Brain Function. Proceedings of the First Conference, 1961: Cortical Excitability and Steady Potentials. Relations of Basic Research to Space Biology [Reprint 2020 ed.] 9780520310605

Citation preview

BRAIN FUNCTION CORTICAL EXCITABILITY AND STEADY POTENTIALS RELATIONS OF BASIC RESEARCH TO SPACE BIOLOGY

ADOLF BECK 1863-1942 In rectorial robes, wearing the ring given him in honor of forty years of service to the University of Jan Kasimir in Lvov (Poland), and holding the book he wrote with Cybulski on The Physiology of Man.

UCLA FORUM IN MEDICAL SCIENCES NUMBER 1

BRAIN FUNCTION Proceedings of the First Conference, 1961 CORTICAL EXCITABILITY AND STEADY POTENTIALS RELATIONS OF BASIC RESEARCH TO SPACE BIOLOGY Sponsored by the Brain Research Institute, University of California, Los Angeles, in collaboration with the American Institute of Biological Sciences and with the support of the U.S. Air Force Systems Command EDITOR

MARY A. B. BRAZIER

UNIVERSITY

OF C A L I F O R N I A

BERKELEY AND LOS ANGELES 1963

PRESS

CITATION

FORM

Brazier, M. A. B. (Ed.), Brain Function, Vol. I: Cortical Excitability and Steady Potentials; Relations of Basic Research to Space Biology. UCLA Forum Med. Sei. No. 1, Univ. of Calif. Press, Los Angeles, 1963.

University of California Press Berkeley and Los Angeles, California Cambridge University Press London, England Library of Congress Catalog Card Number: 64-22268 Printed in the United States of America

PARTICIPANTS IN THE CONFERENCE H . W . MAGOUN, Co-Chairman Brain Research Institute, University of California Medical Center Los Angeles 24, California

Co-Chairman American Institute of Biological Sciences Time and Life Building, Rockefeller Center New York 20, New York F.

FREMONT-SMITH,

M. A. B. B R A Z I E R , Editor Brain Research Institute, University of California Medical Center Los Angeles 24, California

W.

R.

ADEY

Space Biology Laboratory, Brain Research Institute University of California Medical Center Los Angeles 24, California J. A. V .

BATES

The National Hospital, Queen Square London, W. C. 1, England J.

M.

BROOKHART

Department of Physiology, University of Oregon Medical School Portland 1, Oregon N.

BUCHWALD

Brain Research Institute, University of California Medical Center Los Angeles 24, California T.

H.

BULLOCK

Department of Zoology, University of California Los Angeles 24, California H.

CASPERS

Physiologisches Institut der Universität Münster Münster (Westf.), Germany

M . R. D E

LUCCHI*

U. S. Air Force, Space Systems Division Inglewood, California E . DE ROBERTIS

Instituto de Anatomía General y Embriología, Universidad Nacional Buenos Aires, Argentina E.

EIDELBERG

Barrow Neurological Institute Phoenix, Arizona J. D.

FRENCH

Brain Research Institute, University of California Medical Center Los Angeles 24, California S . GOLDRING

Division of Neurosurgery, Washington University School of Medicine St. Louis 10, Missouri B . GRAFSTEIN F

Department of Physiology, McGill University Montreal 2, Canada R. J.

GUMNIT

Department of Neurology, State University of Iowa Hospitals Iowa City, Iowa W.

HAYMAKER

National Aeronautics and Space Administration, Ames Research Center Moffett Field, California J. P.

HENRY

Department of Physiology, University of Southern California School of Medicine Los Angeles 7, California A . A . P . LEÄO

Instituto de Biofísica, Universidade do Brasil Rio de Janeiro, Brasil R. B.

LIVINGSTON

Laboratory of Neurobiology, National Institutes of Health Bethesda 14, Maryland

* Currently assigned to the Brain Research Institute, University of California, Los Angeles, f PRESENT ADDRESS: Rockefeller Institute New York 21, N.Y.

S.

LUSE

Department of Anatomy, Washington University School of Medicine St. Louis 10, Missouri H.

MCILWAIN

The Maudsley Hospital, Denmark Hill London, S.E. 5, England F.

MOBRELL

Division of Neurology, Stanford Medical Center Palo Alto, California D . P.

PURPURA

Department of Neurological Surgery Columbia University College of Physicians and Surgeons New York 32, New York O. E.

REYNOLDS

National Aeronautics and Space Administration Federal Office Building 6 Washington 25, D.C. V.

ROWLAND

Department of Psychiatry, School of Medicine Western Reserve University Cleveland, Ohio V. S.

RUSINOV*

Institute of Higher Nervous Activity, Academy of Sciences of the USSR Moscow, USSR A. VAN

HARREVELD

Department of Biology, California Institute of Technology Pasadena, California A . A . WARD,

JR.

Department of Neurosurgery, School of Medicine University of Washington Seattle, Washington * Not present.

UCLA F O R U M IN M E D I C A L

SCIENCES

EDITORIAL STAFF VICTOR E . H A L L , M A R T H A BASCOPÉ-VARGAS,

Editor Assistant

Editor

EDITORIAL BOARD

Forrest H. Adams

William P. Longmire

Mary A. B. Brazier

H. W. Magoun

Louise L. Darling

Sidney Roberts

Morton I. Grossman

Emil L. Smith

John S. Lawrence

Reidar F. Sognnaes

UNIVERSITY

OF C A L I F O R N I A ,

LOS

ANGELES

UCLA FORUM IN MEDICAL SCIENCES A Preface to the Series

When recently in Los Angeles as a Visiting Professor, Dr. Jan Waldenstrom observed the irony of one of the pendular swings in the history of science. In medieval times, he pointed out, scientific publications were so few that interested scholars could learn of foreign accomplishments chiefly through personal visits. Now the vast number of burgeoning journals is sometimes a formidable barrier between the inquirer and the exposed heart of a large problem—thus does ultra-modernity lead to medieval methods of learning! Whether for medieval or modern reasons, the visits of those who bring a point of view, a provocative question and, above all, an answer, are invigorating and welcome occasions. Sometimes convergent visits, planned as symposia, seem to organize, if only temporarily, and therefore challengingly, a shifting assortment of facts and ideas otherwise difficult to hold in focus. The editing and publication of such meetings which are periodically held at UCLA seemed to us to be one small part of an answer to a large need accentuated by the "information explosion". This need was described by President Kennedy's Science Advisory Committee in a report entitled "Science, Government, and Information" and editorially pin-pointed in Science as follows: "One recommendation which could be implemented is that some scientists and engineers 'commit themselves deeply to the job of sifting, reviewing, and synthesizing information' since 'reviewing, writing books, criticizing, and synthesizing are as much a part of science as is traditional research.'" Out of these considerations, and with this initial volume, the UCLA Forum in Medical Sciences has been created, to review, to synthesize and to analyze rather than to serve as another outlet for original papers. The Forum will be published irregularly as the spirit moves the Editorial Board and as those from our own school and from afar gather under its aegis for discussion. The topics will vary widely from the deep roots of medicine in biology, chemistry and physics to the applied medical arts. It is our hope that each volume, whatever its subject, will in its sphere provide that broad view which stands between atomistic surfeit on the one hand and a formless void on the other. Even more we hope that the Forum will reflect among participants, auditors and readers alike, a certain warmth felt by those whose labors are related by content and aspiration to works in distant lands and, indeed, related to life itself. SHERMAN M .

MELLINKOFF

Dean, UCLA School of Medicine

NOTE

The conference, the proceedings of which occupy this volume, is the first of a series supported by grants made to Dr. H. W. Magoun of the Brain Research Institute of the University of California, Los Angeles, by the United States Air Force Systems Command. The American Institute of Biological Sciences acted as co-sponsor. The second conference of the series, held in November 1962, was on the subject of RNA and Brain Function. The third one (November 1963) dealt with Speech, Language and Communication. VICTOR E .

HALL

Forum Editor

PREFACE

Magoun: In an age in which man is going into space, his brain will encounter stresses whose effects are, as yet, little known. Imperative for an understanding of how his brain will meet these new experiences is a fundamental study of its most basic reactions. In each of this series of conferences it is proposed to focus on some facet of the brain's function which needs elucidation in this context. Clearly cortical excitability is one of the most important of the factors that deserves understanding in a situation in which man must maintain performance capability in an alien environment. This can be studied in many ways; this first conference will explore a possible sign of cortical excitability that has received less consideration than many others, namely a steady potential change that can be recorded between the cortical surface and indifferent structures. Recent study has led some observers to regard this electrical sign as indeed a potent indicator of cortical excitability. The shift to negativity in alerting situations, such as alarm, awakening, orienting reactions or peripheral stimulation, stands out in strong contrast to the positive shift that is seen in sleep or anesthesia or tranquilization with chlorpromazine. What these shifts mean in terms of intimate mechanisms such as ion transport, transmitter substances, membrane changes or other fine grain processes, necessitates, among other things, an understanding of the microstructure and microchemistry of the brain. A synthesis of this magnitude cannot be managed in one conference, but hopefully the presentations to follow will form a first approach to this goal. I now introduce Colonel James Henry of the Space Task Group at Langley Field, in charge of the animal phase of the Mercury Project. Henry: D.C. potential shifts are an expression of some change, perhaps in the set and physiology of the brain stem. If for example you recognize, with Dr. Caspers, the significance of D.C. shifts in the relationships between sleep and wakefulness and if, as Dr. Adey points out, we can relate tissue impedance to D.C. changes, and to other associated physiological changes, then we can see that the central question of man's role in space missions, that of the over-all competence of performance of the brain, may be very closely related to the subject of the present conference. The question as to how classical basic neurological and behavioral research can relate to space science is not usually taken seriously enough. So I would just like to say a few words on the general philosophy of the relationship between biology and flight. In the past, biologists have helped engineers with the hardware that they designed to fly within the atmosphere, and with the so-called aviation medicine problems resulting from such flight. Concern has been with flight in the atmosphere, with the gradual loss of its blanketing effects, with the loss Xlii

of gas pressure, of oxygen tension, and with the accelerations that are induced when turns are made at high speeds. Classical aviation medicine has also involved work on physiological psychology. It has been concerned with vestibular disorientation and, of course, with selection with regard to aptitudes for flight and with man-machine adaptation. But there is a critical difference between the aviation medicine of the past and the "space medicine" or "space biology" that we are now facing. In aviation, exposures were always within the earth's atmospheric blanket, and they have always been of limited duration. The men did not have to maintain the machines in flight. The maximum period of any single flight exposure has been a matter of hours. In the future, we are going to be faced with progressively increasing durations of exposure. Men will have to make the same shift that was made with regard to transoceanic traveling as compared to the early voyaging, when the navigators kept in sight of land and drew up their ships in the evenings and maintained them on the shore. It will be necessary to maintain the vehicles in flight. This imposes stringent new requirements on the skill and versatility of the occupants of such vehicles. Further new conditions in space flight are the absence of gravitational effects, the so-called weightlessness, and the radiation that constitutes such a potential hazard. In space flight as in aviation we are required to do research, not merely in these new areas, but in selection, from the point of view of the motivation and skill needed to handle extremely complex situations for prolonged periods in the face of a threatening environment. The ultimate basis of such selection rests on better understanding of the functioning of the central nervous system. Men will certainly be pushed to the limit in the competition to achieve the missions and goals they set themselves. We have then today to face the question of weightlessness and possible disorientation, and the problem of adaptation of the individual to it. This is, perhaps, the one that concerns us most at the moment. It is always possible that, in practice, adaptation to this situation may develop in the same way as to the motions encountered in oceanic voyages. But even if the matter of weightlessness is satisfactorily settled there will always be the question of developing individuals who will give the highest level of performance in spite of very great demands for prolonged periods. This, in turn, is really a problem which has long been the concern of society. To this extent then the requirements for space crew members do not differ from those for any other exacting profession. We are concerned in fact with methods of determining and improving long-term motivation: with investigating, for example, the possible effects of early experience upon motivation; the problem of improving methods of training, and of conditioning individuals. This training does not solely involve enduring environmental effects, such as labyrinthine stimuli. It involves the whole question of optimum response in the manmachine situation. There is concern too, with the mechanisms underlying diurnal rhythms xiv

and sleep-wakefulness cycles, because it is possible that the space environment may disturb these rhythms. In addition, the demands of the task may be such that it may not always be possible to adjust the work to rhythms as they exist in a terrestrial environment. Finally, the question comes up of controlling responses to the stress of long-term isolation, and to other threats. Contributions to this problem can be expected as a result of the application of the techniques of neurophysiological investigation. So, to summarize, the major role of the neurological sciences in space biology is one of defining the mechanisms underlying the adaptation of the individual to a complex man-machine relationship. It is not feasible to provide biological protection for the individual in the form of increased tolerance for the physical stresses of space. The individual can be protected from the environment by the engineering of his capsule, but he will still have tremendous demands made on him within the machine. He will be required to perform as a vital part of the complex instrumentation-computer assembly, and to maintain his performance for prolonged periods in an environment which, while it may not present any remarkable physical stresses, will surely impose very considerable psychological strains. Magoun: I might add that, while the engineers are going to take care of the hardware, and the astronauts are going to be the pioneers who do the traveling, the responsibility for working out, in any fundamental way, the brain mechanisms that form the basic substratum of performance, lies with the scientists who study the brain. Their contributions should be of great significance and relevance, and it is with the purpose of encouraging these contributions that these conferences are designed. Fremont-Smith: On behalf of the American Institute of Biological Sciences, I am very happy to welcome all of you to this conference. The American Institute of Biological Sciences has, among its many activities, a program of developing better interdisciplinary communication, and is therefore pleased to have the privilege of co-sponsoring this conference.

