International Review of Neurobiology Volume 2

INTERNATIONAL REVIEW OF

Neurobiology VOLUME 2

This Page Intentionally Left Blank

INTERNATIONAL REVIEW OF

NeurobioIouv Edited by CARL C. PFEIFFER New Jersey Neuropsychiatric Institute Princeton, New Jersey

J O H N R. SMYTHIES The Maudsley Hospital, London, England

Associate Editors V. Amassian J. A. Bain D. Bovet Sir Russell Brain Sir John Eccles

VOLUME

E. V. Evarts H. J. Eysenck F. Georgi G. W. Harris R. G. Heath

C. Hebb A. Hoffer K. Killam S. MBrtens

2

@ 1960 ACADEMIC PRESS, New York and London

Copyright 0, 1960, by Academic Press Inc. ALL RIGHTS RESERVED

NO PART O F THIS BOOK MAY B E REPRODUCED IN ANY F O R M , BY PHOTOSTAT, MICROFILM, OR ANY OTHER MEANS, WITHOUT WRITTEN PERMISSION FROM THE PUBLISHERS.

ACADEMIC PRESS INC. 111 FIFTHAVENUE NEWYORK 3, N. Y. United Kingdom Edition

Published by ACADEMIC PRESS INC. (LONDON) LTD, 17 OLD QUEEN STREET,LONDONS.W. 1

Library of Congress Catalog Card Number 59-13822

PRINTED IN THE UNITED STATES O F AMERICA

CONTRIBUTORS ERMINIOCOSTA,Thudichum Psychiatric Research Laboratory, Galesburg State Research Hospital, Galesburg, Illinois EDUARDO EIDELBERG, Brain Research Program, University of California, Los Angeles, California H. J. EYSENCK, Institute of Psychiatry, University of London, London, England R. M. GAZE,Department of Physiology, University of Edinburgh, Edinburgh, Scotland LOWELL E. HOKIN,Department of Physiological Chemistry, University of Wisconsin, Madison, Wisconsin MABEL R. HOKIN,Department of Physiological Chemistry, University of Wisconsin, Madison, Wisconsin ALBERTHERZ,Department of Pharmacology, University of Munich, Munich, Germuny JORGE MARDONES, Institute of Pharmacology and Institute of Research on Alcoholism, University of Chile, Santiago, Chile EUGENEROBERTS,Department of Biochemistry, Medical Research Institute, City of Hope Medical Center, Duarte, California F. W. SCHUELER,Department of Pharmacology, Tulane University School of Medicine, New Orleans, Louisiana EDWARD J. WALASZEK, Department of Pharmacology, University of Kansas Medical Center, Kansas City, Kansas

V

This Page Intentionally Left Blank

PREFACE Progress in neurobiological research must maintain a delicate balance between the fascination of basic explanation of clinical and physiological phenomena by means of chemical and physical concepts on the one hand and the pressing needs for the development of new and effective treatments of disease on the other. Advances in basic biochemistry and biophysics often give rise to developments in the clinical field, but mature judgment is required to select from the vast detail of biochemistry and biophysics, those parts which are likely to apply to human disease. The aim of this review is to enable active workers in the fields of neurobiology, neurochemistry, neuroanatomy, neuropharmacology, neurophysiology, psychopharmacology, psychology, etc., as well as those in biological psychiatry and neurology to give an account of recent progress in their fields. The review covers the whole field of neurobiology and includes work within a particular basic science as well as in neurology and psychiatry. Particular emphasis has been laid on the recent development of ideas that are of fundamental importance and general interest and also those that are likely to further our understanding of nervous and mental disease. In the past the basic neurobiological sciences have played no little part in progress toward these ends. They are most active at present and they hold great promise for the future. These reviews and summaries ordinarily are by invitation, with a deadline for receipt of the manuscript by June 1. The editors, however, will be happy to review unsolicited manuscripts if submitted in outline form. CARLC. PFEIFFER JOHN R. SMYTHIES

May, 1960

vii

This Page Intentionally Left Blank

CONTENTS CONTRIBUTORS ................................................. PREFACE......................................................

V

vii

Regeneration of the Optic Nerve in Amphibia

R . M . GAZE I. I1. I11. IV . V. VI . VII . VIII .

Introduction ............................................ Visual Behavior in Amphibia .............................. Visual Organization in Amphibia .......................... Studies on Regeneration of the Optic Nerve . . . . . . . . . . . . . . . . . . Implications of Visual Recovery ............................ The Hypothesis of Neuronal Specificity ...................... Some Other Aspects of Form and Function . . . . . . . . . . . . . . . . . . Conclusions ............................................. References ..............................................

1 2 4 10 14 22 32 35 38

Experimentally Induced Changes in the Free Selection of Ethanol JORGE

I. I1. I11. IV. V. VI . VII . VIII . IX. X.

MARDONES

Introduction ............................................ General Procedure ....................................... Individual Variations ..................................... Effect of Diet ........................................... Influence of a “Third Choice” .............................. Effect of Previous Ingestion of Ethanol ...................... Endocrine Influences ..................................... Effect of Drugs ........................................ Significance of the Reviewed Facts on Human Alcoholism ...... Summary ............................................... References ..............................................

42 43 44 50 59 61 64 68 71 74

74

The Mechanism of Action of the Hemicholiniums

F. W . SCHUELER

I. I1. I11. IV . V.

Introduction ............................................ Origin and Development of the Hemicholiniums . . . . . . . . . . . . . . Pharmacological Actions of Hemicholinium No . 3 (HC-3) . . . . . . Studies on Other Hemicholiniums and Hemicholinium-like Agents Summary of Effects of HC-3 at Various Cholinergic Sites .... References .............................................. 1x

77 77 80 94

95 96

X

CONTENTS

The Role of Phosphatidic Acid and Phosphoinositide in Transmembrane Transport Elicited by Acetylcholine and Other Humoral Agents

LOWELL E. HOKINand MABELR. HOKIN I. Background on Phospholipid Metabolism in Neuronal Tissue . . . . 11. Effects of Acetylcholine on Phospholipid Metabolism . . . . . . . . . . 111. Phospholipid Turnover and Sodium Chloride Secretion in the Salt Gland of the Albatross . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV . Physiological Significance of the Phospholipid Effect in Brain Cortex Slices and in Sympathetic Ganglia . . . . . . . . . . . . . . . . . . . . V. Effects of Acetylcholine on the Incorporation of P32 from Various Precursors into Phosphatidic Acid in Cell-Free Preparations from Brain .................................................. v1. Enzymes Concerned in the Acetylcholine-Stimulated Exchange of Phosphate in Phosphatidic Acid . . . . . . . . . . . . . . . . . . . . . . . . . . VII. Explanation for the Discrepancy between the Stimulation of Glycerol-l-C14 Incorporation and P32 Incorporation into Phosphatides ................................................... VIII. The Endoplasmic Reticulum as the Site of Phosphatidic Acid and Phosphoinositide Turnover in the Salt Gland . . . . . . . . . . . . . . . . IX. Theoretical Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . X. Summary ............................................... References ..............................................

100 103 110

113

113

118

119 120 121 133 133

Brain Neurohormones and Cortical Epinephrine Pressor Responses As Affected by Schizophrenic Serum

EDWARD J. WALASZEK I. Introduction ............................................

11. The Cortical Epinephrine Pressor Response ( CEPR ) . . . . . . . . . . 111. Effect of Various Centrally Acting Drugs on the CEPR . . . . . . . . IV. Effect of Serum on the CEPR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Influence of Schizophrenic Serum upon the Content of Various Neurohumoral Substances in the Rabbit Hypothalamus . . . . . . . . VI. Correlation of the Inhibitory Effects of Schizophrenic Serum on the CEPR with the Catechol Amine Content of the Hypothalamus VII. Presence of an Abnormal Factor or Factors in Serum of Schizophrenics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII. Summary ............................................... References ..............................................

138 139

144 151 156 166

169 170 171

The Role of Serotonin in Neurobiology

ERMINIO COSTA I. Introduction ........................................... 11. Evidence for a Neurophysiological Role of Serotonin . . . . . . . . .

175 177

CONTENTS

.

111 IV . V. VI VII VIII .

. .

Biosynthesis of Serotonin .................................. Mechanism of Action of Neurotransmitters . . . . . . . . . . . . . . . . . . Mechanism of Serotonin Action at Cellular Level ............ Methods for Serotonin Bioassay ............................ Importance of Serotonin in Psychopharmacology .............. Summary ............................................... References ..............................................

xi 185 190 193 205 207 219 221

Drugs and the Conditioned Avoidance Response

ALBERT HERZ I . Introduction ............................................ 11. Methods for Studying the Conditioned Avoidance Response (CAR) ................................................ 111. Tranquilizing Agents ..................................... 1V. Hypnotics and Sedatives .................................. V . Morphine and Other Analgesics ............................ VI . Adrenergic and Adrenergic-Blocking Substances . . . . . . . . . . . . . . VII . Cholinergic and Anticholinergic Drugs ...................... VIII . Serotonin and Monoamine Oxidase Inhibitors ................ IX . Psychotomimetic Drugs ................................... X . Miscellaneous ........................................... XI . General Discussion ....................................... XI1 . Conclusions ............................................. References ..............................................

229 230 234 250 252 254 256 260 262 264 265 272 272

Metabolic and Neurophysioiogical Roles of y-Aminobutyric

Acid EUGENEROBERTS and EDUARDO EIDELBERG I . Introduction ............................................ I1. Some Aspects of Distribution of Free Amino Acids in Nervous Tissue ................................................. I11. Detection and Quantitative Determination of y-Aminobutyric Acid (GABA) in Tissue Extracts .......................... 1v. Distribution of GABA in Animal Tissues .................... V . Metabolic Relations of Some of the Free Amino Acids in the Central Nervous System (CNS) .............................. VI . Metabolic Relations of GABA in the CNS .................... VII . GABA as a Possible Precursor of Other Substances .......... VIII . Developmental Features of Glutamic Acid Decarboxylase ( GAD ) and GABA ............................................. 1X. Factors Which Might be Involved in Regulating the Levels of GABA in the CNS ....................................... X . Lack of Correlation of GABA Levels and Seizure Susceptibility in Areas of Normal Brain .................................. XI . Physiological Mechanism of Action of GABA at the Cellular Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

280 281 281 283 285 288 295 295 300 318 319

xii

CONTENTS

XI1. Pharmacological Studies with GABA ........................ XI11. General Comments and Summary ......................... References ..............................................

Objective Psychological Tests and the Assessment Effects

322 326 327

of Drug

H . J . EYSENCK I. I1. I11. IV. V. VI. VII .

VIII. IX .

Introduction ............................................ A Critique of Current Methodology ........................ A Theory of Depressant and Stimulant Drug Action . . . . . . . . . . Experimental Tests in the Assessment of Drug Effects . . . . . . . . A Second Paradigm for the Testing of Psychopharmacological Hypotheses ............................................. A Mathematical Model of a Psychopharmacological Hypothesis The Place of Animal Work in a Hypothetico-deductive System A General Dimensional System of Psychopharmacology . . . . . . . . Summary ............................................... References ..............................................

AUTHORINDEX SUBJECTINDEX

................................................ ................................................

333 334 342 348 366 370 375 379 382 383 385 399

REGENERATION OF THE OPTIC NERVE IN AMPHlBlA By R. M. Gaze Department of Physiology, University of Edinburgh, Scotland

I. Introduction ........................................... 11. Visual Behavior in Amphibia ............................ 111. Visual Organization in Amphibia .......................... IV. Studies on Regeneration of the Optic Nerve . . . . . . . . . . . . . . . . V. Implications of Visual Recovery .......................... VI. The Hypothesis of Neuronal Specificity .................... VII. Some Other Aspects of Form and Function . . . . . . . . . . . . . . . . VIII. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.

1 2 4

10 14 22 32

35 38

Introduction

In this review I intend to discuss regeneration of the optic nerve with particular reference to the relationship between structure and function in the visual system. To make the story of regeneration more intelligible and to provide a background for various speculations to be made later, I shall start with a description of some aspects of visual behavior in amphibians and an outline of the known facts of visual organization in these animals; then will follow a review of the phenomena of regeneration and lastly a discussion of some of the problems brought out by the experiments and their possible significance for the understanding of normal vision. The regenerative powers of lower vertebrates are well known; these animals have an almost proverbial ability to develop new tails and limbs after loss of the originals. This ability extends to the central nervous system as well in urodele amphibia (newts and salamanders) and anuran amphibia (frogs and toads). Whereas in mammals regeneration of cut or otherwise damaged axons within the central nervous system occurs to a very limited extent, central regeneration takes place freely in the amphibia, and particularly 1

2

R. M. GAZE

well in the larval or young adult stages. The optic nerve is embryologically part of the central nervous system, and its fibers will grow again following section in these animals; and when the eye has become reconnected to the optic lobe, the animals redevelop what appears to be normal vision. It is possible therefore to follow what happens to the fibers of the nerve as they grow back, and to attempt to correlate anatomical restitution with return of function.

II.

Visual Behavior in Amphibia

Most amphibia are not truly amphibious in the sense that they are equally at home in the water or on the land. In'general, the larval stage of life is spent in the water, when the animals are truly aquatic, and the adult stage on land, with occasional return to water, for instance at the breeding season. Since no amphibian has more than 5 diopters of accommodation (Walls, 1942), if we consider frogs to be emmetropic in air, they will be strongly hypermetropic under water. In a similar way, the aquatic newts which are emmetropic in water will become strongly myopic on land. The eyes of frogs are relatively enormous in comparison with their body size and, as might be expected from their arrhythmic habit of life (most amphibia are nocturnal in habit) and the size of their eyes, they are largely visual animals. The urodeles are more secretive in their life habits and in keeping with this their eyes are not so well developed as those of the frogs. On land, frogs move by jumping, a mode of progression which involves successive periods when the visual field is stationary and when it is in violent motion. The animals appear to respond only to moving objects and the nature of the response seems to vary with the size of the stimulus. For instance a man may stand in shallow water surrounded by hundreds of frogs but unless he moves the animals pay no attention. The slightest movement on the part of the observer will lead to a sudden change in the behavior of the animals-they will cease croaking and sit still; any further movement of the observer will usually cause a sudden and general disappearance of frogs from the scene-mostly into the water. If the stimulus moving in the visual field is small, of the order of size of a fly or other small insect then the response of the animal is attack

OPTIC NERVE REGENERATION IN AMPHIBIA

3

rather than avoidance. The sequence of events during the capture of prey is of special interest since it is this reaction which is commonly used by investigators to determine the adequacy of visual function after optic nerve regeneration. In some animals, notably tree frogs, food may be captured with a single leap; Sperry (1944) reported that Hylu cineria frequently captured with a single jump houseflies walking at a distance of 35 cm. I have watched Raw tempmaria and R a w esculenta feeding on meal worms in the laboratory, and here the sequence is rather different. If the worm is some distance away from the animal, the frog will turn to face it, jump the intervening distance, then bend its head and shoulders down, and finally flick out its tongue to grasp the food. If the worm is close to the animal, the latter first turns to face the prey, bends down to “peer” at it, then captures it. This lowering of the head just before the final capture is made has the effect of bringing the object right into the middle of the binocular field. Thus there appears to be a clear difference between the effects of a small moving stimulus in the posterior field and one in the anterior field. In the former case the stimulus sets up an orientation response, or “taxis” in Tinbergen’s terminology (1951); in the anterior, binocular, field the same moving stimulus may eventually release the capturing behavior. Thus there is ample evidence that frogs respond to gross differences in size of a visual stimulus and that they respond appropriately to variations in retinal local sign. It is doubtful, however, whether they are able to distinguish shapes and whether they have anything like color vision. Both these abilities are attributed to the frogs by Noble (1931) but more recently the evidence for color vision has been reassessed by Walls (1942) who concludes that for amphibia it is not proven. These animals are very difficult to train in any way and this may account for the difficulty in demonstrating such discriminations. It has been reported, however, that some anurans can be trained to distinguish between various shapes (Biederman, 1927). The main factors in the visual environment, then, to which frogs respond are movement and direction. The ability of many frogs to catch insects in flight argues the existence of considerabIe computing ability within the visual system of these animals, since to be effective a strike must be made at some estimated future position of the prey.

4

R. M. GAZE

Another aspect of normal (experimental) visual behavior of these animals is the optokinetic response; and this is frequently used to test the return of visual function following regeneration of the optic nerve. If a normal animal is placed on a stationary platform and a striped drum is rotated round it, the animal will turn its head in the direction of the moving stripes. This is a visual response and in the absence of the eyes or if both optic nerves are destroyed the response does not occur. If one eye is removed, or one optic nerve is destroyed, then in general an animal will only respond in one direction, that is, to rotation of the drum temporonasally across the intact eye. Summary. Little experimental work has been done on visual behavior in amphibians; what there is shows that the animals respond to gross differences in size of stimulus, to movement, and to position in the visual field. Ill. Visual Organization in Amphibia

In urodeles and in anurans eye movements apart from retraction are negligible. Orientation of the eyes is controlled by movements of the head; accommodation is effected by forward movement of the lens. Reports on the accommodative state of the eye in these animals are very conflicting. According to estimates reported by Walls ( 1942), Duke-Elder ( 1958), and Rochon-Duvigneaud (1943), the frogs eye has been found to be anything from 5 diopters hypermetropic to 8 diopters myopic. Perhaps for the purpose of this review we are justified in assuming with Walls that the frog is emmetropic in air and also that the large eye, with its beautifully clear cornea (the main refracting surface) and clear media is capable of projecting a good image onto the retina. Such an assumption would certainly be borne out by the excellence of the visual behavior of these animals. In the frog the ratio of the number of rods per unit area to the number of cones is approximately 2:1 and this ratio is much the same over almost the entire extent of the retina (Rochon-Duvigneaud, 1943). The frog retina has no fovea or macula acuta but it does have a very poorly defined area centralis; otherwise the entire extent of the retina is fairly uniform. It has been stated (Walls,

OPTIC NERVE REGENERATION I N AMPHIBIA

5

1942) that the slight excess of ganglion cells in the region of the area centralis has the effect of making this an area of acute vision; and while the first maps of the retinal projection to the optic lobe in the frog failed to show any significant difference in resolving power between the various parts of the retina (Gaze, 1958a), more recently it has been found that there is actually a difference of about 2: 1 in “magnification factor” between peripheral retina and area centralis (Jacobson, personal communication). One would expect the posterior region of the retina to be more specialized than the anterior, since it is onto posterior retina that the anterior, binocular, field projects. So far there is nothing to support this, except perhaps the observations of Barlow (1953a) that off units are more common in the posterior retina than elsewhere. Four types of visual element are commonly found in the retinae of frogs: red and green rods and single and double cones. Essentially the same types are to be seen in the urodele retina, only here the individual elements are both larger and sparser. There is also a higher visualto-ganglion cell ratio than in frogs. The optic nerve arises from the layer of retinal ganglion cells and passes backward from the eye across the optic chiasma to reach the optic lobe on the other side. Some fibers are also apparently given off before the lobe is reached, to the geniculate region and various other parts of the midbrain. All authors who have described the optic chiasma of the frog and salamander state that it is totally crossed-fibers from the one eye all go to the other side of the brain. Breusch and Arey (1942) made counts of the optic nerve fibers in various animals including some amphibia; they found that in Bufo amricanus the optic nerve contained 5300 unmyelinated fibers and 10,200 myelinated ones, while the nerve of Rana pipkns showed 13,700 unmyelinated fibers to 15,300 myelinated ones. More recently, work with the electron microscope has shown that these figures for the unmyelinated fibers in anurans are far too small; Maturana (1959) has reported finding some 470,000 of these in the optic nerve of the frog and about 320,000 in the toad. In urodeles the count is considerably smaller; Herrick (1941) estimated that the optic nerve of adult Amblystonuz tigrinurn contained a total of about 8000 fibers. Bishop, in 1933, found evidence of three distinct groups of fibers in the optic nerve of the bullfrog. The fastest conducting group showed velocities of 8-16 meters per

6

R. M. GAZE

second (mps) while the second group conducted at 2-5 mps. On the analogy of somatic sensory nerves one is tempted to look for differential functions in the different groups of optic nerve fibers, but so far there is no evidence about the physiological significance of the different conduction velocities. The suggestion made by Herrick in 1941 that the faster conducting fibers (Arnblystoma) might be a form of activating mechanism for the visual brain while the slower conducting fibers might be a localized projection from the retina is interesting, but remains speculation on1y.l Most of the fibers in the optic nerve are, of course, afferent, but there are also some efferent fibers running from the optic lobes to the retina, the function of which is uncertain. The efferent fibers play a part in the control of retinal pigment migration in fishes ( Arey, 1916), but this is more doubtful in amphibia. Nor has it yet been shown that these fibers exercise control over the retinal response to light, as appears to be the case in cats (Granit, 1955). The main visual centers in amphibia are the optic lobes. These are paired, ovoid structures situated between the diencephalon and the cerebellum; the optic lobes are the most conspicuous parts of the brain in the frog, being enveloped in pigmented membranes. The roof of the optic lobe shows marked lamination, especially so in the anura. If we adopt the terminology used by Larsell (1931) and Kollros (1953) there are nine layers, numbered outward from the ventricle to the surface; this layered structure is complex to a degree that allows comparison with the mammalian cerebral cortex. The outermost or ninth layer is called the stratum opticum or opticus layer and it is here that most of the optic tract fibers run. According to Wlassak (1893) some of the optic fibers also run in the seventh or deep medullary layer, and some, the basal bundle, run more ventrally still to end in Herrick's nucleus opticus tegmenti, near the nucleus of the third nerve. Those optic fibers which ramify in the roof of the tectum, however, do so mostly in the superficial or opticus layer. The optic fibers turn ventrally to terminate in or around layers 8 and 7 (Kollros, 1953). Apart from the axons of the optic tract, the opticus layer contains axons and dendrites from cells in the deeper layers; there are also cells in the opticus layer whose processes run for some distance within this layer (Larsell, 1931). 1

But see Lettvin et

d. (1959).

OPTIC NERVE REGENERATION IN AMPHIBIA

7

In early embryonic life the optic lobe consists of a mass of undifferentiated cells. With the arrival of the first retinal axons, cells become detached from the periventricular gray and migrate outward to form the beginnings of the various layers (Kollros, 1953). This differentiation of the layered structure of the lobe begins at the anterior pole, where the optic fibers first arrive, and spreads gradually posteriorly. In the absence of optic nerve fibers, as for instance when the opposite eye is remaved in an early stage, the lobe fails to develop properly. The hypoplasia is most marked in the outer layers (56% in layers 7-9) but still appreciable in the inner layers (38% in layers 1-6; Kollros, 1953). This diminution in the size of a deprived lobe is due to absence of optic fibers, reduction in the number of cells in the affected layers, and poor development of some of the remaining cells. The optic nerve fibers from the retinal ganglion cells are not distributed to the opticus layer in a random manner. There is an orderly topographical projection from retina to optic lobe, such that each area of retina is specifically connected to a corresponding region of lobe surface. In Triturus this orderly projection has been found by Stroer (1940) using anatomical techniques; and Sperry (1944), by ablating parts of the lobe, has shown in the frog that anterior, posterior, and inferior parts of the retina each project to separate areas of lobe surface. The representation of the retina on the lobe of the frog has also been investigated electrophysiologically by Gaze (1958a), who showed, by stimulating the intact eye with a small light and recording the arrival of action potentials at the optic lobe, that there is a point to point (or rather area to area) connection between retina and lobe. All these investigations point to the same conclusion, which is that the posterior retina (anterior visual field) projects to the anterior pole of the lobe, the anterior retina (posterior visual field) to the posterior pole, with inferior retina (superior field) near the mid-line and superior retina (inferior field) tucked round the lateral edge of the lobe. Frogs have a considerable degree of overlap of the visual fields anteriorly, and by means of the evoked potential technique it has been shown (Gaze, 1958b) that, within the binocular field, a point in space is represented binocularly on each optic lobe. Since the chiasma is reported as being completely crossed in the amphibia, it was to be expected that each eye would project only to the con-

8

R. M. GAZE

tralateral lobe; however, that part of the retina which receives stimuli from the binocular field sends impulses to both optic lobes. It is not known so far what paths these impulses take. The electrical activity of the optic lobe in response to stimulation of the eye by light has been investigated by various workers. Buser (1949a) recorded from the frog lobe a series of potential changes after a flash of light to the eye: the fast initial deflections (with a latency of 60 msec and lasting 80-100 msec) he attributed to activity in the optic fibers themselves, and the following slow waves (latency 250 msec or more; duration 200 msec) he thought due to the activity of the postsynaptic neurons of the tectum. These two components, fast and slow, could be dissociated to some extent experimentally; application of pentobarbital to the surface of the lobe rapidly abolished the slow waves while the fast spikes were more resistant to this treatment. Somewhat similar fast and slow components have been recorded from the tectum of fish following electrical stimulation of the optic nerve; in this case (Buser, 1949b) the latency of the responses was, of course, very much shorter, and high-frequency stimulation caused the slow waves to decrease markedly while the rapid spikes were unaffected. As the recording electrode was moved from tectal surface to the incoming optic tract the slow wave disappeared leaving only the fast component (Buser, 1950). Responses from the optic lobe to stimulation of the eye by light have also been recorded by Gaze (1958a, 1959); in these experiments a small light (subtending less than 1" at the eye) was mainly used, and the recording system had a time constant of less than 10 msec; the responses obtained consisted in most cases of bursts of action potentials from groups of units rather than slow wave potentials. The latency of the responses was enormousof the order of 100-200 msec-and this is at present difficult to account for. The conduction distances involved, assuming that the impulses travel by the shortest route from eye to lobe, amount to only about a centimeter. The fastest fibers in the frog's optic nerve conduct at approximately 8-16 mps (Bishop, 1933) which would give a minimum conduction time from retina to lobe of less than a millisecond. Even assuming (what may well be true) that the majority of the optic fibers conduct at less than 2 mps, this would still only account for a delay of 10 msec or so. A considerable amount of the latency of the responses to small light flash must

OPTIC NERVE REGENERATION I N AMPHIBIA

9

presumably be due to intraretinal delay. Hartline (1938) recorded the activity of single fibers near the head of the optic nerve in the frog’s eye; his published illustrations show “on latencies” ranging from about 50 msec to 220 msec, depending on the intensity of light and size of stimulating spot on the retina. In these experiments Hartline used an excised eye preparation; with a stimulating light spot of 0.1mm diameter, the latency of the optic nerve fiber response was not reduced below about 80 msec even when the intensity was some 5 log units above threshold. In the experiments of Gaze (1958a, 1959) the animal’s eye was in situ and intact and its dioptric apparatus was used to transmit the light to the retina; with a perimeter of radius 330mm and a perimetric stimulating light spot of 10 mm diameter (the largest used), the size of the image on the retina was probably of the order of 0.1 mm diameter. If this is so, then the stimulus situations are to some extent comparable, and the long latency of the optic lobe responses may be attributed largely to intraretinal delay. The unit responses that have been investigated by Hartline (1935, 1938, 1940a, b ) and Barlow (1953a, b ) from the retina were “on units,” “off units,” and “on-off units.” Andrew (1955), recording from the optic lobe, found also the presence of units which responded to movement of the stimulus in the visual field; the movement response could not be accounted for by summation of on and off effects. The optic lobe in the lower vertebrates is not merely a visual receiving center. According to many investigators it is also the main sensorimotor coordinating area of the brain. In the anura, electrical stimulation of cerebrum, diencephalon, or cerebellum gives little observable motor effect, provided the stimulation is kept localized. StimuIation of the optic lobes, however, always produces motor effects. Abbie and Adey (1950) found that there was considerable variation in the response to stimulating various parts of the surface of the optic lobe (Hyla and Bufo) but that the overall tendency was for the anterior lobe to give head movementscontraction of the retractor bulbi-while the lateral edge of the lobe gave movements of the forelimbs and the posterior part of the lobe gave trunk and hind-limb movements; these effects were seen in lightly anesthetized animals. This evidence for a crude motor localization on the optic lobe is difficult to interpret as it stands.

10

R. M. GAZE

It would be of great interest if further analysis of these motor effects revealed something comparable to the “visual grasp reflex” described in trout by Akert (1949); here, stimulation of various parts of the lobe surface caused coordinated movements of the eyes (and body musculature) which appeared to have the effect of directing the animal toward the appropriate part of its visual field. The experiments of Apter (1946) on cats showed that comparable reactions are obtainable from localized stimulation of the superior colliculus, which is homologous with the optic lobe in the lower vertebrates. Summary. The optic nerve in anura contains many more fibers than it does in urodeles; in both there is apparently a total crossing at the chiasma although an ipsilateral visual projection has been demonstrated in anurans. The laminated structure of the tectal roof is more marked in anurans than in urodeles; in both, the differentiation of the structure during first development proceeds from before backward. There is a localized projection of retinal fibers onto the surface of the tectum. The electrical response of the tectal roof to stimulation of the eye by light shows two components, fast and slow; there is reason to believe that the fast component indicates activity of the optic fibers themselves while the slow waves are due to activity of postsynaptic elements in the lobe.

IV. Studies on Regeneration of the Optic Nerve The starting point of modem investigations of optic nerve regeneration may be said to be the work of Matthey (1925, 1927). This author (1927) also gives a historical summary of much of the earlier work in this field. Using adult urodeles (Triton cristatus) Matthey resected 1-2mm of optic nerve on both sides, just distal to the optic chiasma, in thirty-seven animals. The mortality following the intracranial operation was high, and he was eventually left with only five animals. In these he found definite signs of return of vision after a period varying from 3-5 months. The animals were tested when swimming in a jar of water; they consistently attacked a worm which was lowered, in a separate glass container, into their water. Attack movements were noted when the worm was brought within about 10cm of the animals. It was

’

OPTIC NERVE REGENERATION IN AMPHIBIA

11

also shown that this response was not due to vibration, when later the animals tried to attack a finger held above the surface of the water; they would also follow a worm, used as a lure, when it was moved round the outside of their jar. Only three of these animals were afterward examined histologically, and of these two were stained with hematoxylin. However, in all five cases the optic nerves could be seen making their way back to, and apparently joining up with, the brain. These observations of Matthey have since been confirmed by numerous reports. Stone and Chace (1941) found good regeneration following simple section of the nerve in Triturus uiridescens; in these experiments, vision returned in some cases within 63 days. Sperry (1943b), using adult T.viridescens, cut the left optic nerve through the roof of the mouth in twelve animals which he then tested for visual recovery (after removal of the right normal eye) beginning 10 days after operation. In most cases vision was found to return some 30 days after operation, and the animals then made accurate localizing reactions involving the eye. Optokinetic responses also returned to movement in the one direction (temporonasally across the remaining eye). The experiments mentioned so far were on urodeles, which have markedly greater regenerative abilities than the anura. In 1944 however, Sperry managed to obtain regeneration following simple section of the optic nerve in various anurans (Bufo terrestris, Hyla cinerea, H . crucifer, H . squirella, R a m clamitam, and R. pipiens), by careful attention to technique of operation. This was a major advance in the study of visual recovery, since vision in the anura, especially the tree frogs, is greatly superior to that in urodeles. Following section of the nerve in tadpoles, optokinetic reactions were used to detect visual function; recovery was seen after periods of 11-23 days. When the normal eye was still intact, the responses of the tadpoles after regeneration of the nerve were identical to those in normal animals. When the normal eye had been removed, optokinetic reactions could be obtained strongly in one directiontemporonasally across the operated ( regenerated nerve) eye-and only weakly in the other direction. After section of the nerve in the adult animals, vision was tested by means of a small lure-a housefly-impaled on a wire. This was moved at about 15cm from the animals and the frogs with regenerated nerves were able to lo-

12

R. M. GAZE

calize the stimulus accurately. In some animals, in fact, attempts to approach the lure were made from a distance of 40cm. Sperry remarks that, while no exhaustive tests were made of the acuity of these operated animals, the crude tests used showed results no different from normal. Visual recovery in these adult anurans took on the average 25 days, the extremes being 2 1 3 3 days. Return of visual function, as tested by a lure, was obtained following section of the optic nerve in Bufo bufo (Maturana, 1958); the time taken for visual recovery in this case was about 50 days, but even after as long as 150 days recovery could still occur. Regeneration of the optic nerve may also take place in certain teleost fishes. Using Bathygobius soporator, Thlassoma bifmciatum, Halichoeres biwittntus, Abudefduf saxatilk, and A. amlagous, Sperry (1948) cut both optic nerves and tested visual function by observing optokinetic responses, avoidance of obstacles while swimming at speed, escape reactions, and positive localizing reactions. In the smaller fish vision returned sooner than in the larger ones. In small B. soporator and H . biuittatus, reactions to startle and optokinetic reactions returned by 6-7 days after operation. By 20 days all the fish had recovered vision except the A. amlugous, which recovered by 38 days. After recovery the visual behavior in these animals was similar to that in the normal fish. Even more remarkable than the evidence for optic nerve regeneration mentioned above, is the fact that in many cases the retina itself may regenerate and so give rise to an entirely new optic nerve. In some amphibia the eye can be autografted (completely removed from the orbit and then reimplanted or transplanted to the other orbit) or it can be homografted (the eye from one animal transplanted into the orbit of another of the same species) and in both cases the eye may recover visual function. In urodeles it has even been shown (Stone, 1930) that if the eye is heteroplastically transplanted from one species to another closely related species, the graft will take and regeneration of the neural retina and optic nerve will be followed by return of vision. Matthey (1926, 1927) used the urodele Triton crktatus for experiments in which he exchanged left eyes between pairs of animals. From a total of eighty-four operated animals he was only able to keep four, and three of these showed return of vision in the transplanted eye when tested after removal of the normal right

OPTIC NERVE REGENERATION I N AMPHIBIA

13

eye. The visual recovery was demonstrated in one case 8 months after operation, in the other two 10 months after. Autoplastic, homoplastic, and heteroplastic grafting of urodele eyes have all been studied in great detail by Stone and his associates (Stone and Ussher, 1927; Beers, 1929; Stone, 1930; Stone et nl., 1937; Stone and Zaur, 1940; Stone and Chace, 1941; Stone and Farthing, 1942; Stone, 1944, 1950a, b, 1953), with attention given mostly to the sequence of histological events following the operation and to the general facts of visuaI recovery, rather than to the quality of the recovered vision. When right eyes were exchanged between larvae of Amblystom punctutum and A. tigrinurn (Stone, 1930) the stump of the optic nerve connected to the brain underwent resorption until 10 days postoperatively when nothing could be seen of it; by 7 days postoperatively the nerve stump connected to the eye had begun to thicken with new growth and by 14 days a slender nerve had grown back to the chiasma. Return of visual function was proved in these animals, and although the time taken for recovery was not noted, in one case visual function was demonstrated in the transplanted eye while the host was still in the larval stage. Following excision and reimplantation ( or transplantation ) of an eye in the adults, severe degenerative changes take place. Stone and Zaur (1940) reimplanted and transplanted the eyes in adult TTiturus viridescens. For the first 3 weeks after the operation the retina was seen to disintegrate. About 20 days after operation the residual undestroyed cells in the ciliary margin of the retina began to show rapid mitosis, and from these cells the new retina was formed, growing back toward the fundus. The regenerating optic nerve was usually connected with the brain before the end of the second month and during the third month vision began to return. In those cases where the retina degenerated completely after operation, it was regenerated from the surviving retinal pigment epithelium, as shown by Stone (1950b). Summary. The optic nerve will regenerate after section in urodeles, anura, and certain fishes. In urodeles the eye may be grafted autoplastically, homoplastically, or even heteroplastically with success; in such grafts the retina usually degenerates and regenerates after a while, giving rise to a new optic nerve. In all these cases there is return of apparently normal spatial vision and movement response.

14

R. M. GAZE

V. Implications of Visual Recovery The ability of the retina and optic nerve to regenerate, as illustrated in the experiments already mentioned, is of interest as a major difference between amphibia and higher vertebrates; what makes optic nerve regeneration uniquely worth studying, however, is the fact that after regeneration has occurred normal vision appears to be restored. While no one has so far investigated form perception in animals with regenerated nerves, it is obvious that two of the main attributes of amphibian vision are recovered to a large extent: visual local sign, or the ability to respond appropriately to stimuli in different parts of the visual field, and movement perception. In normal animals, the ability to strike accurately at an object anywhere in the visual field requires that an image formed on any part of the retina shall have a central effect different from that caused by the image on any other part of the retina. This differential central effect of stimulation of different parts of the retina, or retinal local sign, is associated in normal animals with the orderly projection of fibers from the different parts of the retina to the appropriate regions of the optic lobe surface. Retinal local sign is signaled to the brain mainly by spatial (and temporal) patterning of impulse arrival at the lobe surface, Excitation of the anterior retina (posterior visual field) activates fibers which run mainly to the posterior pole of the lobe; excitation of posterior retina (anterior visual field) sends impulses mainly to the anterior pole, and so forth. And it is reasonable to suppose that in the optic lobe, the main visuomotor coordination center of the amphibian brain, there exists a form of organized motor localization such that arrival of an appropriately shaped parcel of impulses at a point anteriorly on the lobe will initiate a forward strike, while arrival of such impulses at the posterior pole of the lobe will initiate a turning reaction and so on. This supposition is supported by the small amount of evidence existing on the motor activities of the lobe, notably the work of Abbie and Adey (1950: anura), Akert (1949: fish), and Apter (1946: cat). The situation can perhaps be summed up briefly by the statement that for an animal to be able to display differentiated motor responses directed toward the various parts of its visual field,

OPTIC NERVE REGENERATION IN AMPHIBIA

15

stimulation of any effective part of the retina must set up a unique pattern of excitation in the visual centers. It has been shown experimentally in many animals that this uniqueness, or retinal local sign, is achieved in the simplest possible way: by reduplication of the retinal surface on the surface of the visual brain. Often there is considerable distortion of the representation of the retina on the brain, particularly in foveate animals; but in all cases examined so far this representation has been found. It thus appears that an orderly retinal projection to the visual brain is the main neurological factor in visual local sign. What then is to be made of the results of the regeneration experiments on the optic nerve in the lower vertebrates? In these animals visual local sign is restored after regeneration; each part of the retina regains its unique central effect. The visual field of an animal with regenerated optic nerves appears to be properly organized. The animal’s “visual space” is correctly arranged and the evidence shows that the relationship between different points in the visual field is also correct. So far as we can tell, these animals have normal spatial vision. The simplest explanation that can be put forward to account for the visual recovery found after optic nerve regeneration is that all the fibers in the nerve have grown back to their original positions on the optic lobes and have there re-established their original synaptic connections. If this is what happens, then the return of normal vision would be accounted for. This hypothesis, however, brings many problems of a difficult nature; it seems almost certain that regenerating nerve fibers are not guided to their destination in any systematic way (Sperry, 1945a; Weiss, 1955), but that where they end up depends on the statistical sum of many factors. This aspect of the problem will be considered further later on. Another possibility is that the fibers grow back in a disorderly fashion, ending up higgledy-piggledy on the lobe surface. In this case, to account for the return of retinal local sign, one must assume that there occurs a massive central reallocation of synapses, so that fibers from each specific part of the retina can still be routed through to their appropriate motor regions despite the abnormal retinal projection to the lobe surface; otherwise one would have to assume that information about retinal local sign was transmitted to the optic lobe in a totally chaotic manner, and in view of the fact

16

R. M. GAZE

that an organized retinal projection exists in all normal animals so far investigated, this seems unlikely. The reallocation of central synapses required by this theory would be a form of functional adjustment on the part of the central nervous system; the animal would, as it were, “relearn” its visual field. This hypothesis would necessitate considerable structural plasticity within the nervous system, an ability to form and reform synaptic connections according to functional requirements. In an attempt to find out how much plasticity there was in an amphibian central nervous system, Sperry (1943a) rotated one or both eyes in a series of newts (Triturus viridescens), leaving the optic nerve intact. The object of the experiments was to see whether some adjustment on the part of the central nervous system could bring about reintegration of the normal visuomotor coordination following the operative upset in the relationship between eye and brain. Normally, these animals show quite creditable visually directed behavior; they will turn and swim from a distance of at least 25cm to a lure oscillating within a space of 1.5cm outside the aquarium. They also show optokinetic reactions to a rotating visual field. If one eye is absent, the animals will only give this response in one direction, to movement of the field temporonasally across the remaining eye. Sperry found that after 180” rotation of both eyes about the optic axes, the optokinetic reactions of the animals were reversed; if the visual field was rotated round the animals in a clockwise direction the animals turned their heads counterclockwise and vice versa. If one eye was rotated and the other absent, the optokinetic reaction was only obtained in one direction, and this was the reverse of normal. If one eye was rotated and the other normal, the animal gave no optokinetic reaction at all. Visual localizing reactions mediated through the rotated eye were also inverted; a lure presented in the superior visual field would initiate attack directed at the appropriate part of the inferior field and so on. Some of the operated animals were kept under observation for 4% months and during this time there was no evidence of any correction of these maladaptive responses. When, in six cases, the eyes were eventually restored to the normal position, the animals at once showed normal visuomotor reactions. These experiments thus revealed no evidence for functional plasticity in the newt’s visual system; the central nervous system

OPTIC NERVE REGENERATION IN AMPHIBIA

17

appeared to be unable to readjust to the altered eye-lobe relationship. The next step was to see what happened if the optic nerve were cut and the eye rotated; under these conditions the breakdown of the previously existing synaptic connections in the optic lobe, and the eventual establishment of new ones, might give the central nervous system a better chance to effect a functional readjustment if the regenerating fibers grew back in a disorganized, random fashion. Experiments to test this possibility were done on adult T . viridescens (Sperry, 19431,). In some animals the only operative interference was section of the left optic nerve; in others the left eye was rotated 180" and 2 days later the left optic nerve was cut. Starting 10 days after operation the animals were regularly tested for return of visual function in the left eye, final tests of vision being made after removal of the right, normal eye. The results were quite consistent; in all cases tested return of visual function on the operated side was found. In most animals this recovery was noted some 30 days after operation, but in some vision did not reappear for 75-95 days; Sperry supposed that in these latter cases degeneration of the neural retina had occurred. In a11 the animals the recovered vision was orderly: visual local sign was restored. The visual responses of the animals with normally positioned eyes were in all respects apparently normal. But the animals with rotated eyes behaved as if the visual field was inverted (Fig. 1).The spatial localization of small objects in the field was consistent but erroneous, and in direct correlation with the amount of retinal rotation. In both groups of animals, retinal local sign was correctly signaled to the optic lobes, but in the animals with rotated eyes this orderly restoration of relations between retina and lobe was functionally maladaptive. An indication of the disadvantageous nature of visual function with rotated visual field is given by the finding that, following visual recovery, the animals with inverted vision took longer to find food than they did when totally blind. NO sign of correction of these maladaptive visual reactions was seen in animals that were kept for a period of 50 days after return of vision. In a series of similar experiments with rotated eyes, Stone (1953) found that the abnormal tendency of the animals to swim in circles and also the reversed visual responses persisted for over 4 5 years. At the end of this period normal reactions were imme-

18

R. M. GAZE

diately restored when the eye was brought back to the normal orientation. These experiments seem to rule out the possibility of an adjustment on the part of the central nervous system to randomly regenerated fibers: no animal, it is reasonable to suppose, will "adjust" so as to give itself an orderly but upside-down and

Normal

Reversed

FIG.1. Illustration of the reversal of optokinetic reactions that follows 180" rotation of the eyes. Top, rotation around the dorsoventral axis of the body; bottom, rotation around the dextrosinistral axis. In each diagram the large arrow indicates the direction of movement of the visual field and the small arrow shows the movement of the head. These reversed reactions are obtained both when the optic nerve is intact and after regeneration. (After Sperry, 1956.)

back-to-front visual field. Consequently Sperry was led to conclude that during regeneration the optic axons must somehow have managed to re-establish their original central connections. Rotation of the eye by 180" leads to complete inversion of the visual field on both (dorsoventral and nasotemporal) major axes of visual space. If the eye is transplanted to the other orbit, this can be done in such a way that the orientation of the eye is correct on

OPTIC NERVE REGENERATION IN AMPHIBIA

19

one of these axes and inverted on the other. If, for instance, the transplanted eye is correctly oriented on the optic axis, it may also be correctly oriented on either, but not both, the dorsoventral or

Right-ielt transplont wh$ inversion OF NOSO-Temporal AXIS

-

Rlqht left tronsplont with rotation OF eye ond thus inversion OF Dooo-ventroi Ax15

FIG. 2. Transplantation of right eye to left orbit. The right eye is divided into quadrants in order to show the inversions of the axes. D = dorsal, V = ventraI, T = temporal, N = nasal.

the nasotemporal axis (Fig. 2). Yet another variation in the relationship between eye and brain can be achieved if the optic chiasma is uncrossed and each eye connected with the lobe on the same side. In this case, the nasotemporal and dorsoventral axes of the

20

R. M. GAZE

eye are still correct, but the eye is on the wrong side of the midsagittal body axis, with relation to the optic lobe. All these operations were carried out by Sperry (1945b) in a very elegant series of experiments on various anurans (Hyla, Ram, Bufo) and in addition the contralateral eye transplantations were also done on urodeles ( T . viridescem). The results provided support for the hypothesis that the ganglion cells of the retina re-established central reflex connections with the appropriate parts of the lobe. In the frogs with contralateral cross-union of the optic nerves ( uncrossed chiasma), signs of visual recovery were seen after intervals varying from 12 to 48 days. Following regeneration the optokinetic reactions were normal to rotation around the dextrosinistral axis of the body and reversed to rotation round the dorsoventral and rostrocaudal axes. Spatial localization of a lure by these animals was correct only when the lure was presented in the midsagittal plane; in all other positions the lure was erroneously localized at a corresponding point on the opposite side of the mid-plane (Fig. 3 ) . These errors of localization were quite precise with relation to the sagittal plane. Contralateral transplantation of the eye was successful in two anurans ( R u m clumituns) and twelve urodeles ( T . 0. symmetricus and T . 0. viridescens). In the frogs the first signs of visual recovery were seen 41 and 46 days after operation while the first newts to recover vision did so by the forty-seventh day and most could see by the sixty-second day. In all these cases of transplantation the functional result was that expected from the altered polarity of the eye: with the dorsoventral axis of the eye inverted, optokinetic reactions to rotation around the dorsoventral axis of the body were normal while reactions to rotation around the dextrosinistral axis of the body were reversed; with the nasotemporal axis of the eye inverted, optokinetic reactions to rotation around the dorsoventral axis of the body were reversed, while reactions to rotation around the dextrosinistral axis of the body were normal. When tested by their abilities to locate small objects in space, each group of animals made consistent errors according to its own type of retinal disorientation. Animals with inversion of the dorsoventral axis of the eye located objects wrongly with respect to the dorsoventral dimensions of the visual field but located them correctly with respect to the anteroposterior dimensions. Animals with inversion of the nasotemporal axis of the eye located objects wrongly

OPTIC NERVE REGENERATION IN AMPHIBIA

21

with respect to the nasotemporal dimensions of the field but located them correctly with respect to the dorsoventral dimensions. The results of all these experiments seemed to point to one conclusion: that during regeneration of the optic nerve there is a selective re-establishment of the original functional connections of the retinal fibers at the optic lobe. Sperry (1944) attempted to A

B

FIG.3. The normal optic chiasma is completely crossed ( A ) ; when the chiasma is uncrossed and each optic nerve made to connect with the ipsilateral optic lobe ( B ) , visual localization is disturbed as shown in ( C ) . A lure presented at position X causes the animal to jump to X' and a lure at position Y causes a jump to Y'. (After Sperry, 1956. )

provide direct proof of this in a study of the retinal projection to the lobe by observation of the scotomata resulting from lesions in various parts of the optic lobe. In normal frogs he was able to show that lesions of the anterior part of the lobe caused blindness in the anterior field, while lesions of the posterior part of the lobe caused blindness in the posterior field; if the dorsal part of the lobe was injured, the blindness was restricted to the upper part of the field.

