Neural Communication and Control: Satellite Symposium of the 28th International Congress of Physiological Science, Debrecen, Hungary, 1980

Advances in Physiological Sciences, Volume 30: Neural Communication and Control is a collection of papers presented at the 1980 satellite symposium of the 28th International Congress of Physiological Science, held in Visegrá Hungary. This volume is composed of 26 chapters and begins with a description of nervous elements and systems on the phylogenetic scale. The succeeding chapters review studies on the excitable membrane, the properties of a single neuron, of small and large neuronal ensembles and of systems of increasing complexity, considering physiological and anatomical aspects, as well as experimenting and modeling. Other chapters explore the whole-brain function based on a conscious experience. The remaining chapters examine the understanding the neural basis of cognitive experience through experiment on evaluative cognitive agency in "split-brain" patients. This book is of value to physiologists, neurologists, and researchers.

• Language: English
• Publisher: Elsevier Science
• Release date: Oct 22, 2013
• ISBN: 9781483190198

Book preview

Damjanovich

OPENING REMARKS

Donald M. MacKay

In opening our proceedings, let me first express the indebtedness of us all to the University of Debrecen, and to the initiative of Professors Gy. Székely and S. Damjanovich, for making it possible for us to meet in such agreeable surroundings and with such excellent facilities for this somewhat unconventional experiment in cross-disciplinary communication. When the I. U. P. A. B. Commission for the Biophysics of Communication and Control decided to seek co-sponsorship from I. U. P. S. for such a meeting, it was a great pleasure to find that our ideas had already been largely anticipated by these colleagues; and it is to them, together with Professor P. Johannesma of our I. U. P. A. B. Commission, that the bulk of the credit must go to the design and implementation of our programme.

The purpose of our meeting is twofold. First, of course, we want to acquaint one another, as well as our disparities of discipline may permit, with what we see as good examples of the interstimulation of experiment and theory at our various levels of concern with neural communication and control. But secondly, and throughout this process, we want to ask ourselves what we can learn from this experience about the best ways to make experiment and theory interfertile. In this respect, then, our aim might be called metascientific. We must all have been dismayed by the proliferation of theoretical models of neural function over the past 30 years which have seemed to evoke little or no interest among experimental neuroscientists, and by the vast tracts of experimental data that have so far defied or failed to attract insightful theoretical analysis. Why has this been so? Are we perhaps in danger, on both sides, of being tempted by the availability of tools and funds into answering too many inadequately posed questions? Can we help one another to spot some of the more relevant questions we should be asking? Can we recognize any pointers towards a better working partnership between theory and experiment in our needy field?

I do not suggest for a moment that brain research is peculiar among the sciences in this respect; but we are perhaps more sharply aware than in some of the longer-established disciplines of the need for an adequate conceptual framework, within which to design both our experiments and our theories. What does create problems of this kind for brain research, more than for most other sciences, is its multi-level structure. In this respect it is often compared with computer science. A computer chip, for example, can be analysed at the levels of atomic and molecular physics, or crystallography, or transistor circuitry, or information-processing logic. It can also be understood in terms of its function as a component of a central processor, or as part of the embodiment of an artificially intelligent agent. Between some of these levels (the molecular and the transistor levels for example) there are intimate practical relationships. Between others (such as the crystallographic and the programming levels) there are virtually none.

The analogy is valid in so far as it brings out a distinction between two different kinds of traffic between experimentalists and theorists in brain research. First, at each given level (biophysical, neuronal, psychological), there is the usual need for theory to guide experiment and to be guided in return by results, with the same problems of securing interfertility as in any other sciences. Over and above these, however, we have in brain research the problem of deciding with which other levels (if any) a given level of experiment or theory should seek to interact.

The computer analogy is sometimes used to suggest that the higher (more psychological) levels of analysis of brain function need have no more interest in the lower (physiological) than a computer programmer has in the electronics or physics of his machine. This however is an over-simplification that neglects the functional relationships which can be obtained between factors at widely separated levels. Think for example of the biochemistry of mood-control, or the variety of levels at which we can identify functional parameters of interaction in a plexiform neuronal system. If the brain is to be compared at all to a computer, it must be to one whose programming can be affected (in ways not necessarily destructive of function) by a host of variables such as local changes in temperature or conductance, physical proximity of related patterns of activity, overall balance of various supplies and the like. Certainly there can be no excuse for a neuroscientist working at any level to discount a priori the relevance of experiments or theories at another.