XV

CONTENTS PART I HISTORICAL INTRODUCTION. T H E DISCOVERERS OF THE STEADY POTENTIALS OF THE B R A I N : CATON AND B E C K

M. A. B. Brazier

1

ULTRASTRUCTURE AND C H E M I C A L ORGANIZATION OF SYNAPSES IN THE CENTRAL NERVOUS S Y S T E M

E. de Robertis

15

METABOLIC AND ELECTRICAL MEASUREMENTS WITH ISOLATED CEREBRAL T I S SUES: T H E I R CONTRIBUTION TO STUDY OF THE ACTION OF DRUGS ON CORTICAL EXCITABILITY

H. Mcllwain

49

O N THE SPREAD OF SPREADING DEPRESSION

A. A. P. Leao

73

NEURONAL R E L E A S E OF POTASSIUM DURING SPREADING DEPRESSION

B. Grafstein

S7

STUDIES ON LEARNING

F. Morrell: Effect of Transcortical Polarizing Currents V. Rowland: Steady Potential Shifts in Cortex Discussion: W. R. Adey, R. Livingston, R. J. Gumnit

125 136 148

RELATIONS OF STEADY POTENTIAL SHIFTS IN THE C O R T E X TO THE WAKEFULNESSS L E E P SPECTRUM

H. Caspers

177

NEGATIVE STEADY POTENTIAL SHIFTS WHICH LEAD TO SEIZURE DISCHARGE

S. Goldring

215

T H E UNIDIRECTIONAL POTENTIAL CHANGES IN P E T I T M A L E P I L E P S Y

J. A. V . BATES

Group Interchange

237

R E V I E W AND CRITIQUE

D . P . PURPURA

Commentary: O. E. Reynolds

281 xvii

PART II ASPECTS OF B R A I N PHYSIOLOGY IN T H E S P A C E E N V I R O N M E N T

W. R. Adey

321

STUDIES OF B R A I N S E X P O S E D TO C O S M I C R A Y S AND TO ACCELERATED A L P H A PARTICLES

W. Haymaker

347

R E L A T I O N S O F B A S I C RESEARCH TO S P A C E SCIENCE

J. P. Henry

379

INDEX OF S U B J E C T S

387

I N D E X OF N A M E S

392

xviii

HISTORICAL INTRODUCTION THE DISCOVERERS OF THE STEADY POTENTIALS OF THE BRAIN: CATON AND BECK MARY A. B. BRAZIER Brain Research Institute University of California Los Angeles

The major theme of the conference reported in this volume was the possible relationship of level of cortical excitability to shifts in the steady potential difference that can be recorded between cortical surface and indifferent structures. The existence of this steady potential has been known for nearly a century but its relationship to excitability has remained obscure. In the first part of the last century, even the question whether or not the cortex was in any way excitable was the subject of a major controversy. These early polemics focused on whether the cortex was excitable by externally applied stimuli; the more subtle question (with which this conference deals) as to whether the brain's internal electrical characteristics reflected its state of excitability came later. The attempts to explore the effects on the cortex of external stimuli had begun in the 18th century. Haller, searching for irritability, had pricked the brain and applied irritating fluids and concluded that the grey matter was insensitive to stimulation and that the white matter was the seat of sensation and the source of movement (15). The Italian physiologists had been more successful. The Abbé Fontana and Caldani (Galvani's predecessor in the chair of anatomy at Bologna) had convulsed their frogs by electrical stimulation inside their brains (13, 4). Rolando, following their lead, extended his experiments to pigs, goats, sheep, dogs and also to birds (18). The influential Magendie, however, had failed and had proclaimed the cortex electrically inexcitable, an opinion in which he was backed by Flourens (12). In those days before the neuron had been recognized as the unit of the nervous system and before the pyramidal fibers were known to be processes of cortical cells, there was no a priori reason to expect electrical stimulation of the cortical surface to have a motor effect, but soon an incontrovertible proof was to be given. For a new technique was to invade the field of brain localization. This was elec1

2

BRAIN

FUNCTION

Figure 1. Gustav Fritsch (1839-1891) and Eduard Hitzig (1838-1907) who, in their youth, first demonstrated the electrical excitability of the motor cortex.

trical stimulation. As everyone knows, the pioneers were the two young doctors in Berlin, Fritsch and Hitzig (Figure 1). They demonstrated that certain regions of the cortex were excitable by electricity, as evidenced by elicited movements (14). They did not have plain sailing, for though Ferrier (Figure 2) followed up (9, 10) and expanded their original finding in his classic book, The Functions of the Brain (11), acceptance even then was far from general. George Henry Lewes, whose name has almost faded from scientific memory, was then a respected authority, for his Physiologij of Common Life (16) was the best account of the nervous system available at the time. Lewes, better remembered for his liaison with George Eliot, wrote a pungent attack (17) both on Ferrier and on the original Fritsch-Hitzig concepts of functional localization. Lewes did not spare his fire. He deplored "the increasingly popular but thoroughly unphysiological conception of Localisation." . . We should marvel," he wrote, "to witness so many eminent investigators cheering each other on in the wild-goose chase of a function localised in a cerebral convolution." Just because stimulation of a cortical area evoked a movement, that did not, in Lewes' opinion, prove it to be a motor center. "We do not," he wrote, "consider the centre of laughter to be located in the sole of the foot, because tickling the sole causes laughter." It was Lewes' view that the electrical current passed through the grey

HISTORICAL

INTRODUCTION

3

matter as through any conductor and evoked a movement by exciting the white matter. In support of this view, he quoted the fact that, after removal of the cortex, electrical stimulation of the underlying white substance did indeed provoke movement, presumably via the basal ganglia. But these were all experiments and controversies concerning the excitability of the cortex to externally applied currents. It is only in 1875, when the brain was found to have electrical properties of its own, that the suggestion that these may have some relationship to cortical excitability is first encountered. The idea came quite independently to two men, one in England and one in Poland, and in each case it derived from their knowledge of peripheral nerve. In the mid-nineteenth century, physiologists in many countries were focusing their interest on the electrical activity of nerve. The excitement was caused by the realization that the electrical activity of the nervous system could be used as a sign of its excitability. In the eighteen-forties, Du BoisReymond (Figure 3), the great physiologist at Berlin, had finally demonstrated unequivocally that activity in a peripheral nerve was invariably accompanied by an electrical change, a "negative variation" in the standing potential that had been found between a cut end of the nerve and its longitudinal surface (7, 8). This demonstration was the climax to a longdrawn out struggle to confirm or deny the existence of Galvani's "Animal

Figure 2. Sir David Ferrier (1843-1928), author of the classic book The Function of the Brain.

Figure 3. Emil du Bois-Reymond (1818-1896) first demonstrated the action potential of peripheral nerve.

4

BRAIN

FUNCTION

Electricity". Du Bois-Reymond did not underestimate the importance of his demonstration. He said, If I do not greatly deceive myself I have succeeded in realizing in full actuality (albeit under a slightly different aspect) the hundred years' dream of physicists and physiologists, to wit, the identification of the nervous principle with electricity (7). Many workers soon confirmed that activity caused a negative variation in the potential difference between outside and cut surface of peripheral nerve, and in due time the idea came to some workers in distant countries, England, Russia, Poland and Austria, that the passage of sensory impulses in the brain might similarly be detectable by an electrical change. The first to experiment with this idea was Richard Caton (Figure 4), a young lecturer in physiology at the Royal Infirmary in Liverpool. Following, analogously, the technique used for peripheral nerve, Caton put one recording electrode on the exposed cortex of an animal and the other on a cut surface. For stimulus he used the light from an oxy-hydrogen lamp, for he had no electric light. To his delight he found the change in potential he was looking for, but in addition he found something unexpected: when he had both electrodes on the surface of the cortex or one on the cortex and one on the skull, he found an incessant, though feeble, waxing and waning of current in the absence of all stimulation. This was the discovery of the electroencephalogram. It was to be many decades before the electroencephalogram would be used to localize brain function, and Caton's own interest certainly turned more to the phenomenon of the major potential shift that he found on sensory stimulation. In his first brief report, published in the British Medical Journal in 1875 (5), Caton included the following description:

Figure 4. Richard Caton in his thirties—at the period of his discovery of the electrical activity of the brain.

HISTORICAL

INTRODUCTION

5

The electric currents of the grey matter appear to have a relation to its function. When any part of the grey matter is in a state of functional activity, its electric current usually exhibits negative variation. For example, on the areas shown by Dr. Ferrier to be related to rotation of the head and to mastication, negative variation of the current was observed to occur whenever those two acts respectively were performed. Impressions through the senses were found to influence the currents of certain areas; e.g., the currents of that part of the rabbit's brain which Dr. Ferrier has shown to be related to movements of the eyelids, were found to be markedly influenced by stimulation of the opposite retina by light. In this early report one notices the marked impact of Ferrier's work and a tendency to expect related localizations for sensory responses and peripheral motor effects. Two years later, however, Caton reported further experiments that gave clearer results (6). Caton summed up his experiments in this second report as follows: The investigation thus far tends to indicate that the electrical currents of the grey matter have a relation to its function similar to that known to exist in peripheral nerves, and that the study of these currents may prove a means of throwing further light on the functions of the hemispheres. These findings had little impact on the scientific world; no one followed the lead that Caton had given and, as a consequence, we find the same discovery being made again fifteen years later. In Poland, a young assistant in the physiology department of the University of Jagiellonski in Krakow, Adolf Beck (Figure 5), not knowing of Caton's earlier work, was searching initially for the same phenomenon,

Figure 5. Adolf Beck (1863-1942) as a young man, at the period of his work on the electrical activity of the brain.

6

BRAIN

FUNCTION

namely for electrical signs in the brain of impulses reaching it from the periphery. Like Caton before him, he succeeded, and he too found the brain wave.* His animals were mostly dogs and rabbits, and he published the protocols of all his experiments in the Polish language as the thesis for his doctorate (1). As this was a doctoral thesis (Figure 6), we get the experimental procedures and results reported in far more detail than in Caton's briefer publications in the medical journals. Beck found, as had Caton, that he could evoke a potential swing as a response to light, provided at least one of his electrodes lay on the occipital cortex. In two of the experiments he described, he found a small deviation in response to a shout when one of the electrodes was on the temporal cortex. An additional observation of Beck's holds great interest for those familiar with the blocking action of peripheral stimuli on the E E G . This phenomOznaczcnic

w mozgu

l< >kuliz;icyi

i

rdzeniu

za pomoca zjawisk clektrycznych. Xajiixitf

l~>r.

Adolflleck,

asvatent ZakFadu fiivjolofricznepo I'niw. Jag w Krakowie. Figure 6. T h e title page of Beck's thesis, in which he described the steady potentials of the brain and the influence on them of peripheral stimulation.

enon, confirmed later by many, remained an empirical observation for almost sixty years, until the elucidation of the desynchronizing action on cortical potentials of the ascending reticular system. Beck's own description reads as follows: In addition to the increase or decrease in the original deviation during stimulation of the eye with light, rhythmic oscillations that have been previously described disappeared. However, this phenomenon was not the consequence of light stimulation specifically for it appeared with every kind of stimulation of other afferent nerves.f This is the clearest statement, based on experimental work, that comes to us from the last century of the question being examined in the present volume. In suggesting this theory, Beck relates that he was influenced by the * A fuller account of the lives and of the discoveries of Caton and of Beck that relate specifically to the EEG will be found in: Brazier, M.A.B., A History of the Electrical Activity of the Brain—The First Half-Century (3). f Thesis (1): p. 229, translated from the Polish.

HISTORICAL

INTRODUCTION

7

work of Sechenov, the great Russian physiologist (Figure 7), and by reading the account in the medical journal Vrach* of a report presented by Verigo at the Third Congress of Russian Physicians held in St. Petersburg in 1889 (19). Verigo (Figure 8), a pupil in Sechenov's laboratory, reported similar changes in demarcation current in the lumbar cord of frogs on stimulation of the leg, and also reported a negative variation in potential of the anterior part of the frog's brain when its hind legs were moved. The report in Vrach stated, in part, as follows (translated): On connecting the hemispheres of the frog with the circuit of the galvanometer, Dr. Verigo obtained a continuous oscillation of the needle that did not lend itself to explanation; however, it was nevertheless possible to be certain that on each

Figure 7. I. M . Sechenov ( 1 8 2 9 - 1 9 0 5 ) , in his laboratory in the Medico-Surgical Academy at St. Petersburg.

Figure 8. B. F . Verigo, pupil of Tarkanov and of Sechenov, and early student of the electrical potentials of the frog's brain.

9 Report of the Third Congress of Russian Physicians, St. Petersburg, 1889. Vrach, 10:• .145

1889,

s

BRAIN

FUNCTION

movement of the legs, the anterior part of the hemisphere became relatively negative, electrically, to the posterior part. It is interesting to note that, when he had his electrodes on the brain, Verigo was puzzled by the "continuous oscillation of the needle that did not lend itself to explanation." This was, of course, the E E G of the animal, though not recognized by him as such. This led Wedensky, who was present, to suggest that this method might be used to delineate localization in the cortex. These suggestions were, in fact, being made fourteen years after Caton had first applied the method successfully, though his work was apparently unknown to Wedensky. Later in his thesis, Beck returns again to a discussion of this phenomenon and explains it as "suppression or blocking". Beck felt the challenge of his results and of his proposal that the focal shift of the steady potential on specific sensory stimulation signified increased activity in that cortical center. He argued that were this so, a locally produced increase of activity in that center should produce the same potential shift. Consequently, he stimulated the cortex of a dog directly with a weak induction current and obtained a marked deviation. If my assumption concerning the relationship of the change in action current to the origin of the active state in certain centers is correct, namely that the development of electro-negativity in an area of cortex really indicates the creation of an active state in centers located there, then on direct stimulation of that site an electro-negative swing should result. . . . This was theoretical reasoning; the experiments proved that it was right. . . . Since during stimulation of the eye by light there was a positive deviation of 21 mm, and on direct stimulation of the occipital cortex close to the negative electrode there was deviation of 80 mm in the same direction, does this not prove that this same cortical region went into an active state during stimulation by light? I obtained similar results on stimulating the leg and the corresponding part of the cerebral cortex, by which I mean that both gave a deviation in the direction of negativity.* Beck at this time had no camera, but he drew many sketches to illustrate the protocols on which he recorded the movements of his galvanometer in numerical units. A page from his thesis illustrating a graph from one of his experiments is reproduced in Figure 9. This illustrates his claim that direct stimulation of a cortical area produces a shift in potential in the same direction as is caused by peripheral stimulation of its afferent supply. In the experiment illustrated, the same recording linkage is used both for stimulation of the visual cortex and of the motor cortex. Hence the galvanometer swings in an opposite direction for increased negativity at the occiput from that for increased negativity at the motor cortex. In his thesis, Beck says that he was unable to find any reference in the literature, other than the short report of Verigo's describing a shift of the * Thesis,

(1): p. 230, translated.

HISTORICAL

INTRODUCTION

9

standing potential in the brain having a relationship to its active state (19). For measuring these steady potential shifts, what were the techniques used in 1890? Beck describes his (Figure 10) as follows: To conduct the current away from the cord or brain I used nonpolarising electrodes, as described by Du Bois-Reymond, which I modified a little; they were made from a fine clay saturated with 1% solution of sodium chloride, and set in the form of a suppository on a glass tube filled with a concentrated solution of ZnS (zinc sulfide). The glass tube contained an amalgamated zinc wire and all this was supported on a solid tripod insulated with rubber. One wire from the electrode went direct to the galvanometer, the second wire went to a movable switch of a rheostat, one end of which was connected with the second wire of the galvanometer. Both ends of the rheostat were connected with wires to the Daniel cell for balancing current. Needless to add, by means of keys and commutator the direction of the compensating current could be changed, or the connection interrupted between the electrodes (nonpolarizing) and the galvanometer. Opposite the galvanometer mirror, at a distance of 330 centimeters, was a tele-

chairvUHXiBu .'

Figure 9. Beck's graph of shifts in steady potential caused by peripheral stimulation in an uncurarized dog. The galvanometer reading before stimulation oscillated in a range of 150 to 165 mm to the negative side of his scale (see text). One electrode is on the visual cortex; the other, on the motor area for the foreleg. Stimulation with light (Draznienie okaswialtem) moves the galvanometer reading even more to negativity (i.e., 172). Stimulation of the foreleg (Draznienie konczyny przedniej) swings it in the opposite direction for the same electrode linkage. Electrical stimulation of the occipital cortex increases the galvanometer deflection to a reading over 200 (i.e., about 55 mm to the negative side of the baseline), whereas stimulation of the frontal motor cortex swings it to a reading of 75 mm (i.e., to the positive side of the original baseline). Beck makes the note that he has not attempted to represent the time factor in these charts (in other words, the abscissa is arbitrary).