22

R. M. GAZE

In frogs with regenerated optic nerves and normally oriented eyes the results of such lesions were similar to those in the control animals, while in 2 animals with regenerated nerves and rotated eyes, the retinal projection found by this method was reversed. Investigation of the frog’s retinal projection by electrophysiological methods (Gaze, 1958a) gave results that agreed with those of Sperry, and also showed the existence in the normal animal of a well localized point to point relationship between the retinal surface and the surface of the lobe. The electrical method has also been used recently (Gaze, 1959) to investigate the retinal projection following regeneration in Xenopus laeuis. The normal projection in these animals is somewhat similar to that in Raw, and after regeneration of the nerve had taken place the representation of the retina on the lobe was found to be restored. Some of the animals used in these experiments had the eye rotated at the same time that the nerve was cut, and after regeneration in these cases the retinal representation on the lobe was shown to be rotated to a similar extent (Fig. 4 ) . Summary. The existence of visual local sign in normal animals is related to the systematic projection of retinal fibers on to the surface of the optic lobe. The restitution of local sign properties after optic nerve regeneration indicates either recovery of the orderly retinal projection to the lobe or random regeneration followed by functional adjustment on the part of the central nervous system; the evidence available shows that during regeneration the functional projection to the optic lobe is restored.

VI. The Hypothesis of Neuronal Specificity When taken together, the results of the experiments described in the previous sections suggest strongly that, during regeneration, the fibers of the optic nerve grow back to make synaptic contact in the appropriate regions of the optic lobe, thus recreating the original functional relationship between each part of the retina and each corresponding part of the lobe surface. The problem to be discussed at this point is how this selective restitution could be achieved. In the normal optic nerve each fiber runs along its own “channel” to reach the optic lobe; in some animals (Triturus; see

23

OPTIC NERVE REGENERATION I N AMPHIBIA

Stroer, 1940) it has been shown that fibers from the various parts of the retina are segregated into fascicles which retain their identity from retina to lobe; in others (for example Amblystoma; see Herrick, 1941) this fasciculation has not been found. It may be that fasciculation of this sort does not exist in these cases and that there L.Field

L. Retina

L Field

L. Retina

R.Lobe

N m l Eye

Eye Rdated 90’ counterclockwise

FIG.4. The rotation of the retinal representation on the optic lobe following rotation of the eye and section and regeneration of the nerve. Upper T O W . An arrow in the visual field (left diagram) forms an inverted image on the retina (middle diagram). The quadrants of the retina, identified by symbols, project to the optic lobe as shown on the right. A “representation” of the arrow is thus formed on the optic lobe. Lower row. The eye has been rotated 90” counterclockwise and the optic nerve cut and allowed to regenerate. An arrow in the visual field forms an image in the same way on the retina, but the parts of the image now fall on different quadrants. Since each quadrant of retina has reconnected with its original place on the optic lobe, the “representation” of the arrow on the lobe is also rotated.

is no spatial localization within the optic nerve of fibers from the different parts of the retina; or perhaps the anatomical methods used by Herrick were inadequate for the task of demonstrating it. However this may be, if it could be shown that, after section, each optic nerve fiber finds again and grows down its original “channel” in the proximal stump of optic nerve, then the restitution of retinal local sign would be explained without further hypothesis. This

24

A. M. GAZE

however is not the case. Stone (1930) found that following transplantation of the eye in AmbZystoma the proximal stump of optic nerve had degenerated and completely disappeared by 10 days after the operation, whereas the new nerve growing out from the eye had only reached the optic foramen by the same time. In these experiments, therefore, there was no possibility that the fibers had regained their original pathways-the pathways were no longer present. In tadpoles of Xenopus laevis degeneration is also very rapid; McMurray (1954) showed that after crush of the nerve, advanced degeneration was present within 2 days whereas regeneration was in progress only after 5 days. Maturana (1958) reported rather dif; found that degeneration here was a very ferent results in B u ~ Q he slow process, degenerating myelin still being present in the nerve 200 days after section. So in this case, at least, there might be a possibility that the axons could regrow into their previous channels. However, Sperry (1943b) has shown that following section of the optic nerve the regenerating fibers become a tangled mass at the site of the lesion as they penetrate the scar tissue. In a normal nerve the fibers are more or less parallel to each other and may run straight for considerable distances; the mix-up that occurs at the site of the section is such that the original spatial distribution of the fibers within the nerve must be grossly disarranged when they reach the proximal side of the scar. To enable these intertangled fibers each to regain its original channel beyond the scar, a form of specific guidance would have to exist, such that each fiber was “attracted back to its rightful place in the nerve. But this would be flying in the face of the evidence: it is well established (see Weiss, 1955) that nerve fibers bridging a gap in a nerve are not specifically attracted back down the right pathway but will with equal facility grow down a wrong pathway if one is presented. This lack of a distinctive attractive force exerted on optic fibers has been demonstrated by experiments of Ferreira-Berrutti ( 1951) on the visual system of chick embryos. In these embryos regeneration of the optic axons may occur if the section is done at a sufficiently early stage of development. Ferreira-Berrutti cut and deflected the optic nerve in early life (before 70 hours), before the fibers had reached the brain, to see if the pre-existing epithelial tube of optic stalk served as a necessary guide to the fibers. In

OPTIC NERVE REGENERATION IN AMPHIBIA

25

successful cases the optic nerve so treated grew askew and ended up in various abnormal places. When, in Amblystomu, the eye vesicle was transplanted to the position of the excised ear vesicle (May and Detwiler, 19W), the developing optic nerve made contact either with a cranial nerve ganglion or with the medulla. The ability of the optic nerve to grow in an abnormal direction is also shown by some experiments of Weiss (1941) where, after removal of the brain, the optic fibers grew along the olfactory pathway to reach the nose. Some unpublished work of Sperry (reported 1955) suggests that the distribution of fibers in the optic nerve is not a critical factor in the restitution of optic nerve connections; this author cut the “medial optic tract” in the normal frog (Hyla cinerea) and thus produced a blind area in the dorsal part of the visual field over the animal’s head; similar lesions after regeneration tended to be less effective and in some cases left fairly good vision in the same dorsal part of the visual field. The work of Gaze (1959) on Xerwpus laevis also indicates the lack of specificity in the path followed by regenerating fibers; in one animal (Xellopus A3) the cut optic nerve grew back into the brain via the root of the oculomotor nerve, ventral to the optic lobe. Despite this abnormal pathway and the fact that the fibers had to approach the visual area of lobe from the wrong direction and in unusual grouping, the axons appeared to end up in the right place. While the greater part of the available evidence shows no form of guidance by “attraction,” there are some observations on directional nerve growth which do not fit any theory so far advanced; the work of Hooker (1930) on regeneration in sectioned, rotated spinal cord, and the findings of Piatt (1942, 1957) on the development of normal patterns of innervation when fibers grow into limb transplants, show the existence of some form of direction-finding ability as yet unexplained. It seems likely that in many cases the degenerated cordlike remnant of the proximal optic nerve stump remains intact and acts as a general guide to the regenerating fibers. This guide would then suffice to direct (by “contact guidance”) the fibers back as far as the chiasma, after which point they would be on their own, as it were. Such a guide could play no part, however, in allocating the individual fibers to their rightful places on the surface of the lobe. The end result of regeneration in these animals is that each individual part of the retina reconnects with its own individual part of

26

R. M. GAZE

the surface of the optic lobe; if the original nerve fiber pattern is not reconstructed beyond the scar, then whereabouts between site of section and optic lobe does the restitution take place? Sperry (summarized 1951) has put forward a hypothesis invoking neuronal specificity in an attempt to explain the findings in his experiments. According to this theory each ganglion cell of the retina, together with its optic nerve fiber, possesses unique biochemical properties; and each neuron in the corresponding part of the optic lobe possesses a comparable and similar specificity; only when “like meets like” does an effective and stable synaptic union take place. In the reformation of synaptic contacts in the optic lobes, Sperry envisages a “shotgun” effect: each fiber puts forth many branches as it grows into the brain and the cells of the lobe itself have widespread dendrites. There would, therefore, be a good chance that any given fiber would eventually make contact with its appropriate lobe cells. Sperry (1956) pictures “the advancing tip of a fibre making a host of contacts, the great majority of which come to nothing; but eventually the growing tip encounters a type of cell surface for which it has a specific chemical affinity and to which it adheres.” The simplest way in which this neuronal specificity can be conceived is that the cellular differentiation should be orderly and continuous (Sperry, 1943b) such that neurons further apart in the retina should differ more than those closer together. In this way complete identification of any retinal point could be achieved on the basis of only two differential gradients, one in each major axis of the retina. To oversimplify for the sake of clarity, if there exist regular differences of two parameters A and B, distributed at right angles to one another on the retina, then any point on the retinal area can be completely specified by two quantities only, one measuring A and the other measuring B. As Sperry points out, there are good grounds for postulating the existence of such differentiation fields in experimental embryology. These retinal gradients would of course have to be paralleled by similar gradients on the lobes. This attractive hypothesis brings several attendant difficulties, apart from the obvious one that the biochemical specificity has not been demonstrated, only inferred. There is, for instance, the result of the experiments on uncrossing the optic chiasma ( Sperry, 1945b). In a normal amphibian visual system the one eye projects mainly to the other optic lobe. If the eye is transplanted to the other orbit,

OPTIC NERVE REGENERATION IN AMPHIBIA

27

the regenerating nerve still crosses the chiasma and joins the contralateral lobe-contralateral, that is, to the eye in its new position; the balanced relationship of the retina and lobe about the midsagittal line of the body is preserved. However, when the eyes are left in their normal places and the chiasma is uncrossed, each eye then connects with the ipsilateral lobe. Both kinds of experimentuncrossing the chiasma and contralaterally transplanting the eyeshow that if specificity of the retinal ganglion cells exists, it is paralleled not merely by a similar specificity of the central cells of the contralateral lobe, but also by a mirror-image specificity of central cells on the same side as the eye. Optic nerve fibers can, if experimentally induced to do so, make orderly synaptic contact with either lobe; the resulting visual field is properly organized within itself in both cases, but where the connection is from eye to contralateral lobe the field is inverted about one of the axes of visual space and where it is from eye to ipsilateral lobe the field is inverted about the midsagittal plane of the body. The situation is further complicated by the fact that each eye in the anura projects to both optic lobes, not merely to the contralateral one as was formerly thought (Gaze, 19558b). The eye projects in the normal maplike fashion to the contralateral lobe and that part of the retina which receives stimulation from the binocular field projects also to the anterior region of the ipsilateral lobe. The ipsilateral projection is also localized in a point-to-point fashion and is so arranged that a point in space, within the binocular field, is represented via both eyes at one place on the lobe (and at a different place on the other lobe; see Fig. 5). Up to the present time it is not known whether this ipsilateral projection is direct or multisynaptic. But since the ipsilateral, as well as the contralateral, projection is restored in an orderly way following regeneration of the optic nerve (Gaze, 1959), the system of proposed neuronal specificities becomes more difficult to interpret. As yet there is no experimental evidence that the ipsilateral fibers regenerate following contralateral eye transplantation or uncrossing of the chiasma; it seems reasonable to suppose that they do. Fibers from the posterior retina must be distinguishable in two different ways; those heading for the contralateral lobe must be differentiated from those going to the ipsilateral lobe, since the fibers from one retinal locus go to dissimilar places on each lobe. It is interesting to speculate

28

R. M. GAZE

upon the fate of the ipsilateral fibers when an eye is transplanted to the other orbit; do they then grow in an orderly fashion to the new ipsilateral lobe, which was previously the contralateral one? At this stage the situation becomes rather complex and the mind begins to boggle.

FIG.5. Binocular vision in the frog. The arrow points down the midsagittal plane of the animal. X is a point within the binocular field. From the left eye, X sets up impulses that travel across the chiasma (continuous line) to reach a spot on the right lobe; from the right eye impulses go to the same place (dotted line-pathway unknown). X is binocularly represented on the left lobe in a comparable way, but at a dissimilar point.

Another obstacle to the specificity theory is provided by the fact of visual recovery following heteroplastic transplantation of the eye. According to Sperry’s hypothesis, such restitution of orderly vision indicates that the pattern of retinal-lobe specificities is the same in the various species of animals used in these experiments. Not merely the same in general, but superimposable, a direct fit. Within the range of species that can be used for heteroplastic grafting, the specificity is an anatomical, not a species pattern. At

OPTIC NERVE REGENERATION IN AMPHIBIA

29

first sight this widespread distribution of the same highly organized pattern of specificities seems to provide a formidable objection to the hypothesis. This difficulty is more apparent than real, however; if the patterns are thought of as being formed by crossed gradients measuring some (as yet unknown) parameters, for instance, there is nothing unlikely in the existence of interspecific transferability of specificity. Consideration of certain other related fields of experimental embryology lends some support to the concept of specificity in the relations between sensory center and periphery. The restoration of retinal local sign after regeneration of the optic nerve is not a unique phenomenon; it is presumably similar in nature to the recovery of cutaneous local sign that may follow section of cutaneous afferent fibers proximal to the sensory ganglion. Sperry and Miner (1949) showed that, in newts of the eft and adult aquatic stages, crossunion of the root of cranial V to the root of cranial VII allowed the fibers of V to regenerate into the brain over the pathway of VII; the result was an orderly recovery of cutaneous local sign. Further evidence of a different sort was obtained by Miner (1956) from experiments on frog tadpoles: a dorsoventral strip of skin was totally removed from the body, rotated 180" and reapplied to the animal; the autograft took well and underwent self-differentiation so that after metamorphosis the animal had a patch of light-colored belly skin on its back and a patch of dark-colored dorsal skin on its belly. When the dark skin on the belly was gently stimulated the frog aimed a wiping response at its back and vice versa. The results of these and other experiments led Miner to conclude that the pattern of central synaptic associations formed by the cutaneous fibers was partly determined by the quality of the skin with which they connected. There are other observations (for example the development of a lid-closure reflex from transplanted eyes; Weiss, 1942) which point in the same direction.2 However, the evidence is rather conflicting in that, for instance, Sperry and Miner (1949) failed to obtain any respecification of sensory fibers after peripheral nerve crossing in postlarval Triturus; nor did they find it in the trigeminal afferent system of frog tadpoles. There is 2 In a recent paper Szkkely (1959) has reported experiments which suggest a different mechanism from the one proposed by Weiss to explain his results.

30

R. M. GAZE

therefore an urgent need for further experiments along these lines; at present it seems likely that the capacity of nerve tissue to become respecified (“modulated) varies both with the developmental stage of the animal and with the part of the nervous system concerned. The operation of some specific influence is suggested by Eccles et al. (1960) to account for their finding of altered monosynaptic connections in cat motoneurons, following experimental alteration in early postnatal life of the destination of the axons. The hypothesis of neuronal specificity has considerable advantages from a practical point of view; it would account for the facts and it is compatible with recent neurological and embryological data. The word “specificity” in this context brings to mind immediately the study of the antigenic differences between tissues from different animals; and it is possible that the specificities we seek are to be found in a form of immunological differentiation. This approach would be well worth exploring; the difficulty lies in the very level of specificity required. It is known that tissues of one animal differ immunologically from those of another, although this differentiation between individuals of the same species is not well marked in the amphibia. It has even been shown (Waksman and Adams, 1955) that a detectable antigenic difference exists between the central nervous system and peripheral nerve tissue in rabbits. The technical difficulties involved in a demonstration of the possible immunological specification of the various parts of the retina and the optic lobe would be enormous. Even the detection of the comparatively crude differences between central nervous system and peripheral nerve by Waksman and Adams required a formidable neuropathological technique; yet for our purposes differences would have to be shown, not merely between gross subdivisions of the central nervous system such as optic lobe and cerebellum, but between different parts of the optic lobe and even between different cells within parts of the lobe. Up to the present time no technique sufficiently selective for this purpose has been used, but it would be worth while applying the gel-diffusion method (for instance the micromodsification as used by Mansi, 1958) to the differentiation of different parts of the optic lobe. Another possible factor, in the selective restitution of synaptic contacts that occurs in regeneration, is the possibility that the orderly termination of fibers from different parts of the retina is an

OPTIC NERVE REGENERATION IN AMPHIBIA

31

expression of the time sequence of events during fiber growth. It is known that the optic lobe differentiates from the anterior pole towards the posterior pole (Nieuwkoop and Faber, 1956; Kollros, 1953), and also that the development of cholinesterase activity in the lobe proceeds in the same direction (Boell et al., 1955). The optic axons enter the lobe from the anterior end. Perhaps the order of arrival of the fibers determines their systematic distribution over the surface of the lobe (suggestion of Dr. Marcus Jacobson). If this were so, one could imagine the first wave of fibers arriving at the lobe making synaptic contacts with lobe neurons at the anterior pole and thus filling up all the “vacancies” in this region; the next wave of fibers to arrive would thus have to grow further posterior on the lobe before they could find sites for the formation of synapses, and so forth. Under these circumstances, one would expect to find a projection from posterior retina (anterior field) to the anterior regions of the lobe while the rest of the lobe was still unconnected to the retina, and the last part of the lobe to become visually active would be the posterior pole. In this way, if fibers from different parts of the retina each had specific rates of growth, or if the fibers started to develop in an orderly time sequence, the anteroposterior distribution on the lobe could be accounted for. Up to the present time there is inadequate information on this matter. It would be most interesting to know whether there is any systematic sequence of arrival of the fibers at the lobe surface during regeneration. With present electrophysiological methods (Gaze, 1958a, 1959) this question should be answerable; experiments directed to this end are at present in progress. A difficulty encountered by this hypothesis is that, while it could account satisfactorily for the anteroposterior distribution of fibers on the lobe, it could not so simply provide an adequate explanation of the mediolateral fiber distribution. The orderly recovery of fiber projection following regeneration of the optic nerve via the root of the oculomotor nerve, as seems to have happened in one of the Xenopus used by Gaze (1959), would be impossible on this theory. It would be appropriate at this stage in the discussion to underline certain differences that exist between regeneration of the optic nerve and its initial development. During this phase of development of the visual system the optic lobe is an undifferentiated mass of cells until the arrival of the optic afferents. The eye is connected

32

R. M. GAZE

to the brain by the optic stalk long before the retinal ganglion cells give rise to optic nerve fibers. Consequently these fibers, during their first outgrowth from the eye, are guided to the appropriate part of the brain along this preformed pathway. So far as is known the optic stalk provides only a general directional guidance; there is no evidence to suggest that the point of termination of any one fiber is determined in this way. Regenerating optic nerve fibers have to face a very different situation; here, all connection between the distal stump of nerve and the brain may have been severed. The fibers have nothing more precise to guide them than the general orientation of the contents of the posterior orbit. In spite of this disadvantage, the remarkable fact remains that the reconstituted nerve usually follows an approximately normal path and forms a chiasma in the right place. Thus the circumstances attending the formation of the optic nerve are different in the two cases. Difference is also to be seen in the condition of the optic lobe that receives the fibers when they arrive. In embryonic development the differentiation of the complex layered structure of the lobe is called forth by the arrival of the optic nerve axons. When the nerve regenerates, on the other hand, the lobe is already largely developed. Once the optic nerve connection has been made, section of the nerve does not lead to complete regression of the lobe to its previous undifferentiated state. If the lobe is deprived of its optic afferents it will show considerable involution, loss of cell and fiber content, as mentioned in a previous section of this review, but the basic layered structure remains intact. Summary. The recovery of visual local sign after regeneration of the optic nerve is paralleled by similar observations on cutaneous local sign; reasons are advanced for accepting the theory that there are constitutional differences between ganglion cells in different parts of the retina and that there exist appropriately fitting differences between central cells as well.

Vll. Some Other Aspects of Form and Function

The results of the experiments on optic nerve regeneration and rotation of the eye show that the different quadrants of the retina are not equipotential insofar as their central effect is concerned;

OPTIC NERVE REGENERATION IN AMPHIBIA

33

the eye shows polarity. Sperry (1943b) suggested that if the retinal ganglion cell “field” or its anhge were rotated sufficiently early during development, it would compensate, and normal vision result rather than inverted vision. This was shown to be so by Stone (1944, 1947, 1948) who excised, rotated, and reimplanted an eye in embryos of Amblystoma pnctatum at various stages between the closure of the neural folds and just before the beginning of feeding (stages 2 0 4 2 of Harrison). Visual tests on these animals during larval and adult life showed that up to the late tailbud stages the eye could be rotated 180” without affecting normal vision, but if the eye was rotated at stage 36 or after, inverted vision resulted. Stone (1947) thought that stage 36 was a critical period in the development of retinal polarity, since eyes rotated just before this gave confused visuomotor reactions. This observation is of special interest since it suggests that the fiber projection from retina to lobe may have been disorderly in these cases, a point which could be decided fairly simply by the focal recording method. SzBkely (1957) has reported that the functional polarity of the eye anluge is determined earlier in the anteroposterior than in the dorsoventral direction, which would lend support to the idea that retinal specificity involves a differentiation along two crossed axes. As an appropriate counterpart to this work of Stone and of SzBkely, attempts have been made to investigate the effects of rotation of the entire optic tectum about the dorsoventral axis of the body, Crelin (1952) excised one tectum and rotated the other in larvae of Amblystoma at different stages and afterward tested for recovery of visual function. He found that, provided the rotation was performed before Harrison’s stage 30, the animals recovered normal vision, but that rotation after this resulted in progressive impairment of visual function. Rotation of the lobe up to stage 35, however, did not result in reversed vision. Thus the lobe was not polarized before stage 35. Since the first optic fibers to grow into the lobe do so at about stages 38-39 ( Herrick, 1942, 1948) and since in all probability the related polarity of retina and lobe only develops when retinal fibers contact the lobe, it would be expected that the polarity of the lobe would be indeterminate until between stages 36-38. It seems likely that the failure to obtain good visual results when the lobe was rotated later than stage 30 was due to the gradual decline in the ability of the central nervous system to survive the reimplantation.

34

R. M. GAZE

The marked tolerance of some amphibians to heteroplastic transplantation of tissues affords the most interesting possibilities for study of the relative contributions made by eye and brain to visual performance. Other things being equal, a large eye containing many photoreceptors and ganglion cells will be more effective than a small eye containing fewer such elements. The possibility of altering an animal’s visual performance by substituting a larger or smaller eye in place of its own was investigated by Stone and Ellison (1945), who exchanged eyes between pairs of Amblystoma punctatum and Triturus viridescens. These animals are different in size and in visual habit; T . viridescens is a free swimming animal with good visual function whereas A. punctatum is secretive and appears to have eyes of lower visual acuity; Triturus is also considerably larger than Amblystoma. Stone and Ellison found that Triturus ( T ) eyes transplanted to Amblystoma ( A ) hosts did not do well; they eventually sloughed. But eyes from A grafted into the orbits of T survived much better and eventually gave return of visual function. The visual performance of these T hosts with A eyes was better than that of A eyes in A hosts but was worse than that of T eyes in T hosts. This finding suggests that in the T hosts with A eyes the normal functioning of the A eyes had been enhanced by the central activity of the larger T central nervous system. The visual tests used by Stone and Ellison were “rough and ready” but the method is most promising and should be extended to cover further combinations of center and peripheral receptors. The visual function of the anuran eye is far superior to that of the urodeles and it would be of great interest to graft anuran eyes onto urodele hosts and vice versa; since limbs can apparently be transferred between anurans and urodeles (Guyenot, 1927) there should be no insuperable difficulty in effecting the transplantation of eyes. The return of apparently normal function after regeneration of the optic nerve suggests many other questions; the normal anuran nerve has about half a million fibers and the receptive areas of individual ganglion cells (which may be about 1 mm in diameter, according to Hartline, 1938) overlap considerably. What proportion of these optic nerve fibers must regenerate to appropriate terminations in order to account for the observed return of vision? What happens, during regeneration, to the fast and slow fibers of the nerve? If there is more rapid regeneration of one of these

OPTIC NERVE REGENERATION IN AMPHIBIA

35

groups then tests of visual performance may provide a clue as to their differential functions. Is it possible that the “failure of prediction,” observed by Maturana (1958) in toads with regenerated nerves, is related to differential regeneration rates of these fibers? What happens, during regeneration, to the organization of the lobe in depth? The experiments of Gaze (1959) showed that the retinal projection to the surface of the optic lobe is restored after regeneration of the nerve; this has since been confirmed by Lettvin et al. (1959), who also found that a similar projection exists in four layers of the tectal roof, that the four projections are all in register one with another, and that each projection regenerates appropriately. A different function is allotted by these authors to each of the four layers. Summary. Experiments involving rotation of the retina at various stages of development have shown that retinal polarity develops at approximately stage 36 in urodeles. Similar experiments showed that the optic lobe was unpolarized up to stage 35. The tolerance shown by some amphibians to heteroplastic grafting has permitted the investigation of the relative role of retina and brain in visual performance. VIII.

Conclusions

How then are we to account for the return of orderly vision following optic nerve regeneration? It seems reasonable to assume that visual local sign is in normal animals a function of the orderly retinal projection to the visual centers. The experiments on regenerated nerves and rotated eyes show that the functional relationship between each part of the retina and each corresponding part of the optic lobe is re-established, regardless of the usefulness of the restitution; electrical recording from the optic lobe after regeneration of the nerve shows that the original point-to-point projection tends to be regained. In all probability the optic nerve fibers find their way back eventually to their “correct” positions on the optic lobe. If this is so, three hypotheses need be considered: (1) the fibers are directed back to their rightful destinations, ( 2 ) they are not directed at all but must in any case find their way back by accident of development, ( 3 ) the fibers approach the lobe with random dis-

36

R. M. GAZE

tribution in space and the correct synaptic linkages are established eventually on the basis of neuronal specificity of some kind. The first theory is not tenable since it is in total disagreement with all the experimental observations. The second theory, that of the “historical accident of development,” may provide a possible explanation of the facts; but until further investigations have revealed something of the spatiotemporal pattern of regeneration little can usefully be added to what has been said already. The third theory, invoking neuronal specificity, seems to me to provide the simplest explanation, not only of the findings after regeneration of the optic nerve, but of the observations from related fields as well. The presentation of these three hypotheses as entirely separate and mutually exclusive is an oversimplification; the specificity theory, for instance, requires some form of guidance for the growing axons. This guidance is often provided by the proximal stump of the severed nerve. In the absence of any guiding structure, the nerve will probably fail to regenerate. And, as was pointed out by Sperry (1946, 1958), the inherent organization of the parts of the central nervous system is not presumed to be dependent entirely on refined inductive effects involving specificity of neurons; other factors such as the proper timing of the developmental sequence are necessary as well. Regenerating fibers grow back to the optic lobe in whatever order and at whatever time circumstances allow; the postulated specificities of the different neurons come into play when the fibers reach the lobe, and aid them to form appropriate synaptic connections. During the process of differentiation of the embryo, future nerve cells are separated from cells destined to form other tissues. Within the forerunner of the central nervous system this process of differentiation continues, giving eventually the various distinct parts of the hervous system such as cerebral hemispheres, diencephalon, optic lobes, retinae, cerebellum, and so forth. Neurons from one part of the brain are in many cases different in size and shape from those in other parts. Biochemical as well as morphological differentiation takes place during development (and in the last analysis, of course, the latter depends on the former); examples of this are the differential distribution of cholinesterase found in the brain (Shen et al., 1955) and the antigenic differences demon-

OPTIC NERVE REGENERATION IN AMPHIBIA

37

strable between central and peripheral nervous tissue. The selective actions of certain poisons on various regions of the nervous system and the localized attacks made by some viruses are also indicative of this process of “invisible” cytodiff erentiation. The adult retina does not contain a layer of equipotential ganglion cells; at a certain stage of development the ganglion cell layer becomes polarized and after this stage the relationship between each part of the retina and the surface of the optic lobe is fixed. It seems reasonable to suppose that the differentiation of the retina continues to the point where each minute area, or even each ganglion cell, becomes endowed with a certain unique biochemical specificity. On this hypothesis the cells of the optic lobe must also possess a corresponding specificity. This could conceivably be the result of a parallel self-differentiation of the central neurons; or, what is perhaps more likely, it could be imposed on them following the arrival of the retinal fibers, by the process which has been called “modulation.” Since the environment of the growing fibers is so much simpler in first development than it is in later regeneration, and the retina develops almost in apposition to the visual brain, it is conceivable that orderliness of the growth process together with selective fasciculation enables the first fibers to establish the beginnings of a normal pattern on the lobe surface. The axons which arrive at the appropriate part of the lobe could then modulate or induce specificity in the central tissue and after this the retinalobe relationship would be permanently established. This theory could be most usefully tested by detailed investigation of the spatial pattern and time course of retina-lobe connection during first development. It would also be interesting to know how the fiber count in the optic nerve of the anuran tadpole compares with that in the adult; Herrick (1948) has shown that in urodeles there occurs a great increase in number of fibers at about the time of metamorphosis. This hypothesis of neuronal specificity has met and will continue to meet with considerable resistance from many neurophysiologists; the concept of specificity on the required level is contrary to established teaching. As Weiss (1950) has pointed out, most physiological concepts of central nervous system activity assume that all the neurons are essentially alike, and that the sig-

38

R. M. GAZE

nificant variations are in the electrical state of the membrane and the configuration of the neural network. Representative of this approach is the tendency to treat the various retinal ganglion cells as a series of similar units. During development the eye vesicle differentiates until the layers of the retina are properly formed; then, since no further changes can be seen, it is assumed that differentiation has stopped. This rather naive interpretation must sooner or later give way to one more in accordance with the accumulating evidence from the field of experimental embryology. In recent years most neurophysiological investigation has been directed towards the study of the nervous system as it exists, formed and mature, at the time of experiment. This has of course greatly increased our knowledge of how many adult nervous mechanisms function. But the more fundamental problem is not so much how the nervous system works but how it came to develop into such a state that it could work. The experiments on regeneration of the optic nerve serve to emphasize what should be obvious: that we require an understanding not only of static structure but also of the way the structure developed. At present we are hardly at the beginning of such an understanding.

REFERENCES Abbie, A. A., and Adey, W. R. (1950). J. Comp. Neurol. 92, 241. Akert, K. (1949). Helv. Physiol. et Pharmacol. Acta 7 , 112. Andrew, A. M. (1955). J. Physiol. (London) 130, 25. Apter, J. T. (1946). J. Neurophysiol. 9, 73. Arey, L. B. (1916). J. Comp. Neurol. 26, 213. Barlow, H. B. (1953a). J. Physiol. (London) 119, 58. Barlow, H.B. (1953b). J. Physiol. (London) 119, 69. Beers, D. N. (1929). PTOC. SOC. Exptl. Biol. Med 26, 477. Biederman, S. (1927). Prace lnst. im. Nenckiego 4, No. 56, 1. Bishop, G. H. (1933). Am. J. Physiol. 106, 460. Boell, E. J., Greenfield, P., and Shen, S. C. (1955). J. Exptl. 2001.129, 415. Breusch, R. S., and Arey, L. B. (1942). J. Comp. Newol. 77, 631. Buser, P. (1949a). Compt. rend. SOC. biol. 143, 30. Buser, P. (194913). Compt. rend. SOC. biol. 143, 817. Buser, P. (1950). 1. Physiol. (Paris) 42, 557. Crelin, E. S. ( 1952). J. Exptl. Zool. 120, 547. Duke-Elder, S. (1958). “System of Ophthalmology,” Vol. 1. Henry Kimpton, London, England.

OPTIC NERVE REGENERATION IN AMPHIBIA

39

Eccles, J. C., Eccles, R. M., and Magni, F. (1960). J. Physiol. (London) 160, 29 p. Ferreira-Berrutti, P. (1951). Proc. SOC. Exptl. Biol. Med. 76, 302. Gaze, R. M. (1958a). Quart. 1. Exptl. Physiol. 43, 209. Gaze, R. M. ( 195813). J . Physiol. (London) 143, 20 p. Gaze, R. M. (1959). Quart. J. Exptl. Physiol. 44, 290. Granit, R. (1955). J. Neurophysiol. 18, 388. Guyenot, E. (1927). Rev. suisse zool. 34, 1. Hartline, H. K. (1935). Am. J. Physiol. 113, 59. Hartline, H. K. (1938). Am. J. Physiol. 121, 400. Hartline, H. K. (1940a). Am. J. Physiol. 130, 690. Hartline, H. K. (1940b). Am. J. Physiol. 130, 700. Herrick, C. J. (1941). J. Comp. Neurol. 74, 473. Herrick, C. J. (1942). J. Comp. Neurol. 77,191. Herrick, C. J. (1948). “The Brain of the Tiger Salamander.” Univ. Chicago Press, Chicago, Illinois. Hooker, D. (1930). J. Exptl. Zool. 55, 23. Kollros, J. J. (1953). J. Exptl. Zool. 123, 153. Larsell, 0. ( 1931). J. Exptl. Zool. 68, 1. Lettvin, J. Y., Maturana, H. R., McCulloch, W. S., and Pitts, W. H. (1959). Proc. I.R.E. (Inst. Radio Engrs.) 47, 1940. McMurray, V. M. (1954). J. Exptl. Zool. 125, 247. Mansi, W. (1958). Nature, 181, 1289. Matthey, R. (1925). Compt. rend. SOC. biol. 93, 904. Matthey, R. (1926). Compt. rend. soc. biol. 94, 4. Matthey, R. ( 1927). Arch. Entwicklungsmech. Organ. 109, 326. Maturana, H. R. (1958). J. Anut. 92, 21. Maturana, H. R. (1959). Nature 183, 1406. May, R. M., and Detwiler, S. R. (1925). J. Exptl. Zool. 43, 83. Miner, N. (1956). J. Comp. Neurol. 105, 161. Nieuwkoop, P. D., and Faber, J. (1956). “Normal Table of Xenopus laeois ( Daudin ) .” North-Holland Publ., Amsterdam, Holland. Noble, G. K. (1931). “The Biology of the Amphibia.” McGraw-Hill, New York. Piatt, J. (1942). 1. Exptl. Zool. 91, 79. Piatt, J. (1957). J. Exptl. Zool. 134, 103. Rochon-Duvigneaud, A. ( 1943). “Les Yeux et la Vision des VertBbrCs.” Masson, Paris, France. Shen, S. C., Greenfield, P., and Boell, E. J. (1955). J. Comp. Neurol. 102, 717. Sperry, R. W. (1943a). J. Exptl. Zool. 92, 263. Sperry, R. W. (1943b). 1. Comp. Neurol. 79, 33. Sperry, R. W. (1944). J. Neurophysiol. 7, 57. Sperry, R. W. (1945a). Quart. Rev. Biol. 20, 311. Sperry, R. W. (1945b). J. Neurophysiol. 8, 15. Sperry, R. W. (1946). J. Comp. Psychol. 39, 321. Sperry, R. W. (1948). Physwl. Zool. 21, 351.

40

R. M. GAZE

Sperry, R. W. (1951). In “Handbook of Experimental Psychology” ( S . S. Stevens, ed.), p. 238. Wiley, New York. Sperry, R. W. (1955). In “Regeneration in the Central Nervous System” ( W. F. Windle, ed. ), p. 66. C . C Thomas, Springfield, Illinois. Sperry, R. W. (1956). Sci. Am. 194, 48. Sperry, R. W. (1958). In “Behaviour and Evolution” (A. Roe, and G. G. Simpson, eds.), p. 128. Yale Univ. Press, New Haven, Connecticut. Sperry, R. W., and Miner, N. (1949). J. Comp. Neurol. 90, 403. Stone, L. S. (1930). J . Exptl. Zool. 65, 193. Stone, L. S. (1944). Proc. SOC. Exptl. Biol. Med. 67, 13. Stone, L. S. (1947). Trans. Ophthalmol. SOC. United Kingdom 67, 349. Stone, L. S. (1948). Ann. N . Y. Acad. Sci. 49, 856. Stone, L. S. (1950a). Acta XVI Concilium Ophthalmologicum (Britannia) p. 644. Stone, L. S. (195Ob). J . Exptl. Zool. 113, 9. Stone, L. S. (1953). A.M.A. Arch. Ophthalmol. 49, 28. Stone, L. S., and Chace, R. R. (1941). Anat. Record 79, 333. Stone, L. S., and Ellison, F. S. (1945). J . Exptl. Zaol. 100, 217. Stone, L. S., and Farthing, T. E. (1942). 1. Exptl. Zool. 91, 265. Stone, L. S., and Ussher, N. T. (1927). Proc. SOC.Exptl. Biol. Med. 25, 213. Stone, L. S., Ussher, N. T., and Beers, D. N. (1937). J. Exptl. Zoul. 77, 13. Stone, L. S., and Zaur, I. S. (1940). J. Exptl. 2001.86, 243. Stroer, W. F. H. (1940). Acta Neerl. Morphol. 3, 178. SzBkely, G. (1957). Acta Biol. Acad. Sci. Hung. Suppl. 1, 20. Szbkely, G. (1959). J. Embryol. Exptl. Morphol. 7, 375. Tinbergen, N. (1951). “The Study of Instinct.” Oxford Univ. Press, London and New York. Waksman, B. H., and Adams, R. D. (1955). J . Exptl. Med. 102, 213. Walls, G. L. (1942). “The Vertebrate Eye and Its Adaptive Radiation,” Bull. No. 19, Cranbrook Institute of Science, Michigan. Weiss, P. (1941). Growth 6, 163. Weiss, P. (1942). J . C m p . Neurol. 77, 131. Weiss, P. (1950). In “Genetic Neurology,” (P. Weiss, ed.), p. 1. Univ. Chicago Press, Chicago, Illinois. Weiss, P. (1955). In “Analysis of Development” (B. H. Willier, P. Weiss, and V. Hamburger, eds. 1, p. 346. Saunders, Philadelphia, Pennsylvania. Wlassak, R. (1893). Arch. Pathol. Anat. u. Physiol. Virchow’s Suppl. Bd. 1.

EXPERIMENTALLY INDUCED CHANGES IN THE FREE SELECTION OF ETHANOL By Jorge Mardones Institute of Pharmacology and Institute of Research on Alcoholism. University of Chile. Santiago. Chile

I. Introduction ........................................... I1. General Procedure ...................................... I11. Individual Variations ................................... A. Animal Differences .................................. B. Interindividual Variation of Rats ....................... C . Genetic Origin of Individual Variations in Rats . . . . . . . . . . . D. “Drinker” and “Nondrinker” Rats ..................... E . Intraindividual Variations in Rats ...................... F. Summary .......................................... IV Effect of Diet .......................................... A. Organic Composition of Diet ......................... B. Thiamine Deprivation ............................... C . Factor N, ......................................... D. Thioctic Acid ...................................... E . Other Single Vitamin Deficiencies ..................... F. Multiple Vitamin Deficiencies ........................ G . Glutamine ......................................... H . Restriction of Food ................................. I . Summary .......................................... V. Influence of a “Third Choice” ............................ A . Influence of Sugar Solutions .......................... B . Influence of Fat .................................... C. Influence of a “Palatable” Fluid ...................... D . Summary .......................................... VI . Effect of Previous Ingestion of Ethanol .................... A. Effect of ad libitum Administration of Ethanol .......... B. Effect of Forced Administration of Ethanol .............. C . Summary .......................................... VII Endocrine Influences .................................... A. Influence of the Thyroid ............................. B . Influence of the Endocrine Pancreas ................... C . Influence of the Gonads .............................. D. Influence of Adrenal Glands .......................... E . Summary ..........................................

.

.

41

42 43 44 44 45 45 48 49 49 50 50 52 53 55 56 56 57 58 58 59 59 60 60 61 61 61 62 64 64 64 65 66 67 67

42

JORGE MARDONES

VIII. Effect of Drugs ........................................ A. Effect of Drugs Acting on the Central Nervous System . . . . B. Drugs Blocking Ethanol Metabolism ................... C. Experimental Liver Damage .......................... D. Other Drugs ....................................... E. Summary .......................................... IX. Significance of the Reviewed Facts on Human Alcoholism . . . . . A. Difference between High Level Free Choice of Ethanol and Pathological Desire ................................ B. Difference between High Level Free Choice of Ethanol and Inebriation ....................................... C . Free Selection of Ethanol and Withdrawal Symptoms . . . . . D. Determinative Causes of Drinking Alcoholic Beverages . . . . E. Indifference versus Aversion .......................... X. Summary ............................................. References .............................................

1.

68 68 69 69 70 70

71 71 71 72 72 73 74 74

Introduction

In several papers published from 1936 to 1941, Richter and co-workers reported the results obtained in experiments of selfselection of food by rats on normal and various pathological conditions. In one of these papers, Richter and Campbell (1940) reported on the free choice of ethanol solution by rats. Their results show that these animals are able to recognize ethanol in solutions of 1.8% and higher and that when they have a free choice of water and an ethanol solution, they prefer 5.1% concentration. Since other experiments show that rats refuse solutions of some alkaloids (morphine, strychnine), even when the doses that could eventually be absorbed were below threshold, the authors conclude that an important difference could be established between ethanol and other drugs (Richter, 1941). The general conclusion from the experiments of Richter and co-workers on self-selection of food was that rats are able to select the best combination of food for normal growth and reproduction (Richter et al., 1938) and food selections compensate for each of the pathological conditions studied, for instance: rats increase salt intake after adrenalectomy (Richter, 1936) and decrease carbohydrate consumption after pancreatectomy (Richter and Schmidt, 1941) and during vitamin B complex deprivation (Richter and Hawkes, 1940).

CHANGES IN FREE SELECTION OF ETHANOL

43

Following these ideas, Mardones and Onfray ( 1942) tested whether thiamine deprivation, which inhibits carbohydrate metabolism between CB and Cz compounds, could increase the alcohol preference exhibited by rats. The experiments showed that rats fed with a diet containing autoclaved yeast (source of the B complex free from thiamine) had an increased preference for ethanol; but when thiamine was supplemented, these rats did not decrease their voluntary alcohol intake. In contradistinction, when untreated yeast or dried liver were supplemented, a significant decrease of the alcohol intake was observed. The authors suggested then that a thermolabile compound different from thiamine is present in yeast and liver; they called this compound factor N. Afterward the same group of workers demonstrated that this factor is indeed a mixture of thiamine and a partially thermostable compound different from all others known at that time, which they called factor N1 (Mardones et al., 1948). Williams et al. (1949) emphasized individual variations of the voluntary intake of ethanol among rats and mice, either under normal or deficient diets, which led them to a concept that alcoholism is a “genetotropic disease.” Since then, it has been shown that alcohol preference of rats and mice could be experimentally modified by different means. The present review deals with the changes of the preference for ethanol that have been induced experimentally in rats and mice. The influence of experimental neuroses has been excluded intentionally from this review because this field is beyond the competence of the author.

II. General Procedure

The general procedure followed in the experiments is to offer rats and mice a free choice of water, various ethanol solutions, and a solid food. After 1 or 2 weeks the consumption of alcohol is sufficiently stabilized and the animals are ready for experiment. The experimental changes can affect either the animal or the selection conditions, i.e., free selection of ethanol, taste of the diet, offering a third choice. Obviously the experimentally induced changes must be considered in the interpretation of the results. When a supplement is added, the taste or smell of the diet can be changed, altering the conditions of the free selection. The

44

JORGE MARDONES

results thus obtained cannot be attributed to a systemic effect of the substance added without critical analysis. Water and ethanol solutions were offered in identical containers. Some workers interchange daily the position of water and alcohol containers in the cage, but this precaution proved unnecessary. The water and ethanol consumptions were usually measured daily. For practical reasons, the ethanol solutions were made by adding water to a known volume of 95% ethanol. Different solutions have been employed, but the most commonly used has been 10% v/v of 95% ethanol in water. The ethanol consumption was expressed usually in milliliters of 95% ethanol consumed daily per 100gm of body weight. Therefore, in this review, ethanol consumption will be expressed in this unit, by recalculation when necessary of the published data. During the experiments, animals were housed either individually or in groups. In the last ten years, due to the knowledge of the importance of individual variation, the grouping of animals in cages was less frequently used.

111.

Individual Variations

A. ANIMAL DIFFERENCES The amount of ethanol consumed under conditions of free choice varies in the different laboratory animals studied. Table I summarizes the range of alcohol intake of white rats (Rattus nmvegicus) and mice (Mus musculus) reported by several authors. It can be seen that the range of variation among rats is practically the same in different climates and strains. Referring the intake to body weight, it appears that mice consume higher amounts of alcohol than rats. This higher intake by mice has been also observed by Loiseleur and Petit (1947). The higher alcohol intake of this animal is correlated with a higher rate of ethanol oxidation. Alcohol metabolism in rats is of the order of 300 mg/kg/hour ( Segovia-Riquelme et al., 1956) while that of mice is about 700 mg/kg/hour (Marshall and Owens, 1955). Emerson et al. (1952) have also observed a great variation in the alcohol intake of rats (Sprague Dawley ), cotton rats (Sigmodon hispidus hispidus), hamsters ( Cricetus aumtus ), and deer mice ( Peromyscus californicus californicus) . They reported that ham-

45

CHANGES IN FREE SELECTION OF ETHANOL

sters and deer mice drink much more alcohol than rats and cotton rats. Forsander et al. (1958) have also reported that hamsters drink only ethanol solution and no water when they have a free choice of both. TABLE I RANGE OF VARIATION OF FREE-CHOICE ETHANOL CONSUMPTION BY RATS AND MICEFEDWITH STOCK DIET

Animal

Ethanol consumption, range (m1/100 gm b.w./day)

References

Rats Strain “0” Strain “H” White White

0.03-1.00 0.02-0.70 0.02-0.72

Williams et al. (1950) Williams et al. (1950) Mardones et al. (1949) Forsander et al. (1958)

Mice CF, and dba, 8 CF, and dba, 0

0.78-1.24 0.56-1.06

Mirone ( 1952)a Mirone ( 1952 ) a

0.07-0.75

0 Calculated from the reported data assuming that the weight of each mouse was 20.0 gm.

B. INTERINDIVIDUAL VARIATION OF RATS The interindividual fluctuations of the voluntary alcohol intake of rats has been observed independently by Williams et al. (1949) and by us. Mardones et al. (1949) studied these variations systematically in rats fed either a stock diet or a purified diet lacking factor N1 (see Section IV, C). Rats were grouped according to their free-choice alcohol intake by class intervals of 0.20 ml 95% ethanol per 100gm of body weight per day. The frequency distribution obtained is represented by the histogram (Fig. 1 ) . The asymmetry of this distribution suggests that the population of rats in our colony is not homogeneous, i.e., it may include rats of different genetic structure.