Very well; but how does this work out in practice? What established theoretical skills (if any) are worth acquiring by someone interested in neural communication and control, and at what levels? Where (if anywhere) have experimental data accumulated in enough quantity and solidity to expose a theoretical model to a crucial test? Where (if at all) has theoretical or experimental work at one level led to insightful experimental design at another? These are some of the questions that I hope we will address, explicitly or implicitly, in our adventure the next few days.

A COMPARATIVE NEUROLOGIST’S VIEW OF SIGNALS AND SIGNS IN THE NERVOUS SYSTEM

Theodore Holmes Bullock,     Neurobiology Unit, Scripps Institution of Oceanography and Department of Neurosciences, School of Medicine, A-001 University of California, San Diego, La Jolla, CA 92093, USA

Publisher Summary

This chapter presents a comparative neurologist’s view of signals and signs in the nervous system. All neural systems consist of small components as the actual sensing and reacting elements. Chemical signs are generally the release of something if it’s an organic molecule or else movement in either direction if it’s an ion. Neurons have a lot of ways of responding. This reflects two different kinds of diversity;(1) diversity of types of neurons, and (2) multiplicity of substances released from single neurons. Neurons can release more than one transmitter, along with one or more modulators, several metabolites and sometimes proteins, neurosecretory, or other special products. There is a variety of synaptic potentials; not only excitatory and inhibitory consequences distinguish them but also several other properties. Some have a passive decay and are monophasic, others have a convex, partly active, decremental falling phase, and some are distinctly biphasic. Some are facilitating, others anti-facilitating and there can be fast and slow phases of these effects of history. There are three levels of interaction of neurons recognized, at least in mammals and believed to be general among vertebrates and higher invertebrates. In Floyd Bloom’s formulation, there are levels of (i) macrocircuits of sensory, motor, and association connections, (ii) microcircuits or local circuits operating within the macrocircuits, and (iii) modulatory macrosystems superimposed on both, such as the noradrenaline locus coeruleus system, the 5-hydroxytryptamine raphe system, and the dopamine medial forebrain bundle system.

I INTRODUCTION: AIMS, SLANT AND SCOPE

I take it one of the prime questions in the communication aspect of neurobiology is what are the signals actually employed in neural systems?

I take it we must look for them by measuring signs that something has been communicated. Therefore, an essential question that we must answer first, in order to come to the previous one is what are the signs given by small components of the system? The qualification small components is necessary merely to remind us that all neural systems, as far as we know, consist of small components as the actual sensing and reacting elements. I am not ruling out, you’ll note, that a massed potential for example from a piece of cortex could be a signal, but only reminding you that what detects and responds is not the cortex but cells of the cortex.

The aims of this modest piece are to ask these two questions, in sequence.

I take it that our brain is the product of a lot of evolution. Indeed it would seem obvious that no other system has come such a long way from the level exemplified in coelenterates, flatworms or even insects and gastropods to that of Einstein or Shakespeare, as has the nervous system. Therefore, it behooves us to maintain perspective and I propose to examine the two questions What are the signs of response in small components of the nervous system? and What are the signals actually employed in neural systems? from a comparative standpoint. It is not only that we can expect clues and leads by studying simpler systems, or that some favorable material like the squid giant axon may help, but in addition the perspective itself, the act of comparing, the effort to discern trends or at any rate differences is sure to reveal insights we would miss otherwise.

Further to expose my biases at the outset, the slant here will be pluralist, eclectic and empirical. That means I would rather notice and list than to overlook a phenomenon that could be a relevant sign or signal even if I can not explain it or fit it into a theoretical framework such as the sodium theory. I want to encourage theory and the development of a systematic frame of reference but even more I want to be sure we don’t overlook relevant phenomena when we erect such structures. I believe that as scientists we commonly exhibit our human limitations, one of which is a willingness to pay attention only to certain aspects of reality that impress us for some reason and to overlook or to take for granted, as not worth remark, a great deal that may be quite germane but is outside of our normal domain of discourse.