10

BRAIN

FUNCTION

scope through which a suspended scale was read. The sensitivity of the Wiedemann galvanometer, modified by Hermann, was calculated by the physics assistant of the Jagiellonski University. One centimeter on the scale (when the distance from the galvanometer mirror was equal to 288 centimeters) corresponds to 43/10,000,000 ampères.* His primary objective was to relate these potential shifts to localization of areas on the central nervous system activated by peripheral stimuli, and in this he was very successful. He used the following formula for expressing the electrical sign of the potential shifts that he found: I have to explain the numbers given here; the scale is one meter long with the zero point in the middle and 500 millimeters division on each side. The part of the scale situated to the left of person looking in the telescope I consider positive, the other side—negative.f Using this definition of "negativity", the following generalization can be made of his findings in the many experiments he detailed: in all his experiments, he found (before stimulation) the more rostral parts of the nerve axis negative (in his terminology) to the caudal ones. The more rostral parts became even more negative on sciatic stimulation. Beck adopted Sechenov's term for what we now call the resting potential (as differentiated from the demarcation potential)—he called it the "active autonomous current". Beck found the extent to which the standing potential could be caused to shift by peripheral stimulation, and specifically related this to the excitability of the central nervous system : I have to agree with Sechenov that the autonomous current and the constancy

Figure 10. The electrodes used by Beck for recording steady potentials from the brain. These non-polarizable electrodes were made of cotton threads embedded in clay, or alternatively in the pith from birch trees. * Thesis (1): p. 195-196, translated. In other words, he used an opposing current to balance out the resting potential, for sometimes the resting potential gave a deflection over 500 mm from zero and therefore off his galvanometer scale. For those interested in E E G potentials, he also found these and gives a measurement of 12 mm deflection for them (superimposed on the standing potential). f Thesis (1): p. 197, translated.

HISTORICAL

INTRODUCTION

II

of occurrence of the negative shift are very changeable, depending upon the excitability of the central nervous system. The difference was striking between the frogs kept in the institute during the winter and fresh, spring frogs; in the former, the deflection caused by the autonomous current was negligible—in the latter, however, one had to use a fairly strong current to compensate for the autonomous current." As the conference reported in this volume gives consideration to the effect anesthesia has on the standing potential of the mammalian brain, it is of interest to see what this young Polish scientist reported in 1890. He gives the protocol of experiments on a rabbit and a sketch of his electrode placements (Figure 1 1 ) . With electrodes on the occipital lobe, his galvanometer gave a deflection (oscillating with the E E G ) initially on the positive side of zero by his arbitrary terminology. I narcotized the animal with chloroform to see what happens to the independent oscillations during narcosis. As soon as I started to give chloroform, the deviation which read 140, changed to the negative side as far as — 330, where it began to oscillate, until the narcosis deepened. After the corneal reflex disappeared, the oscillations also ceased, f

Figure 11. Beck's diagram of a rabbit's brain showing the electrical placements used in the experiment on the effects of anesthesia on the steady potential and the superimposed E E G (see text). " Thesis (1): p. 210, translated, f Thesis (1): p. 216, translated.

Figure 12. Beck's diagram of a dog's brain in which he studied the effect of sound stimuli and of chloroform (see text).

12

BRAIN

FUNCTION

In attempting to assess the true sign of the change (called negative by Beck's arbitrary terms), it may be noted that, in this same rabbit, the effect of peripheral stimulation (light, sciatic stimulation) was a change in the opposite direction to that caused by anesthesia. In another experiment, this time in a "large non-curarized dog" (Figure 12), chloroform brought the galvanometer reading from 100 mm (plus or minus an EEG oscillation of 10 mm) on the positive side of zero down to 55. Just prior to being given the anesthetic the dog had been tested for the response to sound. A deflection in the opposite direction to that later caused by anesthesia resulted. In the light of today's technology it is indeed impressive to find in this doctoral thesis of 1890 a clear description of the EEG; of the blocking effect on the EEG of afferent stimulation by all modalities; of the standing (D.C.) potential, together with the localized changes in it evoked by specific sensory stimuli; and the effect on both these electrical phenomena of anesthesia. This young man went on to become Poland's most outstanding physiologist and the co-author with his old Professor, Cybulski, of the standard Polish textbook on the physiology of man (2). Five years after obtaining his doctorate, he was appointed to the Chair of Physiology at the University of Jan Kasimir in Lvov, and there he spent the rest of his life. While still continuing (and publishing) laboratory experiments, Beck rose to be Dean and finally Rector, standing by his University through the years of Poland's stormy history until the final tragedy overtook him, in the form of Hitler's program for the extermination of all Jews. As the Germans closed in on Lvov the danger to Beck increased, for he was Jewish. An old man now, rather than go into hiding he chose to stay in the shelter of the University to which he had given so many years of his life. Just before his eightieth birthday, he became unwell, and while he was in the hospital for an ailment, the Germans came to take him to the extermination camp. Beck's son, a physician, had supplied all members of the family with capsules of potassium cyanide. Beck took his capsule, and saved himself from the gas chamber. The memory of this man, as a scientist and a humanist, is honored by his countrymen today, as it was in the years of his service to the University he loved so well. In 1934 he had been presented with a gold signet ring to mark forty years of scientific work and, the following year, his portrait had been painted for the University by Stanislaw Batowski. This portrait, that appears as frontispiece to this volume, and in which he can be seen wearing the ring and holding a copy of the book he wrote with Cybulski, is one of the few material traces of Beck to survive the Occupation. The ring, hidden by his daughter under the floor of her home in Warsaw, was found by her after the war in the ashes of the house, which like the rest of the city had been burned to the ground by the Germans after the unsuccessful Warsaw

HISTORICAL

INTRODUCTION

13

uprising of 1944. The enamel had turned from red to black, but still legible on the ring are the words : Bene merenti facultas medica, a fitting tribute to a fine scientist. REFERENCES 1.

2. 3.

4. 5. 6. 7. 8. 9. 10. 11. 12.

13. 14. 15.

16. 17. 18.

19.

A . , Oznaczenie lokalizacyi w mözgu i rdzeniu za pomoca zjawisk elektrycznych (Determination of localization in the brain and spinal cord by means of electrical phenomena). Thesis, Univ. Jagiellonski, Krakow, 1890. Rozpr. Wydz. Mat-Przyr. Polsk. Akad. Um., Series II, 1891, 1: 186-232. B E C K , A . , and CYBULSKI, N . , Fizyologia Czlowieka (The Physiology of Man), 2 vols. Krakow, 1915. BRAZIER, M . A . B . , A History of the Electrical Activity of the Brain—The First Half-Century. Pitman, London; Macmillan, New York, 1961. CALDANI, L., Institutiones Physiologicse et Pathologicse. Luchtmans, Leyden, 1784. CATON, R . , The electric currents of the brain. Brit. Med. J., 1875(2): 278. , Interim report on investigation of the electric currents of the brain. Brit. Med. J., 1877(1), Supp.: 62. Du B O I S - R E Y M O N D , E., Untersuchungen über thierische Elektricität, Vols. I and II. Reimer, Berlin, 1848, 1849. , Gesammelte Abhandlungen zur allgemeinen Muskel- und Nervenphysik. Veit, Leipzig, 1877. F E R R I E R , D., The localization of function in the brain. Proc. Roy. Soc., 18731874, 22: 229-232. , Experiments on the brain of monkeys. Phil. Trans., 1875, 165: 433488. , The Functions of the Brain. Smith, Elder, London, 1876. FLOURENS, J . P . M . , Recherches Expérimentales sur les Propriétés et les Fonctions de Système Nerveux dans les Animaux Vertébrés. Crevot, Paris, 1824. FONTANA, F . , Accad. Sc. Ist. Bologna, 1757. FRITSCH, G., and HITZIG, E. Über die elektrische Erregbarkeit des Grosshirns. Arch. Anat. Physiol., 1870, 37: 300-332. H A L L E R , A . , Mémoires sur les Parties Sensibles et Irritables du Corps Animal. D'Arnay, Lausanne, 1760. L E W E S , G. H., The Physiology of Common Life. Blackwood, Edinburgh, 18591860; Appleton, New York, 1871. , Book review of The Functions of the Brain by D. Ferrier. Nature, 1876, 15: 73-74 and 93-95. ROLANDO, L . , Saggio sopra la Vera Struttura del Cervello delVUomo e degV Animali e sopra le Funzioni del Sistema Nervoso. Stamperia S.S.R.M., Sassari, 1809. VERIGO, B . F . , The action currents of the frog's brain. In: Proceedings of the Third Congress of Russian Physicians and Biologists, Vol. II. St. Petersburg, 1889 (In Russian). BECK,

ULTRASTRUCTURE AND CHEMICAL ORGANIZATION OF SYNAPSES IN THE CENTRAL NERVOUS SYSTEM EDUARDO DE ROBERTIS Instituto de Anatomía General y Embriología Universidad de Buenos Aires

I want to bring to your attention some recent results of our work on synapses, which we have been studying now for eight years. This work reveals some new details of the synaptic region which show it to be more complex and more interesting from the physiological and neurochemical point of view than had been thought before. The second part of this talk will be related to some work attempting to isolate nerve endings from the brain. You are probably familiar with the appearance of the synaptic region as revealed by the optical microscope and the many synapses that can be observed on the surface of the single motor neuron, both on the soma and the dendrite. These synapses are very small, generally 1 to 2 microns in size. In the cerebral cortex they are even smaller, and that is why it is very difficult to stain them. They are usually of the order of 0.5 micron and are not stained by most silver stains. The advantage in having such small synapses is the decrease in resistance of the contact of the synaptic cleft, thus permitting better adjustment between the action potential when it reaches the terminal and the postsynaptic component. With the electron microscope, one has a much more complex view of the synaptic region (see Figure 13). In addition to some mitochondria which may be present in the ending (and, indeed, sometimes they are very numerous), the most conspicuous component is the vesicles which we described, together with Dr. Bennett, in 1953 as 'synaptic vesicles'. This is a constant component of synapses. We can say that all synapses so far studied in the central and peripheral nervous system, and both the excitatory and inhibitory synapses, contain a similar component (11, 13). The other fact brought out by this diagram is that the synaptic vesicles, although they are dispersed within the endings, tend to become attached to certain regional points of the membrane (8, 9, 10). This may also make a difference, physiologically, in the transmission of the nerve impulse. Palay (31), who made similar observations, described them as the active points of the synapses, indicating that they may be the regions in which physiological 15

16

BRAIN

FUNCTION

activity takes place. The fact that only this region may be active and not the whole contact is of importance, because there are synapses in certain regions which may be rather large. Indeed, the resistance there would be much greater, but not in those cases where the contact is effective only at small zones of the synaptic membrane. Recently we have observed new details in synapses of the cerebral cortex. When one looks at the cerebral cortex of a rat, in regions between the neurons, between the dendrites of the pyramidal cells, one is impressed by the fact that there are numerous synaptic endings which are very small. These can be recognized especially by the presence of the vesicles inside. The whole tissue is very compact and there are practically no spaces in between the contacting membranes. There is only a cleft of the order of 100 A, which surrounds all of these synaptic structures. Magoun: May I ask if you have ever observed a synapse upon a synapse? Eccles has recently proposed a presynaptic category of inhibition in which excitatory influences are blocked before reaching the postsynaptic membrane (18). A synapse of an inhibiting, hyperpolarizing, or in some manner excitation-neutralizing nature upon the presynaptic ending might form a possible mechanism. De Robertis: This would be a rather difficult thing to see in the cerebral cortex because the material is so tangled up and can be seen in only two dimensions. One should probably reconstruct it in a three-dimensional way in order to be able to detect those presynaptic endings. In general, one sees only the relationship of the synapses with the postsynaptic component, but I think that one might indeed find those contacts. Luse: In the medulla, I have seen, by accident, a synapse upon a synapse which was quite distinct. It was, actually, one of the first endings that we observed. I have not been able to find another, but neither have I really looked for this with any degree of determination. De Robertis: The interpretation of the synaptic region may be rather difficult in a material as complex as that. For example, if you see synaptic vesicles on one side and on the other, that does not mean that there are two synapses in contact. They may be two endings, side by side, which are not making a true synaptic contact. I think that the contact is something more complex, and this is what I want to show now. Figure 13 is a diagram which we have drawn from our results in the cerebral cortex (15). This diagram introduces two new elements in the synapse which I think may be rather interesting. One can see the ending containing the synaptic vesicles and mitochondria, and the postsynaptic component, which in general corresponds to the so-called spine coming from a dendrite in the neighborhood. There are other interesting details. One is that the presynaptic and postsynaptic membranes are thicker and have greater electron density than the membranes in the other regions. They are also separated by a longer distance. This is about 300 A. So that there is a

SYNAPTIC ULTRASTRUCTURE

AND ORGANIZATION

Figure 13. Diagram of a synapse of the brain cortex. At the top the presynaptic component shows one mitochondrion (mi), a vacuole (V) and numerous synaptic vesicles (SV), some of which are adjacent to the presynaptic membrane. The synaptic cleft (SC) is crossed by parallel intersynaptic filaments (if) of about 50 A and separated by 100 A intervals. These filaments are fixed to both the pre- and subsynaptic membranes, which are slightly thickened and denser than the other surface membranes. Within the postsynaptic component there is a web of filaments (or canaliculi) of about 80 A, that is implanted on the subsynaptic membrane on one side and extends at a varying distance into the postsynaptic cytoplasm. This is the so-called subsynaptic web (SSW) of De Robertis et al. (15). G: glial processes that surround the synapse. (From De Robertis et al., 15.)

17

18

BRAIN

FUNCTION

difference also in distances. The most interesting thing is the presence of the filaments which join the two membranes together. There is a system that we have called the intersynaptic filaments, about 50 A thick, which connects the two membranes tightly together. Then there is another system of filaments going into the postsynaptic component. This can penetrate at different distances, varying from one synapse to another. The development of this subsynaptic web, as we have called it, differs from one type of synapse to another. These two components seem to be unique for the synaptic region. Figure 13 illustrates particularly this part of the synaptic region, this complex made by the two membranes, the intersynaptic filaments and the subsynaptic web. Brookhart: May I ask whether these are elements of the dendrites or whether they are present in other synaptic relations? De Robertis: This is a rather general characteristic. We have found it not only in cerebral cortex but also in the hippocampal cortex and in some of the nuclei, such as the amygdaloid nucleus and, also, in the cerebellar cortex. I think that synapses on the soma may have a similar but less welldeveloped apparatus. Probably there is also a difference in the degree of development of this subsynaptic web. I do not know whether Figure 14 may answer your question. In this, one can see one synaptic region in the hippocampal cortex. The presynaptic and the postsynaptic membranes can be seen, and the subsynaptic material coming out from the postsynaptic membrane. One can recognize the vesicles, some of them becoming attached to the presynaptic membrane. One can see the intersynaptic filaments running across the cleft. The cleft is rather wide, wider than the other spaces in between the different structures. The new structures of the synaptic contact were first found in isolated nerve endings, and I will describe later how this isolation can be made. One can clearly see the two membranes and the filament going across the cleft and, also, the so-called subsynaptic web (16). Then we went back to the sections and found we could recognize this structure in all of them. These findings indicate that the filaments bind tightly the synaptic membranes so that, when the nerve endings are isolated, there is a breaking of the structure at the point of least resistance, and the whole postsynaptic membrane comes away attached to the ending. Professor Hyden in Goteborg has shown that, if you dissect nerve cells with their dendrites and stain them with methylene blue, all the nerve endings remain attached to the nerve cell (27). This was very well known, also, in the old literature. Carpenter in 1911 had already shown that the endings may stay attached to the surface of the cell (5). This indicates that there is a mechanism attaching the nerve endings to the surface of the cell. I think this system of intersynaptic filaments may well be related to that kind of attachment. Perhaps it may be important in the determination of the spe-

SYNAPTIC ULTRASTRUCTURE

AND

ORGANIZATION

19

Figure 14. Electron micrograph of a synapse from the hippocampal cortex of the rat showing two active points at the synaptic membranes. The subsynaptic web (SSW) and the intersynaptic filaments crossing the cleft are clearly visible. SV: synaptic vesicles; mi: mitochondria. X 120,000.