C. GENETICORIGINOF INDIVIDUAL VARIATIONS IN RATS The genetic origin of the individual variations of voluntary alcohol intake has been studied in rats fed with a purified diet in our laboratory since 1948. Since we desired to increase the proportion of rats exhibiting a high level of alcohol intake in

46

JORGE MARDONES

a

W

a t-

2 W

V

a W a

0.0

0.2

0.4

0.6

0.8

1.0

DALILY ALCOHOL INTAKE (m1/100q b.W.1

FIG. 1. Histogram showing the distribution of a colony of rats classified according to the free choice of ethanol. White area: rats fed with stock diet; dotted area: rats fed with purified diet lacking factor N,.

U

FIG.2. Pedigree of strain A of rats (nondrinkers). Squares represent males and circles females. Open ones: alcohol intake less than 0.20ml 95% ethanol by 100 gm body weight per day. Black ones: alcohol intake 0.60 ml or higher. Crossed ones: 0.20 to 0.59ml.

CHANGES IN FREE SELECTION OF ETHANOL

47

order to accelerate our study on nutritional factors, we selected “drinker” rats for inbreeding. The first evidence of a genetic origin of differences in alcohol preference was obtained by the study of the offspring of a female rat drinking daily an average of 0.73 t 0.03 ml of 95% alcohol per 100gm of body weight

FIG.3. Pedigree of strain B of rats (drinkers). Same key as Fig. 2.

mated with a male consuming 0.43 z k 0.04 ml. The distribution of the four litters (14 males and 12 females) according to their alcohol consumption, differed significantly from the standard distribution of the colony (Mardones et al., 1950a). Figure 2 represents the pedigree of the strain A (nondrinker) and Fig. 3 those of the strain B (drinker). The statistical analysis of the results obtained in the generations 3 to 7 in both strains was studied by calculation of the coefficient of heredity (Mardones et al., 1953a).

48

JORGE MARWNES

This was +0.406, with t = 8.72, demonstrating that it is extremely improbable that this distribution resulted by chance. From these figures it is evident that heredity has an important role in alcohol preference of rats. Soon we shall have enough genetically pure adults belonging to the alcoholic and nonalcoholic strains to begin hybridization.

D. “DRINKER” AND “NONDRINKER” RATS A comparison of our strains A and B has yielded no other characteristic to account for this preference or nonpreference for ethanol as shown in the following studies. Growth rates are the same for both strains as shown in Table 11. The rats of both TABLE I1 ADULTWEIGHT OF MALEAND FEMALE RATS OF STRAIN A AND B

Sex

Strain

Number of cases

A 295 B 157 A 304 B 202 * Arithmetic mean 2 its standard error. f Standard deviation.

Males Males Females Females

Adult weight in grams (age 6 months) m 2 sm*

S.D.f

279.0 k 3.22 273.4 & 4.49 201.0 & 1.96 197.8 & 2.17

57.1 56.3 34.2 30.8

strains were fed with stock diet from weaning till 45 to 60 days, and then with purified diet and free access to ethanol solution 10% v/v. The rate of oxidation of ethanol in both strains has been studied by Segovia-Riquelme et al. ( 1956). Ethanol-l-Cl4 was injected in doses of 0.5, 2, and 3 gm per kilogram of body weight, and hourly activity of the excreted COs was measured during a 6-hour period. No difference in the rate of combustion of ethanol between the groups was observed. The same study was performed in our laboratory (to be published) with ethanol-2-C14 yielding similar results. It was postulated that the increase of voluntary alcohol intake might be due to a block in the metabolic pathway between C3 and Cs compounds. To test this hypothesis the rate of oxidation of pyruvate in both strains was studied by injecting l-C14 labeled pyruvate and measuring the activity of the excreted CO, (Segovia-

CHANGES I N FREE SELECTION OF ETHANOL

49

Riquelme et al., to be published). The results obtained did not show a difference in the oxidation of the carbon 1 of the pyruvate by rats of either strain. Experiments are now proceeding with pyruvate-2-C14. In summary, no metabolic difference between “drinker” and “nondrinker” rats has been found as yet. Dember and Kristofferson (1955) have observed a negative correlation between free-choice alcohol consumption and latency time for the onset of audiogenic seizure in rats. These authors interpret their results by assuming that the “individual differences in alcohol consumption are generated by differences in emotionality.”

E. INTRAINDIVIDUAL VARIATIONS IN RATS In most of the experiments performed on the voluntary alcohol intake each rat acts as its own control. Therefore it is important to analyze the magnitude of the spontaneous fluctuations observed in each individual. For this purpose the daily variation of alcohol intake (Table 111) as well as the variation of the mean of consecutive periods of 10 days (Table IV) was analyzed. These data obtained at random from our files show that the amplitude of spontaneous daily variation is important and should be statistically evaluated. The means of three consecutive periods of 10 days (Table IV) show little variability, but the standard deviations are large. These results emphasize the necessity for large numbers of animals per group.

F. SUMMARY The amount of alcohol consumed per body weight is dependent upon the laboratory animal used. Mice and hamsters consume greater amounts than rats and cotton rats. This difference appears to be related to the rate of combustion of ethanol. Among rats, there are wide interindividual variations, even when they are submitted to the same environmental conditions. This variation is of genetic origin. By artificial inbreeding two strains of rats, A (“nondrinker”) and B (“drinker”) have been isolated. No metabolic differences between the “drinker” and “nondrinker” rats have been found as yet. The intraindividual variations in rats are important enough to make statistical analysis of data necessary.

50

JORGE MARDONES

TABLE I11 DAILYINTRAINDIVIDUAL VARIATIONS OF THE VOLUNTARY OF RATS ALconoL INTAKE Statistical data Mean alcohol intake (ml/lOO @day 1 0.00-0.19 0.20-0.39 0.20-0.39 0.40-0.59 0.40-0.59 0.60-0.79

Diet

Number Mean of of rats means

Stock Stock Purified Stock Purified Stock

48 12 22 8 9 4

0.06 0.27 0.30 0.51 0.49 0.71

Mean of standard deviation

Mean of coefficient variation

2 0.06 2 0.13 2 0.10 f 0.18 f 0.12 2 0.20

(%) f 111

47 f 37 2 34 & 25 ~ f :27

&

TABLE IV SPONTANEOUS VARIATIONS OF THE MEANOF THE ALCOHOLINTAKE ON RATS FEDWITH WRIFXED DIET OBSERVED Range of alcohol intake (m1/100 gm/day)

Number of rats

0.20-0.39 0.40-0.59 0.60- up

79 27 5

m2-m1* Mean 4 . 0 1 1 f 0.011 -0.019 2 0.024 -0.034 f 0.078

m3-m1*

S.D.t f0.095 f0.123 f0.175

Mean -0.016 -0.036 -0.138

k 0.012

2 0.027 f 0.023

S.D.t 20.105 20.141 20.051

* ml, m2, and m3 represent the means of the alcohol intake during each of 3 consecutive periods of 10 days in each rat, expressed in milliliters of 9570 ethanol per 100 gm body weight per day. Standard deviation.

+

IV. Effect of Diet Studies concerning the effect of diet on free choice of ethanol deals with organic composition, vitamin deprivation, presence of other nutrients, and restriction of amount given.

A. ORGANIC COMPOSITION OF DIET The results of a study of the influence of changes in organic composition of the diet (from a basic one to a carbohydrate-rich or to a carbohydrate-poor one) on the voluntary alcohol intake of rats were reported by Mardones et al. (1955). No significant

51

CHANGES IN FREE SELECTION OF ETHANOL

change of alcohol intake was observed by any of these changes. All the diets were given ad libitum. The composition of the tested diets appears in Table V. The differences are not only in the TABLE V OF TESTED DIETS COMPOSITION

Dieta Sucrose Commercial casein Vegetable oil Salt mixture

Basic

Low carbohydrate

High carbohydrate

(%I

(%)

(%I

64 18 14

29 39 28 4

81 9 6 4

4

Hydrosoluble vitamins were given separately and liposoluble vitamins were incorporated in the vegetable oil. @

carbohydrate proportion but also in protein and fat. After a period on the basic diet, the rats were maintained for 20 days on the experimental diet and then returned to the basic one. The results summarized in Table VI show that these changes did not modify significantly the voluntary alcohol intake. TABLE VI INFLUENCE OF CHANGES OF THE DIETON THE VOLUNTARY INTAKE OF RATS ALCOHOL

Changes of the diet

Number of rats

Changes of the alcohol intake (mV100 gm/day)

Basic to high carbohydrate Basic to low carbohydrate High carbohydrate to basic Low carbohydrate to basic

11 6 11 6

- 0.057 +- 0.020

+ 0.029 t 0.031

- 0.084 2 0.036 - 0.020 +- 0.014

The conclusion of Lester and Greenberg (1952) that rats consumed more alcohol on a 46% than on a 63% carbohydrate diet was based on a comparison of ethanol consumption in a group of 10 rats. The two trials were separated by a period of a few months during which several changes in self-selection conditions were introduced. The presence of intraindividual spontaneous variations of alcohol intake of rats, discussed in section 111, E would appear to weaken this conclusion.

52

JORGE MARDONES

The effect of the organic composition of the diet on voluntary consumption of ethanol was studied by Mirone (1957) on black C57 mice of both sexes submitted to similar conditions. The analysis of the results was complicated by the introduction of an intermediary period during which mice received only 5% ethanol. To simplify the analysis, it is convenient to consider only the differences observed by changing the basic diet to the experimental one. This analysis shows that the alcohol consumption increases significantly when the basic diet is changed to high protein, low carbohydrate (60% casein, 18% sucrose). It appears also that alcohol intake decreases with a high-fat, low-carbohydrate diet (45% lard, 23% sucrose). The changes observed with highcarbohydrate, protein-free, or fat-free diets seem nonsignificant; however, the statistical data reported are insufficient to allow definite conclusions.

B. THIAMINEDEPRIVATION An increase of voluntary alcohol consumption as a consequence of thiamine deprivation has been observed by Mardones and Onfray (1942) in rats fed with a purified diet containing autoclaved yeast as vitamin B supplement. This fact has been confirmed by Beerstecher et al. (1951) using pursed diet supplemented with pure vitamins excluding thiamine. An indirect evidence for the influence of thiamine deprivation on alcohol intake is given by the observations of Brady and Westerfeld (1947) and of Williams et al. (1950). Both groups of workers observed that when the alcohol intake was increased in rats fed with a diet lacking thiamine and other vitamins, the supplement of a mixture of pure vitamins including thiamine induced a decrease of alcohol consumption. Mardones and Onfray (1942) reported that no decrease of ethanol consumption was observed when rats fed with the aforementioned diet received a supplement of thiamine. The difference between these results and those reported by other authors should be explained by the fact that Mardones and Onfray gave thiamine after an experimental period which was long enough to obtain the depletion of other factor( s ) (see Section IV, C ) . Mirone (1957)did not observe a change of alcohol consumption in black CS7mice after changing from a basal diet to a

53

CHANGES IN FREE SELECTION OF ETHANOL

thiamine-free diet. However these results are not conclusive because there is no indication that coprophagy was prevented-a necessary factor for the induction of thiamine deficiency.

C. FACTOR N, The absence of an effect of thiamine on the alcohol intake of rats fed with purified diet containing a supplement of autoclaved yeast supported the assumption that a thermolabile factor other

2

'

L

-

0

g 100 v)

O

2 W 0 K

-

w

-

E

50-

W Y

-

a.

U

+ 5

-1

-

0

-

0

-

I

0

a

07

0

0

0 8

I

,

I

FIG. 4. Relationship between the doses of a liver extract on the voluntary alcohol intake of rats fed with a purified diet lacking factor N, (semilogarithmic).

than thiamine played a role in the free selection of ethanol. This factor was called factor N (Mardones and Onfray, 1942). The relationship between the doses of a liver extract and the decrease of the voluntary alcohol intake of rats fed with this deprived diet was studied by Mardones et al. (1946). Their results show that the effect rises with increase in dosage. Figure 4 exhibits the relationship between this effect and the logarithm of the liver extract doses calculated from the data reported on the afore-mentioned paper.

54

JORGE MARDONES

Shortly after, Mardones et al. (1947), reported that when the diet of rats deprived of factor N was supplemented with thiamine, a slow decrease of voluntary alcohol intake was observed. This decrease began around 5 weeks after starting the supplement and reached the control level 7 weeks later. Furthermore, when rats fed with a stock diet were shifted to a purified diet containing autoclaved yeast plus thiamine, they did not increase their voluntary alcohol intake. As a result, it was realized that the factor we had called N was not really a new thermolabile factor, but a mixture of thiamine ( thermolabile ) and another substance which is only partially destroyed by autoclaving the yeast. In order to avoid confusion, the last one was called factor N1 (Mardones et al., 1948). The evidence that N1 is a new factor that prevents an increase in alcohol consumption by rats, is given by the higher alcohol intake observed in rats fed with a purified diet supplemented with a mixture of pure vitamin B complex, including thiamine. Table VII summarizes the data on alcohol intake observed in rats fed with a purified diet concerning different supplements. TABLE VII TO A PURIFIED DIET^ EFFECTOF VARIOUSSUPPLEMENTS VOLUNTARY ALCOHOLINTAKE OF RATS?

Supplement Untreated yeastd Autoclaved yeaste Autoclaved yeast Pure B vitamins9

+ thiaminef

Number of groups0 50 164 9 16

ON THE

Alcohol intake 95% ml ethanol per day (mean k standard error) Per rat

Per 1 O O g m b.w.

0.18 -c- 0.02 0.70 rt 0.02 0.21 k 0.05 0.57 -+ 0.04

0.13 -c- 0.01 0.64 -c- 0.02 0.13 rt 0.03 0.35 k 0.03

0 Casein, 20; sucrose, 60; vegetable oil enriched with vitamins A, D, and E, 15; Osborne and Mendel salt mixture, 5, per cent. b Mardones et al. (1948). c Groups of 4 to 6 rats. d Commercial dry brewers' yeast, 10 gm per 100 gm of basic diet. e Commercial dry brewers' yeast autoclaved 90 min, 125"C, pH 9, 10 gm per 1OOgm of basic diet. f Four micrograms per 100 gm body weight per day. 0 Daily per 100 gm body weight: thiamine, 4 pg; riboflavine, 25 pg; calcium pantothenate, 10 pg; pyridoxine, 10 pg; niacin, 0.5 mg; and choline chloride, 1.0 mg.

CHANGES IN FREE SELECTION OF ETHANOL

55

Rats depleted of factor N1 do not always decrease their voluntary alcohol intake after receiving a supplement of liver or yeast. About one-third of the rats of our strain “B” did not respond, and represent another kind of individual variation. No clear hereditary transmission of this behavior has been observed. This fact is a handicap for the bioassay of factor N1, since each rat that does not exhibit a response to a testing product must be treated with liver in order to determine whether or not its voluntary alcohol consumption is refractory to liver supplement. Our attempts to isolate factor N1 have been unfruitful as yet. The richest sources are yeast ( Sacchuromyces cerevisae and Torulopsis utilis), liver, and meat.

D. THIOCTIC ACID The 6,8-dithiooctanoic acid, known by the names of thioctic or a-lipoic acid, is a factor needed by some microorganisms. Functionally it acts as a coenzyme in the decarboxylation of pyruvic acid, resulting in an unstable combination with acetaldehyde ( active acetaldehyde). The susceptible microorganisms are not able to grow in this factor-free medium when the metabolite is glucose or pyruvate, but they grow freely when the metabolite is acetate. Since the action of thioctic acid is similar to that of thiamine on the pyruvate metabolism and its natural sources are the same as factor N1, we tested the effect of this substance on the free choice of ethanol of rats fed with a factor N1 free diet (Mardones et ol., 1954). The results show that thioctic acid given during a period of 4 to 10 days in total doses of 62.5 to 750 pg per 100 gm body weight, induces a slow decrease of the alcohol consumption, which was significant at the 2% level in a period from 11 to 20 days after starting treatment. The subsequent administration of dry liver induced a more marked decrease in alcohol consumption. Since the thioctic acid present in liver supplement is lower than the amount given as a pure substance, it is clear that the effect of liver is not due solely to its thioctic acid content. The latency observed in the onset of action of thioctic acid suggests the possibility that the effect could be mediated through the intestinal flora. This is supported by the observation that prior and simultaneous administration of sulfasuxidine prevents

56

JORGE MARDONES

the effects of thioctic acid, but not that of liver (Mardones et al., 195313).

E. OTHERSINGLEVITAMINDEFICIENCIES Beerstecher let al. (1951) have studied the effect of deprivation of single vitamins on the free selection of ethanol by rats. They observed that the single deprivation of riboflavine, pyridoxine, or pantothenate induced an increase of alcohol intake, which was re-established to the basic level after restoration of the vitamin. No clear results were obtaked with vitamin A, biotin, and choline chloride. The same kind of experiments were performed by Mirone (1957) on black C57 mice. She observed no significant change in the free selection of ethanol by shifting the animals from basic diet to another deficient in pyridoxine or pantothenic acid. Her results are subject to the same criticism discussed in Section IV,B. Williams et al. (1950) observed that in some rats fed with purified diet supplemented with autoclaved yeast, the administration of an antipernicious anemia liver preparation, after, or with, a supplement of 10 pure vitamins, induced a decrease of the alcohol intake. The effect of vitamin BIZ on rats fed with a purified diet supplemented with pure vitamins was studied systematically by Mardones et al. (1952). These authors observed that cyanocobalamine in doses as high as 10 pg every other day for 20 days, did not alter significantly the alcohol consumption. These results suggest that the effect observed by Williams et al. could be the consequence of factor N1 present in the liver preparations.

F. MULTIPLE VITAMINDEFICIENCIES The effect of a basic diet free from all the hydrosoluble vitamins and vitamins E and K was studied by Brady and Westerfeld (1947). They observed that the alcohol intake of rats fed with this diet increased at a higher rate than with any other deficient diet, Similar results were reported by Mirone (1957) on black CS7mice fed with a diet deficient in all the B complex vitamins, in contrast to the absence of effect that she observed with single deficiencies, Delore and Berry (1955) observed that rats fed with a diet composed of purified casein, 10%; starch, 30%; peanut oil, 7.5%; salt mixture, 2.5%; and water, 50%, increased their

57

CHANGES IN FREE SELECTION OF ETHANOL

consumption of 4% v/v ethanol solution offered as a free choice. The alcohol intake reverted to the basic level when the rats returned to the stock diet.

G. GLTJTAMINE In 1955 Ravel et al. reported that the toxicity of alcohols on Streptococcus faecal& was prevented by some liver extracts. These authors identified glutamine as the substance responsible for this effect. Rogers et al. (1955) tested the effect of glutamine on the voluntary intake of ethanol by rats. They reported that a diet adjusted to provide 100 mg of glutamine per day significantly reduced the alcohol intake of rats. The subcutaneous injection of glutamine in equivalent doses did not, however, change the ethanol consumption. This difference could be explained on the basis that the food was more palatable. To test this possibility Mardones, Segovia-Riquelme, Hederra, and Alcaino ( 1956, unpublished) studied the effect of glutamine given orally, but separate from the food, in daily doses of 10 mg and 20 mg per 100 gm of body weight. The decrease of alcohol intake was not as marked as that obtained by the above mentioned authors, but the difference was statistically significant. As far as dose is concerned, our experiments employed lower doses than those given by Rogers et al., since 100 mg per day represents about 40 mg per 100 gm of body weight. Table VIII summarizes the results obtained by Rogers et al., and by ourselves. These data show that the effect of glutamine rises linearly with the logarithm of the dose, and they support the idea that the effect of this substance is not the consequence of a change in the taste of the food. TABLE VIII EFFECTOF GLUTAMINE ON THE VOLUNTARY ALCOHOL INTAKE OF

RATS

~

Days Change of the No. of Days Daily dose of alcohol intake of supple- of Basic glutamine m1/100plm b.w. % Ref. rats ment study diet (mg) 1 20 -0.074 i:0.038 -14 10 Purified 10/100 gm b.w. 11 1 20 -0.111 k 0.032 -25 10 Purified 20/100 gm b.w. 18 2 10 26 26 -0.202 t 0.067 -38 Stock 100hat 3 4 3 42 42 -0.15 10 Stock lOOJrat REFERENCES: 1. Mardones, Segovia-Rquelme, Hederra, and Alcaino ( unpublished data); 2. Rogers et al. (1955); 3. Rogers et al. (1956).

58

JORGE MARM)NES

Other amino acids closely related to glutamine do not induce the same effect on the voluntary alcohol intake. Rogers et al. (1956), using the same method, obtained negative results with glutamic acid, monosodium glutamate, asparagine, and glycine.

H. RESTRICTIONOF FOOD Westerfeld and Lawrow (1953) studied the changes in voluntary alcohol consumption in rats on restricted food intake. They reported that with a restriction to 75% of normal no significant change in alcohol intake occurred; but with a restriction to 50% of the normally consumed food, a marked increase of alcohol intake was observed. This extra ethanol supplied about 40% of the withheld calories. Since thiamine deficient rats ate much less than normal, these authors did paired-feeding experiments on rats with a “complete diet” (minus factor N1)and a thiamine-free “complete diet.” Results show that alcohol intake was increased in both groups. These results suggest that increased alcohol consumption induced by thiamine deficiency is the consequence of a decreased food intake; however, the experimental findings give no evidence that this is the only mechanism.

I. SUMMARY Some dietary changes alter the free choice of ethanol in laboratory animals. Diets with different organic composition induce only slight changes in voluntary alcohol consumption. Single deprivation of thiamine, riboflavine, pyridoxine, pantothenic acid, and unknown factor N1,result in an increased alcohol intake. A greater effect is obtained when the diet is deficient in more than one of these vitamins. Vitamin A, biotin, choline, or Blz deficiencies do not induce clear changes in ethanol preference. The administration of thioctic acid to rats fed with factor N1 free diet induces a decreased alcohol intake. This effect is not observed when sulfasuxidine is given simultaneously. Glutamine given orally in doses of 10 to 40 mg per 100 gm of body weight decrease ethanol consumption. The magnitude of this effect appears to be correlated with the logarithm of the dose. Some other amino acids related to glutamine (glutamic acid, asparagine, glycine) do not effect alcohol preference. The restriction of food to 50% of normal results in an increased ethanol intake which supplies approximately 40% of the calories withheld.

59

CHANGES IN FREE SELECTION OF ETHANOL

V.

influence of a

“Third Choice”

In 1952 Lester and Greenberg reported that a “third choice” of sugar, fat, or saccharin solution offered to rats having a free choice of water and ethanol solution immediately reduced aIcohol consumption. A. INFLUENCE OF SUGARSOLUTIONS The results reported by Lester and Greenberg (1952) and by Mardones et al. (1955) regarding the influence of sugar on voluntary alcohol intake of rats are summarized in Table IX. The deTABLE IX EFFECTOF “THIRDCHOICE”OF SUGARSOLUTIONS ON THE VOLUNTARY ALCOHOL INTAKEOF RATS Number

of Sugar Sucrose Sucrose Sucrose Sucrose Sucrose Sucrose

11.5%) 10% ) 30% ) 70%) solid ) solid)

Dextrose ( 10% ) Dextrose ( 30% )

rats

10 10 9

10 10 38

10 9

Alcohol inStatistical take change significance ( % of basic intake) (P)

- 82 - 48 - 59 - 53 -34

+ 19

- 47 - 58

< 0.001 < 0.01 < 0.001 < 0.001 < 0.05 > 0.1 < 0.01 < 0.001

Ref. 1 2 2 2 1 2

2 2

REFERENCES:1. Lester and Greenberg (1952); 2. Mardones et al. (1955).

crease of ethanol consumption induced by offering 10 to 70% sugar solutions (sucrose or dextrose) is an established fact. The effect of solid sucrose as a third choice appears controversial. The difference observed by Lester and Greenberg was obtained by comparing alcohol intake during two periods separated by an interval of 45 days during which fat and sucrose solution were offered. The intraindividual variation discussed in Section II1,E as well as the low level of statistical significance diminishes the degree of accuracy of these figures. Our results have been obtained by comparing two consecutive periods. If the same value is given to all data reported and all cases are considered as a single group, the difference does not appear

60

JORGE MARDONES

significant. It was important to establish whether or not there is a real difference in alcohol consumption after the addition of a free choice of solid sucrose. At first glance, it could be assumed that the basic diet is poor in carbohydrates and consequently the rats increase that proportion by drinking the sugar solutions. But this supposition is contradicted not only by the absence of an effect of solid sucrose, but also by the fact that rats receiving their diets separated in two parts (sucrose and nonsucrose moiety) select a combination of both with a carbohydrate proportion of around 30% instead of the 64% present in the basic diet (Mardones et al., 1955). In addition, the amount of sugar solution consumed depended upon the sugar concentration. The daily intake of 30% sugar solution was lower than the 10% and higher than the 70% solution. The total sugar ingested daily was the same when solutions of either 30 or 70% sucrose were offered, 2.9 f 0.02 and 2.9 A 0.35 gm per 100 gm of body weight, respectively, and 2.7 & 0.55 gm for dextrose 30%. The decrease of alcohol intake was not correlated with the volume of fluid consumed from the third choice ( T = 0.066, not significant); but it is positively correlated with the amount of sugar ingested ( T = +0.512, P < O . O l ) .

B. INFLUENCE OF FAT The effect of a third choice of an emulsion of 5.8% Wesson oil homogenized in 1% gum acacia was studied by Lester and Greenberg (1952). The decrease in alcohol intake was less marked than when a 11.5% sucrose solution was offered. The intake diminished to 49% of the basic level; this difference was statistically significant ( P < 0.001).

c.

INFLUENCE OF A “PALATABLE’’

FLUID

In order to study whether decreased alcohol consumption, following the third choice of sucrose solutions, was the consequence of the extra calories absorbed, Lester and Greenberg (1952) tested the effect of a 0.09% saccharin solution which had the same taste but was without any energy value. The result was a signscant decrease of alcohol intake to 41% of the basic intake ( P < 0.001). This result does not agree with that of Mardones et al. (1955),

CHANGES I N FREE SELECTION OF ETHANOL

61

who used as a “palatable” fluid a solution of pure B vitamins. The ingestion of 20 or 40 ml of this solution per day did not change significantly the alcohol intake. This fact is in agreement with the absence of a correlation between fluid intake and the decreased alcohol consumption reported in the same paper. The difference in the results obtained with two types of fluid, both liked by rats (one with and the other without a sweet taste), suggests that the sweet taste of saccharin given after sucrose solution evokes a different mechanism (conditioned?) which is absent when a nonsweet fluid is given to a rat as the unique third choice.

D. SUMMARY The offering of a “third choice” of sugar solutions or fat emulsion to rats having free choice of water and an ethanol solution, induces a decreased alcohol intake that is positively correlated with the amount of nutrient ingested and not with the volume of fluid taken as the third choice. A difference has been observed between the effect induced by nonnutritive saccharin given to rats that have received sugar solution previously and a solution of pure B vitamins without energy value. The former induced a significant decrease of alcohol intake, while the latter did not change it significantly.

VI.

Effect of Previous Ingestion of Ethanol

Since human alcoholism can be considered a pathological condition created by the continuous use of alcoholic beverages, it is important to determine whether an abnormal preference for alcohol can be induced in laboratory animals by continuous administration. A. EFFECTOF ad libitum ADMINISTRATION OF ETHANOL

The question that rats maintained for a long time under ad libitum ethanol conditions would increase their preference for this drug, was raised by Brady and Westerfeld (1947). They observed that a diet which prevented an increased alcohol intake, did not induce a permanently decreased alcohol consumption

62

JORGE MARDONES

when it was given to rats whose intake had reached a high level due to a deficient diet. They pointed out that: “Once a high level alcohol consumption has been established, it may be that nonnutritional factors complicate the picture.” Another explanation for the difference may be the fact that a preventive dose of a certain factor is significantly lower than a curative dose. To test this hypothesis, Mardones et al. (1950b) studied the effect of a basic diet supplemented separately with untreated dried yeast (Torubpsis utilis) given ad tibiturn for 60 days. They observed that a curative effect was maintained throughout this period. This experiment shows that high doses of factor N1 induce a permanent decrease of alcohol intake. Thus, there is no reason to believe that, in the afore-mentioned conditions, nonnutritional factors complicate the picture.

B. E F F EOF ~ FORCED ADMINISTRATION OF ETHANOL In a report in which he summarized all his experiments with alcohol intake by rats, Richter (1957) pointed out that his attempts to produce addiction by forcing domesticated rats to drink alcohol in concentrations of 8, 16, and 24% for periods of 9 to 15 months were unsuccessful. Neither an increase of voluntary alcohol intake nor any sort of withdrawal symptoms were observed. He believed that alcohol addiction was produced in 3 wild Norwegian rats; however, a figure representing the water, alcohol, and food consumption of these rats was not convincing. Prieto et al. (1958) studied the influence of forced administration of ethanol 10, 20, and 30% during periods of 1 to 5 weeks, on voluntary alcohol intake of white rats fed with stock diet. The effect was evaluated by a comparison of the alcohol consumption in the period that followed the forced ingestion with the period that preceded it. The results are summarized in Table X. The data show that a period of forced ingestion of ethanol did not induce an increased preference. Rats tolerate 10 or 20% alcohol as their only source of fluid, but not 30% solutions. For that reason the period of forced ingestion with this concentration lasted only 1 week. During the periods with forced alcohol at 20 and 30% the rats drank significantly less fluid than previously, contracting some sort of fluid debt, because in the subsequent period during which they had free access to water, they drank an extra volume of fluid.

TABLE X BY RATS SUBMITTED TO A PERIODOF FORCED INGESTION OF 10, 20, ETHANOL CONSUMPTION 30% v/v 95% ETHANOL Experimental Number

of Strain A A A A B B

B

Ethanol solution

12

10 20 20 30

2 4 1

8 11 9

20 20 30

2 4 1

0.42 f 0.07 0.44 2 0.06 0.39 f 0.04

7

( % v/v)

5

z

c)

!2

Alcohol intake (mVl00 gm b.w. ) 10 Preceding days 0.08 & 0.01 0.06 & 0.02 0.09 & 0.03 0.06 2 0.02

rats 18 7

Duration (weeks)

AND

Experimental 0.87 f 0.02 0.94 f 0.14 1.21 2 0.08 0.76 f 0.12

10 Subsequent days 0.04 2 0.01 0.22 -c 0.09 0.17 t 0.06 0.06 f 0.01

1.21 f 0.14 1.34 f 0.07 1.10 f 0.06

0.28 2 0.10 0.25 t 0.06 0.20 0.09

*

2 irn2

M

4!

g 2

E9

Q,

w

64

JORGE M A R W N E S

Mirone (1952) studied the effect of offering 5% ethanol as the fluid source on the subsequent voluntary alcohol intake by mice ( CFr and dba). She observed that these mice exhibited a higher alcohol intake than their littermates fed with water. Since no data of interindividual and intraindividual variations of these mice were reported, no statistical evaluations can be made.

C. SUMMARY No increase in ethanol intake has been induced in domesticated rats by giving this drug either freely or forced for long periods. The changes in voluntary alcohol intake induced by this procedure in wild Norwegian rats and in mice are questionable. VII. Endocrine Influences The influence of some endocrine factors on the free choice of alcohol consumption has been studied in rats and other laboratory animals. Extensive studies have been made on the thyroid, pancreas, and gonads. A. INFLUENCE OF THE THYROID

The most controversial results are those related to the influence of the thyroid. Mardones and Segovia (1951) reported that thyroxine in daily doses up to 50 pg did not change significantly the alcohol intake of rats; but doses of 100 pg per day produced a significant increase. Rosenberg and Zarrow (1952) and Zarrow and Rosenberg (1953) studied the effect of propylthiouracil added to the diet. They observed that 0.8% concentration of this drug resulted in immediate increase in the voluntary alcohol consumption of rats. Lower concentrations (0.2 and 0.4%) had the same effect but with a gradual onset and a less pronounced increase. The results led the authors to believe that the thyroid played no role in this effect. In the first place, the change of the alcohol intake appeared before any effect on thyroid function could be detected; second, the effect disappeared immediately after the discontinuation of the drug, before the reestablishment of the thyroid activity could be expected; third, the administration of thyroxine during the period with propylthiouracil treat-

CHANGES I N FREE SELECTION OF ETHANOL

65

ment did not change the high level of alcohol intake, and fourth, thyroidectomy did not induce the same effect as propylthiouracil. Since propylthiouracil is a bitter substance the observed effects might be due to the animals’ refusal of unpalatable food. Richter (1957) has shown that when the diet was made bitter with quinine, the voluntary alcohol intake of rats increased significantly. However in this instance with propylthiouracil, the increase of alcohol intake was more marked than could be explained by a decreased food intake. Richter (1956) reported some cases of rats in which the administration of a diet containing 0.04 to 0.1% of desiccated thyroid decreased the alcohol consumption. Concentrations of 0.02% or lower did not change it. Because of the spontaneous intraindividual variations observed in rats (see Section II1,E) this kind of change could be observed spontaneously in a certain proportion of cases. The influence of the return to the basic diet has not been studied in any of the reported cases. The influence of adding thyroid powder (0.5%) to a stock diet was studied by Prieto et al. (1957) in 18 “nondrinker” and in 3 “drinker” rats. The changes observed during the third week of thyroid treatment in the “nondrinker” rats were an increase from 0.05 .t 0.007 to 0.22 f. 0.060 ml 95% ethanol per 100 gm of body weight. This difference is statistically significant at the 2% level. The changes observed in the 3 “drinker” rats in the same periods were 0.23 to 0.27, 0.25 to 0.57, and 0.66 to 1.20. These changes were associated with an increased water intake. As it can be seen, these results do not agree with those reported by Richter ( 1956). Concerning the effect of thyroidectomy on the alcohol intake of rats, Richter (1957) reported one case in which thyroidectomy seemed to increase the alcohol intake. Unfortunately the effect of thyroxine or of thyroid feeding in this case was not studied. The effect of shifting black mice C5, from a basic diet to an iodine deficient diet was studied by Mirone (1957) who observed a significant decrease of alcohol intake. It is possible that this diet could affect the thyroid function.

B. INFLUENCE OF THE ENDOCRINE PANCREAS Mardones and Segovia (1951) reported that partial pancreatectomy induced in rats a slow decrease of alcohol intake follow-

66

JORGE MARDONES

ing which the administration of insulin (2.5 I.U. per 100 gm of body weight) produced a rise whereas the subcutaneous injection of 2.5 to 10.0 I.U. of insulin per 100 gm of body weight in normal rats produced a nonsignificant increase. Independently, Forsander et al. (1958) studied the role of the pancreas in the voluntary intake of ethanol by rats. In 4 rats exhibiting a daily spontaneous alcohol intake lower than 0.30 ml 95% ethanol per 100 gm of body weight the administration of an antidiabetic sulfa ( N1-Sulfanilyl-N2-butylcarbamide-Nadisan ) added to the water and to a 15% v/v ethanol solution in concentration of 2 gm per liter, increased significantly the alcohol intake. This consumption rose gradually and, after stopping the drug, decreased slowly. In one rat exhibiting a basic consumption of 0.69 ml per gm of body weight per day, no significant change was observed. Insulin ( 1 I.U. of protamine zinc insulin daily) increased significantly the alcohol intake ( P < 0.001) in a group of 25 rats whose basic consumption ranged from 0.02 to 0.46 ml per 100 gm of body weight per day with a mean of 0.15. In the same paper the author reported that alloxan (15 mg per 100 gm of body weight, intraperitoneally) induced a significantly decreased consumption of ethanol. This effect could be correlated with the fact that the toxicity of alcohol was augmented during alloxan diabetes, as observed in mice by Hiestand et nl. (1953).

C . INFLUENCE OF THE GONADS Shadewald et nl. (1953) comparing the preference for alcohol in a group of 40 Sprague Dawley rats, observed that males exhibited a higher preference than females, estimated either by the ratio (ethanol so1ution:water) or by the alcohol intake per body weight. Even though this difference is statistically significant ( t = 2.7, P < 0.02), due to the importance of interindividual variations observed in rats (see Section II1,B) it is difficult to decide whether or not the groups were genetically comparable. We have studied (unpublished data) whether a sex difference existed in groups of littermates from our inbreeding experiments, and observed no significant difference between the alcohol intake per body weight of males and females. Mardones and Segovia (1951) reported that gonadectomy did not change significantly the alcohol consumption of either sex.

CHANGES I N FREE SELECTION O F ETHANOL

67

Shadewald et al. (1953) performed gonadectomy in 5 males exhibiting high ethanol intake and in 5 females exhibiting a low one. They observed a slight tendency toward decreased consumption by the males and an increased consumption by the females. Mardones and Segovia (1951) observed that testosterone did not change alcohol intake in gonadectomized males and females. However, diethylstilbestrol induced a significantly decreased alcohol intake in either normal or gonadectomized males and females. Emerson et d. (1952) reported that high doses of estradiol benzoate induced a very slow decrease in alcohol consumption of deer mice, while massive doses of progesterone did not change the preference for ethanol in the same species. The influence of pregnancy and lactation on the voluntary alcohol intake was studied in hamsters by Carver et al. (1953). They observed that 1 to 3 days before the end of gestation alcohol as well as total fluid intake decreased significantly. Furthermore, during lactation the alcohol preference was also reduced, expressed either by the ratio (ethano1:water) or by the total ethanol intake.

D. INFLUENCE OF ADRENALGLANDS The influence of adrenal glands on voluntary consumption of ethanol has received very little attention. Mardones and Segovia (1951) reported that adrenalectomy performed in rats induced a significant decrease in alcohol intake from the third to twelfth day after the operation.

E. SUMMARY The influence of the thyroid gland on free choice of ethanol has not been clearly established. Probably the influence of spontaneous variations, differences in strains, and food and environmental conditions are responsible for the discrepancies in collected data. Insulin, as well as an antidiabetic sulfa, increased ethanol consumption; while pancreatectomy and alloxan diabetes decrease it. The gonads appear to play no important role in ethanol preference. The sex difference is questionable. The administration of diethylstilbestrol results in a decrease of alcohol intake. An extensive study of the effect of adrenals on free choice of ethanol must be done. Adrenalectomy induces a transitory decrease of alcohol consumption; as far as the author knows, the

68

JORGE MARDONES

effect of administration of corticoid hormones has not been studied. VIII.

Effect of Drugs

The effect of various drugs on the free choice of ethanol in laboratory animals has been studied by many authors. To review this matter it seems convenient to analyze separately the drugs acting on the central nervous system, those that block metabolism of ethanol, and poisons acting through liver damage.

A. EFFECTOF DRUGS ACTING ON

THE

CENTRAL NERVOUS SYSTEM

1. CNS stimulants The effect of amphetamine (0.4 mg per 100 gm of body weight, subcutaneously) on free choice of ethanol, was studied in 18 rats by Moore ,etul. (1952). The results obtained were not significant because the responses exhibited a wide individual variation. There was a trend toward a decrease in alcohol intake. Methylphenidate (Ritalin) (in daily doses of 25 mg mixed with the diet) was tested by Rogers and Pelton (1958) in a group of 10 rats. The result observed was an average, nonsignificant increase from 0.28 to 0.35 ml of 95% ethanol per 100 gm of body weight. The effect of deanol ( 2-dimethylaminoethanol) a psychic energizer, was studied in rats by Prieto, Pfeiffer, and Mardones (1958, unpublished). The tested doses ranged from 5 to 10 mg per 100 gm of body weight given subcutaneously during a period of 6 days. No significant changes in alcohol consumption were observed. 2. Aturmtic drugs The effect of some ataractic drugs mixed with the diet were studied by Rogers and Pelton (1958). The voluntary alcohol consumption observed during an experimental period of 3 weeks was compared with the one observed 3 weeks before and 4 weeks after. Each drug was tested in a group of 10 rats. The following drugs were examined: reserpine, 0.3 mg/day; chlorpromazine (Thorazine), 15 mg/day; promazine (Sparine), 15 mg/day; azacyclonol ( Frenquel), 6 mg/day; meprobamate ( Miltown ) (Equanil), 480 mg/day; pipradrol ( Meratran) , 1.5 mg/day.

CHANGES I N FREE SELECTION OF ETHANOL

69

Promazine was the only drug that induced a significant change in alcohol consumption. The free choice of ethanol rose from 0.28 to 0.56 ml 95% ethanol per 100 gm of body weight per day. This effect was associated with some refusal of the diet, perhaps due to taste of the drug. The authors pointed out that although the diet containing meprobamate was also refused, the increase of alcohol intake induced by this drug was nonsignificant (0.28 to 0.34).

3. Psychtomimetic drugs Lysergic acid diethylamide (LSD) in daily doses of 10 pg was tested under the same conditions by Rogers and Pelton (1958) during a 5-week period in a group of 10 rats. No significant difference from the control group was obtained. B. DRUGSBLOCKING ETHANOL METABOLISM Disulfiram ( tetraethylthiuram disulfide, Antabuse) is a drug employed extensively in the treatment of alcoholism because it blocks ethanol oxidation at the stage of acetaldehyde which induces discomfort and severe symptoms of syncope and vasodilatation. When this drug is given by stomach tube to rats having a free choice of water and 10% ethanol solution, a decreased ethanol intake is observed. Harkness et al. (1953) have used this decrease in intake as a test to measure the potency of drugs related to disulfiram. They have observed that this effect was induced also by tetramethylthiuram disulfide, dipentamethylenethiuram disuKde, tetraisobutylthiuram disulfide, and sodium diethyldithiocarbamate.

C. EXPERIMENTAL LIVERDAMAGE Sirnes (1953) reported that carbon tetrachloride given subcutaneously in doses of 0.2 ml twice a week for 4 months induced a liver cirrhosis in 11 rats and caused no lesions in 3. The free choice of ethanol observed in the cirrhotic rats was considerably higher than that in the control group. The 3 noncirrhotic rats exhibited an alcohol consumption that did not differ from the control. The author obtained similar results with rats suffering from chronic liver injury induced by prolonged exposure to

70

JORGE MARDONES

phosphorus. It is surprising that rats suffering from extensive lesions of the liver did not show any decrease in the rate of ethanol oxidation. The metabolic disturbance associated with liver damage responsible for increased alcohol consumption remains an open question. Sirnes (1953) assumed that there was some interference with the utilization of thiamine or factor N1 (see Section IV) which brought about this effect. D. OTHERDRUGS Rogers et aZ. (1957) have tested the effect of some glutamint: antagonists on the free choice of ethanol by rats. Azaserine, given in doses of 0.75 mg per day, induced a slight decrease of alcohol intake. Methionine sulfoximine in two daily doses of 50 mg induced an immediate increase of alcohol consumption that lasted for about 1 week and afterward a decrease to a lower than basal level. Glutamine was not able to counteract the effect of these drugs. Moore and Emerson (1952) treated a group of “free choice” rats suffering from pneumonitis with sodium sulfadiazine (200 mg/kg intramuscularly) and observed that even though the disease itself did not alter alcohol consumption, 4 of the 13 treated rats exhibited a significant increase of alcohol intake. Mardones et ,aZ. (1953b) reported that sulfasuxidine, a poorly absorbed sulfa drug (2% in a purified diet lacking factor N1) did not change significantly the alcohol intake. Potassium tellurite did not change the alcohol consumption of deer mice (Emerson et al., 1952).

E. SUMMARY Ataractic and stimulant drugs have no active effect on the free choice of ethanol. The only effect that has been reported as significant is that of promazine, but it is possible that this effect is produced by a change in the taste of the food. Disulfiram and other drugs blocking ethanol metabolism induce a clear decrease in alcohol consumption. Chronic liver damage induced in rats by some poisons (carbon tetrachloride, phosphorus) results in a marked and significant increase of the alcohol intake. Methionine sulfoximine induces a biphasic response with an ultimate decrease of ethanol consumption. The mechanism of this effect remains unknown. The effect of sulfa drugs is questionable.

CHANGES IN FREE SELECTION OF ETHANOL

71

IX. Significance of the Reviewed Facts on Human Alcoholism The consumption by laboratory animals of daily amounts of alcohol as high as 7 gm per kilogram of body weight approximates that of human alcoholics. Several authors, including ourselves, have considered the increase of free choice of ethanol attained by laboratory animals as “experimental alcoholism.” However, an accurate comparative analysis shows that there are striking differences between these two phenomena. In order to discuss the relationships between the changes of voluntary consumption of ethanol in animals and in man, it seems useful first to analyze the differences, and then to apply the results to the knowledge and management of human alcoholism. BETWEEN HIGH LEVELFREECHOICEOF ETHANOL A. DIFFERENCE AND PATHOLOGICAL DESIRE

The desire for alcohol observed in human alcoholics is usually urgent and overpowering. Patients experiencing withdrawal symptoms, such as tremor, insomnia, hyperreflexia, and anxiety, or suffering from psychological tension, have an urgent desire to drink alcoholic beverages in order to alleviate their symptoms or to relieve their tension. In both conditions they seek liquor in spite of any barrier that society or their relatives have imposed. This urgent and overpowering desire for alcohol characterizes the behavior of the human alcoholic. The desire for ethanol observed in the laboratory animal seems to be less intense. As far as the author is informed, there are no reported estimates of the difficulties that rats or other laboratory animals will surmount in order to gain access to ethanol. The fact that the offering of a “third choice” of sugar solutions induces a significant decrease of the alcohol intake, gives supporting evidence that the desire is neither urgent nor overpowering. From our point of view this difference is important.

