they innervate. The acceptance of the concept of chemical transmission across that gap proceeded in stages with transitional periods between them marked by elevated levels of controversy. The first hint that a chemical process mediated neural transmission came from the observation that acetylcholine and adrenaline appeared to mimic the effects of stimulating the parasympathetic and sympathetic nervous system, respectively. The idea that these substances might be secreted by nerves was not seriously entertained, however, until Loewi’s seminal experiment in 1921, which triggered both interest and controversy.
Despite the awarding of the Nobel Prize to Otto Loewi and Henry Dale in 1936, most considered chemical transmission to be a special case, applicable only to the relatively slow–responding smooth muscles of visceral organs. Any suggestion of chemical transmission at central nervous synapses was vigorously disputed. For a number of years, the dispute centered on the question of whether the innervation of skeletal muscles by spinal nerves is chemical or electrical. A breakthrough occurred when John Eccles, a major opponent of chemical transmission, became convinced by his own experimental evidence that transmission between spinal motor nerves and skeletal muscles is chemical. However, the possibility that transmission at brain synapses is chemical was still not so much disputed as ignored.
The last stage, that of proving chemical transmission at brain synapses, can be characterized as a period of persistent skepticism leading up to the time when the techniques became available that provided evidence that could no longer be ignored. Today it is accepted that transmission at over 99 percent of all synapses is chemical, mediated by one or another of an estimated 50–100 different chemical neurotransmitters that act on an even greater number of receptors.71
The universal acceptance of the existence of neurotransmitters stimulated some to think about their origin. Did neurotransmitters originate with the nervous system, or are their roots to be found in chemical communication that existed before the nervous system evolved? How are responses integrated in organisms that have no nervous system? Herbert Spencer Jennings, a close friend of Walter Cannon from their student days at Harvard, became a leading authority on the behavior of the one-celled animals (protozoa) that have no nervous systems. Jennings described the large range of adaptive behaviors that protozoa display, and he even argued that their behavior is modified by experience in ways that reflect a capacity for memory and learning.72 All of this is accomplished by chemical communication without a nervous system.
A principal adaptive advantage of the early nervous system was that it sped up chemical communication. As multicellular animals became larger and more differentiated, the nervous system made it possible for distal parts of such animals to respond more rapidly and in a more integrated pattern than was possible by means of chemical diffusion alone.73 Evolution tends to conserve what has been found useful, however, and it is now known that the primitive nervous system secretes some of the same chemicals that serve as “messenger molecules” in animals without a nervous system. These “messenger molecules” have apparently had a long evolutionary history, as receptors for many of the neurotransmitters found in higher animals have been discovered in bacteria and yeast as well as in protozoa.74
If this evolutionary history had been known earlier there probably would have been less resistance to the idea of chemical neurotransmitters, and the “war of the soups and sparks” might have been only a brief skirmish between disciplines defending their own turf. However, none of this was known at the time, for interest in the evolution of neurotransmitters arose only after the concept of chemical transmission became more acceptable.
In the brief epilogue that follows I offer some personal reflections on the value of history in general and this one in particular.