20

BRAIN

FUNCTION

cific synaptic pattern of the cell. We know this synaptic pattern is established rather early at different stages of development and remains fixed throughout life. Probably these filaments may provide a mechanism for the establishment of such synaptic patterns. So, to sum up what I have said so far, we think that the synaptic region is a rather complex structure in which there are the synaptic vesicles on one side, and the two membranes, which are attached to each other by this system of filaments, and then the subsynaptic web. This makes a whole complex which probably has some physiological implications. We think that the vesicles are probably connected with the transmitter substances. We postulated from the very beginning that these synaptic vesicles might be the basis of the quantal units of transmitter substance in the ending. We have the system of intersynaptic filaments which may be a way for the ending to recognize the other surface and so to become attached to a specific locus on the postsynaptic cell. This is at present pure speculation. Recently we have tried to isolate nerve endings from the cortex and to study some of their neurochemical components, with the idea of demonstrating that the synaptic vesicles are really the containers of the transmitter substances. This is done by homogenization of the whole brain. We take the rat brain and homogenize it in iso-osmolar sucrose solution. We get three primary fractions: the nuclear, the mitochondrial and the supernatant, which corresponds to the microsomal fraction. The mitochondrial fraction of the brain is a very complex material and one has really to look at it with the electron microscope to recognize how complex it is. When one fractionates liver cells or kidney cells, one can very easily obtain pure mitochondria. From the brain, this is very difficult. The next figures will show how very different and very complex is this mitochondrial fraction. For example, if one isolates the mitochondrial fraction from the cortex or from the total brain, one can recognize several structures. Figure 15 shows that there are three main components: one represented by the nerve endings, another by free mitochondria, and finally the myelin. The nerve ending is attached to the membrane. The mitochondrion is free. Figure 16 gives the data on the protein content, the cholinesterase activity, the succinodehydrogenase activity which accounts for the mitochondrial content, and the acetylcholine content of the three primary fractions. One can see that all the important substances which we are assaying are in the mitochondrial fraction. Nearly 90 per cent of the succinodehydrogenase (SDH) and about 50 per cent of the true cholinesterase are in the mitochondrial fraction. More than 60 per cent of the bound acetylcholine is also in this fraction. If now one subfractionates the mitochondrial fraction M in order to isolate the nerve endings, one can obtain five subfractions. What was done

SYNAPTIC

ULTRASTRUCTURE

AND

21

ORGANIZATION

\ \ e \ .^-«•--

a

•S

M

« I-H ~

w *

w W .

s o ^H

ti

o s

as

00 o

S»

ïï

STUDIES

ON L E A R N I N G TABLE

133

4

Session after cathodal visual polarization

Session after anodal visual polarization

Session after break in training

SESSIONS

Session before anodal and cathodal visual polarization

SUMMARY OF D A T A IN CRITICAL T R A I N I N G

Number of observations

23

10

14

11

14

24

13

11

22

Median

15

16.5

15.5

14

5

16

8

18

8

11-18

12-20

12-17

12-17

3-13

10-18

3-16

15-20

3-17

Polarization of Ear

Motor cortex

Visual cortex

Anodal Catho- Anodal Cathodal dal

Range

cathodal (visual) polarization; session after anodal (visual) polarization; session after break in training. Statistical analysis of these findings leads to the following conclusions. 1. Performance under the condition of visual cathodal polarization differs from that of visual anodal, motor anodal and cathodal, and ear (anodal and cathodal) at better than the 1% level of confidence. 2. There is no significant difference in performance between any one of the polarization conditions, except visual cathodal, and any other. 3. Performance on the day following visual cathodal polarization differs from that on the day preceding it at the 1% level. 4. There was no significant difference between performance on the day following visual cathodal polarization and the day after a break in training. 5. Comparison of performance on the day after visual anodal polarization with that on the day preceding (visual anodal) polarization yields a difference (improvement!) significant at better than the 1% level of confidence. With respect to the last "conclusion" listed, one must hasten to add that there is no justification for attributing the improvement to an effect of anodal polarization. The difference may simply reflect the rising learning curve or the gain normally expected in two days of practice. Only if the experiment were performed at a point on the learning curve where the gain to be expected in two days of practice was negligible, could one attribute the change to the neurological intervention. Such was not the case in these experiments and therefore the influence (if any) of anodal polarization of the cortical receiving area for the conditional signal remains uncertain. In summary, it is clear that the imposition of a surface-negative potential gradient along the axis of the main neural elements of the cortical receiving area for the conditional signal interferes with conditioned performance and prevents retention of the experience acquired during such polarization. Although the evidence is much less conclusive it seems possible that surface

134

BRAIN

FUNCTION

positive currents, while not producing any improvement in performance, may lead to increased retention of the information transmitted during the period of current flow. Ward: As I understand it, with an epileptic focus involving the visual cortex in a conditioning situation analogous to the last series of experiments, spontaneous epileptic activity decreases the level of performance, and this is similar to cathodal polarization in the intact animal. In the intact animal, with cathodal polarization, there is hyperactivity of neural elements such as you showed in the homologous situation in the epileptic preparation. How much activity in visual cortex do you stir up with your levels of cathodal polarization? I thought that you obtained about 200 per cent more driving with cathodal polarization than without it. Is that correct? Morrell: No, it was actually during the anodal polarization that the maximum driving occurred. During the cathodal polarization there was less activity. Cathodal polarization impairs performance. Reynolds: Would it be possible to do both cathodal and anodal polarization on the same day, with half as many trials of each, and see whether they extinguish each other? Morrell: We have not done that, but in some preliminary explorations we did place the anode on one hemisphere and the cathode on the other. We observed no deviation from normal in the animals' behavior. Van Harreveld: Do you have any values of the actual current applied in microamperes per millimeter? Morrell: I cannot give the exact voltage in every instance, but the current was maintained at a constant level of ten microamperes per square millimeter. The voltage therefore varied, of course, with the different diameters of the capillary pore electrode. However, we could easily calculate the value of voltage with fair accuracy because the electrode diameter is known and all except about one per cent of the resistance in the circuit was fixed (source resistance). We chose the capillary pore electrode because it provides a more defined surface area than does, for example, the wick of a calomel half-cell, and therefore allows a better estimate of the actual area of electrode contact. Nevertheless, even if one assumes that there is little or no shunting by surface electrolyte, there must be some dispersion of current as it passes through the dura. Although the magnitude of the shunt is probably small, it sets a distinct limit to the accuracy of the calculation. Bates: Can you say whether, in the rabbit, ablation of the same small area impairs conditioning in the same way that cathodal polarization does? Morrell: Not from my own experience. As a matter of fact, I do not know any study of the rabbit in which this has specifically been done. Dr. PintoHamuy (43) has recently shown in rats that bilateral ablation of visual cortex markedly interferes with acquisition of a conditioned avoidance response based upon a visual cue. I suspect that an excision of visual cortex equal in area to that influenced by the polarizing electrodes would have a

STUDIES

ON

LEARNING

135

similar detrimental effect if performed at a similar stage of learning. It is important to emphasize that these experiments were performed when the animals were at a very early stage in the acquisition of a conditioned response. If the effect of cathodal polarization is similar to that occurring in the presence of an epileptogenic lesion, one would expect that the highly trained or overtrained animal would be much less sensitive than those I have just described. Magoun: Do you think your negative polarization paralyzed the visual receiving cortex? Was there a failure of perception rather than one of retention during such polarization? Is there any indication the animal was able to perceive the conditional signal, which was a photic one? Morrell: In almost all cases the animals made at least a few correct responses during the passage of visual cathodal current (Figure 48C). Usually there was a change in respiratory rate even when the specific motor response did not occur. Apparently there was a reaction to the stimulus even when there was no evidence that the stimulus was "recognized" as a conditional signal. Magoun: The failure to respond was not due to impairment of perception, in your view? Morrell: The animals were certainly not blind, but I cannot exclude the possibility of subtle distortions of the visual environment. Purpura: As I recall, Bishop and O'Leary (9) found that during cathodal polarization of striate cortex positive phases of the primary evoked responses were augmented. This would suggest that geniculo-cortical projection activity was somewhat enhanced. For this reason I suspect that activity was still arriving over the primary pathway to visual cortex in your animals, Dr. Morrell. However, cathodal polarization might be blocking transmission of activity along cortico-cortical pathways, i.e., from visual to motor cortex. What interests me is whether you can detect a residual D.C. change in the visual cortex during the phase of enhanced performance after anodal polarization. That information would be crucial for your argument. Morrell: I would like very much to know the answer to that question but the appropriate measurements were not made in this series of experiments. Rowland: Do you feel it is critical whether your stimulus is above or below 10 microamperes as you have previously shown? Morrell: I am not sure. The current level selected was based upon previous findings (39) on single unit responses. The parameters were tested in that study but not in the present one. Magoun: There was one other question I wondered about as you described your experiments. Does this effect result from negative polarization of areas of the cortex besides the analyzer for the conditional signal and the motor cortex? Did you try some other areas which you did not mention? Morrell: No; in this experiment those were the only areas tried. Polarization of the ears served as a partial control.

136

BRAIN

FUNCTION

I I . STEADY POTENTIAL SHIFTS IN

CORTEX

Rowland: Our interest in D.C. work was stimulated by hearing Dr. Goldring present some of his material a couple of years ago. We wanted to look at the subject from the standpoint of its possible significance in conditioning studies that were being carried out at that time with ordinary wire electrodes and conventional EEG recording. Our first question was: can this dimension of recording give information about changes in the brain with conditioning? And, for this purpose, we used acute studies in fiaxedilized cats to see whether or not a D.C. change occurred to the conditioned stimulus with respect to its being coupled with an unconditioned stimulus that had the capacity to shift the D.C. (44). Figure 49 is an example taken from this first series of experiments. It shows a ten-second auditory stimulus, consisting of 2/sec. clicks. There was recorded, from the posterior ectosylvian cortex referred against the frontal bone, a very definite tendency of the baseline to be shifted, in this case in the positive direction. Very frequently, after a number of habituation trials of this signal (without any association to an unconditioned stimulus), one saw occasional returns of baseline to the pretrial level. But, on the whole, there was a rather sustained tendency to remain below the baseline that existed before delivery of the clicks. The unconditioned stimulus was a half-second train of 50/sec. shocks delivered to the superficial radial nerve. This produced in some cortical areas a negative shift of 200 to 300 microvolts lasting upwards of 30 to 60 seconds. On introducing the unconditioned stimulus at the end of the conditioned stimulus we found that, gradually, the baseline associated with the conditioned signal was significantly elevated above the pretrial level in association with a very pronounced accentuation and broadening of the evoked response. This was a reversible change. On going on to the extinction process by omitting the shock to the nerve and merely presenting the conditioned signal alone, we saw a gradual return of the record to its preconditioning picture. In Figure 49 it had not fully returned, but was showing a definite tendency in that direction. Figure 50 shows responses of another cat, recorded from anterior ectosylvian gyrus, and reveals a slightly different pattern. The 2/sec. clicks are seen in a first presentation here, with prominent evoked responses which are much less prominent after the repetitions in the control or habituation period. The same unconditioned peripheral nerve stimulus was then given at the end of the ten-second train of 2/sec. clicks for conditioning. We were using equipment at that time which had a galvanometer excursion too limited for our needs. The unconditioned stimulus, the superficial radial

>*

-ra «; . i t 0} X Qí e ï «"O

CVJ

I

o o

O)

00

< er

CVJ


>3 ^ O . t' T3 (M JU ¡¿ v, C 3 bJO s .s Cfl ' * ° g « c ^ CT3 o O. j H p o 3 ai < > .> i • >L H •M a, w T3 C3 2e en t+H te O "G5 a

OI +J

o S OS rt ? ^ « Ü fr £ÌS. 33 n tí •&Í fa

M S

STUDIES

ON

LEARNING

139

nerve stimulus, by itself was capable of producing, again, this very marked negative shift, beyond the limits of excursion of the pen. This experiment also differed from the former one in having a long intertrial interval of 20 minutes. By the second trial we saw evidence of a very definite tendency toward desynchronization, and increase in prominence of the evoked responses. As these proceeded over the course of the next hour, a very profound baseline drop appeared at the onset, with enhancement of the evoked responses (Figure 50). At this point we saw the response to the unconditioned stimulus reverse itself, and now give a positive shift. We have no information as to whether simple repetition of the peripheral nerve stimulus by itself could have been capable of doing this. We can make no statement that this change in response to the unconditioned stimulus is a consequence of, or has any particular connection with the development of new response to the conditioned stimulus. We merely point it out as an item of interest. In this case, also, the extinction process developed very slowly and, after 75 extinction trials, there was only a beginning tendency toward reversal to the preconditioned state. In this particular case we had to make a maneuver to get the response back to its preconditioned type of pattern. This was done by converting the clicks from a 'danger' signal to a 'safety' signal. That is, we used continuous clicking, with ten seconds of silence, and then a shock. The record then looked quite like the original preconditioned response on simple presentation of ten seconds of 2/sec. clicks. We felt fairly definitely, then, that it was worthwhile to try to develop a technique for implanting nonpolarizing electrodes. We were unaware of the work of Gumnit (25), and had, ourselves, tried a small calomel electrode but were unable to make a sufficiently stable electrode of small size. We adopted the silver-silver chloride system (Figure 51), and used an electrode consisting of a 5-inch length of approximately 30-gauge silver wire, coiled into a double or triple helix to reduce its size, and then chlorided (45). This, when immersed in a paste of chemically pure silver chloride, held back by a cotton plug and conducting through a saline column with various tips, gave us an electrode that had a 2 to 5 ijV per minute drift, and could be implanted and remain stable for months. Our data on conditioning in chronic animals is still very limited, but I would just like to present one experiment to show what our major experience has been so far. The data were collected over a period of six weeks of training. The animal in Figure 52 had three implanted electrodes. We lost the use of one, but the ectosylvian position, referred to the skull at the vertex, was recorded. The animal was trained to a ten-second 2/sec. click with liquid food reinforcement. There were no evoked responses from this area. This stimulus was immediately followed by an automatic feeding. After several days of this training, the animal showed, on the basis of a learning

140

BRAIN

FUNCTION

Figure 51. Construction of non-polarizing electrode for chronic implantation. A: basic construction showing 5-inch silver wire, coiled in double helix, chlorided, and immersed in paste of AgCl and NaCl. A cotton plug holds the silver chloride in place and permits electrical continuity with NaCl bridge in pipette. B, C and D indicate modifications for various usages. Right angle bends at C and D permit laying body of electrode along surface of skull. (From Rowland, 45.)

curve, a marked tendency toward a positive shift in the response to the conditioned stimulus, reaching a maximum value of 180 microvolts. The unconditioned stimulus, the taking of the food itself, produced a very stable positive shift of 200 to 300 pV, which is not illustrated in the upper record but is shown on the lower one, for which a slower paper speed was used, still with a ten-second conditioned stimulus. The animal was taking its food after the termination of the conditioned stimulus and a positive shift occurred. This may sometimes be followed by a negative shift, but usually this positive shift endures through half to two-thirds of the animal's feeding time. This kind of shift was not seen in this animal with other licking activities, such as grooming. The coupling, then, of the food reward with 2/sec. clicking was associated with a ten-second steady potential positive shift and, behaviorally, the animal invariably turned toward the cup. A movement artifact in Figure 52 shows the time when the animal faced the cup. This did not disturb the overall baseline shift. As controls, a 2/sec. light flash and a 20/sec. click were delivered, neither of which produced a persisting D.C. shift. After this control had been demonstrated, we made a cross-over, now reinforcing the 2/sec. light flash and stopping the reinforcement of the 2/sec. click. In this way the animal was put in an extinction situation for clicks and an acquisition situation for the flash. Over the next four days, we saw the development of a positive shift to the light flash, with the appearance of spikes. We make no claim that these

• -M ÏH * C u w CS i) ^ ~ in o o o

- i

E

Z

o CL,.2

^ ü g u £ s O W e o ca ^ 3 X. rC c as o « *> • S « w s P. o o o ¿ ¡ " S z -à rrt o ' S . ca ^ s ce jy is o « ' S o o « i. £ "C « -S g (D o £ ûh M ¿ j : ™• .-a^ -g W)l3 .SP ó fe - c 3 " u 5 5m M .2 "G O o o \ 2-3 ^ _ toco C

* 30fiA

" M V ^ f l r — % r i

* 60)tA

• 60)1 A

A^JwJv

"WyW^v^^Wv

| 5OOfiV

I

Ln.V.