The scope of my effort today will be limited to those levels of neural events between major parts of a cell at a lower level and major parts of the brain at my upper limit. That means drawing lines to exclude molecules and organelles and membranes at one end and complex behavior, mentation or conscious experience at the other.

So much for aims, premise, slant and scope. Let me turn to the first question, about signs, that is, forms of response in cells and normal arrays of cells.

II SIGNS: FORMS OF RESPONSE

The main result of a survey of animals, high and low, is that almost every possible sign is either to be found or is probable. I don’t know of a nerve cell that luminesces or moves its pigment granules but the three main classes of response: chemical, electrical and physical are each represented by diverse specific examples.

Chemical signs are generally the release of something if it’s an organic molecule or else movement in either direction if it’s an ion. I wonder whether the former is limited to release just because it would be difficult to detect a transient uptake of a small quantity. What we can say, as a result of recent developments, is that the list of chemicals released in response to stimuli is long. There are not only the half dozen or so transmitters but a longer list of maybe fifteen or more different modulators and in addition a variety of metabolic by-products. Proteins are released by some nerve cells, so that the potential exists for a much longer list of specific substances. It’s not my purpose to discuss the substances, the circumstances that cause their release or the meaning of the release, because the point that deserves emphasis here is that neurons have a lot of ways of responding. It now appears clear that this reflects two different kinds of diversity. One is diversity of types of neurons – a diversity far larger than we used to think. The other is perhaps even farther from the usual textbook view, namely a multiplicity of substances released from single neurons. It is widely agreed now that neurons can release more than one transmitter, plus one or more modulators, plus several metabolites and sometimes proteins, neurosecretory or other special products.

Our knowledge of most of these classes of substances in invertebrates is meager but it does permit us to say that variety of substances is not a monopoly of vertebrates. There may well be a flowering in the vertebrates, increasing the number of substances but we don’t know that for sure.

Electrical signs have classically been given as synaptic potentials leading to spikes which are essentially all alike. The evidence today requires us to paint quite a different picture. First, there is a variety of synaptic potentials; not only excitatory and inhibitory consequences distinguish them but several other properties. Some last about a millisecond, others up to at least ten. Some have a passive decay and are monophasic, others have a convex, partly active, decremental falling phase and some are distinctly biphasic. Some are facilitating others antifacilitating and there can be fast and slow phases of these effects of history. Amplitudes of course can be from vanishingly small up to an overshooting 100 mv.

Then there is a whole spectrum of non-classical potentials such as the ILD’s – inhibitions of long duration. Hyperpolarizations and depolarizations can be associated, not only with increased conductance, but with decreased conductance and possibly increased pump action. There is a class of relatively unfamiliar potentials called plateau potentials in which a neuron acts as though it has two states and can be flipped between them; one input flips it to the depolarizing plateau and another flips it back. There are also other examples of regenerative hyper- or repolarizations, opposite in sign to the classical spike.

We musn’t forget the local potential, a graded, local, active but not regenerative event. This may be important in axonal terminals and other situations, indeed in many axons it is one of the normal forms of action and in some the only form. These are the spikeless neurons. They are not necessarily amacrine, not even necessarily short axon neurons in the usual sense of intrinsic, Golgi type II cells with axons a few hundred microns long. The best studied spikeless neuron is in the legs of crabs and has a large axon from a centimeter to several centimeters long, depending on the size of the animal.

A special form of graded and local potential that might be quite important but is little known even phenomenologically is the potential between points on the same cell, for example between dendrites and axon. How general this is or how large or how it changes with time and activity are hardly even studied since Gesell (1940) claimed its importance more than forty years ago. I for one regard it as a neglected and possibly major cellular state variable which might be both an effect and a cause.

Finally, there is a major category of potentials which is probably not one class in terms of mechanism, but heterogeneous. These are the oscillatory and more or less spontaneous potentials. They vary from extremely rhythmic to highly stochastic, from continuous, on-going autochthonous series to rapidly damped ringing, from nearly sinusoidal to quite spike-like. They may be signs of discrete inputs or of the prevailing steady state. They may act as though a single periodic process is at work or like relaxation oscillators.