B. DIFFERENCE BETWEEN HIGH LEVELFREECHOICEOF ETHANOL AND INEBRIATION Alcoholics usually drink until they experience a certain degree of intoxication, which is characterized by changes in behavior and

72

JORGE MARWNES

motor coordination. “Alcoholism without inebriation” is a clinical rarity, and even the existence of this entity is questionable. In general, the laboratory animals exhibiting a high intake of alcohol do not show any overt sign of intoxication. Even though the amount of ethanol consumed is the highest that the system can oxidize, its ingestion is controlled in such a way that the increase of blood alcohol does not reach a level which seriously disturbs the functions of the central nervous system. Apparently laboratory animals behave in part like the volunteers studied by Isbell et al. (1955). With a regular ingestion of a diluted ethanol solution in many small doses throughout the day, they drank as much as 397 to 466 ml of 95% ethanol per day without signs of intoxication and the alcohol blood levels were less than 50 mg per 100 ml. Thus far liver damage is the only experimental condition that has been able to induce an alcohol intake large enough to increase markedly the blood level of animals. T. B. Sirnes (personal communication, 1959) has observed in rats, dying from liver damage and having a free choice of ethanol, blood levels as high as 300 mg per 100 ml.

C. FREESELECTIONOF ETHANOL AND WITHDRAWAL SYMPTOMS One of the fundamental characteristics of human alcoholism is the presence of withdrawal symptoms after cessation of drinking. These overt symptoms have been produced experimentally in human volunteers by Isbell et al. (1955). The symptoms of withdrawal after a period of intoxication were: tremulousness, nausea, perspiration, insomnia, tremor, marked weakness, fever, hypertension, vomiting, convulsions, hallucinations, and delirium. In contradistinction, no overt symptoms have been observed in laboratory animals afer removal of the ethanol bottle, even though the amount consumed previously (free or forced) was very high.

D. DETERMINATIVE CAUSESOF DRINKING ALCOHOLIC BEVERAGES The differences discussed in the previous paragraphs establish a clear cut separation between human alcoholism and voluntary ethanol intake of laboratory animals. This statement does not exclude the existence of some relationship between both conditions. It is convenient to remember that human alcoholism can be considered as a pathological state created by the continuous use of alcoholic beverages. The primary condition for the establishment

CHANGES IN FREE SELECTION OF ETHANOL

73

of alcoholism is, of course, that the patient likes alcohol. It is impossible for a person refusing alcoholic beverages to become an alcoholic. Therefore, the desire for ethanol-ranging from refusal to pathological desire-is one of the determining factors in alcoholism. In other words, the first steps to alcoholism are determined by the reasons for drinking alcoholic beverages. Our present knowledge of alcoholism leads us to believe that people drink for three principal reasons (Mardones and Varela, 1956; Forsander et al., 1958): ( 1) Because alcohol satisfies some physiological requirements: e.g., supplying energy directly, thereby bypassing some enzymatic systems ( vitamin deficiencies, liver damage, and endocrine diseases). ( 2 ) Because alcohol depresses the central nervous system thereby alleviating psychological tensions induced by every day conflicts. ( 3 ) Because, in order to satisfy social demands, people are expected to drink at celebrations, parties, etc. Of course, the use of alcoholic beverages for any of these reasons is always a step toward alcoholism, and the only sure way to avoid it is not to drink.

E. INDIFFERENCE VERSUS AVERSION To create in man a lack of interest in or refusal for alcohol is the goal that must be reached in order to prevent alcoholism or maintain an effective cure. Relapses always begin by an occasional drink induced by any of the three afore-mentioned causes, and not necessarily by a craving. The methods now used to induce a refusal for alcoholic beverages are psychotherapy, aversion induced by conditioned reflexes, and administration of disulfiram and related drugs. It is obvious that these measures can be used only to prevent relapses in alcoholic patients, and not as preventive measures for the whole population. It is possible that conditions which modify the free choice of ethanol in experimental animals might give LIS a clue toward a means of inducing in man a specific indifference or distaste for alcohol that could be used for the prevention and cure of alcoholism. By this means the physiological and maybe the social causes can be controlled. That the same mechanism may be involved in the use of alcohol for the relief of tensions has been suggested by experiments of Dember and Kristofferson (1955) (see Section 111,D).

74

JORGE MARDONES

X. Summary

The amount of ethanol consumed under free choice conditions varies in different laboratory animals. Smaller animals such as mice metabolize ethanol faster and this may correlate with their higher voluntary intake. Individual variation in consumption of ethanol is a genetic factor in rats and these have been inbred to produce “drinker and nondrinker” strains. No metabolic differences between these two strains have been found. Single deprivation of most of the water soluble vitamins increases ethanol consumption. This also is true of the unidentified factor N1. Thioctic acid or glutamine decrease ethanol intake. When sugar solution or fat emulsion is offered as a third choice, the intake of ethanol decreases and this is correlated with the extra calories provided by the third choice. Saccharin after sugar solution also decreases ethanol intake. No increase in ethanol intake has occurred in laboratory rats either after forced or free choice ethanol administration for long periods. Of the endocrine influences reported, the thyroid effect is questionable, gonads are noncontributory, and the adrenal cortex has been insufficiently studied. Insulin and oral antidiabetic drugs increase and alloxan or pancreatectomy decreases ethanol intake. Adrenalectomy also decreases voluntary intake. Stimulant, ataractic, or psychotomimetic drugs do not change ethanol intake. Disulfiram and related drugs decrease the ethanol intake. Drugs which produce liver damage definitely increase ethanol intake, Methionine sulfoximine produces an increase and ultimately a decrease in ethanol intake. These studies are compared to chronic alcoholism in man. REFERENCES Beerstecher, E., Jr., Reed, J. G., Brown, W. D., and Berry, L. J. (1951). Uniu. Texas Publ. 5109, 115. Brady, R. A., and Westerfeld, W. W. (1947). Quart. j . Studies A k . 7 , 499. Carver, J. W., Nash, J. B., Emerson, G. A., and Moore, W. T. (1953). Federation Proc. 12, 309. Delore, P., and Berry, H. (1955). Presse mdd. 63, 1591. Dember, W. N., and Kristofferson, A. B. (1955). Quart. J . Studies AZc. 16, 86. Emerson, G. A., Brown, R. G., Nash, J. B., and Moore, W. T. (1952). j . Pharmacol. Exptl. Therap. 106, 384.

CHANGES IN FREE SELECTION OF ETHANOL

75

Forsander, O., Kohonen, J., and Soumalainen, H. (1958). Quart. J. Studies Alc. 19, 379. Harkness, W. D., Johnston, C. D., and Woodard, G. (1953). Federation Proc. 12, 328. Hiestand, W. A., Stemler, F. W., Wiebers, J. E., and Rockhold, W. T. (1953). Federation Proc. 12, 67. Isbell, H., Frazer, F., Wikler, A., Belleville, R. E., and Eisenman, A. J. (1955). Quart. J. Studies Alc. 16, 1. Lester, D., and Greenberg, L. A. (1952). Quart. 1. Studies Alc. 13, 553. Loiseleur, J., and Petit, M. (1947). Compt. rend. SOC. biol. 141, 568. Mardones, J., and Onfray, E. (1942). Reu. chilena hig. y med. prevent. 4, 293. Mardones, J., and Segovia, N. ( 1951). Fundacidn Lucas Sierra, Hosp. ViAa del Mar, Jornadas Clin. Verano, 5 a Jornada, p. 378. Mardones, J., and Varela, A. (1956). Rev. Psiquiat. Santiago, Chile 21, 83. Mardones, J. Hederra, A,, and Segovia, N. (1949). Bol. SOC. biol. Santiago, Chile 7, 1. Mardones, J., Segovia, N., and Hederra, A. (1947). Bol. soc. biol. Santiago, Chile 4, 121. Mardones, J., Segovia, N., and Hederra, A. (1948). Bol. soc. biol. Santiago, Chile 6, 27. Mardones, J., Segovia, N., and Hederra, A. (1950a). Bol. soc. biol. Santiago, Chile 7, 61. Mardones, J., Segovia, N., and Hederra, A. (1950b). Bol. SOC. biol. Santiago, Chile 7, 82. Mardones, J., Segovia, N., and Hederra, A. (1952). Acta Physiol. Latinoam. 2, 43. Mardones, J., Segovia, N., and Hederra, A. (1953a). Quart. J. Studies Alc. 14, 1. Mardones, J., Segovia, N., and Onfray, E. (1946). Arch. Biochem. 9, 401. Mardones, J., Segovia, N., Hederra, A., and Alcaino, F. (1953b). Acta Physiol. Latinoam. 3, 140. Mardones, J., Segovia, N., Alcaino, F., and Hederra, A. (1954). Science 119, 735. Mardones, J., Segovia-Riquelme, N., Hederra, A., and Alcaino, F. (1955). Quart. J. Studies Ah. 16, 425. Marshall, E. K., Jr., and Owens, A. H., Jr. (1955). Proc. SOC. Exptl. Bwl. Med. 89, 573. Mirone, L. ( 1952). Quart. 1. Studies AEc. 13, 365. Mirone, L. (1957). Quart. J. Studies Alc. 18, 552. Moore, W. T., and Emerson, G. A. (1952). J. Pharmacol. Exptl. Therap. 106, 408. Moore, W. T., Moore, B. M., Nash, J. B., and Emerson, G. A. (1952). Texas Repts. Biol. arid Med. 10, 406. Prieto, R., Varela, A., and Mardones, J. (1957). Asoc. Latinoam. Cienc. Fisiol. l a Reun. Res. trab. p. 146.

76

JORGE MARDONES

Prieto, R., Varela, A., and Mardones, J. (1958). Acta Physiol. Latinoam. a, 203. Ravel, J. M . , Felsing, B., Landsford, E. M., Jr., Trubey, R. H., and Shive, W. (1955). J . Biol. Chem. 214, 497. Richter, C. P. (1936). Am. J . Physiol. 116, 155. Richter, C. P. ( 1941). Quart. J . Studies AZc. 1,650. Richter, C. P. ( 1956). Endocrinology 69, 472. Richter, C. P. (1957). In “Neuropharmacology,” Transactions of the Third Conference, p. 39. Josiah Macy, Jr. Foundation, New York. Richter, C. P., and Campbell, K. M. (1940). Science 91, 507. Richter, C.P., and Hawkes, C. D. (1940). Am. J . Physiol. 1S1, 639. Richter, C. P., and Schmidt, E. C. H., Jr. (1941). Endocrinology 28, 179. Richter, C. P., Holt, L. E., and Barelare, B., Jr. (1938). Am. J . Physiol. 122, 734. Rogers, L. L., and Pelton, R. B. (1958). Texas R e p . Biol. and Med. 16, 133. Rogers, L. L., Pelton, R. B., and Williams, R. J. (1955). 1. Biol. Chem. 214, 503. Rogers, L. L., Pelton, R. B., and Williams, R. J. (1956). J . Biol. Chem. 220, 321. Rogers, L. L., Pelton, R. B., and Williams, R. J. (1957). Federation Proc. 16, 238. Rosenberg, B., and Zarrow, M. X. (1952). Federation Proc. 11, 131. Segovia-Riquelme, N., Vitale, J. J., Hegsted, D. M., and Mardones, J. (1956). J . Biol. Chem. 223, 399. Shadewald, M., Emerson, G . A,, Moore, W. T., and Moore, B. M. (1953). Federation Proc. 12, 364. Sirnes, T. B. (1953). Quart. J . Studies Alc. 14, 3. Westerfeld, W. W., and Lawrow, J. ( 1953). Quart. J . Studies Alc. 14,378. Williams, R. J., Berry, L. J., and Beerstecher, E., Jr. (1949). Proc. Natl. Acad. Sci. US.35, 265. Williams, R. J., Berry, L. J., and Beerstecher, E., Jr. (1950). Texas Repts. Biol. and Med. 8, 238. Zarrow, M. X., and Rosenberg, B. (1953). Am. 3. Physiol. 172, 141.

THE MECHANISM OF ACTION OF THE HEMlCHOLlNlUMS By F. W.

Schueler

Department of Pharmacology, Tulane University School of Medicine, N e w Orleans, Louisiana

I. Introduction ........................................... 11. Origin and Development of the Hemicholiniums . . . . . . . . . . . . . 111. Pharmacological Actions of Hemicholinium No. 3 (HC-3) . . . . . A. Toxicity ........................................... B. Mechanism of Respiratory Depressant Action of HC-3 . . . . C. Other Actions of HC-3 ............................... IV. Studies on Other Hemicholiniums and Hemicholiniumlike Agents V. Summary of Effects of HC-3 at Various Cholinergic Sites ..... References .............................................

1.

77 77 80 80 82 92 94 95 96

Introduction

The coined name hemicholiniuin was introduced ( Schueler, 1955) to denote a group of hemiacetals containing a cholinelike moiety. These products are notable for their high toxicity which is especially slow in onset in the larger common laboratory animals and which grossly resembles some phases of botulism. Continuing interest in the hemicholiniums has revealed the fact that they are capable of affecting cholinergic transmission at a variety of sites in an apparently unique fashion and that they, therefore, represent a new tool in the investigation of cholinergic mechanisms. It is the objective of this review to summarize the development of these new agents and to point up, insofar as possible, certain salient unresolved problems that have arisen in the course of their study. II. Origin and Development of the Hemicholiniums

During the synthesis and pharmacological evaluation of a series of a,a'-quaternary ammonium salts derived from 4,4'-bisacetophe77

7s

F. W. SCHUELER

none (Long and Schueler, 1954; and Schueler, 1955) it was observed that three distinct types of pharmacological effects were elicited depending upon the nature of the substituent quaternary ammonium radicals. Thus, based upon the quaternization of a,a’dibromo-4,4’-bisacetophenone according to scheme ( I ) , three n -

0

Br -CH2-C “

-!

-CH,-Br

+ tertiary

amine

Scheme ( I )

classes of quaternary ammonium salts were obtained: (1) products derived from simple trialkyl tertiary amines, ( 2 ) products derived from heterocyclic tertiary amines, and ( 3 ) products derived from ethanol tertiary amines. In general, the products of class 1 exhibited the properties of rapidly acting neuromuscular blocking agents antagonizable by edrophonium and neostigmine. The products of class 2 also acted rapidly but toward the production of generalized peripheral cholinergic stimulation in uivo. They were also found to be potent anticholinesterase agents against “true” cholinesterase at EDao concentrations of from to 10-lOM in vitro. Certain of the products of class 3 exhibited the new type of action which forms the basis of the present review. The chemistry of the class 3 products from a,a’-bisacetophenone appears to be crucial to their special action. Chemical, ultraviolet, infrared, and pharmacological studies ( Schueler, 1955) reveal that the characteristic high and latent type of toxicity presented by products of class 3 is correlated with their ability to undergo cyclization through hemiacetal formation according to scheme ( I1 ). The most potent product thus far investigated, hemicholinium No. 3 or HC-3 (Schueler, 1955), was obtained through the quaternization of dimethylaminoethanol in which the radicals ( R ) in scheme (11) are methyl groups. The presence of two cholinelike moieties arising through hemiacetal formation, as emphasized in

MECHANISM OF ACTION OF THE HEMICHOLINIUMS

quateriiiration

I

R

\

f

N- CHs-CHz-OH

R/

2 Br-

CH,

CH2

\&/

R’

‘R

79

unisolatable intermediate

eyclization

CH2

CHI

\ + ,CHz

CH2

\&/

R/N\R

R’

2 Br- bis-hemiacetal

R ‘

Scheme (11)

structure (I11 ), was the basis of the coined name “hemicholinium” for this and all similar products. Since the hemicholinium designated HC-3 continues to be denoted as such by most investigators, and as this product is the material upon which most studies have been based, the abbreviation HC-3 will be used throughout the remainder of this review to denote the above specific product.

80

F. W. SCHUELER

111.

Pharmacological Actions of Hemicholinium No. 3 (HC-3)

A. TOXICITY The primary toxic manifestation of HC-3 is a slowly progressive failure of the respiration (Long and Schueler, 1954; Schueler, 1955; Kas6 and Borison, 1958; Reitzel and Long, 1959a; McPhail, 1959). There is a rough correlation between lethal dose, size of the animal used, and prelethal period. Thus, 10- to 15-kg dogs require doses of HC-3 from 0.075 to 2.500 mg/kg, by the intravenous route of administration, with respiratory failure ensuing after 2 to 6 hours. The larger the dog, in general, the longer the prelethal period. Adult rabbits and cats require doses from 0.050 to 1.200 mg/kg intravenously, respiratory failure occurring after 30120 minutes. Guinea pigs require intraperitoneal doses from 0.30 to 0.75mg/kg with a prelethal period of 30-45 minutes. In rats the LD50is 0.45 mg/kg with a prelethal period of 2030 minutes. In the author’s hands, the LD50 in mice for HC-3 varies widely depending upon the batch of mice used. Thus, on occasion, the LD50for mice has been as low as 0.015 mg/kg or as high as 0.075 mg/kg. It is an important observation by all of the afore-mentioned investigators that high intravenous doses of HC-3 produce immediate death by respiratory paralysis and that the mechanism in this latter case is unrelated to the typical latent toxicity at lower doses. The prompt lethality at high doses is due to the well-known neuromuscular blocking action produced by many quaternary ammonium ions, and is effectively antagonized by neostigmine and edrophonium whereas the latent lethality is antidoted to only a limited degree by such agents (Reitzel and Long, 1959a; Schueler, 1955). The distinction between the primary latent toxic action of the hemicholiniums, which is the action of greatest interest in this review, and the acute neuromuscular blockade effected at high doses is illustrated in Fig. 1. Here it is observed that respiration and contractions of the gastrocnemius muscle fail promptly at high doses followed by recovery, and then at a later time failure of respiration occurs without apparent impairment of gastrocnemius muscle contractions. It has, through such experiments, become increasingly apparent in the case of the hemicholiniums that proper attention to dosage must be kept constantly in mind. Indeed it is true with HC-3 as with all other chemical agents that “enough

MECHANISM OF ACTION OF THE HEMICHOLINNMS

81

of anything will inhibit anything.” The issues at stake become increasingly clouded when the primary effects of a drug (i.e., those that are observable at the lowest dosage that produces any effect at all) become confused with effects observed at higher dosage ranges.

FIG. 1. Rabbit under urethane anesthesia arranged for the recording of respiration (R.) and contractions of the gastrocnemius muscle (M.) under electric stimulation ( l / l O sec) of the sciatic nerve. At each arrow 1 mg/kg of hemicholinium was administered. Artificial respiration ( RA. ) was instituted for a brief period by endotracheally administered oxygen. Note the reinstitution of respiration following (RA.) with recovery of (M.) followed later by failure of (R. ) with continuation of (M.). (From Schueler et al., 1954.)

In recapitulation of this section on toxicity, three problems are outstanding. They are: (1) the reason for the latent period before respiratory failure; ( 2 ) the reason for respiratory failure in the face of apparently intact sciatic nerve-gastrocnemius muscle function as illustrated in Fig. 1; and ( 3 ) the basis of the apparent relationship between animal size, toxic dose, and length of latent period to respiratory failure. It will be the objective of succeeding sections to examine the evidence available toward clarification of these and other problems that have arisen in the course of studies with the hemicholiniums.

82

F. W. SCHUELER

B. MECHANISMOF RESPIRATORY DEPRESSANT ACTIONOF HC-3 In elucidating the mode of action of any drug which exhibits a respiratory depressant effect, at least six sites must be considered. These are: (1) the musculature which performs respiratory movements including the diaphragm and the intercostal muscles; ( 2 ) the phrenic and intercostal nerve-muscle junctions; ( 3 ) the nerves per se which conduct impulses toward the activation of the muscles; ( 4 ) the spinal relay synapses for respiration; (5) the respiratory center and its various components; and ( 6 ) various afferent receptors and their nerves through which impulses impinge upon the central respiratory center. 1. Central uersus Peripheral Site of Action The initial studies with the hemicholiniums (Long and Schueler, 1954; Schueler et al., 1954; Schueler, 1955) which made use of supramaximal square wave shocks (frequency 1/10 sec, duration 1 5 m s e c ) applied to the sciatic nerve in rabbits and dogs revealed no evidence of depression of gastrocnemius muscle contractions after doses of HC-3 which produced delayed respiratory failure ( 0.3-0.5 mg/kg intravenously). Thus, as illustrated in Fig. 1, although doses of 2 mg/kg in rabbits produces transitory immediate respiratory failure and depression of the contraction of the gastrocnemius muscle, the primary effect in such preparations is the latent slowly developing paralysis of respiration without failure of the contractions of the gastrocnemius muscle. The latter effect is observed at lower doses which do not elicit the prompt action produced by high doses. Since all of the well-known neuromuscular blocking agents, e.g., d-tubocurarine, decamethonium, and succinylcholine, produce peripheral neuromuscular blockade in doses that effect respiratory failure under the above conditions, it was conjectured that HC-3 possessed either of two properties to produce its primary action: ( 1) a special predilection for the neuromuscular junctions involved in respiration, or ( 2 ) an action at some other site or sites involved in the activation of respiration. An early experiment (Schueler, 1955) designed to eliminate the first of these two possibilities consisted of stimulating the phrenic nerve of the dog maintained on endotracheally administered oxygen following cessation of respira-

MECHANISM OF ACTION OF THE HEMICHOLINIUMS

83

tion caused by 0.3-0.5 mg/kg of HC-3. After respiratory failure, electrical stimulation of the phrenic nerve w7as successful in activating respiratory movements, and it therefore appeared that the locus of HC-3 action should be sought elsewhere than at the nervediaphragm junction, the body of the phrenic nerve, or the diaphragm muscle per se. In an attempt (Schueler, 1955) to determine a possible central locus of HC-3 action, a head-cross circulation experiment in anesthetized dogs after the manner of Taylor and Page (1951) was employed. In this preparation, one dog serves as a donor via one carotid artery to the head of a recipient dog, and the venous head outflow from the recipient is returned to the donor by way of one of its jugular veins. After elimination of all tissues between the trunk and head of the recipient animal with the exception of the spinal cord, the vagus, phrenic, and intercostal nerves, injection of a dose of HC-3 sufficient to kill both animals was administered to the donor. In this experiment, the donor animal upon cessation of respiration was maintained on endotracheally administered oxygen while the recipient continued to activate its own respiration. Such an experiment, while allowing only an indirect inference suggested that HC-3 does not attack the respiratory center per se but may act via depression of the spinal relay centers for respiration. Since the above initial experiments, the problem concerning HC-3 action against respiration has become increasingly complex. Thus Kask and Borison (1958) have analyzed the effects of HC-3 by means of electrical stimulation and transections of the lower brain stem in unanesthetized, pentobarbital anesthetized, and decerebrate cats. From their studies, they have concluded that the HC-3 respiratory paralyzant action is due mainly to depression of the respiratory regulatory mechanism within the brain stem rather than through depression of the spinal cord. In their study, which was long and elegantly conceived, Kas6 and Borison have described the action of HC-3 in terms of the “modulator,” “oscillator,” and “integrator” functions within their scheme of the processes activating respiration centrally. They have concluded the HC-3 does not, however, depress pacemaker function since cough responses can be elicited after respiratory failure effected by HC-3. While they observed, as have others, that neuromuscular blockade takes place with large doses of HC-3, they have indicated that central depres-

84

F. W. SCHUELER

sion takes precedence over peripheral neuromuscular blockade. In addition, they have described a depressant effect of HC-3 upon medullary vasomotor regulation in doses “too small to cause ganglionic blockade” and have found HC-3 to be significantly more potent by the intracarotid route of injection than by the intravenous route. Very recently Waton (1959) has indicated that HC-3 depresses occipital cortical potentials evoked through rapid photic stimulation of the retina in anesthetized cats. While the case for a central site of action of HC-3 appears strong, the problem is by no means settled. Aside from possible differences in sensitivity of different central sites in cats and dogs (as seem evident from the dog head-cross circulation experiments and the results of Kas6 and Borison in cats), there are cogent reasons for suspecting that the problem of HC-3 toxic action is not so much one of where it acts centrally but whether its site of action is not primarily peripheral. Thus, Longo (1959) has studied the effect of HC-3 upon the efferent impulses recorded from the central stump of the phrenic nerve in the necks of anesthetized rabbits and cats. Blood pressure and respiration were also recorded. HC-3 injected in doses from 0.5 to lmg/kg effected the typical progressive diminution in amplitude of respiratory movements; the phrenic potentials, however, were not concomitantly diminished but, on the contrary, augmented. Moreover, the phrenic potentials were present after respiratory movements failed and until the blood pressure fell to very low levels. Aracial respiration, if instituted before irreversible anoxia had occurred, always restored the phrenic potentials. Further, studies by Schueler and co-workers (1954) indicate H’C-3has no effect upon EEG potentials recorded through frontal, parietal, or occipital electrodes in rabbits provided adequate artificial respiration is maintained following respiratory failure. Such experiences have tended to focus recent attention upon the role of peripheral sites as targets of HC-3 toxicity.

2. Neuromusculur Blockade: Presynaptic, Postsynaptic, or Both Recently Reitzel and Long (1959a) have obtained evidence of blockade of the neuromuscular junction by HC-3; the maximum effect in rabbits is obtained 40 to 60 minutes following administra-

MECHANISM OF ACTION OF THE HEMICHOLINIUMS

85

tion of HC-3 and occurs only when the rate of stimulation of the nerve is 60 per minute or greater. Now an immediate problem which presents itself in any consideration of the neuromuscular blocking action of a drug is whether the given agent acts, broadly speaking, presynaptically, postsynaptically, or both. As outlined by Birks and MacIntosh ( 1957) and MacIntosh (1958), presynaptic processes supporting synaptic transmission involve the formation, storage, and release of the transmitter. In the case where acetylcholine serves as transmitter, formation appears to take place within mitochondria carried from the neuronal cell body by axoplasmic flow. Choline appears to be actively concentrated or sequestered by the presynaptic membrane and is acetylated within the mitochondria through the intermediary action of the enzyme choline acetylase utilizing acetyl coenzyme A. “Active” acetate for the formation of acetyl coenzyme A is supplied, at least in part, by citrate and various enzyme systems requiring ATP, glucose, and probably a special ‘‘plasma factor.” Here, and quite aside from the precise detailed validity of the above scheme, the crucial point is that acetylcholine formation is a complex biochemical process and that there are undoubtedly numerous potential points of attack for any drug which may interrupt formution of this transmitter. Following through, however, with the scheme of MacIntosh and Birks: acetylcholine, once formed, finds its way, possibly via the endoplasmic reticulum (Koelle and Steiner, 1956), toward storuge in synaptic vesicles which approach the presynaptic membrane by Brownian movement. Each vesicle stores some hundreds of thousands of molecules of acetylcholine which are released in a quanta1 fashion upon “effective” collision of a vesicle with the presynaptic membrane. In the resting state, random effective collisions are rare, but upon depolarization of the membrane they discharge the transmitter synchronously in large numbers. What determines “effective” collision is not known, although the concentration of calcium ions in the presynaptic axoplasm has been implicated (Birks and MacIntosh, 1957). Again, omitting details, the crucial point is the undoubted complexity of the processes of storage and release which, together with formation, serve as potential targets for the action of drugs. In the case of antonomic ganglia, MacIntosh and his associates (1956, 1958) presented cogent evidence that HC-3 inhibits the

86

F. W. SCHUELER

formation of acetylcholine at some stage previous to the process of choline acetylation. HC-3 therefore does not appear to act in the manner of botulinus toxin which has been shown to block release and to have no effect upon formation. If both HC-3 and botulinus toxin act presynaptically, although at different phases of this function, this fact would go far toward explaining their gross similarities of action previously indicated ( Schueler, 1955). Following up the observation (Schueler, 1955) that choline is an excellent antidote for HC-3 poisoning (choline chloride, 100 mg/kg intraperitoneally, antidotes completely 0.2 mg/kg of HC-3 i.p. whereas 0.05mg/kg was 100% fatal in a group of 50 mice)this fact was subsequently confirmed by Giovinco ( 1957), Reitzel and Long (1959a)-MacIntosh et al. (1956) tested the effect of HC-3 upon a choline acetylase system derived from acetone dried brain powder. In this latter preparation, HC-3 was found to have negligible potency as an inhibitor of acetylcholine formation, a fact which is particularly interesting because HC-3 had been shown by the same investigators to be an excellent inhibitor of acetylcholine formation in autonomic ganglia, cardiac vagal efferents, and the corda tympanic supply to the salivary gland, all of the latter effects being reversed by choline in doses which have no significant effect upon unpoisoned animals. By way of interpretation, MacIntosh and his co-workers postulated that HC-3 blocked an active concentrating mechanism for choline at the presynaptic membrane. Thus they argued that blood concentrations of choline being very low and the demand for choline for transmitter synthesis being high, it is not unreasonable to assume that choline is concentrated by some special mechanism in cholinergic axonal membranes. While no direct proof of this conjecture is yet available, these authors were able to demonstrate that HC-3 very effectively blocks choline excretion by the avian kidney in doses that have no effect upon phenol red excretion. This observation has been confirmed by Farah (1959) using dog renal slices. Farah also demonstrated that HC-3 is quite a powerful inhibitor of N-methylnicotinamide ( N M N) uptake in renal slices. Since choline, tetraethylammonium, and NMN, as well as some other basic substances, are probably all excreted by the same or similar renal mechanism, Farah has suggested that the inhibition by HC-3 of choline and N M N transport is probably due to the ability of HC-3 to block the renal

MECHANISM OF ACTION OF THE HEMICHOLINNMS

87

transport mechanism for such basic substances in general. Farah‘s observations clearly throw some doubt on the validity of any parallel between transport in the kidney as a model for the highly specific transport of choline postulated for cholinergic nerve. Evidence of another kind, however, does indirectly lend strong support for a specific transport process for choline into nerve. Giovinco (1957) and, very recently, Reitzel and Long (1959b) studied series of substances related to choline including tetramethylammonium, tetraethylammonium, 3-hydroxyprop yl trimethylammonium fl-methylcholine, P-hydroxyethyl dimethylethylammonium, etc. With the single exception of p-hydroxyethyl dimethylethylammonium, which exhibited a low degree of antidotal powers, only choline exhibited marked potency in reversing HC-3 toxicity. It will be of interest here to note that the tertiary amine dimethylaminoethanol which has been postulated to be a choline precursor does not act as an antidote to HC-3 toxicity (Pfeiffer, 1959). Gardiner (1957) has presented further evidence bearing upon the hypothesis of an active transport mechanism for choline into nerve cells. Gardiner studied the effect of HC-3 upon acetylcholine formation in guinea pig brain minces and homogenate brain mitochondrial fractions, as well as activated homogenate preparations after ether treatment according to the method of Hebb and Smallman (1956). HC-3 was found to be a potent inhibitor of acetylcholine formation in minces, much less so in homogenates, and not at all in ether-treated homogenates. As summarized by Gardiner (1957) : “The action of HC-3 on acetylcholine formation is thus shown to be weakened by treatments which disrupt the organization of tissue. Intracellular material prepared by methods normally preserving labile enzymic activities is only weakly inhibited and this effect is abolished by treatment to disorganize mitochondria. The absence of action on endogenous acetylcholine production and on solvent-activated preparations supports the idea that choline-acetylase itself is not affected by HC-3, but that the compound acts on the systems transporting choline into the cell and into the mitochondrion.” The evidence reviewed above presents a strong case for a presynaptic site of action for HC-3. Such evidence does not, of course, rule out the possibility of action at the postsynaptic membrane. Brooks and Thies (1959) have communicated to the writer some

88

F. W. SCHUELER

of their recent experiences with HC-3 utilizing nerve diaphragm and serratus nerve muscle preparations after the method of Brooks (1951, 1956, 1959). They recorded muscle action potentials and end-plate potentials (e.p.p.’s) from the surfaces of muscles as well as intracellular miniature end-plate potentials (min. e.p.p.’s ) . Sensitivity at end-plate zones to acetylcholine was tested by applying acetylcholine to the surfaces of muscle at the end-plate zone. Concentration of 10-4M HC-3 produced blockade of the nerve diaphragm function in 30 to 4Q minutes under a stimulation no greater than 12 per minute and blockade of the serratus muscle in 10 to 12 minutes. As pointed out by Brooks and Thies, increasing the frequency of stimulation briefly (to 30 or 60 per minute) deepens the inhibition of muscle potentials. This is a result which is compatible with a decrease in available transmitter by HC-3, but it does not exclude a postsynaptic site of action of the drug. Indeed, Preston and Van Maanen (1953) have shown that the neuromuscular block effected by d-tubocurarine and other postsynaptically acting agents is deepened by increased rates of stimulation. Thus, as Brooks and Thies have indicated, increasing the frequency of stimulation does not distinguish between pre- and postsynaptic actions of HC-3. Brooks and Thies have suggested that min. e.p.p.’s are better criteria for determining the site of action of HC-3. They have pointed up the studies of del Castillo and Katz (1956) in which factors which act presynaptically change the frequency of min. e.p.p. discharge, while neuromuscular blocking drugs which act postsynaptically alter the amplitude of min. e.p.p.’s. Brooks and Thies have observed that concentrations of HC-3 of 2.5 X M or greater depress significantly min. e.p.p.’s. At 5 x 10-8 M depression in amplitude is more than So%, but there is yet no effect upon frequency. Orkand and Martin (1959), using similar parameters, have observed like effects induced by HC-3 in the frog toe muscle preparation. In addition to their studies relating the effect of HC-3 upon min. e.p.p.’s, Brooks and Thies also found that HC-3 (2-10 X M) depressed the responsiveness to externally applied acetylcholine following HC-3. Moreover, Brooks and Thies were unable to demonstrate that choline antagonized HC-3 block in their preparation. Since choline has been demonstrated repeatedly to be a dramatic antidote in HC-3 poisoned animals and in nerve muscle

89

MECHANISM OF ACTION OF THE HEMICHOLINIUMS

preparations in anesthetized mice, rats, rabbits, cats, dogs, and chickens (Reitzel and Long, 1959a; Schueler, 1955; Giovinco, 1957), as well as in autonomic ganglia (MacIntosh et al., 1956, 1958) and in brain tissue poisoned by HC-3 (Gardiner, 1957), the results of Brooks and Thies with choline present a new and difficult problem for interpretation. Brooks and Thies did observe an antidoting effect by physostigmine which has been observed to be an antidote to HC-3 poisoning in mice (Schueler, 1955). In commenting on the data of Reitzel and Long (1959a), Brooks and Thies regrouped some of the former authors’ data as show in Table I. Thus, for different antagonists and rates of onset EFFECT O F HC-3 Dose (I.V.) of HC-3 (m g / k 1 0.16 4.8 a

ON

TABLE I NEUR~MUSCULAR TRANSMISSION IN h l 3 B I T S

Frequency of Time required indirect for maximum stimulation effect (minutes) 60/minute 70 6Jminute 3

Inhibitor of muscle contraction

Primary antagonist

95% 90%

Choline Neostigmine

From Reitzel and Long (1959a).

in experiments, Brooks and Thies (1959) suggested that the inhibition observed in each case is the result of two distinct processes. More specifically, HC-3 appears to act primarily postsynaptically when high doses are used, while it also blocks presynaptically at low doses, where rapid rates of stimulation are necessary to reveal the presynaptic effect. The interpretation proposed by MacIntosh and his co-workers for transmission at ganglia applies also to the neuromuscular junction. Thus, the crucial factors determining the rate of onset and efficacy of HC-3 are the rate of formation of acetylcholine, the quantity already stored, and the demand as determined by the rate of stimulation. If the rate of stimulation is low, then the supply (in terms of acetylcholine stored and already formed) is quite high and blockade by HC-3 will not occur or only very slowly. To produce blockade, one must either increase the rate of stimulation, thereby exceeding the available stores and rate of formation of the poisoned junction to unveil presynaptic poisoning, or one must increase the dose of HC-3 until concentrations effecting postsynaptic neuromuscular blockade are achieved. All of this recalls

90

F. W. SCHUELER

Fig. 1 described in Section I11 and again re-emphasizes the need to interpret effects in terms of dosage. HC-3 is a “clean acting” drug (MacIntosh, 1958) in low doses in that it attacks only the presynaptic process under conditions that place a high demand upon the source of the transmitter. Desmedt ( 1958) illustrated the distinction between postsynaptic blockade by d-tubocurarine and the effects of HC-3. Following partial curarization or small doses of HC-3 in cats anesthetized with chloralose, 3/sec test trains of supramaximal single shocks were delivered to the peroneal nerve and the integrated electrical responses of the tibialis anticus muscle recorded via belly tendon electrodes. Such test trains were repeated briefly every minute or so. Following each given test train, Desmedt calculated the per cent amplitude of the fifth response ( b ) in comparison with the amplitude of the first response ( a ) of the test train; i.e., ( b / a ) x 100. After several such test trains, the nerve was then tetanized very briefly and the test trains again continued to see the resultant effect of faradization upon the decrement, (b/u) x 100, in the posttetanic period. In partially curarized preparations, he observed the well-known immediate posttetanic facilitation in which the values, ( b / a )x 100, rose immediately to nearly 100% and then gradually fell again to the pretetanic level. On the other hand, in the HC-3 treated preparation, there was but a short-lived posttetanic facilitation followed by a rapid drop of the values, ( b / a ) x 100, well below the pretetanic level (i.e., posttetanic exhaustion) and then a very gradual return to the pretetanic level. By stimulating the adductor pollicis muscle via the ulnar nerve in human myasthenia gravis patients Desmedt obtained results essentially similar to those effected by HC-3 from which he drew the conclusion that myasthenia gravis results from a presynaptic biochemical defect. He also pointed out that this type of experiment does not indicate precisely which presynaptic defect is present in myasthenia so that it is not possible to say that HC-3 disorganizes presynaptic function in the same manner as that occurring in myasthenia (i.e., whether the latter is due to choline acetylase deficiency, impaired choline transport, etc. ) . The distinction between a curare block, however, on the one hand and blocks produced by HC-3 (and myasthenia gravis) on the other hand are quite striking by Desmedt’s method.

MECHANISM OF ACTION OF THE HEMICHOLINIUMS

91

3. Mechanism of Progressive Respiratoy Inhibition and Its Relationship to Animal Size In the light of the above evidence, it is of interest to return to a consideration of the mechanism of the peculiar progressive respiratory inhibition effected by HC-3. One problem in particular stands out. Why, after respiratory paralysis effected by “low” doses of HC-3, is it still possible to effect respiratory movements by electric stimulation of the phrenic nerve (Schueler, 1955) or by electric stimulation of the central respiratory substrate (Kas6 and Borison, 1958). Perhaps the answer is related to the distinction between the fusillade of impulses passing down the phrenic in normal respiration as opposed to the effect of single shocks applied to the nerve or in the CNS. Thus a single concerted and closely synchronized set of action potentials passing down the phrenic nerve axons may momentarily break through and effect diaphragmatic contraction during a neuromuscular block that prevents the more diffuse fusillade of impulses of normal respiration from being effective. In this connection, one recalls that HC-3 produces respiratory failure at doses that show no evidence of neuromuscular block in the gastrocnemius muscle under electric stimulation of the sciatic nerve by single shocks at the rate of 1/10 sec. On the other hand, the latter is slowly but progressively blocked at stimulation rates of l/sec (Reitzel and Long, 1959a). Moreover, a muscle blocked at l/sec recovers from the block if the stimulus is interrupted for a time or choline is administered. If one views the fusillade of impulses of normal respiration as a kind of high frequency form of stimulation whereas single shocks applied to the phrenic nerve or within the CNS are regarded as a low rate form of stimulation, then the parallel between the conditions for respiratory failure and failure of the sciatic-gastrocnemius are essentially similar. Beyond this one sees that whereas an animal must continue to breathe to live and is therefore constantly bombarding its diaphragm with impulses, it is under no such obligation to bombard other muscles. Hence the conscious animal in respiratory failure under HC-3 shows little evidence of neuromuscular blockade as far as his appendages are concerned, for in them he can save accrued acetylcholine for transmission even if the rate of transmitter formation is depressed equally at all such neuromuscular sites. It may also be pointed out here that in general the larger the

92

F. W. SCHUELER

animal, the slower its respiration and presumably, therefore, the less the rate of demand for acetylcholine at the phrenic nerve diaphragm junction. Therefore, the larger the animal, the longer it would take for him to exhaust stored acetylcholine toward respiratory failure. Size of animal also correlates well with the size of the toxic dose since the more slowly the animal depletes his stored supply of transmitter for respiration the larger would a single dose of HC-3 have to be (due to excretion or metabolism) to sustain inhibition of transmitter synthesis over a period sufficiently long for respiratory failure to supervene.

4. Summary HC-3 possesses no significant direct ganglion-blocking ( MacIntosh et al., 1956) or atropinelike activity (Long and Schueler, 1954; Schueler, 1955). In high doses it produces prompt neuromuscular blockade of the postsynaptic type. This latter effect is to be distinguished from the primary action of the drug which is presynaptic and which is observed at “low” doses under the conditions that place a high demand on rapid acetylcholine synthesis. At low but lethal doses H C 3 effects a characteristic slowly progressive inhibition of respiration most strikingly seen in the larger laboratory mammals (rabbits, cats, dogs), Its effect upon respiration is interpretable upon the basis of a presynaptic peripheral neuromuscular blocking effect, but may also involve to some degree action upon the centers for respiration in the CNS. The evidence for involvement of this latter site of action is however inconclusive. Mild sedation, which reduces the rate of respiration, very significantly decreases the toxicity of HC-3, while excitatory drugs or procedures which force animals to be very active and increase the rate of respiration increase toxicity of the drug (Giovinco, 1957). Choline but not dimethylaminoethanol administered chronically to mice affords protection against HC-3 (Pfeiffer, 1959). One may conjecture that differing amounts of choline in various diets may account for the variability of HC-3 toxicity observed with different batches of mice. C. OTHERACTIONSOF HC-3 Waton ( 1959), Shamblin (1957), and Madison (1959) have studied the effects of HC-3 upon the eyes of rabbits and cats.

MECHANISM OF ACTION OF THE HEMICHOLINIUMS

93

HC-3 produces a significant mydriatic effect by both topical application or parenteral administration. This mydriatic effect is observed well before respiratory embarrassment supervenes and also in the presence of artificial respiration. This effect which is not due to atropinelike or sympathomimetic action is reversed by choline administration. Investigation of the aqueous humor dynamics before and after HC-3 administration revealed no alterations in intraocular pressure. HC-3 in contact with the severed edges of conjunctiva produced acute transient edema (Madison, 1959). Preliminary distribution and excretion studies in rats using radioactive carbon (methyl group) labeled HC-3 ( Domer, 1959) indicate this material following intraperitoneal injection is relatively rapidly distributed. Within 20 minutes (the time of onset of obvious toxic manifestations) concentrations of the drug are found to be very high in the blood (53.1%), less in the liver (2.4%), and still less in the spleen (0.7%), kidney (0.5%), and brain (0.4% ). At 40 minutes (the time observed by Domer as that exhibiting maximal symptoms of toxicity) all the above tissues exhibited their maximal contents which were appreciably above the 20-minute levels except for blood (33.4%) and brain (0.3%). At this later time very considerable material had been excreted into the urine (35.1%) and also the feces (2.8%). A balance sheet and chromatographic isolation procedure to account for all injected labeled-HC-3 indicates a material having the same Rt value as C14-HC-3was excreted after 9 hours via the urine (approximately 75%) and the remainder or its metabolite (approximately 25%) in the feces. The form of the labeled material excreted in the feces has not yet been determined. No C14-labeledCOe was expired. Coleman and co-workers (1958) working at the Defence Research Kingston Laboratory in Canada studied HC-3 as a possible antagonist to poisoning by anticholinesterase agents of the “Gosh type.” R,O

R,

0

\/ P

- S - CH,

- CH2

/

-N

/

\

R2O

R4

“GOSH-TYPE” AGENT

R

= alkyl

94

F. W. SCHUELER

They detected no prophylactic or therapeutic effectiveness of HC-3 over a wide range of dosage combinations. They also confirmed that atropine has little value as an antagonist to HC-3 poisoning, whereas physostigmine and choline in particular are quite effective as previously indicated ( Schueler, 1955). Experiments by Farrell (1959) at the Connaught Medical Research Laboratories at the University of Toronto tested the effect of HC-3 as a possible inhibitor of poliomyelitis virus in tissue culture. His concept, most interestingly, is that a drug which acts primarily against some process proceeding in cholinergic nerves might inhibit the growth of a virus which has some special predilection for growth in such nerves. Farrell has reported that “It does appear that proliferation of virus was delayed, though at the expense of some cell destruction” and that “treatment of established tissue cultures clearly interfered with formation of plaques, but not to any extent with adsorption of virus by the cells.” It is hoped that these interesting observations will be followed up on poliomyelitis and perhaps other viruses such as rabies, for the inhibition of plaque formation implies an inhibition of virus replication within cells.

IV. Studies on Other Hernicholiniurns and Hernicholiniurnlike Agents

Thus far, of the initial group of congeners in the hemicholinium series, only HC-3 has been studied to any great degree. Schueler ( 1955) reported that three other closely related products (compounds no. 4, 5, and 14) exhibited the typical progressive respiratory paralysis of interest here, although HC-3 was quite the most potent. Since then, Long ( 1959) has synthesized hemicholiniumtype products in which the diphenyl system is replaced by benzidine, diphenylmethane, and diphenyl ether. Powers (1959) at Tulane has also produced preliminary evidence that a hexamethylene chain may be substituted for the diphenyl system with retention of hemicholiniumlike action. So far, however, no hemicholinium congeners have exhibited potency equal to or greater than HC-3. Gesler and Hoppe ( 1956) reported bis-quaternary salts derived from pyridazine. One of their products in particular 3,6-bis(3-diethylaminopropoxy) pyridazine bismethiodide ( Win 4981 ) Gesler et al., 1959) exhibits a pattern of action strikingly similar to HC-3. Win

MECHANISM OF ACTION OF THE HEMICHOLINIUMS

95

4981 exhibits a slow progressive paralysis of respiration antagonizable by choline. Its blocking action at the neuromuscular junction is markedly facilitated by high frequencies of stimulation and is antagonized by choline and also, but to a lesser degree, by ethanol dimethylethylammonium iodide as is HC-3. Studies by these investigators have lead them to suggest a presynaptic mechanism of action for Win 4981 though it is not yet clear whether Win 4981 acts presynaptically against acetylcholine formation in the same manner as HC-3. Cavallini and his co-workers (1954) have also described ethers of diphenylethane and stilbene which possess a prolonged latent period in onset and are of long duration in action [see also Bovet et al. ( 1959)l. It is yet too early to tell whether any or all of these diverse groups of products act in fundamentally the same fashion on the molecular level. It is however clear that the field of cholinergic pharmacology is still an increasingly complex area of research which is being constantly enriched by new problems arising out of the study of subtly different types of new drugs.

V.