-too >—^ r Cathodal

-tp

i"o Î Q.

polarization

-SO

-30

0

1

.

f /

// J .

-

1 ••

••••

» '' *30

* 60

• tO/jA

Mi . ' " ' . ' J *

I 3*400

J

a-——T-' ' . J 2

-*oo Cj -300

* *

Anodal

polarization

• /• /

/

rf / *aoo

Figure 87. The action of anodal and cathodal polarizations of the cortical surface on the direct cortical response (dendritic potential, D.P.) to single electrical stimuli of increasing intensity (St.I. 1, 2, 3 in I and St.I. 1, 2, 3, 4 in II). I : Original record. II: Diagrammatic presentation summarizing the results of 48 polarization experiments; the arrows L.n.V. denote the approximate range of the physiologically occurring D.C. fluctuations.

194

BRAIN

FUNCTION

DCR (see Figure 87). These correlations are demonstrable in a variety of experimental conditions. Positive D.C. deviations with a corresponding increase of the DCR are to be found with anodal polarization, in natural sleep, during cooling of the cortical surface (see Figure 86) and after a local application of substance P (cf. 17). Negative D.C. shifts with a concurrent decrease of the DCR, on the other hand, occur during cathodal polarization and natural arousal, at higher cortical temperatures (see Figure 86), during the development of convulsive activity, and after local applications of vari-

Injury

potential

Figure 88. Changes of the direct cortical response (dendritic potential, D.P.) and their relations with D.C. deviations (injury potentials) in the immediate surroundings of an acute cortical lesion. I: D.P. alterations following a small cortical lesion below recording point 1 St. denotes the position of the stimulation electrodes. II: T h e strong reduction of the D.P. is temporarily abolished by an anodal polarization of the cortical surface within the focus area (Pol.: + 6 0 ¡j.A). A: diagrammatic presentation based on 3 5 experiments. It demonstrates the development and time-course of the injury potential (I.P.) and of the D.P. changes at recording points 1 and 2 (D.P.I; D.P.2) after an acute cortical lesion (L). In section B the reduction of the D.P. is plotted against the injury potential.

ous substances, such as potassium chloride and gamma-aminobutyric acid. Figure 88 demonstrates that exactly the same correlations are to be found in the immediate surroundings of a cortical lesion which gives rise to an injury potential. Possibly, the extinction of the evoked potentials occurring with the development of Leao's spreading depression may be explained in a similar way. Purpura: Why do you emphasize only the changes in responses at the superficial cortical level when, under your experimental conditions, changes in excitability may be detectable at all levels? For example, during hypo-

POTENTIAL

SHIFTS

AND T H E

SLEEP

CYCLE

195

thermia, both the positive and negative phases of specific evoked responses are enhanced, and I suspect that with local cooling of the cortical surface there are changes in cortical activity that are not reflected in the so-called "direct cortical response". Caspers: These experiments were designed to establish whether the D.C. shifts are produced by the same structures as the A.C. components of the conventional electrocorticogram and might thus be attributed to activity changes of neuronal elements. For this purpose we had to compare the D.C. deviations with an evoked response which originates, at least approximately, in the same cortical layer and which is induced, furthermore, by a stimulus of constant intensity. On that account we preferred the direct cortical response. The generator structures of a peripherally evoked potential, on the other hand, are scattered in different cortical layers and excited, moreover, by afferent nerve impulses, the intensity of which is variable and can hardly be measured. The same holds true with the conventional EEG. For this reason the peripherally evoked potentials, for instance, seemed to be a less suitable indicator for the relations under investigation. Goldring: I think it makes a difference as to how one cools the cortex. If cold is applied to the surface, a temperature gradient results, with cortical surface being affected more than cortical depth. However, if one cools the brain by surface body cooling, the entire cortex is cooled equally. Then one gets the change that Dr. Purpura mentioned: the whole response increases. Cortical excitability at temperatures between 28° to 34° C. is increased (51). I think, effectively, what you have done here is to influence the more superficial elements of the cortical surface. Caspers: That is true, and I think this is an important point. We have to differentiate between the generator structures which produce an evoked response and the stimulus which is releasing it. When deviations in temperature are restricted to the cortical surface, the resultant alterations of the evoked potentials are primarily due to changes in the polarization of the generator structures which are reflected in the D.C. shifts. If we cool or warm the entire brain we change the reactivity of the generator too, but in addition we strongly affect the number and synchronization of afferent nerve impulses, i.e., the strength of the releasing stimulus. These various actions are hardly to be disentangled. For these reasons we preferred to restrict the temperature changes to the more superficial elements of the cerebral cortex. On the same account, we used the direct cortical response rather than a peripherally evoked potential as an indicator of the temperature effects. Goldring: The only point I am emphasizing is that your results and those mentioned by Dr. Purpura are compatible. Caspers: I agree with that but, for the above-mentioned reasons, I think it is more difficult to interpret the effects of cooling or warming the whole brain. On this occasion I should like to stress another interesting point. A surface-positive shift of the steady potential is nearly always associated with an increase of a (supra-threshold) direct cortical response and vice versa,

196

BRAIN

FUNCTION

though the basic mechanisms producing the D.C. displacements may be entirely different. According to my opinion, these findings prove that the alterations of the so-called dendritic potential are actually related to the D.C. shifts and represent no accidental parallel. Eidelberg: There is one case in which this does not happen. If you inject xylocaine, a very marked positive D.C. shift, of the order of 2 to 5 mV in amplitude, develops very quickly, in about a minute and one-half or so. During the D.C. shift, however, the direct cortical responses are not increased. Caspers: There are, in fact, some rare exceptions. A parenteral injection of hydantoin, for instance, produces a positive shift of the cortical steady potential which is not associated with an increase of the direct cortical response. I am, however, not quite sure that these findings are actually contradictory to the above-mentioned results. The interpretation of these effects has been discussed in greater detail in a paper, written with Baedeker, published in Pflugers Archiv (14). As a whole, all of these findings suggest that the direct cortical response is produced by the same generator structures as the surface steady potential, and may thus be regarded as a locally restricted amplitude modulation of the D.C. component. The results, furthermore, show that at least most of the natural fluctuations of the steady potential can be actually interpreted in terms of polarization changes of neuronal elements located in the outer cortical layer. The anatomical shape as well as the functional properties of these structures is, as yet, rather obscure. Considering the findings of Chang (18), Clare and Bishop (19), Purpura and Grundfest (41) and Tasaki et al. (47), one may, however, assume that dendritic activities are substantially involved in producing the D.C. component. The simple hypothesis concerning the origin of the cortical steady potential is, without doubt, preliminary. More experimental data will be needed to test its reliability. Magoun: May I ask whether the positive shift observed during sleep occurred only during light sleep as this is now differentiated, or did it persist also when sleep passed into the stage which is now called "deep sleep", in which the cortical EEG becomes low voltage-fast, apparently like that in EEG arousal? If positivity continued during deep sleep, it would seem possible, despite the similar EEG tracings, to differentiate features of cortical neuronal activity in deep sleep from those in EEG arousal, in that in the first instance you have a positive shift of the steady potential while, in EEG arousal, the shift is negative. Is this generalization one to which you would subscribe? Caspers: That is quite correct. According to our experience, the positive shift of the cortical steady potential becomes regularly more pronounced with increasing depth of sleep. In more recent experiments on rats we observed short-lasting phases of the "paradoxical" sleep described by Jouvet (31) and others. During these periods, muscle tone and heart rate clearly

POTENTIAL

SHIFTS

AND T H E

SLEEP

CYCLE

197

decreased while the usual EEG waves became accelerated and reduced in voltage. In all these experiments the cortical steady potential showed a distinct positive shift relative to the preceding level. This finding confirms that the D.C. component reflects the actual activity state of the animal more precisely than the conventional EEG. Magoun: Of further interest, in man this stage of deep sleep provides the background for subjective dreams. While we never really know if animals dream, sometimes twitching of the eyeballs or extremities, or pawing movements, suggest that the animal is dreaming. Do you have any observations during such inferred dreaming? Caspers: According to our present experience, the occurrence of sleep movements is nearly always associated with a slight negative shift of the D.C. baseline. Provided the results obtained in man may be transferred to animals, these movements could actually be characterized as dream movements. They consist in twitchings of the lids, slight turnings of the head and in slow movements of the extremities. In this connection it seems to be of particular interest that the occurrence of every pronounced spindle series in the EEG, as already observed by Goldring and O'Leary (25), is also associated with a slight negative D.C. shift. Eidelberg: Dr. Caspers has described the relationships between steady potential levels and the "sleep" and "arousal" electrocorticogram. Dr. N. A. Buchwald and I were interested, a year ago or so, in the generation of "tripped spindles" in the cortex. These spindles, which appear following single shocks to the cortex, caudate nucleus or thalamus, seem to be closely related to the spindles that appear in the ECG during the intermediate stages of sleep. Dr. Buchwald had been studying the anatomical pathways involved in spindling induced by caudate stimulation. We were both puzzled by the fact that a period of time, no shorter than 100 msec., always intervened between the stimulus and the beginning of the spindle. This was far too long to be explained away by even the most circuitous pathways we could think of. We felt that it was necessary to postulate that either an inhibitory mechanism in the cortex delayed the appearance of spindles or that—conversely—their development was the result of an active process preceding them. Conventional EEG records failed to demonstrate any consistent changes in the period between the primary evoked response complex and the beginning of the spindle train. Records taken with calomel-cell electrodes and D.C. amplification showed that a negative "plateau" of up to 0.5 mV in amplitude and lasting about 100 msec, immediately preceded the onset of the tripped spindles. The distribution of these plateaus on the cortical surface coincided with the area of distribution of spindling and of the maximal amplitude of recruiting responses (cat and rabbit). Both the plateaus and the spindles were suppressed by high-frequency reticular, caudate or thalamic stimulation. This seemed to indicate that the appearance of the spindles was causally related to that of these plateaus or long-

198

BRAIN

FUNCTION

lasting late negative waves or short-lasting D.C. shifts. It seems interesting to mention here that sleep spindles in the ECG are related to transient changes in the D.C. baseline. This preliminary work has been followed up by an electrophysiological analysis of the properties of these plateaus, in search for the mechanism of their production. They have been found to follow, in time, the classical evoked potential complex characterizing responses in the cortex to local, transcallosal, caudate or thalamic stimulation. Their area of distribution is, however, larger than that of, say, the transcallosal response. They are not abolished, in this last case, by callosal section, but are eliminated in all these cases by electrocoagulation of the anterior pole of the thalamus. In sharp contrast to classical dendritic responses, these late components showed gradation in relation to stimulus strength only in a very limited span above threshold intensity, becoming maximal rather rapidly. Often they even decreased in size with supramaximal shocks. Recovery-cycle analysis showed at least partial refractoriness at stimulus intervals under 200 msec. They did not interact with dendritic components. They did not reverse in polarity on penetration of the cortex until the deepest cortical layers had been reached. On repetitive stimulation these plateaus could be seen to summate smoothly into the rising steady potential baseline when the frequencies of stimulation exceeded 4 cps. It seems rather clear from this description that these late negative components must be regarded as different from dendritic responses, being generated therefore either by different structures or through different mechanisms than the dendritic responses. In looking for analogous phenomena in less complex systems it seems to us that they may be comparable in their electrical behavior to the late after-potentials of squid or insect nerve. Of course, any relationship beyond similarity remains to be demonstrated but, if I may be allowed to speculate, it seems rather interesting that Shanes (45) has proposed that these late after-potentials are related to potassium release, and that Frankenhaeuser and Hodgkin (20) have indicated that for these after-potentials to be explained it is necessary to postulate that K+ is released into a limited space around the active membrane, no wider than some 300 Á. This value is remarkably close to the dimensions given by the electron-microscopists for the width of the extracellular space in the cortex. Be as it may, it is at least probable that these late negative plateaus or waves constitute the quantal elements in the production of cortical steady potential shifts. Goldring: We have just the opposite results, Dr. Eidelberg. In studying the spindles incident to caudate stimulation we found a positive shift preceding each spindle. Occasionally, when the stimulus was just at threshold, a positive shift occurred without the spindle. This was in cat, with one recording electrode on the surface and the other in the white matter, just beneath the cortex. Eidelberg: We were recording from the suprasylvian cortex.

POTENTIAL

SHIFTS

AND

THE

SLEEP

CYCLE

199

Goldring: We were recording from the motor cortex, from the anterior cruciate in the cat, using transcortical recording; there a positive shift precedes this spindle. It may occur without the spindle at threshold. It is quite different from what one gets with the barbiturate spindle. There a negative shift occurs. The other point I would like to make is this: in animals paralyzed with gallamine triethiodide, artificially respired, and being monitored for blood pressure and pH, the intravenous injection of pentobarbital sodium just sufficient to create spindling produces a positive shift at the cortical surface (23). So, an anesthetic produces a positive steady potential shift similar to the one which occurs with sleep. Bates: Dr. Caspers, could I ask whether you can tell us the sensitivity of your equipment to pH changes? If your apparatus is a good pH-measuring machine, then changes of pH of the order of 0.004 of a pH unit are going to cause swings of 200 nV. This is a very small change in pH, and may account for your findings on C0 2 administration. I think it is in the right direction. Caspers: We did not measure the concomitant changes of pH values during the administration of C0 2 . Such measurements are scarcely feasible in small animals such as rats, but I admit that they are most important in order to clarify whether the observed D.C. shifts are produced by unspecific variations of the pH values or by specific C 0 2 actions. * Bates: Referring to Figure 84, in which there is a corticogram on the top and a single trace of the steady potential at the bottom: as there is no change at all in the corticogram on this occasion, one must conclude there is not necessarily any detectable change in the corticogram during a quite severe and abrupt change in steady potential. Caspers: With some fantasy one might detect a slight acceleration and reduction in the corticogram here, but this change if present may be accidental. Bates: I think that would be pushing it too far to say there is any slight change in that record when C 0 2 is administered. I could point out twenty other places where there are similar changes. Caspers: I readily admit that the changes of the conventional EEG in this record are not convincing. Actually, I am most satisfied by this comment. According to my opinion, too, this is another example which demonstrates that the D.C. component reflects the various activity changes of the cerebral cortex more precisely than the conventional EEG. The same conclusion may be drawn from simultaneous recordings in waking and active animals. If C 0 2 is applied for a longer period the alternations of the EEG waves may become, however, more marked. Bates: If you are going to relate C0 2 to dendritic potential changes, you ° Dr. Casper's note, after the Conference: With respect to the pH sensitivity of our recording device, pH changes of the order of 2-4 would be required to reproduce the D.C. shifts which are released by 1-3% C0 2 . This finding excludes, I think, an essential admixture of technical D.C. components.