In short, the variety of electrical signs is formidable. Like the chemical, they differentiate a variety of distinct cell types but at the same time, a given neuron can use several of these forms of electrical signs.

Notice that I have often used the word cell instead of neurons. This is to include the possibilities that glial cells may participate in some of the responses or signs of stimulation.

Evoked potentials and ongoing potentials recorded from organized arrays of cells and gross brain structures are also signs of response and of activity states. They are presumably the volume conducted sum of cellular events of all the just mentioned kinds dependent not only on the mix of kinds but also on the relative timing or synchrony and on the geometry of the cells and processes. There might also be contributions from other sources such as vascular streaming potentials, potentials due to accumulation of ions in intercellular spaces, glial cell membrane potentials, and potentials between cerebrospinal fluid and intercellular fluids. These sources are primarily steady or only slowly changing and hence may contribute little directly to the conventionally filtered evoked or ongoing potentials of the brain. But some of them might on occasion change rapidly enough and the slow and infraslow potentials might be indirectly signs of brain states because the events of higher frequency content might depend on the level of standing potential. I would like to underline that evoked potentials are facultative, or may I say optional products of a mass of cells, not an obligate or predictable ouput. We have various examples where no e.p. is seen although cells are very active – e.g. the inferior colliculus of dolphins stimulated with sonic frequencies; dolphins have nevertheless good e.p.’s in the cortex.

Mechanical signs of neural activity are the least familiar and it is something of a guess on my part that future research will uncover more examples both in respect to locus in the nervous system and to form of response. What we know is that in situations favorable for observation, movements and changes in dimensions can take place in the processes of neurons, as well as of glial cells. This is not the place to review the evidence, which is fragmentary, sparse and cannot in general be confidently extrapolated to gross movement of ordinary brain cells in situ. Suggestive evidence comes from tissue culture; some comes from organ culture or from small autonomic ganglia in the periphery. The ubiquity of cytoplasmic streaming, axoplasmic movement, mitochondrial active movement, changes in optical state of membranes and the like are all suggestive. Optical changes in axons do not always or necessarily mean dimensional changes but at least some are believed to. Dimensional changes in synapses are well established over periods of time adequate for sensory deprivation or enriched experience to act. It may seem maladaptive for a functioning, normal brain to permit cellular movements unless as part of a systematic learning process or lasting effect of environment. Nevertheless, I am betting that various forms of shifts, changes in shape, and in approximation of parts of separate cells are taking place in my brain right now, even if it doesn’t help!

So much for a natural history of signs of response, symptoms of activity, or consequences of stimulation by signals received.

III SIGNALS ADEQUATE TO INDUCE RESPONSE, HENCE TO CARRY INFORMATION

These are the prerequisites for addressing our second question, which was What are the signals actually employed in neural systems? Unless there are other signs of activity, not detectable by the methods used, it is from the foregoing list that we can expect to find the forms of activity that act as signals. Signals will be those signs that can normally influence another cell. We suppose that they are a subset of the list of signs, some of which are presumably epiphenomena like the noise of your automobile – a sign to us but not an effective signal to any part of the automobile.

The results of canvassing the candidate signals for evidence that they can normally be causes and not only effects are by no means all in; we don’t know the answer in the cases of many particular candidates. But there is enough evidence in to permit the conclusion that not a few but many of the substances including transmitters, modulators, ions and many metabolities, as well as many of the forms of current whose potentials we record, are physiological signals. It remains to evaluate where and to what effect some hormones, neurosecretions, CO2, pH, ions and several known as well as suspected transmitters and modulators are actually used normally. Extrapolation from one demonstrated case to others not directly tested is unsafe. The same is true for most field potentials, electroretinograms, evoked potentials and EEG waves.