Summary of Effects of H C - 3 at Various Cholinergic Sites

HC-3 has been studied primarily with reference to the neuromuscular junction, autonomic ganglia, postganglionic parasympathetic endings, and the central nervous system. It blocks effectively at all cholinergic sites studied but only when the synapses involved are forced to transmit at high frequency. The primary actions of the drug (i.e., those observed at lowest effective doses) involve a presynaptic depressant effect resulting from decreased acetylcholine formation. Its effect in this latter regard is not upon the acetylation phase of acetylcholine synthesis, but probably upon some process involved in the acquisition of choline by nerve cells. Choline is a specific and dramatic antidote for HC-3 poisoning. Physostigmine also antidotes HC-3 toxicity but to a lesser degree. Physostigmine presumably acts by protecting the preformed transmitter from hydrolysis by acetylcholinesterase. HC-3 blocks acetylcholine synthesis by minced brain tissue, to a lesser degree the synthesis by isolated brain mitochondria, and not at all the synthesis by ether-

96

F. W. SCHUELER

treated mitochondria. HC-3 therefore appears to depend for its action upon cellular and subcellular integrity. If high doses of HC-3 are administered one observes a prompt postsynaptic type of neuromuscular block characteristic of many quaternary ammonium ions. This latter effect is, however, not of primary interest. It is the action of HC-3 at low doses effecting a latent inhibition of presynaptic processes, in particular inhibition of acetylcholine formation, that makes the substance of value as a new tool in cholinergic pharmacology. REFERENCES Birks, R. I., and MacIntosh, F. C. (1957). Brit. Med. Bull. 13, 157. Bovet, D., Bovet-Nitti, F., and Marini-Bettolo, G. B. (1959). “Curare and Curare-like Agents.” Elsevier, Amsterdam. Brooks, V. G. (1951). Science 113, 300. Brooks, V. G. ( 1956). J . Physbl. (London) 134,264. Brooks, V. G., and Thies, R. E. (1959). Personal communication. Cavallini, G., Costa, E., Ferrari, W., and Masserani, E. (1954). Arch. intern. phurmucodynumie 99, 283. Coleman, I. W., McIvor, R. A., Little, P. E., and Grant. G. A. (1958). Personal communication. del Castillo, J., and Katz, B. (1956). Progr. in Biophys. and Biophys. Chem. 6 , 121. Desmedt, J. E. (1958). Nature 182, 1673. Domer, F. R. ( 1959). Personal communication. Farah, A. ( 1959). Personal communication. Farrell, M. A. ( 1959). Personal communication. Gardiner, J. E. (1957). Proc. Physiol. Soc. Phila. 138, 13. Gesler, R. M., and Hoppe, J. 0. (1956). J. P h u r w o l . Exptl. Therap. 118, 388. Gesler, R. M., Lasher, A. V., Hoppe, J. O., and Steck, E. A. (1959). J. Pharmucol. Exptl. Therap. 126, 323. Giovinco, J. F. (1957). Bull. Tulane Univ. Med. F a . 16, 177. Hebb, C. O., and Smallman, B. N. (1956). Nature 178, 365. KasB, Y., and Borison, H. L. (1958). J . Phurmucol. Exptl. Therap. 122, 215. Koelle, G. B., and Steiner, E. C. (1956). J. Phurmmol. Exptl. Therap. 118, 420. Long, J. P. ( 1959). Personal communication. Long, J, P., and Schueler, F. W. (1954). J . Am. Pharm. Assoc., Sci. Ed. 43, 79. Longo, V. G. ( 1959). Arch. intern. pharmacodynamie 119, 1. MacIntosh, F. C. (1958). Can. J . Biochem. and Physiol. 37, 343. MacIntosh, F. C., Birks, R. I., and Sastry, P. B. (1956). Nature 178, 1181. MacIntosh, F. C., Birks, R. I., and Sastry, P. B. (1958). Neurology 8, 90.

MECHANISM OF ACTION OF THE HEMICHOLINIUMS

97

McPhail, M. K. ( 1959). Personal communication. Madison, J. B. (1959). Senior Thesis. Tulane University School of Medicine, New Orleans, Louisiana. Orkand, R. K., and Martin, A. R. (1959). Federation Proc. 18, 430. PfeiEer, C. C. ( 1959). Personal communication. Powers, M. ( 1959). Personal communication. Preston, J. B., and Van Maanen, E. F. (1953). J . Pharmacol. Erptl. Therap. 107, 165. Reitzel, N. L., and Long, J. P. (1959a). Arch. intern. pharmacodynamie~ 119, 20. Reitzel, N. L., and Long, J. P. (1959b). Personal communication. Schueler, F. W. (1955). J . PhamzacoZ Exptl. Therap. 116, 127. Schueler, F. W., Longo, V. G., and Bovet, D. (1954). Arch. itd. sci. famzacol. 4, [111] 36. Shamblin, J. R. ( 1957). Senior Thesis. Tulane University School of Medicine, New Orleans, Louisiana. Taylor, R. D., and Page, I. H. (1951). Circulation 4, 563. Waton, N. G. (1959). Personal communication.

This Page Intentionally Left Blank

THE ROLE OF PHOSPHATIDIC ACID AND PHOSPHOINOSITIDE IN TRANSMEMBRANE TRANSPORT ELICITED BY ACETYLCHOLINE AND OTHER HUMORAL AGENTS By Lowell E. Hokin and Mabel R. Hokin Department of Physiological Chemistry, University of Wisconsin, Madison, Wisconsin

I. Background on Phospholipid Metabolism in Neuronal Tissue . . . 11. Effects of Acetylcholine on Phospholipid Metabolism . . . , . . . . . A. Cerebral Cortex Slices . .. .. .. . . .. .. .. ... . . . . . .. . . . . . . B. The Turnover of Individual Phospholipids in Cerebral Cortex Slices ............................................. C. Phospholipid Turnover in Various Parts of the Brain . . . . . . D. Turnover of Phospholipids in Slices of Sympathetic Ganglia E. The Phospholipid Effect in Endocrine and Exocrine Glands 111. Phospholipid Turnover and Sodium Chloride Secretion in the Salt Gland of the Albatross . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Physiological Significance of the Phospholipid Effect in Brain Cortex Slices and in Sympathetic Ganglia . . . . . . . . . . . . . . . . . . . V. Effects of Acetylcholine on the Incorporation of P32 from Various Precursors into Phosphatidic Acid in Cell-Free Preparations from Brain . .. .. .. .. .. ...._.... . . . . . .. . . . .. . . . ... . A. From Inorganic Phosphate during Oxidative Phosphorylation B. From a-Glycerophosphate-P32 and from Adenosine Triphosphate-P32 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI. Enzymes Concerned in the Acetylcholine-Stimulated Exchange of Phosphate in Phosphatidic Acid . .. .. . . .. . . . . . . . . . .. .. . VII. Explanation for the Discrepancy between the Stimulation of Glycerol-l-C11 Incorporation and P32 Incorporation into ........... Phosphatides ............................... VIII. The Endoplasmic Reticulum as the Site of Phosphatidic Acid and Phosphoinositide Turnover in the Salt Gland . . . . . . . . . . . . . . . IX. Theoretical Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. A Scheme for the Participation of Phosphatidic Acid as a Carrier in the Sodium Pump ... . . . .. . .. .. . . . . .. .. ... . B. Energy Relationships and the Phosphatidic Acid Cycle . . . . c. Significance of the Stimulation of Phosphoinositide Turnover D. Are There Specific Carriers for Individual Substances? . . . . E. The Phospholipid Effect and Pinocytosis . . . . . , . . . . . . . . . . 99

100 103 103 105 107 108 109 110 112 113 113

116 118

119 120 121 121 126 128 130 131

100

LOWELL E. HOKIN AND MABEL R. HOKIN

F. The Phosphatidic Acid Cycle and Synaptic Transmission ... 131 G . On the Mechanism of Action of Acetylcholine ............ 132 X. Summary ............................................. 133 References ............................................. 133

1.

Background on Phospholipid Metabolism in Neuronal Tissue

Lipids comprise about half of the dry matter of the brain, and the phospholipids account for about half of this total lipid (Deuel, 1955). Most of the individual phospholipids found in the brain are not unique to that organ but are also found distributed widely throughout the various tissues; in general, however, the brain contains a higher proportion of phosphatides than do the other tissues. No attempt will be made here to give a comprehensive discussion of the chemistry, history, metabolism, and functions of the phospholipids. The interested reader is referred to the excellent monographs by Witcoff (1951) and Deuel (1955) and to shorter reviews by Sloane-Stanley ( 1952), Dawson ( 1952, 1957), Zilversmit (1955), and Kennedy (1957). In this review we shall be concerned only with that group of phosphatides known as diacyl glycerophosphatides. The structural feature which all of these compounds share is that they are derivatives of a-glycerophosphoric acid, the two remaining free hydroxyl groups of glycerol being esterified with long chain fatty acids predominantly of the Cla and CIS varieties. The degree of unsaturation of the fatty acids varies with the source of the lipid. The simplest member of this group of diacyl glycerophosphatides is phosphatidic acid (Fig. 1), which is diacyl a-glycerophosphoric acid. Esterification of the phosphate of the phosphatidic acid with any one of a variety of substances which contain an alcoholic group gives rise to the other glycerophosphatides. Thus esterification with choline, ethanolamine, serine, or myo-inositol gives rise to lecithin (phosphatidylcholine), phosphatidylethanolamine, phosphatidylserine, or monophosphoinositide ( phosphatidylinositol ), respectively. An additional glycerophosphatide, called diphosphoinositide because it yields inositol metadiphosphate, glycerol, and fatty acid on acid hydrolysis, has been isolated from brain by Folch (1949). Its precise structure has not been elucidated. So far, it has been demonstrated only in brain tissue. Studies on the specific physiological functions of phosphatides

ACIDIC PHOSPHATIDES AS CARRIERS I N MEMBRANES

101

in brain and other organs are just beginning. Since the phosphatides are important constituents of the cytoplasmic membranes, intracellular organelles, and myelin sheaths, a structural role has been assumed as one of their functions. The structural role of phosphatides has been reviewed by Bloor (1943); other functional roles have been reviewed by Beveridge (1956). It is the purpose of this review to discuss, with reference to the nervous system, one particular aspect of physiological function in which some phospholipids appear to play an active role. H ? HC-O-C-R,

9

I

R2- C- ,O-CH

PHOSPHATIDICACID FIG. 1. Structure of L-a-phosphatidic acid. Esterification of the phosphate with choline, ethanolamine, serine, or myo-inositol yields, respectively, lecithin ( phosphatidylcholine),phosphatidylethanolamine, phosphatidylserine, or monophosphoinositide ( phosphatidylinositol) .

With the advent of isotopic tracer techniques a dynamic role for some of the phosphatides was suspected. When animals were fed or injected with P32 the phosphatides were shown to become labeled; the labeling of brain phosphatides was slower than in other organs [see reviews of this subject by Hevesy (1947), Artom et al. (1938), Chaikoff ( 1942), and Chaikoff and Zilversmit ( 1948)1. The slow incorporation of P32 into brain phosphatides in vioo appears to be due to the slow permeation of phosphate through the blood-brain barrier ( Greenberg et al., 1943). Much greater incorporation of P32 into the cerebral phosphatides is obtained if the isotope is injected directly into the cerebrospinal fluid (Bakay and Lindberg, 1949). If slices of brain tissue are incubated in vitro in the presence of oxygen and a suitable substrate, P32 is rapidly incorporated into the phosphatides (Hahn and Hevesy, 1937; Fries et al., 1942; Taurog et al., 1942; Schachner et al., 1942; Findlay and Strickland, 1953; L. E. Hokin and M. R. Hokin, 1955a). Friedkin and Lehninger

102

LOWELL E. HOKIN AND MABEL R. HOKIN

( 1949) showed that liver particulate fractions incorporated P32 into the total phosphatides. This incorporation was dependent on oxidative phosphorylation. Dawson ( 1954) and, later, McMurray et al. (1957b), found that if brain homogenates were incubated with the appropriate substrates and cofactors for oxidative phosphorylation, P32 was rapidly incorporated into the phosphatides. McMurray et al. (1957b) also found some incorporation of P32 into the phospholipid extracts from water dispersions of rat brain after anaerobic glycolysis. The development of paper chromatographic techniques, first for the separation of the mild alkaline hydrolysis products of the phospholipids (Dawson, 1954) and later for the separation of the intact phosphatides (Marinetti and Stotz, 1956), has enabled the turnover of the individual phosphatides to be followed. Dawson (1954) showed that in brain homogenates most of the P 3 2 incorporated into the total lipid extract was in a phosphoinositide and a compound which yielded glycerophosphate on mild alkaline hydrolysis, believed by Dawson to be phosphatidic acid. The identification of this material as a phosphatidic acid has since been confirmed and it has been shown to occur and to become labeled not only in vitro but also in vivo (L. E. Hokin and M. R. Hokin 1958a). Dawson ( 1954) found negligible incorporation of P32 into phosphatidylcholine, phosphatidylethanolamine, and phosphatidylserine in brain dispersions. L. E. Hokin and M. R. Hokin (1955b, 1958c) found that the pattern of incorporation of Pq2into the various phosphatides in brain cortex slices was very similar to that found by Dawson for brain dispersions, the one exception being that in the slices there was an appreciable incorporation of P32 into phosphatidylcholine. It should also be pointed out that the pattern of incorporation of P32 into the various phosphatides in slices is very similar to that found in vivo with a variety of tissues, including brain (Marinetti et al., 1957; L. E. Hokin and M. R. Hokin, 1958a). With cell-free preparations, various coenzymes, substrates, etc. presumably become limiting, so that the turnover of certain phosphatides becomes suboptimal or negligible.

ACIDIC PHOSPHATIDES AS CARRIERS I N MEMBRANES

11.

Effects

103

of Acetylcholine on Phospholipid Metabolism

A. CEREBRAL CORTEX SLICES Our interest in phospholipid metabolism in brain tissue developed from studies on the pancreas. In the course of studies on the incorporation of P32 into various substances in pancreas slices during different phases of the secretory cycle it was found that stimulation of enzyme secretion with cholinergic agents resulted in a marked increase in the incorporation of P32 into the total phosphatide fraction (M. R. Hokin and L. E. Hokin, 1953). Acetylcholine (plus eserine) or carbamylcholine were far more effective in this respect than choline. The incorporation of P32 into the acidsoluble phosphate ester fraction was unaffected. The stimulatory effects of cholinergic agents on both enzyme secretion and on the incorporation of P32into phospholipids were abolished by atropine. Stimulation of amylase synthesis by addition of an appropriate mixture of amino acids to the incubation medium did not affect the incorporation of P32 into phospholipids. As an extension to these studies, the effects of acetylcholine on the incorporation of P32 into the total phospholipid fraction were studied in a variety of tissue slices incubated in uitro. These tissues included liver, kidney cortex, heart ventricle, smooth muscle (pigeon gizzard), pigeon brain, and guinea pig cerebral cortex. Of these the only tissues which showed a stimulation of P32 incorporation into phospholipids on addition of cholinergic agents were the pigeon brain and the guinea pig cerebral cortex. In view of the high concentrations of choline acetylase, acetylcholine, and cholinesterase in certain parts of the central nervous system (CNS), and in view of the important role of acetylcholine in synaptic transmission in ganglia, neuromuscular junctions, and in parts of the CNS, it was felt that this effect of acetylcholine on phospholipid metabolism in brain warranted further study. A more detailed investigation of guinea pig cerebral cortex slices revealed that very high concentrations of acetylcholine were required to produce a maximal stimulation of P32 incorporation into phospholipids (L. E. Hokin and M. R. Hokin, 1955a). The minimally effective concentration of acetylcholine (in the presence of eserine) was M . With 2-hour incubation periods the incorporation of P32 into the total phosphatides was proportional to the

104

LOWELL E. HOKIN AND MABEL R. HOKIN

logarithm of the acetylcholine concentration up to 10-2 M acetylcholine. Eserine alone enhanced P32 incorporation into phospholipids slightly The stimulatory effect of acetylcholine on P32 incorporation increased with time. It was suggested that this increase in the acetylcholine effect with time may have been due to a high impermeability of the slices to acetylcholine, and this may have accounted for the relatively high concentration of acetylcholine required to give a phospholipid effect. Subsequent studies, discussed in Section V, have borne this out. Atropine ( M ) completely abolished the stimulatory effect of acetylcholine on P32 incorporation into the phospholipids. In fact, this concentration of atropine reduced the incorporation of P32about 15% below the control level. This effect was statistically significant. It suggested that the steady state concentration of endogenous acetylcholine in the cerebral cortex tissue was responsible for some of the phospholipid turnover observed in the absence of added acetylcholine. An interesting observation made during the course of these studies was that higher concentrations of atropine stimulated Ps2 incorporation into phospholipids in cerebral cortex slices. The stimulatory effect of atropine on P32incorporation was also proportional to the logarithm of the atropine concentration. At the higher concentrations atropine was about as effective as acetylcholine in stimulating PS2 incorporation. It is probable, however, that for a given concentration of atropine in the medium a higher concentration is reached within the tissue than is the case with acetylcholine. This follows from the fact that tertiary amines penetrate nervous tissue far more readily than quaternary amines. More specifically, Pfeiffer and Jenney (1957) have shown that quaternary nitrogencontaining cholinergic and anticholinergic agents do not produce their respective effects on the central nervous system, while the tertiary nitrogen-containing agents do. Although Pfeiffer and Jenney attributed this low permeation of quaternary nitrogen compounds to the blood-brain barrier, a comparison of the effects of different concentrations of acetylcholine on phospholipid turnover in cerebral cortex slices and in cell-free preparations of brain (see Section V ) suggests that an impenetrability of brain tissue itself is an additional factor. Most of the motor and electrophysiological effects of application of cholinergic agents to the brain are abolished by relatively

ACIDIC PHOSPHATIDES AS CARRIERS IN MEMBRANES

105

low concentrations of atropine. It is therefore of interest that low concentrations of atropine also abolished the effect of acetylcholine on phospholipid turnover in cerebral cortex. In this respect, cerebral cortex resembles glands, the heart, and smooth muscles, where cholinergic effects are abolished by low concentrations of atropine. The stimulatory effect of higher concentrations of atropine on phospholipid turnover in cerebral cortex slices is of some interest. It is known that high doses of atropine and related agents in animals produce CNS excitation. Whether this pharmacological effect on the whole animal is related to the phospholipid effect in vitro is not known. What does seem probable is that the stimulatory effects of high concentrations of atropine on phospholipid turnover in vitro are related to its antagonism to acetylcholine in low concentrations. The fact that atropine can compete with acetylcholine at the acetylcholine “receptor” so as to block the action of the latter suggests that atropine may also be capable, under certain circumstances, of mimicking the action of acetylcholine. In other words, two events are envisaged to occur when acetylcholine stimulates a physiological process. First, acetylcholine combines with a receptor. Second, the combination of acetylcholine with the receptor leads to physiological changes. Atropine is usually considered to combine with the receptor as does acetylcholine, but not to lead to the same physiological changes. It would appear that in the case of phospholipid turnover in cerebral cortex slices the combination of atropine with the receptor can lead to the same end result as occurs with the combination of acetylcholine with the receptor.

B. THE TURNOVER OF INDIVIDUAL PHOSPHOLIPIDS IN CEREBRAL CORTEX SLICES The method of Dawson (1954) made it possible to study the effect of acetylcholine on the incorporation of various labeled precursors into the individual phospholipids (L. E. Hokin and M. R. Hokin, 1955b). This method is based on mild alkaline hydrolysis of the total phospholipid mixture, followed by twodimensional chromatography of the water-soluble hydrolysis products. The specific activities (counts per minute per microgram of phosphorus) of the individual phosphatides were obtained by

106

LOWELL E. HOKIN AND MABEL R. HOKIN

this method. More recently, L. E. Hokin and M. R. Hokin ( 1 9 5 8 ~ ) have also used the method of Marinetti and Stotz (1956) in which the intact phospholipids from brain cortex slices were separated by chromatography on silicic acid-impregnated paper; the total amount of P32 or other precursor in each type of phosphatide was determined by this method. In guinea pig cerebral cortex slices phosphoinositide and phosphatidic acid reach much higher specific activities with respect to P32 than do lecithin, phosphatidylethanolamine, and phosphatidylserine (L. E. Hokin and M. R. Hokin, 1955b). It is the turnover of phosphatidic acid and phosphoinositide ( the two phosphatides which are the most highly acidic and which are the most active metabolically) that is preferentially stimulated by acetylcholine. In twelve experiments with guinea pig cerebral cortex slices the mean increase in P32 incorporation into phosphoinositide was 61% k 13% (SE), into phosphatidic acid 135 & 2176, and into lecithin 33 t 17% (L. E. Hokin and M. R. Hokin, 1 9 5 8 ~ ) ;in four experiments with cat cerebral cortex slices the mean increases were 62 t 8% for phosphoinositide, 98 t 20% for phosphatidic acid, and 41 -t 9% for lecithin (M. R. Hokin et al., 1960). There was no significant effect of acetylcholine on the incorporation of P32 into phosphatidylethanolamine. Folch ( 1949) has isolated a diphosphoinositide from calf brain. In cerebral cortex slices labeled diphosphoinositide could not be demonstrated. The phosphoinositide which exhibited such a high rate of turnover, which was further augmented by acetylcholine, was shown to be a monophosphoinositide ( phosphatidylinositol ) (L. E. Hokin and M. R. Hokin, 1 9 5 8 ~ ) .Monophosphoinositides from animal tissues have been isolated and characterized by Faure and Morelec-Coulon (1953, 1954), McKibbin ( 1956), and Hanahan and Olley ( 1958). The monophosphoinositide in cerebral cortex tissue also incorporated inositol-2-H3, and the incorporation of this isotope was increased by acetylcholine to the same extent as that of P32. Glycerol-l-C14 was incorporated into all of the glycerophosphatides; i.e., phosphatidic acid, phosphoinositide, lecithin, phosphatidylethanolamine, and phosphatidylserine. However, the incorporation of this precursor was not stimulated by acetylcholine in any of the glycerophosphatides (L. E. Hokin and M. R. Hokin, 1958~).A possible explanation for this is discussed in Section VII.

ACIDIC PHOSPHATIDES AS CARRIERS IN MEMBRANES

107

Neither respiration nor the incorporation of P32 into the acidsoluble phosphate ester fraction (M. R. Hokin and L. E. Hokin, 1953) was increased when cerebral cortex slices were incubated in the presence of acetylcholine. These findings, along with the fact that only certain phospholipids were affected by acetylcholine, indicate that the effect of acetylcholine on phospholipid turnover is specific and not secondary to an increased metabolism of the tissue.

C. PHOSPHOLIPID TURNOVER IN VARIOUS PARTSOF

THE

BRAIN

When the effects of acetylcholine on phosphatide turnover in slices of tissue from various structures of the cat and guinea pig brain were studied (M. R. Hokin et al., 1960), the stimulation by acetylcholine of the incorporation of P32 into phosphoinositide and phosphatidic acid was found not to be a general property of all nervous tissue, Acetylcholine did not stimulate the turnover of either phosphatidic acid or phosphoinositide in the cerebellar cortex or the inferior colliculus. However acetylcholine did stimulate the turnover of phosphatidic acid and phosphoinositide in certain other structures, in addition to the cerebral cortex. The incorporation of P32 into phosphoinositide was stimulated in the corpus striatum, hypothalamus, and thalamus to approximately the same extent as in the cerebral cortex. The incorporation of P32 into phosphatidic acid was stimulated in the corpus striatum, superior corpora quadrigemina, the olfactory bulbs, and the hypothalamus; these stimulations were not as great as those observed in the cerebral cortex slices. These results indicate that the stimulation of P32 incorporation into phosphatidic acid and phosphoinositide in response to acetylcholine, rather than being a general property of the neuron or of nervous tissue, occurs only in certain parts of the nervous system. This lends strong support to the idea that the effect reflects a true physiological responsiveness of specialized nervous tissue rather than a general pharmacological response of neurons or neuroglia. Most structures which contain relatively large amounts of choline acetylase (Hebb and Silver, 1956) showed an acetylcholine 1960). Most structures which contain effect (M. R. Hokin et a?., very little choline acetylase showed no response, as, for example,

108

LOWELL E. HOKIN AND MABEL R. HOKIN

the cerebellar cortex. However, there were exceptions. The hypothalamus showed an acetylcholine effect even though its choline acetylase content is as low as that in the cerebellum. Also, the inferior colliculus, which is relatively rich in choline acetylase, failed to give a response.

D. TURNOVER OF PHOSPHOLIPIDS IN SLICESOF SYMPATHETIC GANGLIA The fact that acetylcholine is a synaptic transmitter in sympathetic ganglia is now fairly firmly established (see Eccles, 1953, 1957 for reviews). For this reason it was of considerable interest to determine whether sympathetic ganglion tissue would also show a phospholipid effect in response to acetylcholine. This was found to be the case (M. R. Hokin et al., 1960). Acetylcholine (5 x lO-6M) stimulated about 100% the incorporation of Psz into both phosphoinositide and phosphatidic acid in stellate ganglion slices and in coeliac ganglion slices from the cat; the incorporation of Ps2 into lecithin, phosphatidylethanolamine, and phosphatidylserine was not stimulated. Preganglionic sympathetic nerve fibers (splanchnic nerve) and the vagus and femoral nerves did not show a stimulation. The postganglionic sympathetic nerve fibers (splenic nerves) from the coeliac ganglion sometimes appeared to show a response, but this effect could not be obtained consistently. It has not been possible to rule out that these occasional stimulations may have been due to the presence of “displaced” cell bodies in the tissue samples. The most likely explanation for the phospholipid effect in sympathetic ganglia is that it is related to synaptic transmission. The finding of a phospholipid effect in relatively simple synaptic structures in which acetylcholine is known to be a synaptic transmitter throws some light on this effect in the cerebral cortex and other structures in the brain. The simplest explanation for the phospholipid effect in these structures in the brain would appear to be that it is also related to synaptic transmission. An alternative explanation for the phospholipid effect in sympathetic ganglia would be that it is concerned with the transport, either intracellular or extracellular, of noradrenaline or of an enzyme concerned in its synthesis. This is mentioned because a phospholipid effect is also observed in adrenal medulla slices on stimulation of adrenaline secretion with acetylcholine. However,

ACIDIC PHOSPHATIDES AS CARRLERS IN MEMBRANES

109

no evidence is so far available that noradrenaline is secreted from the soma of ganglion cells into the blood stream or is transported from the cell body down the axon in response to acetylcholine. With regard to the intracellular transport of substances generally no evidence has so far been obtained which indicates that the phospholipid effect is concerned in such a process.

E. THEPHOSPHOLIPID EFFECT IN ENDOCRINE AND EXOCRINE GLANDS Prior to the studies on nervous tissue it had been found that stimulation of the secretion of digestive enzymes by pigeon pancreas slices was associated with a marked stimulation of the incorporation of P32 into phospholipids ( M . R. Hokin and L. E. Hokin, 1953). Although in the early stages of the work it was thought that the phospholipid effect was a specific response to acetylcholine this view proved to be incorrect when it was found that pancreozymin-a polypeptide hormone-stimulated phospholipid turnover and enzyme secretion in pancreas slices in exactly the same manner as does acetylcholine (L. E. Hokin and M. R. Hokin, 1956a). Previous physiological studies had shown that pancreozymin specifically stimulates enzyme secretion in the pancreas (Harper and Raper, 1943). Pure secretin, which stimulates only water and bicarbonate secretion in the pancreas (Mellanby, 1925; Harper and Vass, 1941; Jorpes and Mutt, 1954) was without effect on phospholipid turnover or enzyme secretion. It should be pointed out, however, that there was no way of testing whether secretin actually stimulated water and bicarbonate secretion in pigeon pancreas slices. The phospholipid effect in pancreas showed a similar pattern to that of brain cortex; i.e., phosphoinositide and phosphatidic acid turnover were stimulated the most (L. E. Hokin and M. R. Hokin, 195513, 1958b). In pancreas phosphatidylethanolamine also showed some stimulation, but phosphatidylcholine was affected very little. The stimulation of phosphoinositide turnover in pancreas was very much greater than in brain cortex slices, being as much as 15-25fold in some experiments. The stimulation of phosphatidic acid turnover was very much less than that of phosphoinositide, being of the same order as observed in brain cortex slices, or even less in some experiments. The phospholipid effect in pigeon pancreas was shown to occur

110

LOWELL E. HOKIN AND MABEL R. HOKIN

in membranous components of the microsome fraction; the role of these membranes in enzyme secretion in the pancreas has been discussed by Redman and Hokin (1959). A stimulation of the turnover of phosphoinositide and phosphatidic acid was also found in salivary glands when protein secretion was stimulated either by acetylcholine or adrenaline (Hokin and Sherwin, 1957; Eggman and Hokin, 1960); in pigeon esophagus on stimulation of pepsin secretion with acetylcholine (Eggman and Hokin, 1960); in the adenohypophysis on stimulation of adrenocorticotropin secretion by corticotropin-releasing factor ( M. R. Hokin et al., 1958b); and in guinea pig adrenal medulla on stimulation of adrenaline secretion by acetylcholine ( M . R. Hokin et al., 1958a). Morton and Schwartz (1953) observed that incubation of beef thyroid slices with thyrotropin stimulated P32 incorporation into phospholipids, and Freinkel (1957) showed that the stimulation was greatest in phosphoinositide and phosphatidic acid. The secretion of thyroid hormones under these conditions has not been measured; however, in view of the response in other glands it seems likely that the phospholipid effect in the thyroid gland may be related to the stimulation of secretion. The specific increase in the turnover of phosphoinositide and phosphatidic acid on stimulation of secretion in a variety of exocrine and endocrine glands, irrespective of the stimulating agent, lends strong support to the hypothesis that the phospholipid effect is concerned with the extrusion of substances from the cell.

111.

Phospholipid Turnover and Sodium Chloride Secretion in the Salt Gland of the Albatross

In view of the relationship between phospholipid turnover and the active transport of organic molecules from glandular cells it is reasonable to assume that the phospholipid effect in brain cortex and in sympathetic ganglia is concerned with the active transport of some substance. Recently, new light has been thrown on this problem from studies on the salt gland of the albatross (L. E. Hokin and M. R. Hokin, 1959c, 1960; M. R. Hokin and L. E. Hokin, 1960a, b). The salt glands of marine birds secrete solutions of sodium chloride in concentrations as high as 0.84 M (SchmidtNielsen, 1959; Schmidt-Nielsen et al., 1958; Schmidt-Nielsen and

ACIDIC PHOSPHATIDES AS CARRIERS IN MEMBRANES

111

Sladen, 1958). The secretion is essentially free of organic substances or of other electrolytes. The secretory activity of the gland can be stimulated by cholinergic agents since it is innervated by a branch of the facial nerve ( cholinergic) which regulates its activity. The salt gland thus provides an excellent system to study the role of phosphatides in ion transport. Incubation of slices of the salt gland of the Black-footed or the Laysan albatross with acetylcholine ( lO-5M) plus eserine ( 10-4M) led to a dramatic 10- to 20-fold stimulation of the incorporation of P32 into phosphatidic acid, with a lesser (3-fold) stimulation of P32 incorporation into phosphoinositide. The incorporation of P32 into phosphatidylcholine and phosphatidylethanolamine was increased relatively slightly or not at all. In slices of the salt gland orthophosphate enters the cell very slowly; after 1 hour’s incubation, the specific activity of the intracellular orthophosphate was only one-tenth of that of the extracellular orthophosphate. The rate of exchange of extracellular with intracellular orthophosphate was not affected by acetylcholine. The incorporation of P32 into ATP ( 7-minute acid-hydrolyzable phosphate), phosphoprotein, and the acid-soluble phosphate ester fraction was not stimulated by acetylcholine, indicating that the effect on the phosphatides was specific. Since ATP is the precursor for phosphatidic acid when the turnover of phosphatidic acid is stimulated by acetylcholine (see Section V, B ) , the fact that the incorporation of P32 into ATP was not stimulated indicates that the increase in the incorporation of P32 into phosphatidic acid could not have been secondary to an increase in ATP32 specific activity. It could be argued that when phosphatidic acid synthesis from ATP is markedly augmented the turnover of ATP would also be increased and thus its specific activity. However, the specific activity of ATP at incubation times of 10 minutes and greater was found to be the same as the specific activity of the intracellular orthophosphate. This is not surprising, since Krebs et al. (1953) had previously shown that the labile phosphate groups of ATP equilibrate with orthophosphate in a similar period of time in mitochondria undergoing oxidative phosphorylation. It follows that any increase in the rate of turnover of the terminal phosphate of ATP could not lead to an increase in its specific activity under the conditions of our experiments. The oxygen uptake of slices of salt gland was doubled in the

112

LOWELL E. HOKIN AND MABEL R. HOKIN

presence of acetylcholine; this provided evidence that the tissue was undergoing increased secretory activity. Deutsch and Raper (1936, 1938) found that incubation of slices of submaxillary and parotid glands of the cat, dog, and rabbit with acetylcholine (plus eserine) or of cat pancreas slices with secretin led to a stimulation of respiration; the increase in respiratory activity was essentially the same as observed in vim on stimulation of secretion in these glands. Other workers have also reported a stimulation by secretin of respiration in isolated pancreatic tissue of some species (Gerard and Still, 1933; Kiyohara, 1934; Davies et al., 1949). Maximal stimulation of P32incorporation into phosphatidic acid M acetylcholine (plus M eserine). was achieved with However, eserine alone led to some stimulation, and the stimulation increased with increasing concentrations of acetylcholine from 10-8 M to 10-4 M . The data indicate that stimulation of salt secretion in slices of the salt gland of the albatross is associated with the stimulation of the turnover of phosphatidic acid and, to a lesser extent, of phosphoinositide. These effects are very similar to those observed in brain cortex and sympathetic ganglia, in response to acetylcholine, except that in the salt gland the stimulations are much greater.

IV. Physiological Significance of the Phospholipid Effect in Brain Cortex Slices and in Sympathetic Ganglia

Based on the fact that synaptic tissue was the only nonglandular tissue which gave a phospholipid effect it was suggested (L. E. Hokin and M. R. Hokin, 195613) that the simplest hypothesis for the phospholipid effect in synaptic tissue was that it was concerned with the secretion of some as yet unidentified neurohumoral agent. Efforts by the authors to identify such a substance have been unsuccessful. The fact that the turnover of phosphatidic acid and phosphoinositide is concerned with sodium transport in the salt gland suggests that the phospholipid effect in the responsive parts of the central nervous system and in sympathetic ganglia is likely to be concerned with sodium transport. The biochemical nature of the phospholipid effect in the various tissues suggest that this is so. In the brain and salt gland phosphatidic acid turnover is stimulated to a greater extent than phosphoinositide turnover (L.

ACIDIC PHOSPHATIDES AS CARRIERS IN MEMBRANES

113

E. Hokin and M. R. Hokin, 1958c, 1960). However, in proteinsecreting tissue, particularly the pancreas, phosphoinositide turnover is stimulated to a much greater extent than phosphatidic acid turnover (L. E. Hokin and M. R. Hokin, 1958b). Furthermore, in brain and salt gland the incorporation of inositol-2-H3 into phosphoinositide is stimulated to the same extent as P32. In pancreas the incorporation of inositol-2-H3 into phosphoinositide is stimulated only about one-fourth as much as the incorporation of P32. These data suggest that the physiological function of the phospholipid effect is the same in the salt gland and in synaptic tissue and that this function is different in pancreas. When tissue slices from the appropriate structures from the brain or slices of sympathetic ganglia are incubated with acetylcholine and eserine the postsynaptic membranes responsive to acetylcholine would become depolarized. This would be associated with a continual influx of sodium ions into the postsynaptic cell body and an increased activity of the “sodium pump” which extrudes sodium. The turnover of phosphatidic acid is envisaged as participating in the flow of cations across the postsynaptic membrane. As discussed in Section IX it is felt that the turnover of phosphatidic acid is most likely to be concerned with the “sodium pump.”

V.

Effects of Acetylcholine on the Incorporation of P32 from Various Precursors into Phosphatidic Acid in Cell-Free Preparations from Brain

A. FROMINORGANIC PHOSPHATE DURING OXIDATIVE

PHOSPHORYLATION Experiments with cell-free systems were carried out in order to study the enzyme catalyzed reactions concerned in the stimulation by acetylcholine of phospholipid turnover in brain (L. E. Hokin and M. R. Hokin, 1958d). When cytoplasmic particulate fractions of guinea pig brain were incubated under conditions which supported oxidative phosphorylation, the only known phosphatide which became appreciably labeled was phosphatidic acid. Dawson ( 1954) found appreciable labeling of both phosphoinositide and phosphatidic acid in brain dispersions. The labeling of phosphoinositide appears to require factors present in the soluble

114

LOWELL E. HOKIN AND MABEL R. HOKIN

fraction; this fraction was removed in our experiments on cytoplasmic particulate fractions. Several other chloroform-soluble, phosphorus-containing substances became rapidly labeled with P32. These substances, which have not yet been identified, do not behave as typical phosphatides. They streak in solvents which separate the phosphatides as discrete spots. They can be separated as three discrete spots, each with a lower R, than the known phosphatides, when run on silicic acid-impregnated paper with phenol-ammonia as the solvent; in this solvent the known phosphatides all run more or less together near the solvent front. Dawson (1954) also found several unidentified P32-labeled hydrolysis products from lipid extracts of brain dispersions. Addition of to M acetylcholine (with eserine) to the incubation medium stimulated the incorporation of P32 into phosphatidic acid in the cytoplasmic particulate fractions. No stimulation was observed with 10-7 M acetylcholine and below. The stimulation with acetylcholine increased linearly with the logarithm of the acetylcholine concentration from 10-0 M to about 5 x M , above which the stimulation remained constant to M . With optimal concentrations of acetylcholine the stimulation of Pa2 incorporation into phosphatidic acid ranged from 20 to 70%. Stimulations of an equal magnitude were observed in the cytoplasmic fraction (whole homogenate minus nuclei and cell debris), washed cytoplasmic particulate fraction, and the mitochondrial fraction. P32 was also incorporated into phosphatidic acid in the microsomal fraction, although only about one-half as effectively as in the mitochondrial fraction. Since the incorporation of P32 from inorganic phosphate into the phosphatides in brain cell-free preparations is dependent on oxidative phosphorylation ( Dawson, 1953), it is likely that some mitochondria were present in the microsome fraction. The other alternative is that brain microsomal material is unique in that it can carry out oxidative phosphorylation. Acetylcholine stimulated the incorporation of P3? into phosphatidic acid in the microsome fraction more than twice as much as in the mitochondrial fraction. The microsomal fraction contains fragments of membranous material, mainly derived from the endoplasmic reticulum. In view of the fact that the membranous material of this fraction was found to be the site of the greatest stimulation in intact cells (Redman and Hokin, 1959) it seems

ACIDIC PHOSPHATIDES AS CARRIERS IN MEMBRANES

115

most probable that the stimulated turnover of phosphatidic acid in the isolated microsomal preparations was also in the membranous material of the fraction. Since numerous electron microscope studies have revealed that the mitochondrial fractions are heavily contaminated with microsomal components (see e.g., Siekevitz and Palade, 1958) it seems most likely that the smaller stimulation observed in the mitochondrial fraction is due to contamination of this fraction with microsomal elements. The fact that 5 X M acetylcholine gave a maximal phospholipid effect in the cell-free system, whereas much higher concentrations were required to give a maximal effect in slices, indicates that cerebral cortex slices are highly impermeable to acetylcholine, thus offering a satisfactory explanation for the fact that relatively high external concentrations of acetylcholine are required to give a maximal effect in cerebral cortex slices; the lowest concentration of acetylcholine which showed a phospholipid effect in cerebral cortex slices was about 10W5 M, as compared with 10-6 M in the cell-free system. It is known that quaternary cholinergic agents enter the brain very little if at all in uivo (Pfeiffer and Jenney, 1957). However, the classic blood-brain barrier is partially interrupted in the tissue slice experiments, and still the permeation of acetylcholine is very poor. It would appear that the brain substance itself offers considerable resistance to the diffusion of acetylcholine. This situation in nerve tissue is different from that in glandular tissues, which appear to be highly permeable to acetylcholine. From the physiological point of view one might expect this, since acetylcholine is probably liberated and acts at specific synapses in the brain; diffusion to other synaptic regions would reduce its specificity of action. There are other reasons for believing that some “barrier” exists in the region of the synapse which prevents the diffusion of acetylcholine ( see Eccles, 1957). In glands there is a homogeneous population of secretory cells all responding to acetylcholine and all carrying out the same function; here, no impairment of function would result from diffusion of acetylcholine from the region of one cell to another. The concentrations of acetylcholine which are effective in the cell-free system are within the physiological range. The concentration of free and bound acetylcholine in the guinea pig brain is about 10-5 M (Mann et al., 1938), which is ten times the con-

116

LOWELL E. HOKIN AND MABEL R. HOKIN

centration required to give some stimulation of phosphatidic acid turnover in the cell-free system. As with slices, very low concentrations of atropine ( M) abolished the stimulatory effect of acetylcholine on the phosphatidic acid turnover in cell-free systems undergoing oxidative phosphorylation. However, unlike the situation in slices, higher concentrations of atropine did not stimulate phosphatidic acid turnover in cell-free preparations of brain.

B. FROM~-GLYCEROPHOSPHATE-P~~ AND FROM ADENOSINE TRI€'HOSPHATE-P32 Kornberg and Pricer (1953a,b) showed that phosphatidic acid is formed from a-glycerophosphate ( a-glycero-P) and fatty acylCoA (CoASR) according to the following reaction: a-Glycero-P + 2 CoASR + phosphatidic acid

+ 2 CoASH

( 1)

Fatty acyl-CoA can be generated from fatty acids and CoA in the presence of ATP and the appropriate enzyme. McMurray et al. ( 1 9 5 7 ~ )found that phosphatidic acid can be formed from aglycero-P in either oxidizing or glycolyzing brain preparations incubated with CoA. The question arose as to whether this was the reaction which was stimulated in brain microsomes in the presence of acetylcholine. Accordingly, microsomes were incubated with a - g l y ~ e r 0 - Pand ~ ~ a nonoxidative ATP-generating system (L. E. Hokin and M. R. Hokin, 1959b). The nonoxidative ATP-generating system consisted of substrate levels of carbamyl phosphate ( carbamyl-P ) , catalytic amounts of adenosine diphosphate (ADP), magnesium ions, and an extract of Streptococcus faecalis containing the enzyme carbamate kinase which catalyzes the formation of ATP from ADP and carbamyl-P. Under these conditions there was an appreciable incorporation of a - g l ~ c e r 0 - P ~ ~ into phosphatidic acid in brain microsomes, but this incorporation was not stimulated by acetylcholine. This indicated that the formation of phosphatidic acid from a-glycero-P and fatty acyl-CoA, according to the mechanism of Kornberg and Pricer (1953a, b), is not increased in the presence of acetylcholine. The effects of acetylcholine on the incorporation of P32 into phosphatidic acid in a system in which ATP2 was the labeled precursor was then tested. Initially, microsomes were incubated with substrate levels of ATP32 (labeled in the y-phosphate).

ACIDIC PHOSPHATIDES AS CARRIERS IN MEMBRANES

117

Phosphatidic acid became weakly labeled. The weak labeling of phosphatidic acid was probably due to the rapid hydrolysis of the ATP32 by a powerful ATPase present in these preparations. Microsomes were then incubated with the same nonoxidative ATPgenerating system as above, but in this case ~arbamyLP3~ was substituted for nonradioactive carbamyl-P, and a-glycero-P was omitted. Under these conditions ATP32 is continuously generated. In the presence of the ATP32-generatingsystem, phosphatidic acid became readily labeled in the brain microsomal preparations; under these conditions the incorporation of P32 into phosphatidic acid was stimulated by acetylcholine. The incorporation of P32 from ATP32 was not via the prior formation of a - g l ~ c e r o - P ~ ~ , since the addition of unlabeled glycero-P did not reduce the incorporation of P32 from ATP32 into phosphatidic acid. These results indicated that the enzyme reaction involved in the stimulation by acetylcholine utilized ATP32 as the radioactive precursor and did not utilize glycerophosphate. The precise reactions involved were further investigated in a solubilized system and are discussed in Section VI. As in the oxidative system, the stimulatory effect of acetylcholine on the incorporation of P32 from ATP32 into phosphatidic acid was abolished by atropine (lo-" M ) ; atropine by itself in this and higher concentrations had no effect on P32 incorporation into phosphatidic acid. Studies on the effect of various manipulations of the microsomal fractions on the responsiveness to acetylcholine as measured by the phospholipid effect yielded information which perhaps has the greatest bearing on the mechanism of action of acetylcholine generally. It was found that the acetylcholine effect was dependent on the structural integrity of the responsive component of the microsome fraction. Any treatment which would be expected to alter the structure of the membranes abolished the stimulating effect of acetylcholine. Such treatments were freezing of the sucrose suspensions of the microsomes prior to incubation, suspension of the microsomes in distilled water, or addition of low concentrations of deoxycholate to sucrose suspensions of microsomes. Deoxycholate is known to solubilize the lipoprotein membranes of the microsome fraction (Palade and Siekevitz, 1956). Another finding was that each of these treatments which abolished the acetylcholine effect actually enhanced the incorporation of P32 into

118

LOWELL E. HOKIN AND MABEL R. HOKIN

phosphatidic acid. This was particularly evident with deoxycholate, which over a certain range of concentrations enhanced the incorporation of P32 into phosphatidic acid ten to twenty times. These studies indicate that the stimulation of phosphatidic acid turnover by acetylcholine is dependent on the structural integrity of the responsive membranes. VI.