200

BRAIN

FUNCTION

agree, without carrying it any further, that the first second or so of that record is very significant? Caspers: Since the baseline of the record clearly shifts to the positive side, we must state that there is a potential change though the superimposed EEG waves are scarcely altered. The stability of the conventional corticogram actually represents some kind of an artifact due to the employment of a condenser-coupled amplifier with a relatively short time-constant. Buchwald: Most of the evidence presented by Dr. Gumnit, Dr. Caspers and Dr. Rowland has indicated that negative D.C. shifts seem to be related to a kind of orienting response or to approach behavior or facilitation of behavior. If one stimulates subcortically, for example in the ventral anterior nucleus of the thalamus or the caudate nucleus, at repetition rates which, I understand from Dr. Rowland and Dr. Eidelberg, lead to negative D.C. shifts, behavioral activities are produced which can be considered as inhibitory. For example, stimulation at frequencies of 6 to 10 pulses per second in the caudate nucleus or the ventral anterior nucleus inhibits bar-pressing for a food reward in animals which have been trained to perform this feat. In addition, there are a number of reports in the literature of the production of sleep at such stimulus repetition rates. With stimulation at frequencies of 20 to 30 pulses per second, "arrest" reactions often occur. It is not until the animal is stimulated at frequencies of more than 30 or 40 pulses per second that behavioral events, which might be termed facilitatory, occur; for example, an animal which has stopped pressing a bar or has become drowsy or fallen asleep is aroused by the faster stimulation and begins to perform again. I wonder if anybody has data on D.C. changes recorded during a behavioral situation in chronic animals as a consequence of stimulating the thalamus or caudate nucleus at low frequencies, which might explain this apparent difference with the results reported with non-stimulated animals. Morrell: May I just add to this difficulty. Experiences in many different laboratories with the use of polarizing currents have yielded surprisingly concordant results. Inwardly directed currents lower the threshold for activation (8) and produce enhancement of the firing rate of single units (5, 32, 37). Surface-negative polarization results in increased threshold for excitation as well as suppression of certain forms of behavior. On the other hand, the correlation between behavioral and electrophysiological indices of excitation and spontaneous shifts in steady potential is much less clear. For example, the EEG activation pattern or desynchronization induced by peripheral sensory stimulation is usually associated with behavioral alerting. Yet, Dr. Gumnit and many other investigators (38) have shown that EEG desynchronization is accompanied by a surface-negative steady potential shift. I do not know of anyone who has independently measured the cortical potential gradient during the application of a polarizing current. I

POTENTIAL

SHIFTS

AND T H E

SLEEP

CYCLE

201

should think that such a measurement would constitute an extremely valuable observation. Eidelberg: This phenomenon of plateaus' has been described before, several times. It follows almost any evoked potential. If one stimulates the cortical surface locally, or transcallosally, one sees the classical evoked response lasting 20-40 msec. (Figure 89). This classical evoked potential is followed, if one records with proper technique, by a much slower plateau that lasts between 100 and 200 msec., and is followed in turn by a spindle. t.c. D.C.

200msec.

t.e.

100 msec.

Figure 89. Rabbit, precentral-agranular cortex. On upper trace of each pair direct cortical response; on lower trace, transcallosal response to the same shock. Left vertical columns: records taken with D.C. amplification; other two at different time-constants, to illustrate the elimination of the late component of the responses by differentiation and the appearance of artifactual—late or secondary—responses.

That plateau is what interested us. At first we assumed it was generated by the same structures that generate the evoked response. As shown in Figure 90, the area of distribution of the plateaus following a stimulus to the opposite hemisphere is far larger than the area of distribution of the classical transcallosal response. The plateaus can be elicited by either local, transcallosal, caudate or thalamic stimulation, and they have the same bilateral distribution and interact with each other. There is another very interesting characteristic: in contrast with the earlier components of the classical evoked response, the plateaus do not show a graded relationship with input. The linear phase of the input-output curve is extremely short. A maximum amplitude is reached, after which increasing the strength of stimulation may often result even in a diminution of response.

202

BRAIN

FUNCTION

If one plots the recovery cycles of these plateaus (Figure 91), a peculiar partial refractoriness at intervals extending between 100 or 250 msec, is seen, which is in complete contrast with the large amount of facilitation one finds for direct cortical and transcallosal responses. They do not interact with the classical evoked potential. Furthermore, they reverse in polarity with penetration of the cortex at a much deeper level than do the classical

Figure 90. Rabbit. Distribution of transcallosally evoked responses.

dendritic components of the responses. We must assume, therefore, that these responses are generated separately from the "dendritic" components. For these and a series of other reasons, I would like to suggest the possibility that these short-lasting D.C. shifts or plateaus may represent the cortical equivalent of the negative after-potential of nerve, and they may represent what Dr. Grafstein has proposed theoretically, namely the leakage or increase in extracellular potassium following or during activity. Purpura: The short-latency transcallosal response is followed, as you know, by a longer-latency, long-duration response that has been studied by Rutledge and Kennedy (43). I think they have obtained evidence that the latter component is generated by activity arising in the brain stem. What I really want to discuss is the relationship of the surface-evoked transcallosal

POTENTIAL

SHIFTS

AND T H E

SLEEP

CYCLE

203

response and unitary activity in the cortex. Latimer and Kennedy (33) have found different units discharging as late as 100 msec, after the contralateral cortical stimulation. When you start referring slow wave phenomena to summation of after-potentials, I cannot understand how such summations could be so temporally distributed as to produce the prominent cortical surface potential when, in fact, units are discharging in different patterns through-

J ~ w

s.s.

so

100

ISO

200

250 Figure 91. 500

200 msec.

I 400 jV

400msac.

Rabbit, precentral-agranular cortex. Transcallosal shocks, records taken at locus homotopic to stimulus. Paired, equal-shock stimulation, records taken at two different sweep speeds. Numbers on left indicate shock intervals.

out the entire cycle. I think you must be very cautious in attributing that long-latency response to an after-potential of neurons. Eidelberg: What I am suggesting is that the plateau is more comparable in its electrophysiological properties to nerve after-potentials than to any other electrophysiological phenomenon I know of. It is probably in the same order of analogy as that between "dendritic" potentials and motoneuron postsynaptic potentials. The responses I described are probably the same as Rutledge and Kennedy described before (43), but with this difference: since they were recording through R.C. amplifiers and with metallic electrodes, they were probably getting a derivative of the plateau. They were using chloralosed

204

BRAIN

FUNCTION

cats too, which is a complicating factor. Essentially, our distribution pattern is the same as the one they found and similar to that of the classical recruiting response. I would agree that the integrity of the diffuse thalamic projection system is necessary for this plateau to appear. A contralateral plateau can be abolished by coagulating n. ventralis anterior, for example, but not by cutting the corpus callosum. This means that the pathway responsible for originating the plateau is probably thalamic and extracallosal. Brookhart: This plateau that Dr. Eidelberg mentions has been seen by several workers. I cannot go along with his notion that this is an aftermath of previous activity. It seems to me to be related to some active process going on in cortical neurons. There is an observation we have made a number of different times, with conventional recording as well as with D.C. recording: in animals where this plateau shows well with low-frequency repetitive thalamic stimulation, the type of timing which converts an ordinary primary response into an augmenting response is one in which a second and a third stimulus fall at the peak of this plateau. There is a cause-and-effect relationship here, of some sort. Every time subsequent stimuli fall in this plateau, the augmenting type of response develops. If the subsequent stimuli fall outside of the plateau, it remains a primary response. Eidelberg: There are two points there. The first one can be answered by saying that the latency of the plateau is always longer than the time-course of the evoked response. So I assume that some activity has preceded the development of the plateau, although of course this does not necessarily mean that it is not a separate active process. Your second point is very good, if one refers only to the response to, let us say, V.P.L. stimulation. However, I do not see augmentation if I stimulate repetitively across the corpus callosum. In that case, one sees absolutely no augmentation of the primary evoked potential, although the plateaus are there. Brookhart: I still feel more comfortable with the notion that the plateau represents neuronal activity going on in the cortex. It is not the result of preceding activity in the sense of a recovery process going on in neurons which are quiescent at the time. Goldring: If one stimulates the midline thalamus (nucleus paracentralis) in the rabbit, each recruiting response is followed frequently by a slower negative potential that lasts anywhere from 150 to 300 msec. It has been found by Gerber (21) that injection of convulsant agents, such as caffeine, picrotoxin or metrazol, causes the slow negativity to increase in size. If the stimulus frequency is 3/sec., a response very similar to the spike-and-dome discharge is produced. These remarks are based on work in which we studied the effects of convulsants and depressants on the DCR in isolated cerebral cortex. There

POTENTIAL

SHIFTS

AND

THE

SLEEP

CYCLE

205

the very same potential phenomena are evoked and the intravenous injection of metrazol produces changes in the slow potentials similar to those seen in the neuronally intact cortex. Along this line, the action of pentobarbital sodium on slow negativity is interesting. If slow negativity is evoked by a stimulus to the cortical surface, pentobarbital sodium causes it to increase in size; concomitantly, the primary negative potential is decreased. However, if slow negativity is activated by midline thalamic stimulation, the administration of pentobarbital sodium then causes slow negativity to disappear. The necessary condition for evoking slow negativity is virtual absence of anesthesia. If you give pentobarbital spdium, slow negativity is reversed to a positive process. I might also mention that Pearlman (40), working with us, found optic nerve stimulation to evoke a similar slow negativity in visual cortex with a duration of about 200 milliseconds. In that situation slow negativity had the lowest stimulus threshold, stronger stimuli bringing in the other components of the optic response. Purpura: The surface record would be different from that obtained in the cortical depths. Are you convinced that weak stimulation of the optic nerve did not activate elements in the cortical depths? You cannot be certain that there is no activation of the geniculo-cortical projection pathway with the development of a response in the cortical depths that might be masked by the negative wave. Goldring: I cannot be sure of its reversal point since we have not recorded with intracortical probes. When one speaks of absence of potential in the cortical record, one always has to consider algebraic summation of competing, concurrent potentials of opposite polarity. Therefore, a cortical surface electrode may record nothing when two or more active processes are occurring simultaneously. Rowland: In regard to the very interesting data that Dr. Caspers has presented concerning arousal, desynchronization and negative shifting, I would like to cite a couple of observations that we have made. One has to do with the significance of the arousing stimulus; this becomes apparent in the conditioning studies. We have seen examples where simple arousal by auditory stimulation is recorded on ectosylvian gyrus with minimal D.C. shifting. However, when we condition that animal to auditory stimuli, the arousal (electrographic desynchronization) then becomes prominently associated with a negative shift. So that, although in the conventional EEG this would look like simple activation, in the D.C. record the two circumstances are differentiated by the shift appearing when a new significance is added to the arousing stimulus by conditioning. In addition to this, we have seen desynchronization occurring with shifts both in the positive and negative directions, depending on the circumstances of the conditioned stimulus. In the particular case shown

206

BRAIN

FUNCTION

in Figure 52, acquisition was associated with positive shifting, extinction with negative shifting, both of them being accompanied by desynchronization of the record. One further point I should like to mention is in confirmation of the positive shifting with natural sleep and with pentobarbital. We have used a technique of cooling the reticular formation by circulating water at 8° C. through hypodermic needle stock implanted at the intercollicular level. This reduces the temperature of the reticular formation locally to 23° C. In a chronic animal, in the course of 30 to 60 seconds, this produces a progressive lowering of the head and insensitivity to arousing stimuli. This is accompanied by a pronounced positive shift of about 200 microvolts in frontal cortex (anterior lateral gyrus) and the development of synchrony. And within 30 seconds after cessation of this cooling, the animal spontaneously arouses and once again becomes attentive to stimuli. This restoration is accompanied by a negative shift. Part of the value of the technique is that it can be used in confirmation of, say, something like anesthesia or pentobarbital, where we are not sure that the drug action is exclusively in the reticular formation. We can use the technique as a localized method for reducing reticular activity. We know it is localized, from experiments in acute animals where we progressively lowered two cooling probes in the reticular formation and we did not see the development of synchrony until they entered the reticular formation. Above this, they did not induce synchrony. So, we presume this was not due to vascular effects but was related to specific take-out of the reticular system. Buchwald: In several different behavioral situations in cats, we have noticed considerable activity in the 12-15 cycle range in the occipital cortex during reinforcement of approach conditioning. This changes to a markedly desynchronized record during extinction. This might produce D.C. changes which agree with Dr. Rowland's D.C. measurements. Rowland: In defensive conditioning, that is, where the reinforcement is painful, we have seen in flaxedilized animals that the extinction is sometimes associated with very marked positive shifting and synchrony, wherein we would expect the extinction procedure to be associated with general relaxation of the animal because the shock is not coming. With approach conditioning, we have seen two animals which, when well trained and given the stimulus indicating no food coming', exhibit a response of going to the empty food cup and licking it. There is no question that these animals are aroused. However, in agreement with your observation, we have, in very recent studies, seen synchrony appear not only with extinction but also to a discriminated non-reinforced signal. In these cases no distinct D.C. shift was present, but we were not recording from occipital cortex. Magoun: I would like to put on the record a method Dr. Albe-Fessard

POTENTIAL

SHIFTS

AND THE

SLEEP

CYCLE

207

and her associates employ for focal cooling of the nervous system. Its basis, as I understand it, is the rapid expansion of inert gas under pressure, in a fine metal container which can be oriented like an electrode within the brain.* Livingston: We are using this method. It is a very simple gas expansion cooling device (36). It is quite small and easy to introduce into the brain in a chronic animal. With it, one can reversibly interrupt activity in a local region. The present instrument is 3 mm in outside diameter. There is very little cooling in the shaft. Cooling takes place in a tiny chamber where a jet releases Freon gas for expansion. The gas is under pressure in an inner tube, and, after expansion, refluxes in an outer tube covered with polyethylene. Rowland: We are very much interested in that study. We are looking into cooling on the basis of the Peltier principle, so localized cooling can be done with electrical currents and we can get away from circulating gas or cold water. This is in the stage of development and still has quite a way to go. For now we find effective two needle stock tube systems (Figure 92)

Figure 92. Cement mass and removed calvarium viewed from underside, showing cooling tubes and tips of recording electrodes (arrow). * D. Albe-Fessard, personal communication, 1961.

208

BRAIN

FUNCTION

implanted on either side of the tegmentum near the aqueduct (Figure 93). These are immobilized with polymethacrylate cement at their emergence from the skull. The relatively large holes produced by these cooling tubes were not associated with any gross behavioral changes and only when the cold water was circulated were the effects seen."

Figure 93. Weil-stained cross-section of midbrain showing region occupied by cooling tubes. Despite relatively large size of defects the animal showed no gross behavioral abnormalities two weeks after surgery.

Stuart at Long Beach and Ott at Hughes Aircraft (46) have done cascading of Peltier junctions. They use 0.5 or 1 mm rod to conduct heat away so that the mass of junctions can be held outside of the head. In respect to the heat dissipation necessary for such an effect as hypothermic reticuloplegia, we have measured the temperature gradients and calculated that the heat dissipation is on the order of 15-30 calories per minute. We are not yet sure whether a miniaturized Peltier junction, that can be inserted into brain like an electrode, can meet this requirement. However, values 8 Note: A brief film was presented at the Conference illustrating the points mentioned, namely: normal spontaneous behavior of the cat prior to cooling; then, in conjunction with reticular cooling, the lowering of the head, eye closure, insensitivity to probing of the ear canal, positive D.C. shifting and synchrony; and prompt reversal of all these changes within 30 seconds of cessation of cooling.