Good evidence implicates field potentials as causative on the Mauthner cell axon hillock, where fine axonal terminals of the axon cap exert a hyperpolarizating effect without any EM synaptic contacts. Most fibers of the cap are several microns or even tens of microns from the Mauthner cell surface and yet are effective. I think it likely that the axon cap of Mauthner’s cell is a precedent for a large class of loci in our gray matter, where a fine textured neuropile makes functional contact with a postsynaptic cell via diffuse but local fields acting over microns of distance. But that is only a gratuitous guess.

This example of finding a possible precedent in a fish leads me to my last chapter.

IV EVOLUTION OF SIGNS AND SIGNALS

The first questions and certainly not the easiest to answer, may be stated What was the primitive condition? We know that the most primitive animals with a true nervous system, namely the Coelenterata, which includes the corals, medusae, sea anemones and tiny hydroids, have good all-or-none nerve impulses. It is not so clear, but probably they have both electric and chemical transmission. With only four or five types of neurons and no specialized glial cells, they must employ a limited list of signals. The most intriguing part of this question is whether the all-or-none impulse is primitive or derived, relative to a graded local activity. George Bishop (1956) speculated about this long ago, and suggested that graded activity was primitive, though we find it in the most integrative higher neurons. It’s a very plausible speculation I must admit and the all-or-none event could be easily be derived from it. However, we must remember that not only jellyfish have all-or-none, regenerative events; it may well be that a good many eggs, protozoans and algal cells have also. In any case, impulses have either evolved in or out a good many times, since neurons without them are being found in increasing numbers, especially in arthropods.

That suggests a related question: What may be the adaptive value of spikes? We can no longer say that spiking is mainly an adaptation to long distance signalling. Amacrine cells in the retina and in the lateral geniculate of the rat, as well as many short-axon intrinsic neurones in many ganglia and nervous centers have spikes. We should remember that most ganglia and major brain structures of most animals are well within the dimensions of the space constant of common axons, so that spikes gain little over decrementally spread potentials, in respect to conduction of signals. Accepting that over long distances – many millimeters or centimeters – spikes have a significance for faithful propagation, it seems clear that we have to look for another significance in the common intrinsic neuron where such distances are not involved. Remember, spikes were already well-known to unicellular organisms and the ancestors of Hydra, flat worms and other small invertebrates.

The proposition I want to put forward here is that there is a value to spikes over and above whatever advantages they offer to unicellular organisms and for long-distance conduction. I am thinking of the value in respect to encoding information. By introducing devices that encode and decode pulse trains, a wide dynamic range of signals becomes available represented in terms of numbers of pulses, intervals, distribution of intervals and derivatives of these with respect to time (Perkel and Bullock, 1968; Sherry and Klemm, 1980). These several forms of candidates codes may be advantageous in comparison with, say, the amplitude of graded potential, perhaps in several ways at once. I can guess some advantages but I’m not prepared to evaluate them quantitatively. For example, there may be more independence of signal transmission from unwanted effects of temperature, of osmotic fluctuation, of d.c. or slowly fluctuating electric field potentials and of other unknown sources of ‘noise’ that would directly alter the amplitude of a graded potential but only indirectly alter intervals between spikes. Trains of spikes might give an advantage merely because the effects of these unwanted agencies are relatively confined to the loci of the encoding and the decoding. In addition it is possible that the spiking loci or decoding devices are less sensitive to these perturbations or have a narrower frequency pass band for them than would the simple amplitude-graded signal. Signal-to-noise ratio and dynamic range might be improved by the voltage-to-spike rate and spike to transmitter-release conversions, or if not simple spike rate, one or a combination of the other parameters inherent in a more or less regular train of impulses.

What changes took place between coelenterates and advanced invertebrates such as arthropods and molluscs? Between invertebrates and primitive vertebrates? Between primitive and advanced vertebrates? During each of these transitions the nervous system was achieving enormous advances in respect to complexity of behavior, repertoire of discriminable stimuli and of motor actions available. It seems plausible from the facts available that a large expansion in the variety of signals and signs took place between coelenterates and advanced invertebrates. I cannot say as much for the other two transitions. Perhaps there was an increase in variety of peptides but the facts just aren’t available to evaluate such guesses. The change from large numbers of identifiable, unique cells in advanced invertebrates to a few such cells in fishes may not have required more than a modest addition of some new