Enzymes Concerned in the Acetylcholine-Stimulated Exchange of Phosphate in Phosphalidic Acid

The fact that deoxycholate solubilized lipoprotein membranes in the microsome fraction and the fact that this treatment enhanced P32 incorporation into phosphatidic acid from ATP32 in brain microsomal preparations suggested that the enzymes responsible for this incorporation were present in the membranous material of the microsomal preparations and that they might be solubilized by deoxycholate. Soluble extracts were prepared by treatment of the microsomal fractions with 0.006 M deoxycholate, followed by centrifuging for 30 minutes at 105,000 g to remove the insoluble residue. The soluble extracts were found to catalyze the formation of PWabeled phosphatidic acid from ATP32 (M. R. Hokin and L. E. Hokin, 1959). The formation of phosphatidic acid was markedly increased by the addition of diglyceride which had been prepared by the removal of phosphate from cabbage phosphatidic acid by the action of prostatic phosphomonoesterase. Previous studies had indicated that the fatty acid composition of cabbage phosphatidic acid is very similar to that of brain phosphatidic acid (L. E. Hokin and M. R. Hokin, 1958a). Channon and Chibnall (1927) showed that cabbage phosphatidic acid contains a very high proportion of linoleic acid. In the deoxycholate-solubilized preparations from brain microsomes, diglycerides containing fullysaturated fatty acids were ineffective, and diolein was much less effective than the diglyceride from cabbage phosphatidic acid as substrate for phosphatidic acid synthesis. This suggests that Iinoleic acid is an important component of the substrate for brain diglyceride kinase. In the soluble extracts, the amount of phosphatidic acid formed from a - g l y ~ e r 0 - Pwas ~ ~ only one three-hundredth of that formed from ATP32 in the absence of added diglyceride substrate and less than one two-thousandth of that formed from ATP32

ACIDIC PHOSPHATIDES AS CARRIERS IN MEMBRANES

119

in the presence of added substrate. This indicated that when the extracts were incubated with ATP32 under the conditions used in these experiments, P32could not have been incorporated into phosphatidic acid in the form of a - g l y ~ e r 0 - P ~No ~ . substances other than diglyceride and ATP could be shown to be required for phosphatidic acid synthesis in dialized soluble extracts of brain microsomes. The observations indicated that when ATP32was the radioactive precursor, phosphatidic acid was formed from diglyceride and ATP by a diglyceride kinase reaction: ATP

diglyceride + diglyceride->phosphatidic kinase

acid

+ ADP

(2)

Cabbage phosphatidic acid became rapidly labeled with P32 when it was added to the soluble deoxycholate extracts of microsomes incubated in the presence of the ATP32-generating system. This indicated that an enzyme was present which cleaved the phosphate from the phosphatidic acid to form diglyceride, and that this in turn became rephosphorylated by the action of diglyceride kinase. Smith et al. (1957) found a phosphatidic acid phosphatase in a variety of tissues. M. R. Hokin and L. E. Hokin (1959) found that deoxycholate extracts of brain microsomes liberated orthophosphate when incubated with cabbage phosphatidic acid. These studies indicate the following reaction: Phosphatidic acid

+ HO,-

phosphatidic acid phosphat ase

diglyceride

+ H3P04

(3)

Very little, if any, phosphate was liberated from distearoyl phosphatidic acid, again indicating a specificity in the brain system for lipid substrates with unsaturated fatty acids. The phosphatidic acid phosphatase from brain microsomal extracts was stimulated by magnesium ions. This is in contrast to the phosphatidic acid phosphatase of Smith et al. (1957), which was inhibited by magnesium ions.

VII.

Explanation for the Discrepancy between the Stimulation of Glycerol- I -C14 Incorporation and P32 Incorporation into Phosphatides

As discussed in Section IIB above, although P32 incorporation into phosphatidic acid and phosphoinositide in brain cortex slices

120

LOWELL E. HOKIN AND MABEL R. HOKIN

was stimulated by acetylcholine, the incorporation of glycerol-l-C14 into these phosphatides was not stimulated. In the case of phosphatidic acid, the finding of two synthetic pathways, only one of which is stimulated by acetylcholine, offers an explanation for this discrepancy. A considerable proportion of the glycerol-l-C12 entering phosphatidic acid in intact cells is likely to be via the mechanism of Kornberg and Pricer (1953a, b ) in which phosphatidic acid is formed by acylation of glycero-P, which in turn may be formed directly or indirectly from glycerol. On the other hand, the C14-labeled precursor for the synthesis of phosphatidic acid by the diglyceride kinase reaction, which is the reaction stimulated by acetylcholine, would be labeled diglyceride. No evidence has so far been obtained that diglycerides are formed from free glycerol. Rather they appear to be derived from phosphatidic acid and presumably also from triglycerides. By the time glycerol-l-C14 has entered diglyceride via these pathways it has undergone much greater isotopic dilution than is likely to occur with the formation of a-glycero-P. A situation must thus exist in which most of the glycerol-l-C14 enters phosphatidic acid via a pathway which is not stimulated by acetylcholine, while much of the P32 enters via a pathway which is stimulated by acetylcholine. This can account for the discrepancy between the stimulation of P32 incorporation and glycerol-l-C14 incorporation into phosphatidic acid on stimulation with acetylcholine. The fact that a similar discrepancy exists in the case of phosphoinositide suggests that this phosphatide is also formed by more than one pathway, and that only one of these pathways is stimulated by acetylcholine.

VIII.

The Endoplasmic Reticulum as the Site of Phosphatidic Acid and Phosphoinositide Turnover in the Salt Gland

Doyle ( 1959) has found an extensive smooth-surfaced endoplasmic reticulum in the salt gland of the gull. Many of the membranes of the endoplasmic reticulum can clearly be seen to be in continuity with the apical and basal surface membranes. The membranes of the endoplasmic reticulum may thus be regarded as secretory membranes. This infolding of the surface membranes provides a very large secretory surface.

ACIDIC PHOSPHATIDES AS CARRIERS IN MEMBRANES

121

Palade and Siekevitz ( 1956 ) , Siekevitz and Palade ( 1958 ) , Hanzon and Toschi (1959), and others have shown that when homogenates of various tissues are subjected to differential centrifugation, fragments of the endoplasmic reticulum sediment in the microsome fraction. If the phospholipid effect in the salt gland is concerned with secretion one would expect this effect to be found primarily in fragments of the endoplasmic reticulum, or the microsome fraction. Such was found to be the case (L. E. Hokin and M. R. Hokin, 1960). Slight stimulations were found in other fractions; these can be accounted for by the presence in these fractions of unbroken cells or a small amount of microsomal membranes, depending on the fraction. The observation that the phospholipid effect occurred in the endoplasmic reticulum is similar to the finding in the pancreas (Redman and Hokin, 1959). The experiments reported in Sections V, B and VI showed that acetylcholine stimulated the turnover of phosphatidic acid in incubated microsomal membranes of the brain and that this stimulated turnover was brought about by the combined action of diglyceride kinase and phosphatidic acid phosphatase. Attempts to obtain a similar stimulation of phosphatidic acid turnover in incubated microsome fractions of the salt gland were unsuccessful (L. E. Hokin and M. R. Hokin, 1960). The salt gland resembles other glands, where it has also not been possible to obtain a phospholipid effect in cell free preparations. The reason why glandular tissues differ from brain in this respect is not clear. However, it was possible to show that diglyceride kinase and phosphatidic acid phosphatase were present in high concentrations in the microsome fraction of the salt gland (L. E. Hokin and M. R. Hokin, 1960). And since it could be shown that the phospholipid effect which occurred in slices could be recovered in the microsome fraction it is reasonable to conclude that the stimulated turnover of phosphatidic acid in the salt gland is brought about by the combined action of diglyceride kinase and phosphatidic acid phosphatase.

IX. Theoretical Considerations A, A SCHEMEFOR THE PARTICIPATION OF PHOSPHATIDIC ACID AS A CARRIER IN THE SODIUMPUMP The usual definition of active ion transport is the movement of ions against an electrochemical gradient ( Rosenberg, 1948). The

122

LOWELL E. HOKIN AND MABEL R. HOKIN

membranes which divide the contents of the secretory cells of the salt gland from acinar lumina-or the apical membranes as we shall call them-secrete sodium chloride against a high concentration gradient. Since the direction and magnitude of the electromotive force across this membrane is not known, it cannot be stated with rigorous physicochemical proof that it is sodium ions rather than chloride ions which are secreted against an electrochemical gradient. However, in other tissues in which the active transport of sodium chloride has been studied, i.e., the frog skin, (Ussing and Zerahn, 1951; Ussing, 1952), the nerve axon membrane (Hodgkin, 1951; Keynes, 1951), and the muscle fiber membrane (Boyle and Conway, 1941), sodium has been shown to be the ion actively transported; the movement of chloride passively follows its electrochemical gradient. In our interpretation of the role of phosphoinositide and phosphatidic acid in active transport in the salt gland we have assumed that these phosphatides participate in the transport of sodium rather than chloride. This assumption is based partly on analogy with the tissues listed above, and partly on other grounds. First, if the turnover of phosphoinositide and phosphatidic acid were concerned in active transport of chloride we would expect to see increased turnover of these phosphatides on stimulation of HC1 secretion in the stomach, where chloride is actively transported against an electrochemical gradient ( Rehm, 1950; Hogben, 1955). Preliminary experiments indicate that no phospholipid effect occurs on stimulation with histamine of the mucosae of frog stomach, pigeon esophagus, or mouse stomach. Second, phosphatidic acid and phosphoinositide are themselves anions so that they qualify as possible cation carriers rather than anion carriers. The combined biochemical, cytological, and physiological observations in the salt gland and in brain tissue suggest a scheme whereby phosphatidic acid participates in the active extrusion of sodium across membranes (Fig. 2 ) . This process is assumed to be the same in the salt gland and in neuronal and other cells which extrude sodium against an electrochemical gradient. In this scheme, which may be termed the phosphatidic acid cycle, ATP reacts with diglyceride ( D G ) at the inner surface of the membrane to form phosphatidic acid ( PA), or more correctly phosphatidate. Phosphatidate picks up two sodium ions to form lipid-soluble (Kates, 1955) sodium phosphatidate ( Na2PA). Sodium phosphati-

ACIDIC PHOSPHATIDES AS CARRIERS IN MEMBRANES

123

date diffuses across the lipoidal membrane to its outer surface, where it is hydrolyzed by phosphatidic acid phosphatase to form sodium phosphate and diglyceride. The sodium ions, being water soluble, enter the aqueous phase of the lumen. Diglyceride, which is lipid soluble, diffuses back through the lipoidal phase of the membrane to its inner surface, where the cycle is repeated. The fate of the phosphate, and a suggested mechanism of chloride transport are discussed later. CYTOPLASM

MEMBRANE

DUCTULE LUMEN

FIG. 2. A scheme for the participation of phosphatidic acid in sodium transport across the apical membrane of the salt gland. ATP = adenosine triphosphate, DG = diglyceride, DG-Kinase = diglyceride kinase, PA = phosphatidic acid, PAP-ase = phosphatidic acid phosphatase.

The following observations support this scheme. ( 1 ) Phosphatidic acid turnover is markedly increased on stimulation of sodium chloride transport in the salt gland. ( 2 ) The stimulation of phosphatidic acid turnover by acetylcholine occurs in washed membrane fragments of the endoplasmic reticulum of brain. ( 3 ) The stimulated turnover of phosphatidic acid in these membranes is brought about by the combined action of diglyceride kinase and phosphatidic acid phosphatase. ( 4 ) The site of stimulation of phosphatidic acid turnover in slices of the salt gland is also the endoplasmic reticulum. ( 5 ) Diglyceride kinase and

124

LOWELL E. HOKIN AND MABEL R. HOKIN

phosphatidic acid phosphatase are present in high concentrations in the membranes of the endoplasmic reticulum of the salt gland. (6) The membranes of the endoplasmic reticulum in the salt gland can be clearly seen to be in continuity with the apical and basal surface membranes, so they may be regarded as secretory membranes. ( 7 ) The phospholipid effect has been observed in association with the secretion of many hydrophilic substances which have cationic groups-proteins, polypeptides, catechol amines, and sodium ions. ( 8 ) A phospholipid effect has not been observed in association with the secretion of neutral lipophilic substances; stimulation of the secretion of corticosteroids in adrenal cortex tissue with adrenocorticotropin is not associated with an increase in phospholipid turnover ( M . R. Hokin et al., 1958b). The secretion of lipid-soluble substances would not require a lipoidal carrier as is the case with hydrophilic substances. In this scheme certain assumptions are made. It is assumed that diglyceride kinase is located at the inner surface of the membrane and that phosphatidic acid phosphatase is located at the outer surface. Actually, no localization of diglyceride kinase need be postulated if it is assumed that the ATP, which reacts with diglyceride, is present in the cytoplasmic fluid but not in the luminal or extracellular fluid-an assumption which seems very likely. However, it must be assumed that phosphatidic acid phosphatase is located primarily at the outer surface of the membrane; otherwise, the phosphatidic acid would be hydrolyzed before it reached the outer surface of the membrane. Rothstein and Meier (1948) have shown that certain phosphatases are localized at the outer surfaces of cells. It is also assumed that sodium phosphatidate diffuses along its concentration gradient from the inner surface of the membrane to its outer surface and that diglyceride in turn diffuses back. Diffusion, as used here, would cover a short distance, since the distances across the membrane are probably only a few times the diameter of sodium phosphatidate. It is unlikely that the diffusion would be free as in a homogeneous system. Diffusion is postulated because it is the simplest explanation of a step in the scheme on which we have no information. Other types of movement are obviously possible; e.g., the rotation of a globular protein containing phosphatidate groups on its surface, or the swinging between the inner surface and outer surface of the ends of protein chains

ACIDIC PHOSPHATIDES AS CARRIERS IN MEMBRANES

125

carrying phosphatidate. But all of these mechanisms are more complicated than simple diffusion and have no evidence to support them. On hydrolysis of sodium phosphatidate by phosphatidic acid phosphatase at the outer surface of the apical membrane of the salt gland, phosphate could accompany the sodium ions into the acinar lumen, to be reabsorbed by the same cell or other cells; alternatively, phosphate could be restrained within the membrane, eventually finding its way back to the cytoplasm. The data support the later alternative. When slices of the salt gland were incubated with P32-orthophosphate, the phosphate was found to enter the cell very slowly (M. R. Hokin and L. E. Hokin, 1960a). In one hour only about one-tenth of the intracellular phosphate had exchanged with the extracellular phosphate. Stimulation of the slices with acetylcholine did not increase this rate of exchange. If phosphate were actually secreted into the acinar lumina of salt gland slices incubated in uitro, followed by its reabsorption, the phosphate would partially or completely mix with the extracellular phosphate in the medium. Stimulation of salt gland slices with acetylcholine would thus increase the rate of exchange of intracellular phosphate with extracellular phosphate. Since this was not observed it must be concluded that, on splitting of sodium phosphatidate, phosphate does not leave the apical membrane. It is therefore postulated that on cleavage of sodium phosphatidate in the region of the outer surface of the lipoidal membrane sodium rapidly diffuses through the pores of a protein sieve, or some such barrier, at the surface of the lipoidal membrane. But phosphate remains in the lipoidal phase of the membrane. As it accumulates it diffuses toward the inner surface of the membrane. It is further postulated that the inner surface of the membrane is permeable to phosphate, permitting it to diffuse into the cytoplasm, where it is eventually recaptured by oxidative phosphorylation to form ATP. The net effect of this mechanism would be that the active transport of sodium ions through the apical membrane of the salt gland would establish an electrical potential gradient across this membrane which would be positive from the lumen to the cytoplasm. The membrane is envisaged as being selectively permeable to chloride ions. The electrical potential established by the sodium transport mechanism could thus transport chloride ions across the

126

LOWELL E. HOKIN AND MABEL R. HOKIN

apical membrane of the salt gland without any additional expenditure of energy. Thus, sodium would be actively transported against an electrochemical gradient, while chloride would be transported against a chemical gradient by the electrical potential established by the sodium pump. This situation would thus be similar to that in the frog skin ( Ussing and Zerahn, 1951; Ussing, 1952), the axon membrane (Hodgkin, 1951; Keynes, 1951), and the muscle membrane (Boyle and Conway, 1941). An alternative explanation is that there is a separate chloride carrier mechanism, although if the situation is as described above, none need be postulated. B. ENERGY RELATIONSHIPS AND

THE

PHOSPHATIDIC ACID CYCLE

The phosphatidic acid cycle can explain in terms of chemical reactions how the so-called high energy phosphate of ATP could be utilized for the active transport of sodium ions against an electrochemical gradient. By injecting ATP into giant axons poisoned with cyanide, Caldwell and Keynes (1957) have obtained direct evidence that ATP can be utilized as an energy source for the sodium pump in nerve. Whittam (1958) has obtained evidence of a more indirect nature that ATP is utilized in the active transport of potassium in human erythrocytes. A requirement which the phosphatidic acid cycle must meet is that it must be able to account for the observed ratios of the equivalents of sodium transported to the moles of oxygen consumed. According to the scheme outlined in Fig. 2 one high energy phosphate from ATP would lead to the transport of two sodium ions, since, based on the pK's of the primary and secondary phosphoryl dissociations of a-glycerophosphate, one phosphatidate would carry essentially two sodium ions at pH 7.4. Assuming the generally accepted over-all P/O ratio of 3 for oxidative phosphorylation, twelve sodium ions would be transported per oxygen molecule. Zerahn (1956) found that the average ratio of sodium ions transported to the over-all oxygen consumption (Na/Oz) in the frog skin, bathed with frog Ringer's saline was 5.7; when sodium transport was stimulated by adding sodium to preparations bathed with distilled water on the outer surface of the skin, the average ratio of the increment of sodium transported to the increment of oxygen consumed (ANa/A02) was 18.5. Leaf and

ACIDIC PHOSPHATIDES AS CARRIERS IN MEMBRANES

127

Renshaw (1957) measured sodium transport in the frog skin before and after stimulation with neurohypophyseal hormones and obtained an over-all Na/Oz ratio of 6.8 in the presence of neurohypophyseal hormone and a ANa/A02 ratio of 18. Leaf et al. (1959) found an average ANa/A02 ratio of 16.5 in the toad bladder when sodium transport was stimulated in a manner similar to that used by Zerahn. As has been pointed out by Zerahn (l956), the ratio of ANa/A02 is probably too high, since oxygen uptake in excess of the increment observed on stimulation of sodium transport is probably used for the extra sodium transport. In other words, the resting respiration is “dipped into” for use in sodium transport on stimulation. The percentage of the resting respiration which would have to be used in addition to the A 0 2 uptake in order to give a Na/OZ ratio of 12 can be calculated from the data of the above workers. In the case of the experiments of Zerahn (1956) this figure was 24%. The data of Leaf and Renshaw for frog skin gave 6.7%, and those of Leaf et al. (1959) for toad bladder gave 31%. In the toad bladder experiments the oxygen uptake and the sodium transport were measured in a separate series of experiments. Since from 6 to 31% of the resting oxygen uptake, utilized together with the increased oxygen uptake, could give an Na/02 ratio of 12, it is clear that this ratio is not incompatible with the available data. There is no doubt that on stimulation of some secretory processes the resting respiration can completely satisfy the increased secretory activity. Stimulation of amylase secretion in pigeon pancreas slices with acetylcholine (L. E. Hokin and M. R. Hokin, 1956a) or of amylase secretion in mouse parotid tissue with adrenaline (Lesser, 1927) is not accompanied by any increase in respiration. Yet, amylase secretion in both of these tissues is dependent on respiration, being abolished by anaerobiosis ( Hokin, 1951; Hagen, 1959). Whether respiration is increased on stimulation of a secretory process seems to vary with the species. This problem has been recently discussed by Hagen (1959). S’ince in some secretory processes the resting respiration can provide sufficient energy for all of the increased secretory activity observed on stimulation, it would be reasonable to expect that in other tissues where a respiratory stimulation is observed that some of the resting respiration is channeled into use for the increased secretory activity. It remains in doubt, therefore, whether the true ratios of sodium transported to oxygen consumed in the experi-

128

LOWELL E. HOKIN AND MABEL R. HOKIN

ments of Zerahn and of Leaf and his associates are higher than 12. It should be emphasized that in all of these experiments the ratio of total sodium transport to total oxygen consumption, even under stimulation with neurohypophyseal preparations, has never exceeded an average of 6.8. This ratio is of course a minimal figure, since some of the total oxygen consumption is not concerned with sodium transport. C. SIGNIFICANCE OF THE STIMULATIONOF PHOSPHOINOSITIDE TURNOVER

There are good reasons to believe that phosphoinositide functions in the secretion of organic molecules, such as proteins (see L. E. Hokin and M. R. Hokin, 1959a). For instance, the stimulation of phosphoinositide turnover accounts for most of the increase in incorporation of P32 into the total phosphatide fraction in pancreas slices on stimulation of secretion of digestive enzymes. With the exception that the enzymes concerned in the stimulation of phosphoinositide turnover have not been demonstrated, the case for phosphoinositide as a sodium carrier in the salt gland is as good as that for phosphatidic acid: (1) Phosphoinositide turnover is stimulated at very low concentrations of acetylcholine (1W8 M), and at these low concentrations of acetylcholine the increment in its turnover is actually greater than that of phosphatidic acid (L. E. Hokin and M. R. Hokin, 1960). ( 2 ) The stimulation of the incorporation of P32 into phosphoinositide is accompanied by an equal stimulation of the incorporation of inositol-2-H3, ruling out the possibility that the increased incorporation of P32 into phosphoinositide is due to phosphatidic acid being its precursor. ( 3 ) The phosphoinositide effect occurs in microsomal membranes. The explanation for the phosphoinositide effect which best fits the available data is that phosphoinositide is also a sodium carrier but that its carrying capacity is more limited than that of phosphatidic acid and it functions primarily at low secretory rates. That phosphoinositide functions at low secretory rates is suggested by the observation that its turnover is stimulated more than that of phosphatidic acid with low concentrations of acetylcholine M ) but that as the concentration of acetyl(approximately choline is increased the stimulation of phosphoinositide turnover

ACIDIC PHOSPHATIDES AS CARRIERS IN MEMBRANES

129

reaches a maximum much sooner than that of phosphatidic acid turnover (L. E. Hokin and M. R. Hokin, 1960). Since the phosphate in phosphoinositide contains one negative charge, phosphoinositide could carry only one sodium, so that assuming that one ATP would be utilized per turn of a phosphoinositide cycle, one would obtain a maximum Na/Oz of 6. However, at very low secretory rates a lower efficiency for sodium transport would be less critical. The calling into play of the phosphatidic acid cycle as the predominant sodium carrier mechanism when the secretory stimulus is increased would be a useful mechanism for increasing the efficiency of sodium transport at higher secretory rates. Phosphoinositide may provide certain advantages as a carrier at low secretory rates and may be a more general mechanism for sodium transport in cells where sodium transport serves the sole purpose of maintaining the ionic gradients. The phosphatidic acid mechanism would probably also operate to some extent at these low rates of transport, as judged by the fact that phosphatidic acid turnover is also stimulated to some extent at very low concentrations of acetylcholine ( 10-8 M ) . One possible mechanism for a phosphoinositide cycle is that an activated form of inositol phosphate, X phosphoinositol, reacts with diglyceride at the inner surface of the membrane to form phosphoinositide. This could form the sodium salt, which is lipid soluble, and diffuse to the outer surface of the membrane, where it might be cleaved by a specific phosphatase to form diglyceride and the sodium salt of inositol phosphate. The sodium would be released into the aqueous phase of the lumen, the diglyceride diffusing back to the inner surface of the membrane. By analogy with the orthophosphate released in the phosphatidic acid cycle, inositol phosphate would diffuse back into the cytoplasm. The net result of such a cycle would be an increased turnover of inositol phosphate in phosphoinositide. The experiments with phosphate and inositol-2-Hs indicate that such is the case. However, to obtain more direct evidence, enzymes must be found which catalyze such a cycle, as has been accomplished for the phosphatidic acid cycle, That such enzymes do exist is shown by the experiments of Dawson (1959) who found an enzyme in pancreas which cleaves monophosphoinositides to inositol phosphate and diglyceride, and by the experiments of Redman and Hokin (1960) who found an enzyme in microsomal membranes of the pancreas H

130

LOWELL E. HOKIN AND MABEL R. HOKIN

which forms phosphoinositide from diglyceride and some as yet unidentified inositol-containing precursor.

D. ARE THERE SPECIFICCARRIERS FOR INDIVIDUAL SUBSTANCES? There has been much discussion about the necessity for the sodium carrier to be specific for sodium, and in studies on the specificity of certain phosphatides for potassium and sodium ions the results have been contradictory (see Solomon d al., 1956; and Kirschner, 1958). No studies have been carried out with phosphatidic acids. Although it would be very desirable to study the relative affinities of sodium and potassium ions for phosphatidic acid it is difficult to see how a simple substance such as phosphatidic acid could be highly specific for a particular cation. We think it more likely that the specificity would reside in a part of a protein structure, as is the case for enzyme catalyzed reactions. In the present case, either the diglyceride kinase or certain proteins of the membrane would provide the specificity. Coenzymes function as carriers of electrons, protons, organic moieties, etc., but the enzymes with which they are loosely combined determine the particular substrates with which the coenzymes react. The data in various digestive glands, endocrine glands, the salt gland, and synaptic tissue suggest that there is an economy of carriers for the transport of a wide variety of molecules. It appears at present that only phosphatidic acid and phosphoinositide, and in the pancreas possibly also phosphatidyl ethanolamine, are concerned in the active transport of inorganic cations and organic molecules containing cationic groups. If it is assumed that specificity is imparted by enzymes or nonenzymatic proteins of the membrane it is not necessary to postulate a large number of specific carriers. In the transport of molecules of the dimensions of proteins many phosphoinositide or phosphatidic acid molecules would probably be required to transport a single protein molecule. The studies discussed in this review throw no light on the carrier mechanisms for the active transport of anions or of neutral molecules. One can hopefully predict that the general mechanism for the transport of anions will be similar in principle to that which appears to be concerned in the transport of cations or of substances with cationic groups. However, one would expect the anion carriers to be positively charged lipid-soluble substances.

ACIDIC PHOSPHATIDES AS CARRIERS IN MEMBRANES

131

E. THE PHOSPHOLIPID EFFECTAND PINOCYTOSIS In recent years the processes of pinocytosis and reverse pinocytosis have been claimed as mechanisms for the transport of a wide variety of molecules (see Bennett, 1956). Since the mechanism envisaged by the pinocytosis school involves the formation of new membrane, followed by its breakdown, an increased turnover of phospholipids could be interpreted as evidence for their hypothesis. In the pancreas (Redman and Hokin, 1959) and in the salt gland (L. E. Hokin and M. R. Hokin, 1960) the various glycerophosphatides appear to be concentrated in the microsomal membranes, so that if there were a complete breakdown and resynthesis of whole membranes it would appear likely that all of the phosphatides would show increased turnover. However, it is conceivable that the membranes could be broken into fragments and the breaks could occur only at points involving phosphoinositide and phosphatidic acid. A more serious objection to the pinocytosis hypothesis is its failure to provide sufficient selectivity for the particular ions or molecules to be transported and its failure to explain how such a mechanism can transport substances against electrochemical and concentration gradients.

F. THEPHOSPHATIDIC ACID CYCLEAND SYNAPTIC TRANSMISSION In view of the role of phosphatidic acid turnover in sodium secretion in the salt gland the simplest explanation for the stimulation of this turnover in synaptic tissue is that it is concerned with the active extrusion of sodium from the postsynaptic cell body on continual depolarization by acetylcholine. This process would require energy and would be the closest to the type of transport occurring in the salt gland. In the invertebrate giant axon large accumulations of sodium and losses of potassium occur on repeated stimulation and the ionic gradients are restored relatively slowly by the pumps (Hodgkin, 1951; Keynes, 1951; Hodgkin and Keynes, 1954). In the giant axon the pumps appear to be responding to changes in the ionic composition inside the cell and not directly to the spikes. Assuming that the phospholipid effect in synaptic tissue represents a sodium pump mechanism, it would appear that the pump in this case is activated directly by acetylcholine. This is indicated by the fact that acetylcholine will stimulate phosphatidic acid turnover in cell-free preparations

132

LOWELL E. HOKIN AND MABEL R. HOKIN

of membranous fragments of brain cortex tissue; under these conditions there could be no change in the ionic environment of the membrane. Since ATP is required for the phosphatidic acid cycle (and presumably also for a phosphoinositide cycle) and since the energy of metabolism is not required for the movements of ions along their electrochemical gradients during the spike (Hodgkin and Keynes, 1955), the increased turnover of phosphatidic acid and phosphoinositide is unlikely to be related to the passive transfer of sodium and potassium across the postsynaptic membrane during either depolarization or repolarization, respectively. Phosphatidic acid and phosphoinositide could, however, act as carriers in the membrane during these passive processes, but no breakdown and resynthesis of the carrier, and therefore no increased turnover of phosphate, need be involved, since the process does not require energy. The direction and movement of the revelant cations would in this case be determined by their relative concentrations on either side of the membrane.

G. ON THE MECHANISM OF ACTIONOF ACETYLCHOLINE The data discussed in Section V, B indicate that acetylcholine does not affect the activities of the enzymes which catalyze the turnover of phosphatidic acid. Rather, the acetylcholine effect appears to be dependent on the structure of the responsive membranes. Very little evidence is available on how the acetylcholine effect is related to the structure of the membrane. Acetylcholine could alter the structure of the membrane so as to render diglyceride more accessible to diglyceride kinase, or sodium phosphatidate more accessible to phosphatidic acid phosphatase. It could possibly increase the rate of diffusion of phosphatidate in the membrane by releasing it from bound form; such a mechanism could explain how a single action of acetylcholine might directly stimulate both the influx of sodium along its concentration gradient, as occurs during depolarization, and the pumping of sodium against its concentration gradient. But in the absence of any information on this important area, it is felt that further speculation is unjustified at this time.

ACIDIC PHOSPHATIDES AS CARRIERS I N MEMBRANES

133

X. Summary Stimulation of the secretion of sodium chloride, catechol amines, polypeptides, and proteins in a variety of endocrine and exocrine glands is associated specifically with the stimulation of the turnover of phosphatidic acid and phosphoinositide. A similar effect occurs in cholinergic synaptic tissue in response to acetylcholine. The stimulation in synaptic tissue is believed to be associated with the pumping out of sodium which enters the postsynaptic cell body on depolarization by acetylcholine. The various studies suggest that phosphatidic acid functions as a sodium carrier in what has been termed the phosphatidic acid cycle. Essentially, this cycle consists of the formation of phosphatidate at the inner surface of the membrane from adenosine triphosphate and diglyceride, formation of sodium phosphatidate by ionic linkage, diffusion of sodium phosphatidate across the lipoidal membrane, cleavage of sodium phosphatidate by phosphatidic acid phosphatase at the outer surface of the membrane with the release of sodium ions into the extracellular aqueous phase, and diffusion of diglyceride back to the inner surface of the membrane where the cycle is repeated. The phosphate resulting from cleavage of sodium phophatidate remains in the cell. It is postulated that it diffuses back to the cytoplasm in exchange for two chloride ions. The data indicate that phosphoinositide may also be a sodium carrier, but the enzymatic studies have not advanced sufficiently to indicate the precise mechanism of a phosphoinositide cycle. However, this cycle would presumably be analogous to the phosphatidic acid cycle. The mechanism of the phospholipid effect in the secretion of water-soluble organic substances which contain cationic groupings is assumed to be similar to that concerned with the secretion of sodium ions. The phosphatidic acid cycle can explain in chemical terms how ATP can be used as the energy source for the performance of osmotic work. REFERENCES Artom, C., Sarzana, G., and Segre, E. (1938). Arch. intern. physiol. 47, 245. Bakay, L., and Lindberg, 0. ( 1949). Actu Physiol. Scand. 17, 179. Bennett, H.S. (1956). J. Biophys. Biochem. Cytol. 2, 99. Beveridge, J. M.R. (1956).Can. J. Biochem. and Physiol. 34,361.

134

LOWELL E. HOKLN AND MABEL R. HOKIN

Bloor, W. R. (1943). “Biochemistry of the Fatty Acids.” Reinhold, New York. Boyle, P. I., and Conway, E. J. (1941). J. Physiol. 100, 1. Caldwell, P. C., and Keynes, R. D. (1957). J. Physiol. (London) 197, 12 P. Chaikoff, I. L. (1942). Physio2. Revs. 22, 291. Chaikoff, I. L., and Zilversmit, D. B. (1948). Advances in Biol. and Med Phys. 1, 321. Channon, H. J., and Chibnall, A. C. (1927). Biochem. J. 21, 1112. Davies, R. E., Harper, A. A., and MacKay, I. F. S. (1949). Am. J. Physiol. 167, 278. Dawson, R. M. C. (1952). Biochem. SOC. Symposia (Cambridge, Engl.) 8, 93. Dawson, R. M. C. (1953). Biochem. J . 66, 507. Dawson, R. M. C. (1954). Biochim. et Biophys. Actu 14, 374. Dawson, R. M. C. (1957). Biol. Revs. Cambridge Phil. SOC. 92, 188. Dawson, R. M. C. (1959). Biochim. et Biophys. Acta 33, 68. Deuel, H. J. ( 1955). “The Lipids,” Vol. 11. Interscience, New York. Deutsch, W., and Raper, H. S. (1936). J. Physio2. (London) 87, 275. Deutsch, W., and Raper, H. S. (1938). J. Physiol. (London) 92, 439. Doyle, W. L. (1959). Personal communication. Eccles, J. C. (1953). “The Neurophysiological Basis of Mind.” Oxford Univ. Press, London and New York. Eccles, J. C. (1957). “The Physiology of Nerve Cells.” Johns Hopkins Press, Baltimore, Maryland. Eggman, L. D., and Hokin, L. E. (1960). J. Biol. Chem. In Press. Faure, M., and Morelec-Coulon, J. (1953). Compt. rend. mad. sci. 256, 1104. Faure, M., and Morelec-Coulon, J. (1954). Compt. rend. mad. sci. 238, 411. Findlay, M., and Strickland, K. P. (1953). Chem. in Can. 6, 104. Folch, J. (1949). J. Biol. Chem. 177, 505. Freinkel, N. ( 1957). Endocrinology 61, 448. Friedkin, M., and Lehninger, A. L. ( 1949). J. BW2. Chem. 177,775. Fries, B. A., Schachner, H., and Chaikoff, I. L. (1942). J . BW2. Chem. 146, 281. Gerard, R. W., and Still, E. U. (1933). Am. J. Physiol. lOS, 232. Greenberg, D. M., Aird, R. B., Boelter, M. D. D., Campbell, W. W., Cohn, W. E., and Murayama, M. M. (1943). Am. 1. Physiol. 140,47. Hagen, J. M. (1959). Biochem. Phamzacol. 2,206. Hahn, L., and Hevesy, G. (1937). Skand. Arch. Physiol. 77, 148. Hanahan, D. J., and Olley, J. (1958). J . Biol. Chem. 231, 813. Hanzon, V., and Toschi, G. ( 1959). Exptl. Cell Research 16, 256. Harper, A. A., and Raper, H. S . ( 1943). J. Physiol. (London) 102, 115. Harper, A. A., and Vass, C. C. N. (1941). J. Physiol. (London) 99, 415. Hebb, C . O., and Silver, A. (1956). J . Physwl. (London) 134,718. Hevesy, G. (1947). Adsances in Enzymol. 7, 111. Hodgkin, A. L. (1951). Biol. Revs. Cambridge Phil. SOC. 26, 339. Hodgkin, A. L., and Keynes, R. D. (1954). Symposia SOC. Exptl. Biol. 8, 423.

ACIDIC PHOSPHATIDES AS CARRIERS IN MEMBRANES

135

Hodgkin, A. L., and Keynes, R. D. (1955). J. Physiol. (London) 128, 28. Hogben, C . A. (1955). Am. J. Physiol. 60, 85. Hokin, L. E. (1951). Biochem. J . 48,320. Hokin, L. E., and Hokin, M. R. (1955a). Biochim. et Biophys. Acta 16, 229. Hokin, L. E., and Hokin, M. R. ( 1955b). Biochim. et Biophys. Acta 18, 102. Hokin, L. E., and Hokin, M. R. (1956a). J. Physiol. (London) 132, 442. Hokin, L. E., and Hokin, M. R. (1956b). Can. J. Biochem. and PhysioE. 34, 349. Hokin, L. E., and Hokin, M. R. (1958a). J. Biol. Chem. 253, 800. Hokin, L. E., and Hokin, M. R. (195813). J. Biol. Chem. 233, 805. Hokin, L. E., and Hokin, M. R. (195th). J. Biol. Chem. 233, 818. Hokin, L. E., and Hokin, M. R. ( 1958d). J. Biol. Chem. 233, 822. Hokin, L. E., and Hokin, M. R. (1959a). Gastroenterology 36, 368. Hokin, L. E., and Hokin, M. R. (1959b). J. Biol. Chem. 234, 1387. Hokin, L. E., and Hokin, M. R. ( 1 9 5 9 ~ ) .Nature 184,1068. Hokin, L. E., and Hokin, M. R. (1960). J. Gen. Physiol. In Press. Hokin, L. E., and Sherwin, A. L. (1957). J. Physiol. (London) 136, 18. Hokin, M. R., and Hokin, L. E. (1953). J. Biol. Chem. 203, 967. Hokin, M. R., and Hokin, L. E. (1959). J . Biol. Chem. 234, 1381. Hokin, M. R., and Hokin, L. E. ( 1960a). In preparation. Hokin, M. R., and Hokin, L. E. ( 1960b). In preparation. Hokin, M. R., Benfey, B. G., and Hokin, L. E. (1958a). J. B i d . Chem. 233, 814. Hokin, M. R., H o b , L. E., Saffran, M., Schally, A. V., and Zimmermann, B. U. (1958b). J. Biol. Chem. 253,811. Hokin, hl. R., Hokin, L. E., and Shelp, W. D. (1960). J. Gen. Physiol. In Press. Jorpes, E., and Mutt, V. (1954). Arkiu Kemi 7 , 553. Kates, M. (1955). Can. J. Biochem. and Physiol. 33, 575. Kennedy, E. P. (1957). Ann. Reu. Biochem. 26,119. Keynes, R. D. (1951). “The Role of Electrolytes in Excitable Tissues.” Publ. Inst. Biol., Univ. Brazil, Rio de Janeiro. Kirschner, L. B. (1958). J. Gen. Physiol. 42,231. Kiyohara, K. ( 1934). Compt. rend. SOC. biol. 116, 1166. Komberg, A., and Pricer, W. E., Jr. (1953a). J. Biol. Chem. 204, 239. Komberg, A,, and Pricer, W. E., Jr. (1953b). J. Biol. Chem. 204, 345. Krebs, H. A., Ruffo, A., Johnson, M., Eggleston, L. V., and Hems, R. (1953). Biochem. 1. 64, 107. Leaf, A,, and Renshaw, A. (1957). Biochem. J . 6 6 , 8 2 . Leaf, A., Page, L. B., and Anderson, J. (1959). J. Biol. Chem. 234, 1625. Lesser, E. J. (1927). Biochem. 2.84, 125. McKibbin, J. M. (1956). J. Biol. Chem. 220, 537. McMurray, W. C., Berry, J. F., and Rossiter, R. J. (1957a). Biochem. J. 66, 629. McMurray, W. C., Strickland, K. P., Berry, J. F., and Rossiter, R. J. (1957b). Bwchem. J. 66, 621. McMurray, W. C., Strickland, K. P., Berry, J. F., and Rossiter, R. J. ( 1 9 5 7 ~ ) . Biochem. J. 66, 634.

136

LOWELL E. HOKIN AND MABEL R. HOKIN

Mann, P. J. G., Tennenbaum, M., and Quastel, J. H. (1938). Biochem. J. 92, 243. Marinetti, G. V., and Stotz, E. ( 1956). Biochim. et Biophys. Acta 21, 168. Marinetti, G.V., Witter, F. F., and Stotz, E. (1957). J. BioZ. Chem. 226, 475. Mellanby, J. ( 1925). J . Physiol. (London) 60,85. Morton, M. E., and Schwartz, J. R. (1953). Science 117, 103. Palade, G. E., and Siekevitz, P. (1956). J. Biophys. Biochem. CytoZ. 2, 171. Pfeiffer, C. C., and Jenney, E. H. (1957). Ann. N.Y. Acod. Sci. 66, 753. Redman, C. M., and Hokin, L. E. (1959). J. Biophys. Biochem. Cytol. 6, 207. Redrnan, C. M., and Hokin, L. E. (1960). In preparation. Rehm, W. S. (1950). Gastroenterology 14,401. Rosenberg, T. (1948). Acta Chem. Scand. 2, 14. Rothstein, A., and Meier, R. ( 1948). J. Cellular Comp. Physiol. 32,77. Schachner, H., Fries, B. A., and Chaikoff, I. L. (1942). J. Biol. Chem. 146, 95. Schmidt-Nielsen, K. (1959). Sci. Am. 200, 109. Schmidt-Nielsen, K., and Sladen, W. J. L. ( 1958). Nature 181, 1217. Schmidt-Nielsen, K.,Jorgensen, C. B., and Osaki, H. (1958). Am. J. Physiol. 199, 101. Siekevitz, P., and Palade, G. E. (1958). J. Biophys. Biochem. Cytol. 4, 203. Sloane-Stanley, G. H. ( 1952). Biochem. SOC. Symposia (Cambridge, Engl.) 8, 44. Smith, S . W., Weiss, S . B., and Kennedy, E. P. (1957). J. Bid. Chem. 228, 915. Solomon, A., Lionetti, F., and Curran, P. (1956). Nature 178, 582. Taurog, A,, Chaikoff, I. L., and Perlman, I. (1942). J. Biol. Chem. 146, 281. Ussing, H. H. (1952). Advances in Enzymol. 19,21. Ussing, H . H., and Zerahn, K. ( 1951). Acta Physiol. Scand. 2S, 110. Whittam, R. ( 1958). J . Physiol. (London)140,479. Witcoff, H . (1951). “The Phosphatides,” Reinhold, New York. Zerahn, K. ( 1956). Acta Physiol. Scand. 56,300. Zilversmit, D. B. (1955). Ann. Reu. Biochem. M, 157.

BRAIN NEUROHORMONES AND CORTICAL EPINEPHRINE PRESSOR RESPONSES AS AFFECTED BY SCHIZOPHRENIC SERUM' By

Edward

J . Walaszek

Department of Pharmacology. University of Kansas Medical Center. Kansas City. Kansas

I. Introduction ........................................... I1. The Cortical Epinephrine Pressor Response (CEPR) ......... A . Procedure ......................................... B . Mechanism of the Hypertensive Response .............. I11. Effect of Various Centrally Acting Drugs on the CEPR . . . . . . . A . Tranquilizers ...................................... B . Amine Oxidase Inhibitors ............................ C . Parasympathomimetic Drugs .......................... D . Hallucinogens ...................................... E . Other Pharmacological Agents ........................ IV. Effect of Serum on the CEPR ............................ A . Pretreatment of Animals with Serum from Normal Volunteers ............................................ B . Pretreatment of Animals with Serum from Schizophrenic Patients ............................................ C Pretreatment of Animals with Serum from Pregnant Women and from Patients with Organic Diseases .............. D . Discussion and Summary ............................. V. Influence of Schizophrenic Serum upon the Content of Various Neurohumoral Substances in the Rabbit Hypothalamus ...... A Norepinephrine and Epinephrine ...................... B . Serotonin .......................................... C Substance P ........................................ D. Histamine ......................................... VI . Correlation of the Inhibitory Effects of Schizophrenic Serum on the CEPR with the Catechol Amine Content of the Hypothalamus .............................................. VII . Presence of an Abnormal Factor or Factors in Serum of Schizophrenics .............................................. VIII . Summary ............................................. References .............................................

.

. .

138 139 139 140 144 144 145 146 149 150 151 151 151 153 154 156 158 164 165 165 166 169 170

171

1 This study was supported in part by Grant M-1902 from the National Institutes of Mental Health. U . S. Public Health Service. and Senior Research Fellowship SF-72 from the U . S. Public Health Service.

137

138

EDWARD J, WALASZEK

1.

Introduction

The topical application of a solution of epinephrine on the cerebral cortex of animals elicits a rise in systemic blood pressure. This review will concern itself with this pressor response and the influence of various drugs and serum of schizophrenic patients upon this cortical epinephrine pressor response ( CEPR) . The influence of the cerebral cortex upon the cardiovascular system has been suggested by clinical observations of blood pressure, heart rate, and localized vasomotor disturbances in patients with verified cerebral cortical lesions ( Minkowski, 1933; Popper, 1933; Bucy, 1935). Experimental lesions in animals also yield evidence of vasomotor representation at the cortical level (Pinkston and Rioch, 1938, Kennard, 1935). A great number of investigators have studied the effect of electrical stimulation of various parts of the brain upon blood pressure, and it has been shown that electrical stimulation of the cerebral cortex of dogs, cats, and rabbits produced changes in blood pressure (Dusser De Barenne and Kleinknecht, 1924). These investigators reported blood pressure risen after stimulation of the gyrus sigmoidens and blood pressure falls from stimulation of the anterior portion of the gyrus suprasylvius medius. Langford et al. (1957) reported that the stimulation of the motor cortex of dogs resulted in a systemic hypertension and vasoconstriction. Kaada ( 1951 ) has reviewed this subject and in addition extensively studied this phenomenon. His evidence for the cortical origin of the typical rise in blood pressure was that these effects were diminished after an application of the local anesthetic, procaine, but they returned if the electrodes were plunged a few millimeters below the surface. He also reported depressor responses after stimulation of various areas but these were abolished or reduced following vagotomy. One of the conclusions reached by this study was that the vagi were the principal mediator of the depressor effect but they had no influence on the pressor response. A number of investigators have studied the effect of topically applied acetylcholine on electrical activity of the brain. The direct application of acetylcholine (%lo% solutions) to the cerebral cortex of cats (Brenner and Merritt, 1943) resulted in a profound increase in electrical activity. The electrical changes

139

NEUROHORMONES AND THE CEPR

resembled grand ma1 seizure (Merritt and Brenner, 1941), and were sensitized by neostigmine and abolished by Dilantin. Miller (1937) has shown that the local application of physostigmine and acetylcholine to the motor cortex increased electrical activity which was modified by atropine. Essig et al. (1953) reported that the topical application of acetylcholine to the cerebral cortex of cats under pentobarbital anesthesia produced temporary localized depression in the EEG. Beckett and Gellhorn (1948) concluded that acetylcholine or physostigmine stimulates cortical suppressor areas, abolishes electrical activity, and inhibits response of motor cortex to electrical stimulation. Kremer (1948) made use of 2.5% solutions of methacholine applied topically for the localization of vasodilator areas in the brain. Minz et al. (1953b) have reported that local application of epinephrine on the motor arm area of the monkey cerebral cortex engenders an increase in duration and intensity of the period of over-excitation subsequent to a stimulus eliciting movement. Although studies of cardiovascular changes induced by electrical stimulation of various parts of the central nervous system are quite numerous the study of cardiovascular changes induced by the topical application of drugs to the cerebral cortex are virtually unknown. Minz and his co-workers (1953a) were the first to report a hypertensive effect in rabbits due to the local application of epinephrine on the frontoparietal area of the cerebral cortex. This effect was not due to the diffusion of epinephrine into the general circulation but to a unique central action of this naturally occurring catechol amine. 11.