POTENTIAL

SHIFTS

AND T H E

SLEEP

CYCLE

209

Figure 94. Two pulse generators were used to produce paired 0.1 msec, pulses varied for polarity and sequence. Cortical recording following intralaminar stimulation. A: both pulses positive in sign. B: both negative. C: positive-negative. D: negativepositive. Note that evoked potential only followed pulses of same polarity (A and B). Only with paired negative pulses in B is there a D.C. baseline shift. Horizontal calibration: 500 msec.; vertical calibration: 500 |j.V. Negative is up. (From Mahnke and Ward, 35.)

close to this are likely and we are interested to pursue it further. Unfortunately we cannot escape entirely from a circulating fluid situation, because whenever heat is taken away by this principle it reappears immediately on the other side of the junction, and has to be dissipated by some kind of water or air flow. So, we are presently considering a technique of using a combined Peltier junction and water-current dissipation. Ward: Figure 94 presents a phenomenon which will be familiar to those of you who are engaged in D.C. shifts that are a consequence of intercerebral stimulation. It points out some of the problems and areas of lack of knowledge. In this experiment, carried out by Dr. J. Mahnke (35), a concentric stimulating electrode was placed in the intralaminar nucleus of the thalamus. The recording is from the cortex. The pulse generators delivering these stimuli produce 0.1 msec, pips which can be either positively directed or negatively directed, and are connected so that one can deliver two 0.1 msec, pulses, the polarity of each being independently variable. In line A of Figure 94 the pulses delivered to the thalamus are positivepositive; in line B, negative-negative; in C, positive-negative; in D, nega-

210

BRAIN

FUNCTION

tive-positive. I will call attention first to the large and very dramatic sharp artifacts present in all records and, secondarily, to the fact that, although cortical responses are evoked by both the positive-positive and negativenegative stimuli to the thalamus, a negative D . C . shift of approximately 500 microvolts occurs only when the stimuli are negative in direction. This is presumably an artifactual D . C . shift. Eidelherg: W h e n you change the polarity of the stimulus with bipolar electrodes, the locus of stimulation (cathode) is changed. So that this is not a critical test unless one gets full reversal of polarity of the shift. Every time we see a shift that is not constant, regardless of the polarity of the stimulus, we move an electrode elsewhere. Adey: I am interested in the fact that virtually all the methods of D.C. recording that have been discussed here have hinged on recording at the cortical surface. There are obviously many of these D.C. changes, particularly in chronic states, hinging on behavioral observations, that occur in subcortical structures or in buried cortex. This, parenthetically, is of course one of the advantages of the impedance method I described before. I would like to hear from anybody here who has used D.C. recording in chronic animals in a way that is applicable to subcortical or buried cortical structures, without damage by the insertion of an overly large electrode. Eidelberg: It can be done, in chronic animals, by employing a fine glass pipette electrode filled with agar in saline. Adey: I have been very impressed by the similarities and differences in the impedance record with, for instance, Dr. Caspers' finding. I t is very apparent the behavioral situations in which he saw negative shifts were those in which, in general, we saw an increase in the impedance of dendritic structure. Conversely, the trends toward sleep, and so on, where there is a rise in impedance, were those with an associated D.C. shift. As to the question of overshoot, we do see overshoots in our impedance record. I think the correlation between the two is at least something to be pursued. REFERENCES 1. ARDUINI, A., Enduring potential changes evoked in the cerebral cortex by stimulation of brain stem reticular formation and thalamus. In: Reticular Formation of the Brain (H. H. Jasper et al., Eds.). Little, Brown, Boston, 1958: 333-351. 2. ARDUINI, A., MANCIA, M., and MECHELSE, K., Slow potential changes elicited in the cerebral cortex by sensory and reticular stimulations. Arch. Ital. Biol, 1957, 95: 127-138. 3. BREMER, F., Cerebral and cerebellar potentials. Physiol. Rev., 1958, 38: 357398. 4 . BROOKHART, J . M . , ARDUINI, A . ,

MANCIA, M . ,

and

MORXJZZI, G . ,

Thalamo-

cortical relations as revealed by induced slow potential changes. /. Neurophysiol., 1958, 21: 499-525.

POTENTIAL

5. 6. 7. 8. 9.

10. 11. 12. 13.

14.

15.

16.

17.

18.

SHIFTS

AND T H E

SLEEP

CYCLE

211

J. M., and B L A C H L Y , P. H., The influence of DC potential fields on cerebellar unit activity. In: XIX International Physiological Congress (Abstracts of Communications). Montreal, 1953 : 236-237. B U R E S , J., and BURESOVÄ, O., Die anoxische Terminaldepolarisation als Indicator der Vulnerabilität der Grosshirnrinde bei Anoxie und Ischämie. Pflügers Arch., 1957, 264: 325-334. BÜRGE, W. E., W I C K W I R E , G. C., and SCHAMP, H. M., A study of the effect of different anesthetics on the electrical potential of the brain cortex. Cur. Res. Anes. Analg., 1936, 15: 261-267. BURNS, B . D., The production of after-bursts in isolated unanaesthetized cerebral cortex. J. Physiol., 1954,125: 427-446. B U R R , H . S . , and H A R M A N , J R . , P . J . , Voltage gradients in the nervous system. Trans. Am. Neurol. Ass., 1939, 65: 11-14. CASPERS, H., Uber die Beziehungen zwischen Dendritenpotential und Gleichspannung an der Hirnrinde. Pflügers Arch., 1959, 269: 157-181. , Changes of cortical D.C. potentials in the sleep-wakefulness cycle. In: The Nature of Sleep (G. E. W. Wolstenholme and M. O'Connor, Eds.). Churchill, London, 1961: 237-259. , Die Entstehungsmechanismen des EEG. In: Klinische Electroencephalographie (R. Janzen, Ed.). Springer Verlag, Heidelberg, 1961: 4-26. , The action of substance P on the cerebral cortex and the brain stem reticular formation. In: International Symposium on Substance P (P. Stem, Ed.). Sarajevo, 1961: 29-43. CASPERS, H., and BAEDEKER, W . , Die Verschiebung der corticalen Gleichspannung während der Krampfentstehung und ihre Beeinflussung durch anticonvulsive Substanzen. Pflügers Arch. (In press). CASPERS, H . , and L E R C H E , E., Über die Verwertbarkeit der corticalen Makroaktionspotentiale als Indicatoren einer corticopetalen Erregungsleitung. Pflügers Arch., 1959, 270: 8. CASPERS, H., and SCHULZE, H., Die Veränderungen der corticalen Gleichspannung während der natürlichen Schlaf-Wach-Perioden beim freibeweglichen Tier. Pflügers Arch., 1959, 270: 103-120. CASPERS, H., and STERN, P., Die Wirkung von Substanz P auf das Dendritenpotential und die Gleichspannungskomponente des Neocortex. Pflügers Arch., 1961, 273 : 94-109. CHANG, H. T., Dendritic potential of cortical neurons produced by direct electrical stimulation of the cerebral cortex. ]. Neurophysiol., 1951, 14: BROOKHART,

1-21.

H., and BISHOP, G . H., Properties of dendrites; apical dendrites of the cat cortex. EEG Clin. Neurophysiol., 1955, 7: 85-98. 20. FRANKENHAEUSER, B., and HODGKIN, A. L., The after-effects of impulses in the giant nerve fibres of Loligo. }. Physiol., 1956, 131: 341-376. 21. G E R B E R , C. J., Effect of selected excitant and depressant agents on the cortical response to midline thalamic stimulation. EEG Clin. Neurophysiol., 1961, 13: 345-364. 1 9 . CLARE, M .

BRAIN

212

FUNCTION

S . , SCHOENFELD, R . L . , and H O E F E R , P . F . A . , The slowly changing voltage of the brain and the electrocorticogram. EEG Clin. Neurophysiol., 1951, 3 : 231-236.

2 2 . GOLDENSOHN, E .

2 3 . GOLDRING, S . , M E T C A L F , J . S . , HUANG, S . H . , SHIELDS, J . , a n d O ' L E A R Y , J . L . ,

Pharmacological selectivity manifested by agents acting upon the cortical dendritic spike and its slow after-effects. J. Nerv. Ment. Dis., 1959, 128: 1-11.

24. 25 . 26. 27.

28. 29.

30.

31.

32. 33. 34.

35.

S., and O ' L E A R Y , J. L . , Experimentally derived correlates between EEG and steady cortical potential. J. Neurophysiol., 1951, 14: 275-288. , Correlation between steady transcortical potential and evoked response. I and II. EEG Clin. Neurophysiol., 1954, 6: 189-212. , Cortical D.C. changes incident to midline thalamic stimulation. EEG Clin. Neurophysiol., 1957, 9: 577-584. G U M N I T , R . J . , D . C . potential changes from auditory cortex of cat. J. Neurophysiol1960, 23: 667-675. G U M N I T , R. J., and GROSSMAN, R. G . , Potentials evoked by sound in the auditory cortex of the cat. Am. J. Physiol., 1961, 200: 1219-1225. H A R M A N , P . J . , Anesthesia and the E . M . F . of the nervous system. Yale J. Biol. Med., 1941, 14: 189-200. INGVAR, D . H . , Cortical state of excitability and cortical circulation. In: Reticular Formation of the Brain (H. H. Jasper et al., Eds.). Little, Brown, Boston, 1958: 381-412. JOUVET, M., Telencephalic and rhombencephalic sleep in the cat. In: The Nature of Sleep (G. E. W. Wolstenholme and M. O'Connor, Eds.). Churchill, London, 1961: 188-208. J O Y N T , R . J . , Micro-electrode studies of cerebellar electrical activity in the frog. J. Physiol., 1958,144: 23-37. L A T I M E R , C. N., and KENNEDY, T . T . , Cortical unit activity following transcallosal volleys. J. Neurophysiol., 1961, 24: 66-79. LOESCHCKE, H . H . , Uber Bestandspotentiale im Gebiete der Medulla oblongata. Pftugers Arch., 1956, 262: 517-531. MAHNKE, J . H., and W A R D , JR., A. A., Standing potential characteristics of the epileptogenic focus. Epilepsia, ser. 4, 1961, 2: 161-169. GOLDRING,

3 6 . M A R K , V . H . , CHATO, J . C . , E A S T M A N , F . G . , ARONOW, S . , a n d E R V I N , F .

R.,

Localized cooling in the brain. Science, 1961, 134: 1520-1521. 37. MORHELL, F., Effect of anodal polarization on the firing pattern of single cortical cells. Ann. New York Acad. Sci., 1961, 92: 860-876. 38. , Electrophysiological contributions to the neural basis of learning. Physiol. Rev., 1961, 41: 443-494. 39. O ' L E A R Y , J . L . , and GOLDRING, S., Changes associated with forebrain excitation processes: D.C. potentials of the cerebral cortex. In: Handbook of Physiology; Neurophysiology I (J. Field, H. W. Magoun and V. E. Hall, Eds.). American Physiological Society, Washington, D.C., 1959: 315328.

POTENTIAL

SHIFTS

AND T H E

SLEEP

CYCLE

213

40. PEABLMAN, A. L., GOLDRING, S., and O ' L E A R Y , J. L., Visually evoked slow

41.

42.

43. 44. 45. 46. 47. 48.

49.

negativity in rabbit cortex. Proc. Soc. Exp. Biol. Med., 1960, 103: 600603. PURPURA, D. P., and GRUNDFEST, H., Nature of dendritic potentials and synaptic mechanisms in cerebral cortex of cat. J. Neurophysiol., 1956, 19: 573-595. ROWLAND, V., Electrographic responses in sleeping conditioned animals. In: The Nature of Sleep (G. E. W. Wolstenholme and M. O'Connor, Eds.). Churchill, London, 1961: 284-306. RUTLEDGE, L. T., and KENNEDY, T. T., Extracallosal delayed responses to cortical stimulation in chloralosed cat. ]. Neurophysiol., 1960, 23: 188196. SCHULZE, H., and CASPERS, H., Die physiologischen Schwankungen der corticalen Gleichspannung und ihre Beziehungen zur Reticularisaktivitat. Pflugers Arch., 1961, 274 : 55-56. SHANES, A. M., Potassium movement in relation to nerve activity. J. Gen. Physiol., 1951, 34 : 795-807. STUART, D. G., O T T , L. H., and CHESHIRE, F. C., Thermal electrodes based on "Peltier Effect". EEG Clin. Neurophysiol., 1962, 14: 132-135. TASAKI, I., POLLEY, E. H., and ORREGO, F., Action potentials from individual elements in cat geniculate and striate cortex. ]. Neurophysiol., 1954, 17:454-474. TSCHIRGI, R. D., and TAYLOR, J. L., Slowly changing bioelectric potentials associated with the blood-brain barrier. Am. J. Physiol., 1958, 195: 7-22. VANASUPA, P., GOLDRING, S., O ' L E A R Y , J. L., and WINTER, D., Steady potential changes during cortical activation. J. Neurophysiol., 1959, 22: 273284.

5 0 . WASANO, T . ,

INOKUCHI, S . , INANAGA, K . , NAKAO, H . ,

and

FUCHIWAKI,

H.,

The slowly changing potential of the brain and electrocorticogram under the influence of anesthetics and convulsant drugs. Kyushu Mem. Med. Sci., 1953, 3 : 243-251.

5 1 . WEINSTEIN, W . , KENDIG, J . H . , GOLDRING, S . , O ' L E A R Y , J . L . , a n d LOURIE, H . ,

Hypothermia and electrical activity of cerebral cortex. Arch. Neurol., 1961, 4 : 441-448.

NEGATIVE STEADY POTENTIAL SHIFTS WHICH LEAD TO SEIZURE DISCHARGE* SIDNEY

GOLDRING

Washington University School of Medicine Saint Louis

Recently accumulated evidence establishes steady potential changes as important determinants of paroxysmal states of the nervous system. Some of these are relatively abrupt in appearance, others are slowly progressive. In animals, enduring shifts in potential have been noted during induced seizure activity of cerebrum (3, 22, 29), cerebellum (7, 34) and hippocampus (9, 24, 35). In man, such changes are known to take place during after-discharge bursts (10, 14) and 3/sec. spike-and-dome sequences of petit mal epilepsy (6). Such slow potential changes frequently outlast the seizure period and, since even brief sequences of seizure discharge may last a second or more, faithful recording of the slow components requires a technique different from that ordinarily employed. Recording electrodes must be nonpolarizable and amplifiers must be capable of passing enduring voltage changes. Our recording electrodes are calomel half-cells, and we use D.C. amplifier, oscilloscope and Grass camera to register slow potential phenomena. Such equipment also permits the recording of physiological transients of brief duration, such as evoked, recruiting and direct cortical responses. Evoked responses have been widely studied, and changes in them, which can be induced experimentally, are a convenient reference point for examining the significance of slow potential phenomena. Whereas slow potential changes which are of non-neuronal cause undoubtedly exist, a majority of workers has been willing to accept a neuronal origin for the potentials reported here. However, slow potential changes have been ascribed to the blood-brain barrier (32), and an enduring response from astrocytic glial stimulation in tissue culture (17) suggests a possible glial contribution to steady potential. The slow potentials we describe here have been studied intensively in cerebral cortex. There the cortical pyramid is postulated to show a resting potential gradient which extends downwards ° This research project was aided by grants from USPHS NB04513-01 and the Allen P. and Josephine B. Green Foundation, and the work reported was done in collaboration with Dr. James O'Leary.