The Cortical Epinephrine Pressor Response

(CEPR)

A. PROCEDURE Adult rabbits were anesthetized with urethane ( 1.0-1.5 gm/kg ) . The cerebral hemispheres were exposed, the dura opened, and the mean systemic arterial blood pressure was measured via a cannula in the left femoral artery. The right saphenous vein was cannulated for intravenous injection. A very small piece of filter paper (35 mmz) moistened with a freshly prepared 5% solution of epinephrine bitartrate in Ringer solution (pH 6.0) was placed on the left frontoparietal zone of the cerebral cortex. After a short

140

EDWARD J. WALASZEK

contact (30 seconds) of the well-drained filter paper with the cortex, a typical CEPR occurs; the systemic blood pressure rises (20-50 mm Hg), maintains a maximum for 0.5-2 minutes, and then decreases to its initial value. The applications are repeated every 10 minutes. The filter paper contains approximately 750 pg of epinephrine bitartrate both before and after the application to the cerebral cortex. Solutions of epinephrine bitartrate lower than 2% normally do not elicit a CEPR. Repeated applications of epinephrine to the same cortical area produces a progressively increasing response. This characteristic sensitization has been studied by Chamorro and Minz (1956a). After this period of sensitization the pressor responses become

FIG. 1. Rabbit blood pressure. Four successive topical applications of epinephrine ( E T ) to the cerebral cortex for 30 seconds at 10-minute intervals, illustrating the characteristic sensitization of CEPR.

quite uniform. Figure 1 illustrates four successive applications of the epinephrine solution to the frontoparietal area of the cerebral cortex at 10-minute intervals.

B. MECHANISM OF

THE

HYPERTENSIVE RESPONSE

The application of Ringer solution alone does not produce any changes in blood pressure. The application of a 5% solution of acetylcholine produces a very small depressor response (5-8 mm Hg) on the first application and usually no response thereafter. Topical application of histamine, isoproterenol, and serotonin in the same concentration produces no effect on the systemic blood pressure. Norepinephrine applied topically produces a pressor response, but pressor effects so elicited are much smaller than those due to epinephrine. Chamorro (1957a) reported that 5% solution of naphazoline can elicit a pressor response. Minz and

NEUROHORMONES AND THE CEPR

141

Goldstein ( 1955) electrically stimulated different cortical areas of the rabbit to ascertain whether a pressor response can be obtained. Numerous variations concerning the frequency, voltage, and duration of impulse and placement of electrodes were tried. In no case was a pressor effect recorded. The only effect was a small and transient hypotension. The CEPR can also be elicited from cats and dogs. In the dog (Fig. 2) the response is best studied if the epinephrine remains in contact with the cerebral cortex for 15 seconds rather than 30 seconds. Initially it is useful to apply the epinephrine

FIG.2. Dog blood pressure. Two successive topical applications of epinephrine (ET) to the cerebral cortex for 15 seconds at 10-minute intervals.

for 60-second contact periods since this hastens the onset of the characteristic sensitization of the pressor response in the dog. Chamorro and Minz (195613) applied epinephrine to various areas of the cerebral cortex and have found that the left frontoparietal area was most sensitive for provoking the hypertensive response. A constant finding was that the left frontoparietal area was in almost all cases more sensitive to the application of epinephrine than the right frontoparietal area. It was also shown that the application of epinephrine to one region tended to sensitize other regions to the CEPR. The mechanism of this response was studied by Minz et al. (1953a) and by Chamorro and Minz (1955a). It was evident that the pressor effect, if it were due to a neuronal vasomotor action, would be prevented by an interruption of the cerebrospinal axis. In such experiments this was not the case as the hypertensive

142

EDWARD J. WALASZEK

effect due to the cortical application of epinephrine persisted in full strength after transection of the spinal cord at the Ca level. The pressor response also persisted in adrenalectomized animals. However, the pressor effect disappeared after disconnecting the cortical zone to which epinephrine was applied by undercutting. Consequently the pressor effect probably involved a primary neuronal and a secondary humoral process. The humoral stimulus produced by epinephrine on the cortex was probably transmitted by neuronal elements to a center capable of secreting a substance with hypertensive activity. It seemed logical to think of a possible intervention of the pituitary since the neural lobe secretes two pharmacologically active polypeptide hormones, vasopressin and oxytocin. However, the CEPR could still be obtained in hypophysectomized animals or after sectioning the hypophyseal stalk. ’ The possibility of hypothalamic involvement was considered by Chamorro and Minz (1955a). It was observed that the CEPR completely disappears after a transection at the level of the supraoptic nuclei and after section of the supraoptic hypophyseal tract below the optic chiasma. Thus the humoral stimulus produced by epinephrine on the cerebral cortex was transmitted by neuronal elements to a hypothalamic center capable of secreting a substance with hypertensive activity ( Minz and Chamorro, 1955). It was shown by Chamorro and Minz (195513) that the CEPR was accompanied by a simultaneous contraction of the uterus previously sensitized by an injection of estradiol. The uterine contractions began at the same time as the CEPR and were tetanic in type for a short time, after which there was an augmented tone. This reaction was not modified by hypophysectomy, and the conclusion of the authors was that oxytocin or an oxytocinlike material was also being released from the hypothalamus by the epinephrine applied topically to the cerebral cortex. Chamorro and Minz (1957a) investigated the localization in the hypothalamus of the pressor activity due to the topical appIication of epinephrine to the cerebral cortex. Destruction of the posterior hypothalamus left the CEPR intact and this signified that the vasopressor material was secreted by the anterior hypothalamus. In a further study Chamorro and Minz (1957b) reported that extracts of anterior hypothalamus showed pressor activity while those of the posterior hypothalamus were devoid of pressor effects. This in itself was not valid evidence for the

NEUROHORMONES AND THE CEPR

143

presence of the hypertensive substance secreted by the hypothalamus since many other pressor substances may be present in such extracts of hypothalamic tissue. The CEPR was differentiated from the peripheral effect of epinephrine by Chamorro and Minz ( 1 9 5 6 ~ )After . the intravenous injection of a massive dose of epinephrine (750 pg) the CEPR was greatly diminished or abolished (Fig. 3 ) while the intravenous epinephrine (10 pg) effect was enhanced.

FIG.3. Rabbit blood pressure. At ET, topical application of epinephrine to the cerebral cortex; at El, intravenous administration of 1Okg of epinephrine. At MEl, intravenous administration of a massive dose (750 pg) of epinephrine. The CEPR was blocked after the massive dose of systemically administered epinephrine while the lO+g intravenously-administered dose was enhanced. Timer interval; 10 seconds.

Detailed pharmacodynamic analysis and transection studies at various levels have shown that the CEPR possessed the characteristics of an entirely central phenomenon and was caused by an unknown hypertensive agent released from the anterior hypothalamus. It can then be summarized that the humoral stimulus produced by epinephrine on the cerebral cortex was transmitted by neuronal elements to a hypothalamic center capable of secreting a substance with hypertensive activity. The identity of the pressor substance is unknown. The workers cited above felt that the pressor substance might be vasopressin since this polypeptide is present in the hypothalamus in concentrations of 1 unit/gm of tissue. Furthermore the pressor response to vasopressin administered intravenously was potentiated by reserpine ( Chamorro, 1957b), as was the cortical epinephrine response. However tachyphylaxis develops rapidly to intravenously administered vasopressin but not to the CEPR.

144

EDWARD J. WALASZEK

111.

Effect of Various Centrally Acting Drugs on the CEPR

A. TRANQUILIZERS Chlorpromazine in doses of 5 mg/kg was able to diminish or abolish the CEPR (Minz, 1957). The mechanism of this diminution of the pressor response is unknown but it was suggested that perhaps this apparent block of the cortical application of epinephrine by chlorpromazine was a manifestation of the adrenergic blocking activity of this drug at a site of action in the central nervous system. Ergotamine, a known adrenergic blocking

FIG.4. Rabbit blood pressure. At EI, intravenous administration of 10 pg of epinephrine; at ET, topical application of epinephrine to cerebral cortex; at P, 2 mg/kg piperoxan. The pressor response to the intravenous administration of epinephrine was blocked while the CEPR was not affected.

agent, also diminished the CEPR (Chamorro, 1957a), but piperoxan, another adrenergic blocking agent, was ineffective (Fig. 4 ) . Systemic administration of reserpine in doses of 1.25 mg/kg had an entirely different effect. The CEPR was tremendously potentiated. This augmentation of the pressor response was of such a magnitude that essentially normal pressor responses were obtained with 0.5% and even 0.1% solutions of epinephrine bitartrate ( Minz, 1957). This potentiation manifested itself 4 hours after reserpine administration and persisted for approximately 72 hours. The CEPR was found again to be normal 92 hours after reserpine administration. The augmented CEPR following the administration of reserpine was considerably diminished immediately after the systemic administration of chlorpromazine (from 65 to 13 mm Hg). Rabbits systemically pretreated with Tetrabenazine (40 mg/kg) also showed an augmented pressor response (Fig. 5 ) to cortically applied epinephrine

NEUROHORMONES AND THE CEPR

145

(Minz and Walaszek, 1959). Pletscher et al. (1958) concluded that Tetrabenazine had a short lasting tranquilizing action similar to reserpine and was also capable of releasing catechol amines and serotonin from the central nervous system. Another tranquilizer, azacyclonal, had only a very slight potentiating effect on the CEPR while meprobamate, a tranquilizer and central muscle relaxant, was without any effect on the pressor response.

160

40

I

FIG. 5. Rabbit blood pressure. This illustrates the potentiation of the CEPR in an animal which received 40mg/kg of Tetrabenazine. Topical applications of epinephrine were at 10-minute intervals.

B. AMINE OXIDASE INHIBITORS Antagonism of the tranquilizing effect of reserpine by iproniazid, an amine oxidase inhibitor, has been reported by Brodie and Shore ( 1957). Observations of Spector and co-workers ( 1958) indicated that iproniazid administered systemically to rabbits was capable of increasing the norepinephrine and serotonin content of the brain stem. The mechanism of this increase was considered to be the result of the inhibition of the monoamine oxidase of the brain, which is the enzyme concerned with the normal metabolic destruction of the catechol amines and serotonin. Since reserpine lowers the content of the catechol amines and serotonin in the brain stem and augments the CEPR, it was of interest to study the effect on the CEPR of amine oxidase inhibitors which raise the content of catechol amines and serotonin in the brain stem. Our experiments (Minz and Walaszek, 1958) have shown that rabbits pretreated with iproniazid presented an abolition, diminution, or a reversal of the cardiovascular response to epinephrine applied on the frontoparietal area of the cerebral cortex. In addition to iproniazid we tested two other amine oxidase inhibitors, namely serine N-isopropylhydrazine and acetylmethionylN-isopropylhydrazine ( Minz and Walaszek, 1959).

146

EDWARD J. WALASZEK

Rabbits were injected with the amine oxidase inhibitors subcutaneously and tested 5 and 16 hours later with the cortically applied epinephrine. All animals pretreated with systemically administered iproniazid and the other amine oxidase inhibitors, showed a diminution of the CEPR to the first 6 or 8 applications. Animals tested 16 hours after the administration of the amine oxidase inhibitors consistently showed the greatest diminution of the cortically applied epinephrine pressor response. Approximately 16% of the animals showed a small depressor response to the topical application of epinephrine on the cerebral cortex. This depressor response sometimes coincided with the epinephrine application but often appeared 1 to 2 minutes after the filter paper was removed. In animals in which systemic premedication with iproniazid had effectively inhibited the CEPR, a subsequent intravenous administration of reserpine (I mg/kg) led to the reappearance and to a gradually progressive increase of the CEPR 2 hours later. Moreover, in rabbits pretreated with reserpine and reacting with the usual augmentation of the CEPR, a subsequent intravenous administration of iproniazid ( 100 mg/kg ) produced a considerable decrease of this reaction. The hypertensive effect, which under the influence of premedication with reserpine, had reached values of 110 to 120 mm Hg, dropped to values of 30 to 50 mm Hg 2 hours after the administration of iproniazid. This decrease was particularly characteristic for the sensitizing action of reserpine on cortical responses to epinephrine generally persisted unchanged during a period of 72 hours after the administration of that drug.

DRUGS C. PAFWSYMPATHOMIMETIC Our interest in this group of drugs stems from the work of PfeifYer and Jenney (1957) in which they reported the counteraction of schizophrenia and the inhibition of the conditioned response in the rat by so-called muscarinic stimulation of the brain. In this study they used arecoline, physostigmine, and pilocarpine. We were interested in whether muscarinic stimulation of the brain would affect the CEPR especially if this pressor response was a type of central sympathetic stimulation. The experimental design was as follows: Rabbits were injected subcutaneously with single doses of atropine or methyl atropine

NEUROHORMONES AND THE CEPR

147

( 2 mg/kg ), and the topical epinephrine response was established. Methyl atropine is a quaternary nitrogen analog of atropine which presumably does not cross the blood-brain barrier, while atropine, a tertiary amine, passes the barrier freely. Each of these protected the animals from the peripheral effects of the parasympathomimetic stimulants but probably only atropine would protect the central nervous system. The parasympathomimetic drugs were then injected intravenously and the cortical response measured at 10-minute intervals.

FIG.6. Rabbit blood pressure illustrating the diminished CEPR after the administration of arecoline and the inability of methacholine to affect the response. Animals protected from the peripheral effects of the parasympathomimetic drugs by systemic administration of methyl atropine ( 2 mg/kg). a. At E, topical application of epinephrine to the cerebral cortex; at A, 0.1 mg/kg arecoline administered intravenously. The topical application of epinephrine was at 2 and 12 minutes after the administration of arecoline. b. At E, topical application of epinephrine before and after the intravenous administration of 0.5 mg/kg methacholine (at M ) . Timer interval; 10 seconds.

Arecoline (0.1 mg/kg) decreased the magnitude of the CEPR (Smith et d.,1958). This modifying effect of arecoline appeared 1-2 minutes after its administration and disappeared in 5-10 minutes (Fig. 6). Physostigmine (0.25 mg/kg) also modified the pressor response, but it differed from arecoline in that it required 20 minutes to produce a maximum diminution while arecoline diminished the CEPR immediately. Pilocarpine (0.5 mg/kg ) had minimal effect while neostigmine and methacholine were without any effect on the CEPR. Atropine was able to block this diminution

148

EDWARD J. WALASZEK

in the pressor response only when smaller doses of arecoline were used. With larger doses neither atropine nor methyl atropine were able to block the diminution in the pressor response seen after the administration of arecoline or physostigmine. This diminished pressor response was usually seen only with the first large dose. When small doses were studied a slight enhancement of the pressor response was noted. If, after a number of small doses of arecoline or nicotine, a large dose was administered a diminished pressor response was difficult to obtain. A conclusion was reached that under these conditions arecoline and nicotine were capable of blocking their own effect on the CEPR. The administration of nicotine bitartrate (0.1 mg/kg) first diminished and then abolished the CEPR in animals that were systemically pretreated with atropine or methyl atropine. Furthermore we were able to show that mecamylamine (1 mg/kg), a ganglionic blocking agent, was capable of preventing the usual diminishing effect of arecoline and nicotine on the epinephrine pressor response. Stone and associates (1958) reported that mecamylamine antagonized the convulsive effect of nicotine and postulated an extraganglionic site of action. From these data it was concluded that the ability of arecoline to diminish the pressor response was probably not due to a muscarinic stimulation of the brain by arecoline. Instead, because nicotine diminished the pressor response as did arecoline, and because mecamylamine blocked the diminishing effect, we believe that this diminution by nicotine and arecoline may be caused by a nicotinic stimulation, The terms nicotinic and muscarinic are very poorly defined for the central nervous system and we do not want to transfer to the brain the classic concept of these terms. Instead, since the effect of arecoline was similar to that of nicotine, we would prefer to regard the action of arecoline in this situation as a nicotinelike action. The fact that arecoline possesses considerable nicotinelike action on the peripheral autonomic nervous system has been reported by Euler and Domeij (1945). The presence of a muscarinic component in the action of arecoline and physostigmine in diminishing the CEPR was still possible for the following three reasons: (1) atropine is rapidly metabolized by rabbits and its concentration in the central nervous system may not be sufficient in these experiments, ( 2 ) mecamylamine itself slightly potentiated the CEPR

NEUROHORMONES AND THE CEPR

149

and this may have masked the diminution of the pressor response by arecoline, and ( 3 ) when arecoline and physostigmine were administered to animals, not protected by atropine or methyl atropine, in doses which gave primarily muscarinic effects a diminution in the CEPR was noted. In summary, it has been shown that systemically administered arecoline, physostigmine, and nicotine produce a diminution of: the pressor response normally elicited by topical application of epinephrine to the cerebral cortex. Systemic premedication with mecamylamine completely abolished the diminution in the CEPR that was obtained with systemic administration of arecoline and nicotine. This diminution of the pressor response by systemically administered arecoline may be due to the nicotinelike action of arecoline.

D. HALLUCINOGENS The effect of a number of hallucinogens have been studied on the CEPR (Walaszek and Underwood, 1960). It was found that systemic administration of lysergic acid diethylamide ( LSD-25), bufotenine, and d-adrenochrome produced a potentiation of the CEPR. In some cases the magnitude of this potentiation was 10-20-fold. LSD was studied on both the rabbit and dog cortical preparations. The potentiation in the case of LSD came about quite abruptly at 60-80 minutes after administration of the hallucinogen to the rabbit. Figure 7 illustrates a typical experiment. The mechanism of this potentiation is unknown. LSD has been reported to be an adrenergic blocking agent of both the excitatory and inhibitory actions of epinephrine by Graham and Khalidi (1954). Our investigations seem to illustrate a facilitation of the action of epinephrine applied topically to the cerebra1 cortex. Furthermore, epinephrine, injected intravenously in the rabbits after administration of LSD, was never blocked or diminished. There was no adrenergic blocking activity of LSD in the rabbit at the dose levels (0.05 mg/kg) employed to produce a potentiation of the CEPR. Cerletti (1955) has reported that LSD enhances the effect of epinephrine on the perfused rat kidney, in concentrations which can block the effect of serotonin. There was no potentiation of the pressor response when rabbits were systemically pretreated with mescaline or serotonin. Perhaps

150

EDWARD J. WALASZEK

mescaline was ineffective because the dose was not large enough ( 5 mg/kg) or because the rabbits possess an enzyme (mescaline oxidase ) which can effectively and rapidly detoxify this compound.

FIG.7. The facilitation of the CEPR by LSD. At ET, topical application of epinephrine to the rabbit cerebral cortex at 10-minute intervals. Time in minutes after the intravenous administration of LSD ( 0.05 mg/kg) .

E. OTHERPHARMACOLOGICAL AGENTS Chamorro (1957b) has reported a study on the effect of various systemically administered drugs on the hypertensive response obtained by stimulating the cerebral cortex of rabbits with epinephrine (Table I ) , Pentobarbital and hexobarbital were without TABLE I EFFECTOF VARIOUS SYSTEMICALLY ADMINISTERED DRUGS ON THE CORTICAL EPINEPHRINE PRESSOR RESPONSE (CEPR) Dmg Ethyl ether Hexobarbital Pentobarbital Procaine Morphine Ergotamine Chlorpromazine Thyroxine Histamine Pipradrol Piperoxan

Dose (mg/kg)

3040 15 1040 4-5 0.3 0.5-5 0.3 2 2 2

CEPR Inhibited No change No change No change No change Abolished Abolished Slight potentiation Slight potentiation No change No change

NEUROHORMONES AND THE CEPR

151

any effect on the pressor response as was procaine and morphine. He was able to confirm the previous results of Minz (1957) that reserpine potentiates and chlorpromazine inhibits the CEPR. The administration of ethyl ether as an anesthetic agent partially inhibited the pressor response. A very slight potentiation of the pressor response was reported when thyroxine or histamine were systemically administered to the rabbits. A further interesting observation was that reserpine not only potentiated the cortical response but also the response to an intravenous injection of a posterior pituitary extract.

IV. Effect of Serum on the CEPR Many investigators have reported that schizophrenics have abnormal substances or an excess of normal ones in their body fluids. It thus seemed of interest to study the effects of serum of schizophrenics on the CEPR since physiological and clinical data suggest that the reactivity of central structures may be a major factor in normal function and possibly in cerebral disorders.

A. PRETREATMENT OF ANIMALSWITH SERUMFROM NORMAL VOLUNTEERS Normal serum was obtained from 35 different employees of Qsawatomie State Hospital. Serum was administered to rabbits either subcutaneously or intravenously and the amount of serum was from 1 to 5 ml. The number of injections ranged from 1 to 9 with intervals of 2 to 4 days. The CEPR was studied from 3 hours to 9 days after the last injection of serum. Forty-eight animals injected with serum from normal individuals were found to respond to cortically applied epinephrine with a rise in blood pressure. Such sera thus failed to produce any characteristic modification of the usual blood pressure effects engendered by topical application of epinephrine to the cerebral cortex (Minz and Walaszek, 1957).

B.

PRETREATMENT

OF

ANIMALSWITH SERUMFROM SCHIZOPHRENIC

PATIENTS Serum was obtained from 30 different schizophrenic patients. The mode of injection and amounts were the same as for the

152

EDWARD J. WALASZEK

normal series. The patients were females, all of whom had been hospitalized for at least 5 years. The age group ranged from 25 to 50 years. The diagnosis of “schizophrenia” was established after a very careful psychiatric examination and through a battery of psychological tests. All forms of the disease were represented (simple, undifferentiated, catatonic, paranoid, and hebephrenic ) . The patients had received no medication for a period of at least 4 weeks prior to the vein puncture and were closely supervised by specially trained aids in a separate research ward. Sixty-eight rabbits were systemically pretreated with serum from schizophrenic patients. In these animals the cardiovascular responses obtained by the topical application of epinephrine to the cerebral cortex were divided into four main groups: (1) a usual pressor response, ( 2 ) a diminished pressor response, ( 3 ) an abolished pressor response, and (4) a depressor response. A usual pressor response was obtained in 26.5% of the animals tested. This can be compared with 8S% of animals in the normal series which gave a usual pressor response. Approximately 16% of the animals tested gave a diminished pressor response in the schizophrenic series while 15% of the animals in the normal series gave this response. The pressor response was abolished in 31%, and a depressor response was obtained in 26.5% of the animals tested TABLE I1 EFFECTOF SERUMFROM SCHIZOPHRENIC PATIENTS ON THE CEPRa Pretreatment with serum CEPR

Normal volunteers

Number of animals Usual pressor response Response diminished Response abolished Response reversed a

48 85 15

-

Schizophrenic Pregnant Patients with patients women organic diseases 68 26.5 16 31 26.5

7 86 14 -

-

6 100

-

-

Data expressed in per cent of total number of animals tested.

in the schizophrenic series (Table 11). These last two categories were not represented at all in the normal series. In all instances the animals reacted normally to epinephrine administered intravenously. Inhibitions of the cardiovascular response to cortically applied

NEUROHORMONES AND THE CEPR

153

epinephrine as well as the depressor response occurred in different animals receiving serum from the same patient. Both responses probably represent different degrees of the same inhibitory process and were obtained without any special distinction in rabbits receiving serum from patients that exhibited catatonic, hebephrenic, paranoid, or mixed schizophrenic reactions. In a few experiments when the CEPR was studied 1 to 3 hours after the administration of the serum we were able to observe an exaggerated cardiovascular response. This response is now being studied in more detail. The most favorable conditions for these experiments were as follows: rabbits received intravenously 5 ml of patients’ serum (withdrawn and used the same day), and 16-18 hours later epinephrine was applied topically to the exposed cerebral cortex. Another experimental design which we used (Minz and Walaszek, 1960) was as follows. Rabbits received 2 ml of patients’ serum subcutaneously twice at 4-day intervals. The epinephrine was applied topically to the cerebral cortex within 24 hours after the last injection. OF ANIMALS WITH SERUM FROM PREGNANT C. PRETREATMENT WOMENAND FROM PATIENTS WITH ORGANIC DISEASES

In order to ascertain whether the observed cardiovascular changes obtained with serum from schizophrenic patients were specific or simply due to the presence of abnormal metabolites produced during nonspecific stress in the donors, control experiments were performed with serum from pregnant women and from patients with various organic diseases. Serum from 7 pregnant women (fifth to ninth months of pregnancy) was systemically administered to rabbits under the same conditions as described above for the schizophrenic and normal series, The CEPR was not altered by this pretreatment. It was found that 86% of the animals tested responded with the usual pressor rise to topical application of epinephrine to the exposed cerebral cortex. This compares favorably with the normal series where 85% of the animals responded with the usual hypertensive responses. These results contrast with those of the schizophrenic series where only 265% of the animals responded with the usual pressor rise.

154

EDWARD J. WALASZEK

Serum was obtained from patients with arthritis, asthma, diabetes, liver damage, ulcerative colitis, and pernicious anemia. This serum was administered to rabbits and the cortical epinephrine response measured. In all cases we were able to obtain the usual hypertensive response.

D. DISCUSSION AND SUMMARY The results reported by hilinz and Walaszek (1960) have shown that the hypertensive response to cortically applied epinephrine was typically elicited in rabbits pretreated with serum from normal human volunteers, pregnant women, and from pa--&--,OG

! :

40

mm

Hg

4

4

ET

ET

50

30 r4

ET

ET

FIG. 8. Rabbit blood pressure. Animals pretreated with serum from schizophrenic patients. Upper tracing, two applications of epinephrine to the cerebral cortex at 10-minute intervals illustrating the abolition of the usual CEPR. Lower tracing, same as above except that in this animal a depressor response was recorded.

tients with organic diseases. However this hypertensive response was characteristically modified in rabbits pretreated with serum from schizophrenic patients. Table I1 summarizes all the results. The depressor response (Fig. 8 ) observed in 26.5% of the animals pretreated with schizophrenic serum requires some clarification. At the present time it is difficult to explain this phenomenon. The depressor response sometimes coincided with the application of epinephrine but it often occurred 1 to 6 minutes after the filter paper was removed. It never appeared spontaneously but once the epinephrine solution was applied a series of these depressor responses (Fig. 9) were likely to occur. During this period the response to intravenously administered epinephrine was unchanged.

NEUROHORMONES AND THE CEPR

155

The exaggerated hypertensive response, obtained in a few experiments when the CEPR was studied 1 to 3 hours after the administration of the serum may have been due to the appearance of an initial sensitization possibly followed later on by an inhibition. At the present time, it is not known whether there is any relationship between this exaggerated response and the potentiation of the pressor response which is brought about by systemic administration of some of the hallucinogens. Minz and Walaszek (1960) studied a series of normal and schizophrenic sera in a double-blind experiment. They were able to correctly assign the animals receiving schizophrenic serum 66% of the time. This compares favorably as abnormal pressor responses were found in approximately that number of patients tested. In another double-blind experiment using serum from

+ i.

FIG.9. Rabbit blood pressure. At l., topical application of epinephrine to the cerebral cortex of an animal pretreated with serum from a schizophrenic patient.

patients in another institution, they were able to make the correct diagnosis only 40% of the time. Minz and Walaszek (1960) therefore regard these experiments not as a biological test for schizophrenia but rather as experimental tools for the study of the physiology and biochemistry of this disorder. The results of the inhibitory effects of schizophrenic serum on CEPR were not very uniform. This is not too surprising when the following facts were considered. Schizophrenia often presents phases with alternating spontaneous aggravation and remissions. Therefore, serum from patients studied may contain varying concentrations of abnormal metabolites according to the particular phase at which the serum was withdrawn. Second, different recipient animals may show different degrees of affinity or resistance to the effects of the patients’ serum. Lastly it must be considered that schizophrenia may be a number of different disorders with a common symptomatology. It would seem that effects seen with schizophrenic serum may be due to the presence of an agent or agents which are absent

156

EDWARD J. WALASZEK

from normal serum. Whether this hypothetical substance is the basis of this disease or just a metabolite which may be found in higher concentrations in schizophrenic serum with no relation to the etiology of the disease is not known at this time. In summary, it can be noted that serum from schizophrenic patients diminished, abolished, or reversed the usual hypertensive response which can be obtained with a topical application of epinephrine to the exposed cerebral cortex of rabbits. Serum from normal volunteers, pregnant women, and patients with various organic diseases did not alter the cortical epinephrine pressor response.

V.

Influence of Schizophrenic Serum upon the Content of Various Neurohumoral Substances in the Rabbit Hypothalamus

The results discussed above led us to conclude that serum from schizophrenic patients contains a component or components capable of modifying in a significant manner the reactivity of cerebral elements to epinephrine. In an effort to explain these findings we began to compare the content of pharmacologically active, naturally occurring substances in the brains of rabbits treated with serum from schizophrenic patients and from normal volunteers. Since reserpine greatly enhanced the CEPR and was capable of releasing serotonin from brain depots, it was postulated that the serotonin content of the hypothalamus might be the basis for an explanation of the alteration of the CEPR in animals systemically pretreated with schizophrenic serum. For this reason the concentration of serotonin in the brain was studied first. In all of these studies the serum treatment schedule was as follows: Each rabbit received two subcutaneous injections at intervals of 4 days of 2 ml of serum from either a schizophrenic patient or a normal volunteer. The brain was removed, cleaned, and the hypothalamic area separated for analysis. In some experiments the remainder of the brain, excluding the cerebellum, was also analyzed. The bioassay method (Amin et al., 1954) for the estimation of serotonin utilizing the isolated rat uterus was used. The brain tissue was finely minced and extracted with 90% acetone. The

NEUROHORMONES AND THE CEPR

157

acetone extract was evaporated to dryness, a small amount of water added, and the extract defatted with petroleum ether. The extract was then evaporated to dryness, redissolved in distilled water, and assayed biologically on the isolated rat uterus in estrus. We were able to assay serotonin in normal brains and in brains of animals pretreated with normal volunteer serum, but we were unable to assay the brains of animals treated with schizophrenic serum. This inability to assay serotonin content in brains of animals pretreated with schizophrenic serum was not due to the absence of serotonin in the brain extracts but rather to some interfering substance present in the extract that antagonized serotonin by relaxing the rat uterus. This antagonistic effect to standard serotonin contractions could be washed out and after 15-20 minutes the original reactivity of the rat uterus to serotonin had returned. The antagonistic effect could then be repeated by another dose of the brain extract. When the brain extract was applied to the isolated rabbit intestine as the test object the effect observed was one of relaxation. Injection of the extract into a rabbit produced a rise in systemic blood pressure. Therefore it seemed that the interfering substance had the biological properties of epinephrine. At this point we began to study the catechol amine content of the brains of rabbits pretreated with schizophrenic and normal serum. Throughout this study we made use of various biological assay procedures. In a proper assay of neurohumoral substances it is important that a pharmacological differentiation be obtained. One of the principal difficulties in the biological estimation of a naturally occurring neurohumoral substance is the possibility of interference by other biologically active substances. Tissue extracts generally contain many active principles, and a biological assay is possible only when the extract is pharmacologically pure; that is, when it contains only one active principle effective in the conditions of the experiment, Specific evidence of identification can sometimes be obtained by comparing the tissue extract by more than one method using different biological test objects. If parallel estimates of the potency of the extract agree with one another there is good reason to believe that the active substances have been correctly identified. We have been able to assay or detect the following substances in extracts of rabbit hypothalamus : norepinephrine, epinephrine,

158

EDWARD J. WALASZEK

dopamine, serotonin, histamine, substance P, vasopressin, and oxytocin. The respective biological tests utilized were cat blood pressure, guinea pig blood pressure, isolated guinea pig ileum, isolated rat uterus, and the isolated hen rectal caecum. Pharmacological differentiation was obtained by parallel assay, use of specific antagonists, destruction of interfering substances by chemical means, and by separation of the substances in question by means of paper chromatography. Whenever possible a chemical identification was also obtained using a spectrophotofluorometer. A. NOREPINEPHRINE AND EPINEPHRINE The estimation of norepinephrine and epinephrine was accomplished by using the differential assay procedure of Euler (1956). The two test preparations used were the blood pressure of the

FIG. 10. Differential assay of norepinephrine and epinephrine. a. Cat blood pressure illustrating the equipotent effect of 1 pg of norepinephrine ( N ) and 4 pg of epinephrine ( E ) . b. Effect of hypothalamic extract ( H ) bracketed between 0.5pg and 0.3pg of norepinephrine (N). c. Isolated hen rectal caecum illustrating the equipotent effect of 10 nanograms of epinephrine ( E ) and 400 nanograms of norepinephrine ( N ) . d. Effect of hypothalamic extract ( H ) in milliliters bracketed between two doses of norepinephrine ( N ) in nanograms.

cat and the isolated hen rectal caecum. Norepinephrine is 0.5 to 4 times more potent than epinephrine on the cat blood pressure (Fig. lo), while on the hen rectal caecum epinephrine is 10 to 40 times more potent than norepinephrine. The tissue extract containing both catechol amines was assayed, and the activity ratio of norepinephrine and epinephrine determined on both test

NEUROHORMONES AND THE CEPR

159

preparations. When the effect of the extract on both the blood pressure of the cat and on the hen rectal caecum had been evaluated, the norepinephrine and epinephrine content was computed from the equations of Euler (1956). The cats were anesthetized with pentobarbital sodium (35-40 mg/kg), the femoral vein was cannulated for injections, and the carotid artery was connected to a mercury manometer by means of a glass cannula for recording the blood pressure. Essentially the same combination of drugs as used by Karki (1956) were administered to the cat before the assay. The animals were pretreated with atropine ( 2 mg/kg) (to reduce reflex vagus action), ergotamine (0.15 mg/kg) (to reduce homeostatic pressor receptor reflexes), and chlorprophenpyridamine ( 2 mg/kg) , The latter was used to block possible histamine activity in the extracts and because this compound potentiates the action of norepinephrine and epinephrine on cat blood pressure (Paasonen, 1953). In a few experiments cocaine (8 mg/kg ) was administered in addition to the above mentioned drugs to produce greater sensitivity of the blood pressure to the catechol amines. The isolated hen rectal caecum was suspended in a 10-ml bath at 39°C. The bath fluid was Tyrode solution with half the usual amount of potassium content, aerated with 5% CO, in 0 2 (Barsoum and Gaddum, 1935). A typical assay is illustrated in Fig. 10. The effect of schizophrenic serum upon the catechol amine content of rabbit hypothalamus has been reported by Walaszek and Minz (1958). The hypothalami of 9 rabbits pretreated with serum of normal volunteers showed an average epinephrine content of 0.11 pg/gm of tissue. The hypothalami of 11 rabbits pretreated with serum of schizophrenic patients showed an average epinephrine content of 0.9 pg/gm of tissue (Table 111). The norepinephrine content was 3.3 pg/gm of tissue in the normal series and 7.0 pg/gm of tissue in the series treated with schizophrenic serum. Thus the serum of schizophrenics engendered in the hypothalamus an eightfold increase in epinephrine and a twofold increase in norepinephrine. The values obtained in our assays were in terms of norepinephrine and epinephrine activity on two different biological test systems. A number of pharmacologically active, naturally occurring substances were present in extracts of rabbit hypothalamus, and these substances could interfere with the assay. These are

TABLE I11 NEUROHUMORAL SUBSTANCES IN RABBIT HYPOTHALAMUS Pretreated with normal serum

Neurohumor Norepinephrine Epinephrine Serotonin Histamine Substance P

Number of animals 9 9 9

11 14

Pretreated with schizophrenic serum

Concentrations in pg/gm of tissue 2 S.E.

Number of animals

Concentrations in pg/gm of tissue ? S.E.

P

3.3 rf: 0.4 0.11 f 0.02 2.1 f 0.2 0.11 & 0.01 82 l l a

11 11 9 10 16

7.0 rf: 0.8 0.9 f 0.2 1.7 2 0.17 0.18 f 0.01 57 2 5a

0.001 0.001 0.2 0.005 0.05

a Data in units of substance P activity per gram

of tissue k S.E.

*F

NEUROHORMONES AND THE CEPR

161

dopamine, serotonin, histamine, acetylcholine, substance P, vasopressin, oxytocin, and perhaps other unknown pharmacologically active substances. Acetylcholine and histamine were eliminated by the use of atropine and an antihistaminic in the biological test preparations. Substance P was not a disturbing influence since, if its presence was suspected of interfering with the assay, the extract was incubated with chymotrypsin which destroyed this polypeptide. Dopamine has been reported in rabbit brain by

FIG. 11. Comparative potency of dopamine, norepinephrine, and epinephrine. Upper tracing, spinal cat blood pressure; at N, norepinephrine; at E, epinephrine; at D, dopamine; at HT, serotonin. Amounts in pg administered intravenously. Lower tracing, isolated hen rectal caecum; at E, epinephrine; at D, dopamine. Amounts in nanograms present in 10-ml organ bath. Dopamine was some 1000 times less potent than epinephrine on the hen rectal caecum and 40 times less potent than norepinephrine on cat blood pressure.

Carlsson et al. (1958) in concentrations of 0.4 &gm. Its biological potency as compared to epinephrine and norepinephrine is relatively weak (Fig. 11). However, in large amounts it could contribute slightly toward a higher value for norepinephrine content. Serotonin did not interfere with the cat blood pressure but it was capable of contracting the hen rectal caecum. Vasopressin was probably the substance which contributed the most toward a higher value for norepinephrine content. In order to clarify some of the points mentioned above,

162

EDWARD J. WALASZEK

Walaszek et al. (1958) reinvestigated this problem. Our main objective was to prove or disprove the presence of epinephrine in the hypothalamus and to ascertain what effect schizophrenic serum had on epinephrine content. The hypothalami of 2 animals were pooled and extracted for their catechol amine content by the method developed by Vogt (1954). The extracts were then chromatographed on paper in pheno1:HCl. Using this procedure,

FIG. 12. Separation and localization of norepinephrine, epinephrine, and serotonin by paper chromatography using phenol : HC1 and 3 pg of each substance. Advancing front of solvent was 32cm. One-cm portions of the paper chromatogram were eluted and tested. Tracings labeled as distances in centimeters from point of origin of the paper chromatogram. Upper tracing, cat blood pressure illustrating the separation of norepinephrine (cm 6, 7 , 8) from epinephrine (cm 17, 18). Lower tracing, hen rectal caecum illustrating separation of norepinephrine (cm 7, 8) and epinephrine (cm 17, 18), which relax the muscle, from serotonin (cm 19) which contracts the muscle.

Vogt was able to separate the various pharmacologically active substances in brain extracts. The chromatograms were then divided into l-cm strips, eluted, and assayed biologically on various test systems. An example of such a separation is illustrated in Fig. 12. A clear-cut differentiation can be obtained between epinephrine which relaxes the hen rectal caecum and serotonin which contracts this tissue. Dopamine occurs in the paper chromatogram just below epinephrine while vasopressin and substance P

NEUROHORMONES AND THE CEPR

163

occur at a much higher R, than serotonin. Figure 13 illustrates an experiment on the hen rectal caecum which compares the hypothalamic extract from a rabbit, pretreated with normal serum, with the hypothalamic extract from a rabbit pretreated with schizophrenic serum. The values for norepinephrine and epinephrine obtained by this procedure were one-third to one-half the amounts previously reported. The augmentation of the epinephrine content of the hypothalamus in animals pretreated with schizophrenic

FIG. 13. Isolated hen rectal caecum. Comparison of a hypothalamic extract (0.17 gm of tissue) from an animal pretreated with normal serum to a hypothalamic extract (0.15 gm of tissue) from an animal pretreated with serum from a schizophrenic patient. Chromatography on paper with phenol : HC1. a. Dose response of epinephrine in nanograms in 10-ml bath. b. Hypothalamic extract of animal pretreated with normal serum. The cm 16, 17, and 18 contain epinephrine; the cm 20 contains serotonin. c. Hypothalamic extract of animal pretreated with schizophrenic serum. The cm 16, 17, and 18 contain epinephrine; the cm 20 contains serotonin. This extract contained 3 times more epinephrine per gram of tissue than the hypothalamic extract from the animal pretreated with normal serum.

serum was 35-fold as compared to animals pretreated with normal serum. Thus we were able to verify our previous observations that there was an augmentation in the catechol amine level in the hypothalami of animals pretreated with schizophrenic serum. Dopamine was detected on our chromatograms, and qualitatively there seemed to be an increase in the dopamine content in the schizophrenic series. Figure 14 illustrates the effect of dopamine and epinephrine on the guinea pig blood pressure and the carbachol-stimulated rat uterus, respectively.

164

EDWARD J. WALASZEK

B. SEROTONIN Serotonin was assayed using the spectrophotoflurometric procedure by Bogdanski et al. (1956). A few experiments were performed again using the biological assay method of Amin et al. (1954). In the biological method we followed the procedure of Garven ( 1956 ) which selectively destroyed the catechol amines in the tissue extracts by means of a polyphenoloxidase enzyme.

FIG. 14. Detection of dopamine and epinephrine in the hypothalamic area of rabbits pretreated with serum from schizophrenic patients. Hypothalamic extract chromatographed on paper with phenol : HCl, cut into l-cm portions, eluted, and tested. Tracings labeled in distance in centimeters from point of origin. a. Guinea pig blood pressure; dopamine detected in cm 15 of paper chromatogram by its characteristic depressor response in this species. b. Carbachol-stimulated rat uterus; cm 17 of the same paper strip illustrates the presence of epinephrine by its characteristic relaxant effect on this smooth muscle. Centimeter 18 signifies the presence of serotonin by the contraction of the rat uterus when the extract was added and by the subsequent reinforcement of the carbachol contraction.

The values obtained for serotonin (Table 111) indicate that the group treated with schizophrenic serum had a slightly lower content in the hypothalamus than those pretreated with normal serum. However the P values for the serotonin determination was 0.2. When a small series of animals were tested by the biological method of assay we again found no significant difference between the normal series and the schizophrenic series. The

NEUROHORMONES AND THE CEPR

165

values obtained for serotonin content by the biological assay method were approximately one-third those obtained by the spectrophotofluorometric method. The conclusion reached by this study was that serotonin content of the rabbit hypothalamus was not affected by schizophrenic serum. The augmentation of the CEPR by reserpine and the diminution of this pressor response by iproniazid was not related to the capacity of these drugs in altering serotonin levels but probably was related to their capacity in altering the catechol amine levels in the rabbit hypothalamus. C. SUBSTANCE P The vasodilator polypeptide Substance P was measured by the method of Amin and co-workers (1954) using a standard substance P preparation on the isolated guinea pig ileum in a 3-ml bath with atropine and tripelennamine present. Our assays have shown that the hypothalami of rabbits pretreated with schizophrenic serum have less (Table 111) substance P than the hypothalami of rabbits pretreated with normal serum. It would seem that perhaps some slight significance may be attached to this data. It is interesting to speculate on the possible physiological role of substance P. Its distribution in the central nervous system follows that of norepinephrine and serotonin (Amin et al., 1954). Lembeck (1953) and Zetler and Schlosser (1955) favor the view that Substance P is itself a chemical mediator in the central nervous system. The evidence for the acceptance of this view is not consistent. An interesting suggestion for the physiological role of Substance P was brought forth by Stern and Dobric (1957). These authors found that substance P and mephenesin act synergistically on various preparations. Substance P, mephenesin, and meprobamate can inhibit polysynaptic neurons of the central nervous system. This suggests the possibility that the role of substance P in the central nervous system is that of a physiological tranquilizer. D. HISTAMINE Histamine was assayed biologically by a modification of the method of Code (1937) using the isolated guinea pig ileum in

166

EDWARD J. WALASZEK

a 1.8-ml bath. The results seem to indicate that the histamine content of the hypothalamus was increased in the series of animals pretreated with schizophrenic serum as compared to the normal series (Table 111). The presence of high histamine values for the hypothalamus have been reported by Harris and co-workers ( 1952). Histamine does have some central excitatory actions. At the present time we do not know whether this action of histamine can be correlated with the increase seen in the series of animals pretreated with schizophrenic serum. Lucy (1954) reported that schizophrenics have the capacity for tolerating large doses of histamine. There is also the unexpectedly low rate of allergic disorders (Ehrentheil, 1957) in schizophrenic patients. It would seem that a thorough investigation into the metabolism of histamine by schizophrenic patients would be very worthwhile. Recently we have studied the level of histaminase (diamine oxidase) in serum of schizophrenic patients. We have been able to demonstrate that there is considerably more histaminolytic activity in serum from schizophrenic patients than in serum from normal volunteers. This increased level of histaminase may explain the increased histamine tolerance as well as the low incidence of allergic manifestations in schizophrenic patients. VI.

Correlation of the Inhibitory Effects of Schizophrenic Serum on the CEPR with the Catechol Amine Content of the Hypothalamus

It is interesting to note that drugs (Table IV) which influenced the catechol amine content of the hypothalamus were capable of interfering with the well-balanced steady state condition that apparently regulates the reactivity of cortical effectors to epinephrine. When the content of catechol amines in the hypothalamus was normal, normal CEPR’s were obtained. When the content of the catechol amines was lowered as with reserpine or tetrabenazine, there was an exaggerated cardiovascular response. When the content was raised as with iproniazid or the other amine oxidase inhibitors, there was an inhibition of this pressor response. Animals pretreated with schizophrenic serum show an inhibi-

167

NEUROHORMONES AND THE CEPR

tion of the CEPR and an increase in the catechol amine content of the hypothalamus. It thus seems possible that the mechanism of the inhibition of the pressor response is concerned with the increased catechol amine content of the hypothalamus. With an increased catechol amine content there would be a saturation of the effector sites so that the application of epinephrine to the cortex of such an animal would have little or no effect. Conversely the reduction of the catechol amine content in the hypothalamus would lead to a greater number of unoccupied effector sites which may account for the augmentation of the CEPR. The saturation of TABLE IV RELATIONSHIPOF THE CEPR TO CATECHOL AMINECONTENT OF HYPOTHALAMUS Catechol amine content of hypothalamus

CEPR

40 100 100 100

Decreased Decreased No change Increased Increased

Augmented Augmented No change Diminished Diminished

100 100

Increased Increased

Diminished Diminished

Systemically administered drug Reserpine Tetrabenazine Isoniazid Iproniazid Serine N-isopropylhydrazine Acetylmethionyl N-isopropylhydrazine L-DoDa

THE

1

effector sites may be the reason that the CEPR was blocked with a massive dose (750 pg ) of epinephrine administered intravenously. Epinephrine is unable to penetrate the blood-brain barrier (WeilMalherbe et al., 1959) except to a small extent in the hypothalamus. If only 0.01% of the massive intravenous dose of epinephrine reached the hypothalamus it would be approximately the amount that is normally present in this tissue. Any central effect of epinephrine administered intravenously may be due to its interaction with hypothalamic epinephrine effector sites. Another item of evidence along these lines was that in animals pretreated with systemically administered 3-(3,4-dihydroxypheny1)alanine (L-dopa, 100 mg/kg intraperitoneally ), the precursor of the catechol amines, the CEPR was abolished 3 hours later. The catechol amine content of rabbit hypothalamus (Figs. 15 and 16) can be raised by systemic prernedication with L-dopa.