215

216

BRAIN

FUNCTION

along the apical dendritic shaft from the superficial dendritic plexus to the origin of the axon from the soma. Some composite of the gradients along individual pyramids is presumed to result in a potential difference between surface and white matter. This, plus whatever slow potentials could arise in blood vessels or glial cells, produces a relatively steady baseline from which even prolonged changes in cortical activity can be measured trustworthily. We have called that baseline 'SP' or steady potential. That term only means that an optimally prepared experiment left undisturbed for two to three hours will show no more than a half-millivolt drift over the period. In the material to be reported here, we focus attention on the negative steady potential shifts which usher in seizure discharges of cerebral cortex. Polarity reference is to the electrode situated upon the cortical surface. In different experiments, the other electrode has been situated in subcortical white matter or in contact with nasal periosteum. The initiation of seizure by slow intravenous administration of caffeine, picrotoxin and metrazol was examined in cerebral cortex, and strychnine was similarly studied in cerebellum (34). We also evaluated negative shifts which coincide with seizure discharge activated by repetitive stimulation of either cerebral cortical surface (14, 22) or midline thalamus (13). Based on an analysis of direct cortical response developed hereafter, and upon results of intracellular microelectrode studies during seizure discharge reported by others (19, 23), we postulate a neuronal origin for the negative steady potential shift mentioned above. We view such shifts as reflecting a depolarization of the same loci of neuronal membrane as yield the seizure discharge. It is our belief that, in certain circumstances, such depolarization is a necessary prelude to initiation of paroxysmal activity. The first two figures deal with seizure activity in rabbits. The animal from which each record was obtained is identified on Figure 95. Recording electrodes were arranged transcortically, one placed on the cortical surface, the other in the white matter. In this figure and all subsequent similar ones, the straight horizontal line is the baseline from which potential deflections are read. The polarity refers to the surface electrode and down is negative. An abrupt negative shift of 1 to 2 mV is the first noticeable change in the D.C. electrocorticogram of an animal which has received a sufficient intravenous injection of caffeine sodio-benzoate, picrotoxin or metrazol to activate seizure discharge (Figure 95). For caffeine and metrazol this develops within seconds after the start of the injection, but in the case of picrotoxin the change is delayed for several minutes. Either immediately following the shift or within seconds thereafter, seizure activity also appears; as paroxysm continues, SP recedes towards the baseline and may even continue to shift positively. To show that in other parts of the central nervous system a negative SP shift may precede seizure discharge, the effect of threshold dosage of intravenous strychnine upon the cerebellum is described. The electrodes are again

SHIFTS

LEADING

TO S E I Z U R E

DISCHARGE

217

Caffeine (450mg.i.v. - rabbit)

A ^

Metrazol (isomg. i.v. - rabbit) ,

Metrazol (100mg. i.v. - monkey)

1 mv. I I *ec.

Figure 95. Negative SP shifts ushering in seizure activity induced by intravenous injection of caffeine, picrotoxin and metrazol. Straight line in this and similar subsequent figures is baseline from which potential deflections are read. Up is positive. In A, B and C the breaks in the traces are record omissions of 5, 0.5, and 8 seconds respectively.

arranged transcortically, one on the surface of the cerebellum, the other in the subcortical white matter. There, several seconds after giving 500 pg of the drug, the SP shift appears and grows in amplitude. At the peak of the slow voltage change a rhythmic comb-like discharge 30-50/sec. appears. SP remains shifted negatively while the comb-like discharge persists (Figure 96), but returns to the baseline thereafter. Negative SP shifts also usher in electrically induced seizure discharge. They have been shown to precede after-discharge evoked by repetitive stimulation of cortical surface (3, 10, 14) or midline thalamus (13). The next figure (No. 97) is a record obtained from the exposed human cortex during a neurosurgical procedure. The direct cortical response was recorded from the surface at the site where a repetitive (20/sec.) stimulus was applied. Such a stimulus led to after-discharge. In A, during stimulation, the primary (negative polarity) potentials which follow each succeeding shock diminish in amplitude progressively and then are replaced by corresponding positive polarity potentials. Concomitantly, a striking negative SP shift occurs. The stimulus continues into line B, the positive potentials growing

218

BRAIN

FUNCTION

A

B

1

500yV|

I sec.

Figure 96. Negative SP shift from which rhythmic discharge develops in cerebellum of rabbit following intravenous injection of 5 0 0 ;j.g strychnine. A: record prior to injection; B : 16 sec. after injection. B - D are continuous strips.

in amplitude and the steady potential shift remaining. Cessation of the stimulus (arrow) is followed first by a quiet period during which the SP shift remains evident. Then, upon recession of SP toward the pre-stimulatory baseline, bursts of after-discharge appear, each riding upon a positive plateau of short duration. Thus far the examples shown have been of paroxysm induced by convulsants injected intravenously as well as one evoked by electrical stimulation, the common feature for all being a negatively directed SP shift which ushers in the seizure activity. The next series of figures deals first with a description of the direct cortical response (DCR), and then with the bulbar pyramid response evoked simultaneously with the direct cortical one. Detailed inspection of DCR aids in evaluating the significance of the negative SP shift which introduces these different seizure activities. The response is a complex one and single and repetitive stimuli of graded intensity, as well as recordings on fast and slow time-lines, are necessary to record all components. A single weak stimulus in the squirrel monkey (Figure 98), for example, evokes a 20 msec, negative potential which is followed by a trailing positive one of much longer duration. We call this after-positivity (Figure 98A). Upon increasing the stimulus strength, the primary potential grows in amplitude but after-positivity diminishes and is replaced by an oppositelydirected slow potential, slow negativity. At a stimulus strength near threshold for slow negativity, a series of rhythmic waves at 10-12/sec. may follow

SHIFTS

LEADING

TO S E I Z U R E

DISCHARGE

219

primary potential. As the stimulus intensity is made still stronger, fast positive spikes (1-3 msec.) appear on the starting phase of the primary potential, and with a very strong stimulus one or more such spikes may even precede it (Figure 98D). Brazier: How constant is the interval between these repetitive phenomena? Is that a true "frequency" or do the intervals vary? Goldring: It is fairly regular. I would say it is as constant as that of the barbiturate spindle or the caudate spindle. Repetitive stimulation is especially conducive for displaying slow negativity. Figure 99 is a record obtained in man. Even at a stimulus intensity

\mmf J

500pV | 100 msec.

Figure 97. Human direct cortical response to stimulus sufficiently intense to evoke after-discharge; 2.6 seconds of record omitted between A and B. Arrow indicates end of stimulus period.

which still evokes after-positivity in the response to a single shock to the superior temporal gyrus, after-positivity is replaced by slow negativity during repetitive stimulation at 6/sec. Summation of slow negativity occurs, resulting in a negative SP shift which grows as stimulus frequency is increased (compare amplitude of summed slow negativity in the 6/sec. trace with that shown in the 20/sec. trace). The various DCR components described comprise the total response pattern that can be elicited by stimulation at a point upon the cortical surface immediately neighboring upon the position of the recording electrode. With the exception of the early positive spikes mentioned previously, DCR components have a superficial origin. The slow potential components as well as the primary negative potential are believed indicative of graded responses which are non-conducting or only conduct decrementally, and a dendritic origin has been suggested for the primary potential of DCR (4, 5). Such an interpretation is inferred from the superficial origin of the response in the upper 0.2 mm of cortical thickness, a stratum occupied predominantly by

BRAIN

220

FUNCTION

dendritic plexus. By contrast, the positive spikes are interpreted as all-ornone soma potentials (28). The next few figures are records obtained in the squirrel monkey. Stimulating electrodes were on surface of motor cortex, immediately adjacent to the recording one. In addition, a 20-50 n steel wire was situated within the bulbar pyramid, and led against another steel wire electrode on the adjacent dura. Upon stimulating the motor cortex, the pyramid and cortical responses could be recorded simultaneously. For those not familiar with the response in the pyramid, we follow the ter3 V 1

T

B

(

10 msec.

1 mV J

»

V

' V " V

/

rhythmic waves

1 mV J

t

t

100 msec.

Figure 98. Direct cortical responses to stimuli of increasing intensity, squirrel monkey. Left column shows responses on a time-line sufficiently fast to resolve the detail of the initial part of the response. Right column, responses recorded on slower time-line to demonstrate longer duration components. 1: primary negative potential; 3: after-positivity; 4: slow negativity. Second negative wave labelled 2 in previous communications (28) is not evident in these responses. A, responses to weakest stimulus; B-D, responses to stimuli of progressively stronger intensity. Left time marker is for responses in left column, the other for responses in right column. Larger voltage calibration signal is for responses A-C, smaller one for responses in D. (From Mingrino, Coxe, Katz, Goldring and O'Leary, 28.)

SHIFTS

LEADING

TO S E I Z U R E

DISCHARGE

221

minology of Patton and Amassian (30). Approximately 0.7 msec, following the cortical shock a positive deflection called 'D wave' appears. This in turn is trailed by a series of positive deflections called 'I waves'. The D wave or direct response is believed to be due to direct activation of axons; the I wave, or indirect response, to firing of intercortical neurons. Neither the cortical primary potential nor the still slower potentials which follow have been observed to be related to pyramidal activity; and the thresholds for the medullary pyramidal responses and for these cortical components also differ, being significantly higher for the former. By contrast, the more intense stimuli required to evoke the fast cortical spikes of Weak

Strong St imulus

Stimulus

fysec.

2(ySec.

O

fT

V Y —

500 >N J 100 asec. Figure 99. Human direct cortical response to single and repetitive stimuli: 1: Primary negative potential; 2: second negative wave; 3: after-positivity; 4: slow negativity. Stimulus strength in the 6 and 20,/sec. traces was the same as that used to obtain single response at top left. In 6/sec. trace observe that slow negativity replaces after-positivity during stimulation.

the DCR produce a medullary pyramid response. A response is never recorded from the medullary pyramid unless the fast spikes also appear in the cortical record. Then the number of I waves of the pyramidal response corresponds to the number of positive spikes in the direct cortical response. In A, B and C of Figure 100, observe that at stimulus strengths below threshold for the fast positive spikes of the cortical record, no activity is recorded from the medullary pyramid. When the stimulus intensity is at threshold for fast cortical spikes, the record from the medullary pyramid shows a small response (D). Also, at threshold, the fast cortical spikes are inconstant in the responses to successive stimuli, and when they fail to appear no response is detected in the medullary pyramid either (E). At

222

BRAIN

FUNCTION

stronger stimulus intensities the fast cortical spikes become more conspicuous; and the responses recorded from the medullary pyramid grow commensurately. Then the number of spikes in the pyramidal volley come to correspond to the number of DCR spikes (F, G, H). Another point of interest, and one that I should have mentioned earlier,

A

Figure 100. Simultaneously recorded direct cortical and pyramidal responses, squirrel monkey. Stimulus site is hand area of precentral motor cortex. Upper trace of each pair (A through H) is pyramidal record. In A, stimulus intensity is threshold for direct cortical response. Intensity is increased progressively in A through H with exception of E, which is same as D. Stimulus strength in latter was near threshold for DCR spikes. Note the absence of pyramidal response when stimulus fails to evoke DCR spikes (E), but appearance of it when stimulus is effective (D). Smaller D in H is direct response; /, indirect responses (after Patton and Amassian, 30). Short oblique lines over DCR in H identify the positive spikes. Voltage calibrations on left are for the cortical responses, those on right for pyramidal ones; the smaller one of each pair is for responses in H, the larger one for all others. Positive is up. (From Mingrino, Coxe, Katz, Goldring and O'Leary, 28.)

BJ —r

500 /LtV I

500 tN

I

250 pV

250 /iV

10 m s e c .

is the striking resemblance between the direct cortical response to strong stimulation and the cortical response to activation of a primary sensory projection system; both show a polyphasic positivity followed by a negative wave. Purpura: With the exception that the positive component of the primary response is longer in duration, I think that is a very important difference. Goldring: That is right. The positivity of the response from primary sensory cortex would have longer duration and, perhaps, a higher amplitude. How-

SHIFTS

LEADING

TO S E I Z U R E

DISCHARGE

223

ever, this varies from animal to animal and from species to species. It is much easier to display the polyphasic positivity in the squirrel monkey than in the cat. Brookhart: Is the apparent diminution in amplitude in D of Figure 100 a regular finding? Goldring: As the stimulus is made stronger, it will decrease the amplitude of the negative wave. If the stimulus is strong enough, it may obliterate the negativity. What is happening is that the positivity developing with the stronger stimuli is summing algebraically with the negative potential and cancelling it. Purpura: What is the shortest-latency discharge recorded in the pyramidal tract? Goldring: About 0.7 msec. Purpura: You infer from this that the negative wave is not associated with discharges of pyramidal cells and therefore it is necessary to stimulate with strong currents and activate cells directly in order to obtain the direct tract volley? Goldring: Yes. The pyramidal response appears only when positive spikes become evident in the direct cortical response. Brookhart: In D of Figure 100 the first sign of pyramidal activity appears to have a much longer latency than the direct wave in H. So, do you agree that the pyramidal activity in D must have been part of the so-called 'I wave', having arisen by indirect activation of cortical cells? Goldring: Yes. Yet another sign of activity may appear in the pyramid incident to stimulation of the cortical surface. This is shown in Figure 101 and is the 'unitary' or tonic firing which develops when cortical stimulation becomes sufficiently intense to occasion after-discharge. Such tonic activity appears in the pyramidal record when cortical SP has shifted sufficiently negative, and positive potentials have replaced the primary negative ones during stimulation. The tonic pyramidal activity also continues into the immediate post-stimulatory period. At that time the negative SP shift is the only demonstrable change from the pre-stimulatory trace. As after-discharge bursts appear, tonic activity remains in evidence. They go out of the trace nearly simultaneously. Rowland: Each deflection in the positive direction is accompanied by a clustering of unit activity. When the whole baseline goes further positive, does the unit activity stop? Goldring: Disappearance of unitary activity and recovery of the negative shift to the baseline occur approximately at the same time. It is also important that the cortical after-discharge induced by surface stimulation develops in the most superficial zone of the cortex. In fact, the after-discharge has virtually the same reversal point as the primary potential of the direct cortical response—about 0.2 mm beneath the cortical surface. The data presented thus far support the inference that a negative SP shift,

BRAIN

FUNCTION

such as precedes seizure discharge, can arise out of a membrane depolarization which develops in neural elements of the superficial plexus, probably in dendrites. When a sufficient level of depolarization is reached, our evidence suggests that membrane potential becomes unstable and oscillates, such oscillations (under the conditions of our experiments) being recorded as after-discharge. Similar oscillation of membrane potential has been shown

•«

s. rt c 5 ° «a) -S aC «« .s a « tí wî -G B3 •H 3 ï o " ß

„

¡™É!Í!Í!MP r" sse: SE"

•S ß 13 ~ 13 J) J Ï* 2 J2 t î û 0) -s « SS £g< O ^ DC I §

8 S "S g .3

fis ^ ^ Cd 3ZL ni—^ O "1 «to 1 TJ S,^ .2 s a> cd •B C«J Hl O^JJ cd > ^ O 4) 3 IM I I O > > £ £ J3 J '-g I % -SP S, OT) s S • >s ift.S C U T) 3 cJd 13 (M » T3 3 Q" £ i1 g 13 S I11 o 8 rS U bO a *3 13 u H tfie a„ vì C " J e .2 "g "" rt 3 3 - bo""" "fi C g Js « * J c fr o U "jH C/5 y, S

Ü

a

a a I ^ |

CD

O ^

U O h s ' " 1 ^ rt O J3 C m ^ Ü w CU u a fi '•a ^ fi "3 5 fi ^ a CD H . T3 S i i £ - 0 « o2 ^ N "O fi ^ y, %° |e ^ ^ t-H'^ « ca .t; _ cu a t) • S PQ o Ih T3 CD o a « fi T3 bO CU a c o uo ^ .S3 -fi o 3 T3 t/3 » efi m „ ^ a - •° « "3 fi O T3 o .2 W .. V C D O vi VI VI a _. ä ^ « « o > 05 •S y CD SP'« S f i n g i e S

? -iV-

>

¡5

Jt' ï;


£ o° 15 s i « o^ o •S uT " t-1 -w s z g -o . . . W "Ti g cuz a u o «w S &i IO 5" Ü C8 o « d .5 5




0 >2
>

03 TJ

tí * O ti M

o p -tí C> -O

til t> O

M

fi t 3 O o;

«

g x¡

E

is

I

a

H •