168

EDWARD J. WALASZEK

FIG. 15. Spinal cat blood pressure. Detection of norepinephrine in a hypothalamic extract, chromatographed on paper with phenol : HCl. Tracings labeled in distance in centimeters from point of origin. At N, norepinephrine in pg. a. Rabbit pretreated with isoniazid (100 mg/kg) 16 hours prior to sacrifice. The cm 7 (0.17 gm of tissue) contains less than 0.25 pg of norepinephrine. b. Rabbits pretreated with iproniazid ( 100 mg/kg ) 16 hours prior to sacrifice. The cm 7 contains approximately 0.5 Fg of norepinephrine (0.18 gm of tissue). c. Rabbits pretreated with L-dopa (100 mg/kg) 4 hours prior to sacrifice. The cm 6, 7, and 8 (0.20gm of tissue) contained appreciable amounts of norepinephrine. Pretreatment of rabbits with L-dopa or iproniazid increased the norepinephrine content of the hypothalamus.

FIG. 16. Isolated hen rectal caecum. Detection of epinephrine in hypothalamic extracts, chromatographed on paper with phenol : HCl. Tracings labeled in distance in centimeters from point of origin. a. Rabbits pretreated with D-dopa ( 100 mg/kg); only small amounts of epinephrine were detected in the extract (0.19 gm of tissue). In cm 19 serotonin was present. b. Rabbit pretreated with L-dopa (100 mg/kg), the precursor of the catechol amines. At cm 16 and 17 appreciable amounts of epinephrine were detected (0.20 gm of tissue). At cm 18 a mixture of serotonin and epinephrine was present.

NEUROHORMONES AND THE CEPR

169

The mechanism by which the catechol amine content has been elevated in the hypothalamus of rabbits treated with schizophrenic serum is unknown. There may be an increased production of catechol amines or a decreased destruction. The proposal of decreased destruction could also be applied if there was an abnormal metabolite present in serum which may inhibit the normal metabolic destruction of the catechol amines. We have tested the effect of schizophrenic serum upon the monoamine oxidase activity of the hypothalamus both in vivo and in vitro and found no alteration of enzymatic activity. As yet we have not tested the effect of schizophrenic serum on the O-methylation enzyme which seems to be a more important pathway for the metabolism of the catechol amines (Axelrod et at., 1958). There may be many mechanisms by which various drugs can influence the CEPR. The effects of a certain group of drugs and of schizophrenic serum seem to correlate with the content of catechol amines in the hypothalamus. In summary we may say that the mechanism of the inhibition of the CEPR seems to be concerned with the catechol amine content of the hypothalamus since, when the content is normal, normal pressor responses are obtained. When the content is lowered, as with reserpine 'or Tetrabenazine, there is an augmentation of the usual pressor response. When the content is raised, as with the amine oxidase inhibitors, schizophrenic serum or L-dopa there is an inhibition of the CEPR. These results lead us to the conclusion that schizophrenic serum contains a component or components which are capable of modifying in a characteristic way the reactivity of cerebral elements to epinephrine and in influencing, by some unknown mechanism, the levels of various neurohumoral agents in the hypothalamus.

VII.

Presence of an Abnormal Factor or Factors in Serum of Schizophrenics

The presence of abnormal substances or an excess of normal ones has often been reported in the body fluids of schizophrenic patients. Osmond (1958) has recently reviewed this field. Our results have led us to believe that schizophrenic serum contains a factor or factors which influences the CEPR and the catechol amine

170

EDWARD J. WALASZEK

level of the hypothalamus. We do not know at the present time whether this factor or factors have any relationship with the etiology of schizophrenia or are merely reflections which may have their origin in some secondary metabolic or physiological disturbances. We have attempted to characterize the factor or factors present in serum of schizophrenics. Various extractions with organic solvents failed to concentrate the active material. Using the Cohn plasma protein fractionation technique (Schmid et al., 1956), we were able to find the biologically active material in Fraction 111. Other plasma protein fractions were inactive. The biological activity was very labile and no attempt was made to fractionate larger amounts.

Vlll.

Summary

The pressor response obtained when epinephrine was applied to the exposed cerebral cortex of rabbits has been described. This pressor response was shown to be due not to epinephrine itself but to an unknown hypertensive agent released from the hypothalamus. Transection studies have been cited which led the investigators to the conclusion that the humoral stimulus produced by epinephrine on the cerebral cortex was transmitted by neuronal pathways to a hypothalamic center capable of secreting a substance with hypertensive activity. A number of centrally acting drugs modified the CEPR. Reserpine and Tetrabenazine, tranquilizers which deplete norepinephrine and serotonin from brain depots, markedly augmented the pressor response while chIorpromazine, a tranquilizer with adrenergic blocking activity, blocked the pressor response. Iproniazid and other amine oxidase inhibitors, drugs which elevate the levels of norepinephrine and serotonin in the brain, markedly diminished the CEPR. Parasympathomimetic drugs which pass the blood-brain barrier also diminished the pressor response. The hallucinogens, lysergic acid diethylamide and bufotenine, were able to potentiate the pressor response to cortically applied epinephrine. Pentobarbital, hexobarbital, procaine, and morphine did not modify the pressor response while ethyl ether inhibited the response. Pretreatment of rabbits with schizophrenic serum modified the CEPR in the majority of the cases studied while pretreatment with serum from normal volunteers, pregnant women,

NEUROHORMONES AND THE CEPR

171

and patients with organic diseases had no effect on the response. Pretreatment of animals with schizophrenic serum raised the catechol amine levels in the hypothalamus. Epinephrine and norepinephrine, measured by a biological assay method, was significantly elevated in the schizophrenic series as compared to a series pretreated with normal serum. There was no change in the serotonin level of the hypothalamus. Substance P levels were slightly lowered and histamine levels elevated in the hypothalami of animals pretreated with schizophrenic serum. A tentative explanation of the effects of schizophrenic serum on the CEPR was discussed from the standpoint of catechol amine levels in the hypothalamus. When the catechol amine level of the hypothalamus was normal, normal cardiovascular responses were obtained when epinephrine was applied to the cerebral cortex. When the catechol amine level was lowered as with reserpine or Tetrabenazine, augmented pressor responses were obtained. When the catechol amine level was raised as with the amine oxidase inhibitors, schizophrenic serum, or after the administration of L-dopa (the precursor of the catechol amines), the CEPR was inhibited. A conclusion was reached that schizophrenic serum contains a factor or factors which influences the CEPR and the catechol amine level of the hypothalamus. ACKNOWLEDGMENT The author wishes to express his deep gratitude to Dr. Bruno Minz, professor of physiology at the Sorbonne, Paris, and formerly director of research at Osawatomie State Hospital, Osawatomie, Kansas. Dr. Minz introduced the CEPR and initiated this line of inquiry. The opinions expressed here are mostly also his opinions which were evolved during a fruitful collaboration at which time he very kindly introduced me to this research field.

REFERENCES Amin, A. H., Crawford, T. B. B., and Gaddum, J. H. (1954). J . PhysioZ. (London) 126, 596. Axelrod, J., Inscoe, J. K., Senoh, S., and Witkop, B. (1958). Biochim. et Biophys. Acta 27, 210. Barsoum, G. S., and Gaddum, J. H. (1935). J. Physiol. (London) 86, 1. Beckett, S., and Gellhorn, E. (1948). Am. J. Physiol. 153, 113. Bogdanski, D. F., Pletscher, A., Brodie, B. B., and Udenfriend, S. (1956). J . PharmacoZ. Exptl. Therap. 117, 82.

172

EDWARD J. WALASZEK

Brenner, C., and Merritt, H. H. (1943). A.M.A. Arch. Neural. Psychiat. 48, 382. Brodie, B. B., and Shore, P. A. (1957). Ann. N.Y. Acad. Sci. 66, 631. Bucy, P. C. (1935). A.M.A. Arch. Neurol. Psychiut. SS,30. Carlsson, A., Lindqvist, M., Magnusson, T., and Waldeck, B. (1958). Science 127, 471. Cerletti, A. ( 1955). “Neurophamacology,” Transactions of the Second Conference, p. 51, Josiah Macy, Jr. Foundation, New York. Chamorro, A. (1957a). Compt. rend. ucad. sci. 244, 1413. Chamorro, A. (1957b). Compt. rend. sac. biol. 161, 500. Chamorro, A., and Minz, B. (1955a). Compt. rend. SOC. biol. 149,309. Chamorro, A., and Minz, B. (1955b). Compt. rend. acad. sci. 240, 1368. Chamorro, A., and Minz, B. (l956a). Compt. rend. SOC. biol. 160, 652. Chamorro, A., and Minz, B. (1956b). Compt. rend. SOC. biol. 160, 299. Chamorro, A., and Minz, B. ( 1 9 5 6 ~ )Compt. . rend. sac. biol. 160, 849. Chamorro, A., and Minz, B. (1957a). Compt. rend. SOC. biol. 161, 214. Chamorro, A,, and Minz, B. (1957b). Compt. rend. sac. biol. 161, 272. Code, C. F. (1937). J. Physiol. ( L o n d o n ) 89, 257. Dusser De Barenne, J. G., and Kleinknecht, F. (1924). Z. Biol. 82, 13. Ehrenthiel, 0.F. (1957). A.M.A. Arch. Neurol. Psychiut. 77, 178. Essig, C. F., Adkins, F. J., and Barnard, G. L. (1953). Proc. SOC. Exptl. Biol. Med. 82, 551. Euler, U. S. v. (1956). “Noradrenaline.” C. C Thomas, Springfield, Illinois. Euler, U. S. v., and Domeij, B. (1945). Acta Pharmacol. Toxicol. 1, 263. Garven, I. (1956). Brit. J . Pharmacol. 11, 66. Graham, J. D. P., and Khalidi, A. I. (1954). J. Fac. Med. Baghdad, Iraq 18, 1. Harris, G. W., Jacobsohn, D., and Kahlson, G. (1952). Ciba Foundation Colloquia Endocrinol. 4, 186. Kaada, B. R. ( 1951). Acta Physiol. Scand. 24, Suppl. 83. Karki, N. T. ( 1956). Acta Physiol. Scand. 39, Suppl. 132. Kennard, M. A. (1935). A.M.A. Arch. Neural. Psychiat. 33, 537. Kremer, W. F. (1948). Am. J. Physiol. 162, 314. Langford, H. G., Patterson, J. L., and Porter, R. R. (1957). Circulation Research 6, 268. Lembeck, F. ( 1953). Arch. exptl. Pathol. P h u m k o l . Naunyn-Schmiedeberg’s 219, 197. Lucy, J. D. (1954). A.M.A. Arch. Neurol. Psychiut. 71, 629. Merritt, H. H., and Brenner, C. (1941). Trans. Am. Neurol. Assoc. 67, 152. Miller, F. R. (1937). J. Physiol. ( L o n d o n ) 91, 212. Minkowski, M. (1933). Rev. Neural. 69, 1177. Minz, B. (1957). Compt. rend. sac. biol. 161, 432. Minz, B., and Chamorro, A. (1955). Compt. rend. mad. sci. 240, 454. Minz, B., and Goldstein, L. (1955). Compt. rend. SOC. biol. 149, 1200. Minz, B., and Walaszek, E. J. (1957). Compt. rend. acad. sci. 244, 1974. Minz, B., and Walaszek, E. J. ( 1958). Compt. rend. acad. sci. 246, 1326. Minz, B., and Walaszek, E. J. (1959). Ann. N.Y. Acad. Sci. 80, 617.

NEUROHORMONES AND THE CEPR

173

klinz, B., and Walaszek, E. J. (1960). J. Nervous Mental Disease. In press. Minz, B., Buser, P., and Albe-Fessard, D. (1953a). Compt. rend. SOC. bwl. 147, 1154. Minz, B., Remond, A., and McCulloch, W. S. (195313). Proc. SOC. Exptl. Biol. Med. 83, 306. Osmond, H. ( 1958). “Proceedings of the Symposium on Chemical Concepts of Psychosis,” p. 3. McDowell, Obolensky, New York. Paasonen, M. K. (1953). Ann. Med. Exptl. et B w l . Fenniae (Helsinki) 31, Suppl. 7. Pfeiffer, C . C., and Jenney, E. H. ( 1957). Ann. N.Y. Acad. Sci. 66,753. Pinkston, J. O., and Rioch, D. Mck. (1938). Am. J. Physiol. 121, 49. Pletscher, A., Besendorf, J. and Bachtold, H. P. (1958). Arch. exptl. Pathol. Pharmakol. Naunyn-Schmiederg’s 232, 499. Popper, L. (1933). Deut. med. Wochschr. 69, 1163. Schmid, K.,Rosa, E. C., and McNair, M. B. (1956). 1. Biol. Chem. 219, 769. Smith, C.M., Bemstein J., Walaszek, E. J., and Minz, B. (1958). Am. SOC. Pharmacol. Exptl. Therap., Ann Arbor, Fall meeting, p. 32. Spector, S., Prockop, D. J., Shore, P. A., and Brodie, B. B. (1958). Science 127, 704. Stem, P., and Dobric, V. ( 1957). “International Symposium on Psychotropic Drugs,” p. 448. Elsevier Press, Amsterdam. Stone, C . A., Meckelnburg, K. L., and Torchiana, M. L. (1958). Arch. intern. Pharmacodynamie 117, 419. Vogt, M. ( 1954). 1. Physiol. (London) 123,451. Walaszek, E. J., and Minz, B. (1958). J. Pharm. and Exptl. Therap. 122,80a. Walaszek, E. J., Smith, C. M., and Minz, B. (1958). Federation Proc. 17, 416. Walaszek, E. J., and Underwood, J. ( 1960). Manuscript submitted. Weil-Malherbe, H., Axelrod, J,, and Tomchick, R. (1959). Science 129, 1226. Zetler, G.,and Schlosser, L. (1955). Arch. exptl. Pathol. Phurmukol. Naunyn-Schmiedeberg’s 224, 159.

This Page Intentionally Left Blank

THE ROLE OF SEROTONIN IN NEUROBIOLOGY By

Erminio Costa

Thudichum Psychiatric Research laboratory, Galesburg State Research Hospital, Galesburg, Illinois

Introduction ........................................... Evidence for a Neurophysiological Role of Serotonin ......... Biosynthesis of Serotonin ................................. Mechanism of Action of Neurotransmitters ................. Mechanism of Serotonin Action at Cellular Level . . . . . . . . . . . . A. Importance of Monoamine oxidase ( M O ) ............... B. Action at Receptor Sites .............................. VI. Methods for Serotonin Bioassay ........................... VII. Importance of Serotonin in Psychopharmacology ............. A. Serotonin Metabolism and Reserpine Effects ............. B. Psychotomimetic Drugs and Serotonin . . . . . . . . . . . . . . . . . . C. Antidepressant Drugs and Serotonin .................... VIII. Summary .............................................. References .............................................

1. 11. 111. IV. V.

1.

175 177 185 190 193 193 203 205 207 207 213 215 219 221

Introduction

The work of Bishop, Eccles, Grundfest and Purpura and coworkers has furnished experimental evidence which differentiates conductile from synaptic electrogenesis. The former originates in electrically excitable membranes (Hodgkin, 1951), has a more recent phylogenetic development ( Bishop, 1956), and assures conduction of impulses without decrement, in all-or-none fashion, over the length of a nerve fiber. There are several viewpoints regarding the mechanisms involved. Nachmansohn ( 1955) believes that chemical reactions related to acetylcholine metabolism mediate the conduction of impulses, whereas Hodgkin (1951) considers that conductile electrogenesis is based, essentially, on physical processes. On the basis of Grunfest’s (1957) results one can conclude that the electrophysiological approach to neurobiology supports 175

176

ERMINIO COSTA

the concept of chemical mediation of synaptic transmission. However, the field still remains controversial. For example, recent work on synaptic transmission in crayfish (Furshpan and Potter, 1957) and squid (Hagiwara and Tasaki, 1958) has led to diametrically opposite conclusions; the former supporting electrical transmission, the latter chemical. Acetylcholine (Feldberg, 1945) and choline acetylase (Hebb, 1957) have been found in brain structures and a potential role of the ester in synaptic events has been postulated. However, the presence of acetylcholine and choline acetylase is a property of selected neurons rather than a universal characteristic of nerve cells (Feldberg, 1954). Consequently one must admit lack of knowledge concerning the identity of transmitters at both inhibitory and excitatory neuronal synapses, for acetylcholine mediated transmission is not continuous, even within a single pathway (Eccles et d.,1954). The present trend is toward the analysis of brain for chemical constituents, which, by virtue of their biochemical and physiological characteristics, might play the role of neurotransmitters. Admittedly this approach is only a beginning, but it can contribute to the formulation of testable hypotheses. Therefore, it is not surprising that at present we are confronted with few facts and much speculation, mostly contradictory. The experimental approach employed to elucidate the neurophysiological role of serotonin ( 5-hydroxytryptamine ) has followed the foregoing pattern. Consequently, the evaluation of the experimental data is, as yet, unsettled. Two different interpretations of the role of serotonin in brain mechanisms have been offered: the proposals formulated by Brodie (Brodie and Shore, 1957) and the more conservative suggestions of Erspamer ( 1957). Brodie’s (1958) proposals are summarized by his statement “Serotonin and norepinephrine may qualify as neurohormones” for the two neurovegetative divisions of the hypothalamus. These have been classified by Hess (1954) as trophotropic and ergotropic and Brodie suggests serotonin and norepinephrine, respectively, as neurotransmitters. The strongest support of Brodie’s theory is the selective presence of serotonin (Twarog and Page, 1953; Zetler and Schlosser, 1954; Amin et d.,1954; Bogdanski et al., 1957) in different brain structures. The diencephalon has high concentrations

THE ROLE OF SEROTONIN IN NEUROBIOLOGY

177

of both 5-hydroxytryptophan ( 5-HTP) decarboxylase ( Bogdanski et al., 1958) and serotonin in all mammals tested. On the other hand, Erspamer (1957) states that he would not be surprised if excessive or defective production, as well as liberation, of brain serotonin were entirely unrelated to brain physiological events. However Erspamer does not exclude the possibility that deviation of serotonin metabolism might be either a causal or a facilitatory factor for mental or neurological disturbances, as was first suggested by Woolley and Shaw (1954a, b ) , Gaddum ( 1953), and Amin et al. ( 1954).

II. Evidence for a Neurophysiological Role of Serotonin

A review of the evidence for and against the neurophysiological importance of serotonin will illustrate the contradictory nature of the data mentioned above. 1. In general, any great selective concentrations of a biologically active substance, in one part of an embryologically uniform biological system, such as the nervous system, is an indicator of a specific and potentially important function. The foregoing is true of serotonin, as shown by the following: ( a ) The brain of mammals (Amin ,et at., 1954; Correale, 1956; Garattini et al., 1958) including man (Bogdanski et al., 1957; Costa and Aprison, 1958a) as well as certain ganglia of some invertebrates (Florey and Florey, 1953; Welsh, 1957) contain serotonin. ( b ) In man, the substantia nigra and red nucleus contain greater serotonin concentrations than other parts of the mesencephalon, such as the corpora quadrigemina (Costa and Aprison, 1958a). Among the different diencephalic structures studied in various species of animals, the hypothalamus has the greatest serotonin concentrations (Amin et al., 1954; Paasonen et al., 1957; Bogdanski et al., 1957; Costa and Aprison, 1958a). In man (Costa and Aprison, 1958a) and most mammals (Paasonen et ,ul., 1957) the old phylogenetic telencephalic structures contain greater serotonin concentrations than structures of more recent phylogenetic development. ( c ) The morphological modifications as well as the contractions of neurons and glial cells caused by serotonin added in small

178

ERMINIO COSTA

concentrations to tissue cultures have been regarded as a potential evidence in favor of the hypothesis that serotonin is a brain neurohormone ( Woolley, 1958a). Unfortunately, at present, the importance of this in vitro observation for the physiology of neurons in vivo cannot be evaluated. ( d ) In uitro, different brain structures of dogs, cats (Bogdanski et al., 1957), and rabbits (Fig. 3 ) decarboxylate j-HTP, the biological precursor of serotonin, at differing rates and intensities. Analogous differences were demonstrated in vim for the rate of serotonin synthesis by brain parts of cats (Bogdanski et al., 1958), and rabbits (Costa and Rinaldi, 1958) given 5-HTP. The results obtained in rabbits injected with 75mg/kg 5-HTP are reported in Fig. 1. The brain of these animals was perfused in situ with saline before dissection. Note that the increase of serotonin is greater and endures longer in the midbrain-diencephalon than in any other brain part analyzed. The increase in the cerebellum was delayed and fleeting. The serotonin level found in the hippocampus is greater than that of the other telencephalic structures combined. The serotonin increase shown by the midbrain-diencephalon of rabbits injected with 5-HTP (Fig. 1) becomes more interesting when one considers that in these structures the monoamine oxidase (MO) activity is greater than that of the other brain parts analyzed (Bogdanski and Udenfriend, 1956). ( e ) In the brain the serotonin turnover is very fast. In rats its half-life has been estimated as approximately 10 minutes. This rapid turnover contrasts with that of peripheral tissues, where turnover is several times slower (Udenfriend and Weissbach, 1958) . 2. In peripheral tissues correlations between serotonin and nervous activity have also been found: ( a ) Welsh (1957) analyzed the action of serotonin on the heart of molluscs and concluded that many of his findings support the hypothesis that either this amine or a chemical analog is a product of the cardioregulator nerves. ( b ) In the isolated guinea pig intestine, serotonin inhibits peristalsis when applied to the serous surface (Kosterlitz and Robinson, 1957). On the other hand, serotonin lowers the threshold of the pressure receptors for the peristaltic reflex when applied to the mucosa (Bulbring and Lin, 1958). Serotonin also modifies

* z 1000

8 m c 1

HOURS FOLUmtNG - 0 %TP INJECTION

BRAIN PART

I

2

3

-

TELENCEPHALON

$-P<

0

1

2

5

HIPPOCAMPUS

0

1

2

MIDfZtIN

DlENCEPHAUm

3

0

1

3

MEWLU-WNt

0

I

2

3

cERE(LLuM

.MI COMPARED TO CONTROLS.

I -STAMMRD

ERROR.

FIG. 1. Variations of serotonin content in parts of rabbit brain following 5-hydroxytryptophan ( 7 5 mg/kg intravenously).

180

ERMINIO COSTA

the thresholds for stimulation of sensory endings in the cardiopulmonary regions (Schneider and Yonkman, 1953 and 1954; Mott and Paintal, 1953) and those of the carotid sinus receptors as well ( Ginzel and Kottegoda, 1954). ( c ) Serotonin stimulates the superior cervical sympathetic ganglion and attaches itself to receptors which are different from those for acetylcholine and histamine ( Trendelenburg, 1956b). The nictitating membrane is stimulated by serotonin, and the transmission of impulses through the superior cervical sympathetic ganglion is facilitated ( Trendelenburg, 1956a; 1957; Thompson, 1958). Moreover denervation sensitizes the nictitating membrane to the effects of serotonin (Costa and Zetler, 1959). The nictitating membrane (Costa and Zetler, 1958, 1959) and other smooth muscles (Bulbring and Crema, 1958) are sensitized to the action of physiological neurotransmitters by serotonin. 3. Evidence furnished by psychopharmacological studies indicates that variations of brain serotonin concentrations are associated with behavioral effects. ( a ) Behavioral changes have been observed in rabbits, cats, and dogs given 5-HTP (Udenfriend et al., 1957a, b; Costa and Rinaldi, 1958). A selective increase of brain serotonin content causes locomotor disturbances in dogs (Spector et al., 1959). According to Bogdanski et al. (1958) depression follows small doses of 5-HTP whereas excitation characterizes the behavior of dogs injected with 40-60 mg/kg 5-HTP intravenously. In agreement with these findings, Costa et d. (1959a) have shown that the behavioral pattern evoked by intravenous injection of 5-HTP in dogs may vary according to the dose injected. Five to lOmg/kg caused sedation and tendency to sleep. Increasing the dose to 20 mg/kg caused sedation associated with panting, occasional hyperthermia, and extensor rigidity of the forelimbs. With 40mg/kg the initial sedation ensued for a shorter time, and 20 minutes after the injection the dogs were no longer quiet. They either stood inert or walked restlessly but showed interest in the environment. Later, despite the ataxia and a classic steppage gait, they displayed an obstinate aimless progression. At the climax of this peculiar behavior, the animals appeared indifferent to the environment, did not respond either to sound or light, failed to avoid obstacles as if blind, and finally, as they pushed with the head

THE ROLE OF SEROTONIN IN NEUROBIOLOGY

181

against the walls of the room and the fences of their enclosure, exhausted themselves in incoordinated efforts. Erections have been observed, sometimes with orgasms, as occasional components of this syndrome. More often expiratory dyspnea appeared. All dogs recovered within a few hours. Experiments with injections of 5-HTP into carotid and vertebral loops suggest that these symptoms originate from serotonin actions in the CNS (Costa et al., 1959a). These effects of 5-HTP in dogs are potentiated by premedication with MO inhibitors. ( b ) The tranquilization caused by reserpine injections takes place after brain serotonin concentrations are dramatically decreased (Shore et al., 1955; Shore d al., 1957). When serotonin depletion is reduced, for instance, by premedication with MO inhibitors, the intensity of reserpine tranquilization is weakened in proportion ( Brodie and Shore, 1957). However, serotonin depletion cannot be the only explanation for the sedation caused by this alkaloid, because in animals given reserpine both brain (Holzbauer and Vogt, 1956) and peripheral stores (Carlsson and Hillarp, 1956a) of catechol amines decrease in a similar pattern (Carlsson et al., 1957b). Moreover, certain Rauzoolfia alkaloids deplete selectively brain stores of norepinephrine (Paasonen and Dews, 1958). Pletscher ,et aZ., 1959, compared the effects of some reserpinelike drugs ( benzoquinolizine derivatives ) on brain amines concentration and behavior of mice. They found that 2-oxo-3-isobutyl-9,10-dimethoxy-1,3,4,6,7,-llb-hexahydro-2H-benzo[ a]quinolizine HCl, Tetrabenazine, depressed brain norepinephrine content to a greater extent than brain serotonin. Furthermore, this compound induces a reserpinelike sedation in the absence of a significant depletion of brain serotonin. Recent reports (Quinn et al., 1959) confirmed the greater effect of Tetrabenazine on brain norepinephrine than on brain serotonin, However the sedative effect of the drug parallels the changes in brain serotonin (Quinn et al., 1959). The importance of brain biogenic amine depletion in the mechanism of reserpine sedation is confirmed by the experiments with carboxysyringoylmethylreserpate ( SU 3118) and methyl-18-0( 3N,N-dimethylaminobenzoyl)reserpate ( SU 5171) ( Plummer et al., 1958; Barrett et al., 1958). The depletion of different biogenic amines caused by these drugs has been investigated with regard to peripheral and central pharmacological actions ( Hughes et al.,

182

ERMINIO COSTA

1960; Brodie et al., 1960). These experiments suggest that reserpine sedation is associated with changes in brain serotonin and not in brain norepinephrine (Brodie et al., 1960). The present understanding of the role played by the so-called neurohormones in brain mechanisms does not encourage definite conclusions. Moreover, interrelations between catechol amines and serotonin [compare 2( c ) ] are described in peripheral receptors (Costa and Zetler, 1959). Analogous interactions probably occur also in brain receptors. Thus, on the basis of the experimental evidence available, none of the neurohormones has been proved to have an action in reserpine sedation. This skepticism is encouraged by Zimmerman and Sheppard’s (1959) findings that, in guinea pigs, reserpine sedation occurs in absence of brain catechol amines depletion. Furthermore yohimbine depletes selectively brain catechol amines but does not cause sedation (Karki and Paasonen, 1959). ( c ) Two therapeutic procedures which are valuable in psychiatry cause concomitant modifications of brain serotonin metabolism. Electric shock therapy increases brain serotonin content (Garattini and Valzelli, 1957); insulin coma inhibits 5-HTP decarboxylase (Costa and Himwich, 1958) measured according to Gaddum and Giarman’s method (1956). This inhibition (Costa and Himwich, 1958) found in caudate and hypothalamus of rabbits during deep insulin coma is illustrated in Fig. 2. In these experiments the brain was removed after the hippocampus had shown convulsive activity for about 1837 (S.E., 1662-2031) seconds. However, many other biochemical changes occur in insulin coma. For instance, a marked depletion of catechol amines stored in rabbit brain has been described by Pscheidt and Himwich (1959). Hence the role played by the inhibition of 5-HTP decarboxylase in the mechanisms of the insulin shock therapy awaits further clarification. ( d ) Chlorpromazine antagonizes the peripheral effects of serotonin both in vioo (Table IV) and in vitro (Gyermek, 1955). In vitro the drug concentrations antagonizing serotonin effects on rat uterus are smaller than those able to interfere with the actions of any other known neurohumoral transmitter (Costa, 1956). ( e ) Certain convulsive agents are antagonized when brain serotonin concentrations are increased either by pharmacological inhibition of amine oxidase or by injections of the biological precursor (Kobinger, 1958; Prockop et al., 1959). In rabbits, when

CAUDATE

42.

36.-

36-

24.18 12-

6.

10

I

HYPOTHALAMUS

42-

30.-

x-

NUCLEUS

\ - ;\-

100

1.000

1o.Ooo

100

woo

10

241 18

I )

10

OF

FIG.2. Inhibition of 5-HTP decarboxylase of rabbit tissues after insulin; 0-0 = hypoglycemic (mean of 8 animals).

100

moo

10.1

TISSUE

= controls (mean of 10 animals),

184

ERMINIO COSTA

brain serotonin concentrations (Fig. 1) are increased as a consequence of intravenous injections of 5-HTP (75 mg/kg) the spontaneous EEG pattern becomes uniform: slow high-voltage waves are predominant, 14-per-second spindles tend to disappear, and hippocampal theta rhythm cannot be elicited by peripheral stimulation (Costa and Rinaldi, 1958), all changes indicating a sleep pattern. However, these EEG changes could be aspecific and, for instance, related to changes in cerebral blood flow. In order to exclude some of the possible interferences, a total of 44mg of 5-HTP was given to rabbits in four intracarotidal injections (Costa et al., 1959b). In these animals the vagi were sectioned, the carotid receptors were denervated, and the EEG was recorded from sensory-motor cortex, limbic cortex, and hippocampus. When 44 mg of the serotonin precursor are given by this route the increase of brain serotonin is greater than that obtained when 75mg/kg of the amino acid are injected intravenously. Immediately after the injections slow high voltage waves predominate. Twenty to forty minutes thereafter, when the increase of brain serotonin is maximal, the EEG in bilateral cortical leads is characterized by a lowvoltage fast activity and in the hippocampus by a theta rhythm. Thus the biphasic behavioral effects described in dogs as a function of the dose of 5-HTP injected is paralleled in rabbit by two distinct EEG patterns. As previously stated these bear a certain relationship to the brain serotonin content. Further investigation is in progress to characterize the mechanism of this biphasic EEG variation. Finally, an increase of brain serotonin content reduces the threshold of multisynaptic limbic pathways in pyramidal cats to electrical stimulation ( Costa and Revzin, unpublished observations). Lewis (1958) found that cortical potentials, evoked by peripheral stimulation in pentobarbitalized cats, are depressed by 5-HTP (30 mg/kg). This effect is potentiated by iproniazid and antagonized by reserpine. 4. The evidence that “5 Hydroxytryptamine pretreated erythrocytes lose potassium more slowly on cold storage than do controls,” suggested to Pickles ( 1956) that “. . . it (serotonin) plays some part, as the adrenocortical hormones may do, in the maintenance of the normal intracellular potassium content.” Because high serotonin concentrations were used, more work is needed in order to estab-

THE ROLE OF SEROTONIN IN NEUROBIOLOGY

185

lish the physiological importance of this finding. Nevertheless, one may mention that the effects of serotonin on superior cervical ganglion bear resemblances to those of KCl ( Trendelenburg, 1959). 5. The following remarks may be made in regard to the postulated role of serotonin in brain mechanisms: ( a ) The area postrema, which is formed mostly by glial cells, contains high serotonin concentrations (Amin et al., 1954). Therefore, serotonin is not necessarily concerned alone with physiological events taking place in neurons. It is felt that this objection is weakened since the enzyme synthesizing serotonin is not present in the area postrema (Gaddum and Giarman, 1956). In this particular instance, serotonin might be blood-borne. Moreover, the probable close functional interrelationship of neurons and glia must always be kept in mind. Nevertheless, should the capacity to store serotonin be a generalized property of glial cells, any attempt to interpret the neurophysiological importance of serotonin on the basis of quantitative analysis of serotonin contained in brain structures would be complicated. In nervous tissue the proportions of glial cells to neurons is approximately 10 to 1 (Hyden, 1958). ( b ) Patients with carcinoid tumors which have a high serotonin content (Lembeck, 1953) show a conspicuous increase of blood serotonin concentrations ( Lennard-Jones and Snow, 1956). Despite this fact these patients do not present any psychological or neurological impairment. Interchange between serotonin synthesized in central nervous system and in peripheral tissues does not occur readily (Woolley and Shaw, 1957; Costa and Aprison, 1958b); thus the absence of behavioral disturbances in patients with carcinoid tumors cannot be taken as an argument against the postulated neurophysiological role of serotonin.

111.

Biosynthesis of Serotonin

Serotonin is synthesized at different rates in various brain structures (Bogdanski et al., 1957). This may be an important correlate in favor of the hypothesis that serotonin plays a neurophysiological role. However the significance of this finding has been discussed with regard to the specificity of 5-HTP decarboxylase. With this in mind, a few problems require further explanation:

186

ERMINIO COSTA

1. Sympathetic ganglia which do not contain serotonin (Gaddum and Paasonen, 1955) have been shown to decarboxylate 5-HTP ( Gaddum and Giarman, 1956). 2. Pheochromocytomata contain catechol amines but not serotonin and still decarboxylate 5-HTP ( Westermann &. al., 1958). 3. 5-HTP decarboxylase is inhibited by a-methyldihydroxyphenylalanine ( Westermann et nl., 1958) which is a competitive

VSUBSTRATE (5-HTP)

PER GRAM OF TISSUE

FIG.3. 5-HTP decarboxylase activities of rabbit tissues at different sub= kidney, o- -o = hypothalamus, a-a = strate concentrations; xcaudate, 0- - - -0 = thalamus, A---A = mesencephalon, unfilled hexagons = hippocampus.

inhibitor of the apoenzyme of dihydroxyphenylalanine ( dopa ) decarboxylase (Sourkes, 1954), also present in nervous tissue (Holtz and Westermann, 1956). As shown in Fig. 3, aqueous homogenates of various portions of rabbit brain metabolize 5-HTP at different rates. In this experiment the enzymatic activity was evaluated by analyzing the serotonin formed according to the bioassay technique of Gaddum and Giarman ( 1956). The percentage of substrate metabolized by

THE ROLE OF SEROTONIN I N NEUROBIOLOGY

187

homogenates of different rabbit tissues is calculated by means of the increases of serotonin found after 1hour of incubation (Fig. 3 ) . This value is plotted against the logarithm of the substrate concentrations given in micrograms of 5-HTP per gram of tissue. On a percentage basis the rate of decarboxylation decreases with greater concentrations of substrate. It will be noted, that the amount of serotonin formed actually increases throughout, although the rate of synthesis slowed with the highest concentrations. A high 5-HTP decarboxylase activity occurs in both caudate and hypothalamus. The amine content of the caudate is predominantly dopamine ( Bertler and Rosengren, 1959), that of the hypothalamus, serotonin (Amin et al., 1954), and norepinephrine (Vogt, 1954). The simplest explanation for these findings could be that dopa and 5-HTP decarboxylase are identical but operate on different substrates in the various nuclei of the brain. This hypothesis is suggested by Holtz and his group (Westermann et al., 1958). The ratio between amounts of dopa and 5-HTP decarboxylated by homogenates of various organs is, however, different. Nevertheless, as Westermann et al. (1958) demonstrated, these differences do not support the existence of two different enzymatic systems, for the ratios become uniform when the pyridoxal phosphate concentrations are properly balanced. However, according to Udenfriend ( 1958), the enzyme decarboxylating 5-hydroxytryptophan has a high degree of specificity for the substrate. 2-Carbethoxy-~~-tryptophan, 2-hydroxy-~~-tryptophan, and 2,5-dihydroxy-~~-tryptophan do not serve as substrates for 5-HTP decarboxylase (Freter et al., 1958). Indeed these tryptophan analogs inhibit the decarboxylase in uitro. In conclusion, the distinction between 5-HTP and dopa decarboxylases of mammalian neural tissues appears controversial. The proposals of Clark et al. (1954) is supported by the different sensitivity of the two enzymes to the inhibitory effect of phenylalanine metabolites (Davison and Sandler, 1958) and is opposed by the report of Westermann et al. (1958). A reciprocal inhibition of the two decarboxylases by high concentrations of dopa and 5-HTP has been demonstrated by Yuwiler et al. (1959). Many discrepancies in the field might derive from differences in methods. For example: 1. Manometric and biological methods require different sub-

188

ERMINIO COSTA

strate concentrations. Greater substrate concentrations are required by the former. This fact could account for peculiar enzymatic behavior deriving from differences in the reactivity of the two substrates with coenzyme. 2. As Weissbach et al. (1957) point out, the freezing and thawing of a tissue is important in regard to the pyridoxal requirements of the decarboxylase assayed in vitro. Therefore the variations found by Gaddum and Giarman (1956) in the 5-HTP decar-

..

100

r OF

1o.OOo

1000

5-HTP

PER G

OF TISSUE

5-HTP decarboxylase of brain homogenate of different animals; 0-0 = rabbit caudatus, x= mice brain “in toto.” The two points for each experiment indicate, respectively, the results obtained with 6 x 10-6 and 1.2 x 10-5 gm/ml pyridoxal phosphate. FIG. 4.

A-A

= rat brain “in toto,”

boxylase activity of analogous tissues of different animal species might bear a relationship to the amount of pyridoxal available in the different tissue. We have confirmed the existence of these differences and demonstrated that they are not related to differences in availability of coenzyme. The rate of 5-HTP decarboxylation by homogenized rat brain is slower than that for mouse brain and rabbit caudate nucleus (Fig. 4). These differences are independent of the pyridoxal concentrations present. In these experiments the enzymatic activity was measured by the technique described for Figs. 2 and 3. However, the enzymatic activity in Fig. 4 is reported

THE ROLE OF SEROTONIN IN NEUROBIOLOGY

189

as micrograms of serotonin synthesized in 1 hour by 1gm of tissue. In other experiments not reported in ext,enso in this paper we found that concentrations up to 1 x 10-3 gm/ml of adenosine triphosphate ( ATP) did not modify the 5-HTP decarboxylation ocgm/ml of pyridoxal phosphate. curring in the presence of 6 x Furthermore, in agreement with reports of Buzard and Nytch (1957) on the enzyme of rat kidney, a homogenate of frozen and thawed rat brain did not appreciably decarboxylate 5-HTP if only ATP was added. 3. It is felt that data obtained with purified enzymatic preparations as those of Clark et al. (1954) cannot be compared with experiments performed with tissue homogenates. Any discussion of these experiments must be based on data obtained with identical enzymatic preparations. Another debated problem of serotonin biosynthesis concerns the hydroxylation of the indole nucleus ( Udenfriend, 1958). At present it is certain that tryptamine is not hydroxylated in vitro to form serotonin (Jepson .et al., 1959). Although 5-HTP has never been found in normal body fluids, this amino acid is sometimes present in the urines of patients with carcinoid tumors (Sandler and Snow, 1958). Zbinden ct al. (1958) found that different species of mammals on tryptophan-deficient diets show a decrease of brain and intestine serotonin content. Paasonen and Giarman (1958) found a slight increase of serotonin in the brain of rats injected with L-tryptophan. We found a significant increase of brain serotonin content in rats injected with 1gm/kg of L-tryptophan. The serotonin content of both spleen and intestine were not increased. In both experiments the analysis was performed using bioassay techniques. Paasonen and Giarman (1958) used the heart of Venus mmcwriu, while we employed the rat uterus of spayed rats. Recent reports by Hess et al. ( 1959) question the specificity of the bioassay procedure for analysis of tissues of animals given tryptophan. In fact these authors found that the concentrations of tryptamine in brains of guinea pigs given 800mg/kg of tryptophan, were raised to 1-2 &gm, concentrations which are 100 times greater than control values. Tryptamine may induce uterine contractions but it is 100 to 200 times less active than serotonin (Erspamer, 1954b; Vane, 1959). Furthermore, if brain homogenates are suspended in the

190

ERMINIO COSTA

presence of concentrations of L-tryptophan up to 5 x gm/ml [according to the method of Gaddum and Giarman (1956) for 5-HTP decarboxylase], 1 hour of incubation, at 37" C, produced no substances eliciting contractions of the rat uterus despite the decarboxylation of tryptophan. The conclusion that serotonin brain concentrations are significantly increased in rats given 1 gm/kg of L-tryptophan, would therefore appear possible because the eventual increase of tryptamine does not seem to interfere with the bioassay. Further experimentation is in progress to verify this conclusion. In fact, a differentiation between tryptamine and serotonin is possible; the biological actions of the former are potentiated by MO inhibitors whereas those of serotonin are not (Vane, 1959).

IV.

Mechanism of Action of Neurotransmitters

Synaptic or conductile phenomena in cells are associated with changes of membrane ion flux, detectable as electrical potentials. At peripheral sites ion fluxes are dramatically altered by acetylcholine release. This emphasizes the possibility that chemical agents implicated in transmission of nervous stimuli may act through a modification of membrane permeability to electrolytes ( Shanes, 1958a, b ) . The nature of the changes brought about by the transmitter in postsynaptic membrane is not known. Among the hypothesis formulated is the assumption of the existence of specific membrane receptors for specific neurotransmitters. In the case of acetylcholine, Zupancic ( 1953) theorized that the receptor sites were identical with cholinesterase, but Fatt ( 1954) thinks that this hypothesis is improbable. Although a strict identity as proposed by Zupancic is not yet demonstrated, the cellular location of enzymes implicated in the catabolism of a chemical mediator of neuronal events might be an indicator of its site of action. In fact, according to some recent theories, the rate of destruction of a neurotransmitter is a determinant of the speed of neuronal events. The present status of development of our understanding of neurophysiological problems does not ascribe a definite role to intracellular structures in the phenomena related to the excitability of nerve cells. A great functional significance is instead attributed

THE ROLE OF SEROTONIN IN NEUROBIOLOGY

191

to the surface of neuronal membranes, which are regarded as the exclusive site of action of neurohormones. This concept is the result of a generalization of the findings related to acetylcholine transmission and may not apply to other neurohormones such as serotonin. In fact, should MO have an analogous physiological significance to cholinesterase, it will be difficult to explain why this enzyme is not present in neuronal membranes and instead is exclusively concentrated in mitochondria located inside of nerve cells (Azioka and Tanimukai, 1957). The logical assumption is that serotonin must enter the cell in order to be destroyed by the enzyme which controls the speed of neuronal events mediated by this amine. Consequently intracellular structures, such as mitochondria, would control nervous events in systems mediated by serotonin. Electrical potentials evoked in the basal hippocampus of pyramidal cats by amygdala stimulation are inhibited, reversibly, by injections of 5-HTP (Revzin and Costa, 1959). The intensity of this inhibition is proportional to the amount of amino acid injected and is potentiated by a pretreatment with MO inhibitors. However, when the brain serotonin concentrations are increased by injections of iproniazid the hippocampal evoked potentials remain unchanged. Many hypotheses may be advanced to explain this discrepancy between biochemical and electrical findings. For instance, it is conceivable that MO is not related to serotonin receptors as intimately as cholinesterase is to acetylcholine receptors. This different arrangement might explain why cholinesterase inhibition causes a mimicry of acetylcholine effects whereas MO inhibition does not do the same for serotonin. However, at present, one can only report experimental findings and withhold deductions pending more extensive experimental reports on the effects of different MO inhibitors. In nerve cells the response to stimuli does not occur as a passive phenomenon. Eccles (1957) states that the rate of firing of multipolar neurons does not strictly correspond to the number of depolarizing events occurring in their dendritic synapses. In this respect the location of the synaptic knob is important. The distance of the synapses from the spike-generating membrane is among the factors determining the probability that a synaptic event will trigger a spike discharge. However, this does not mean

192

ERMINIO COSTA

that depolarization of synapses in apical dendrites is of little importance. Synaptic events in apical dendrites, although not capable of triggering a spike, might still be a conditioning factor for cellular responses derived from synaptic stimuli which impinge upon the soma. If this interpretation is correct, apical dendrites exert a direct control on the excitability of neurons for, as suggested by Eccles (1957), they are part of the mechanism which maintains ionic equilibria in an active neuron. Special localized characteristics of the membrane permeability might be involved in this control of the intracellular ion concentration. However, the fine correlations found in the physiological performance of neurons suggest that changes in membrane permeability are not the only regulatory mechanisms. This conclusion is indirectly supported by the generalized reluctance to accept chemical transmission of nervous impulses across the synapses as the only possible mechanism involved in the transmission of information between neurons. Nonsynaptic biochemical factors must be considered as the determinants of neuron excitability. As Bishop ( 1956) suggested the phylogenetic evolution of the characteristics of nerve cells has been accomplished through selective variations in the permeability of portions of the cell membrane. The synaptic portions of the membrane must act on the spike-generating portion, whose excitability is determined, in part, by intracellular ion concentrations. This latter, in turn, may be regulated by the intracellular “pumps” of Eccles (1957) or by some other active process. It is proposed that where serotonin is active in the nervous system, it performs its role by virtue of some intracellular ion regulating mechanism. That is, the intracellular balance of ions in selected neurons is regulated at any moment of physiological activity by a biochemical system in which serotonin is important. This suggestion is supported by the physicochemical characteristic of serotonin which retards passage through cellular membranes (Vane, 1959). Pickles (1957) suggested that serotonin retards potassium movements and that the passive diffusion of sodium and potassium through both toad’s and frogs skin is slowed or accelerated by serotonin according to the previous history of the experimental system. This effect is shared by tryptamine, but not by histamine or bufotenine. Woolley ( 1957) has also recently speculated that serotonin controls cellular ion permeability. Trendelenburg ( 1959) emphasized the similar-

THE ROLE OF SEROTONIN I N NEUROBIOLOGY

193

ities of the effects of potassium and serotonin on superior cervical ganglion. Histochemical evidence suggests that structures with high serotonin concentrations such as the supraoptic, paraventricular, and suprachiasmatic nuclei contain cholinergic cells as well (Abrahams et al., 1957). Moreover, in rats, the importance of cholinergic systems in hypothalamic physiology has been histochemically demonstrated (Pepler and Pearse, 1957). Hence, the peculiarity of the hypothalamus is not the absence of cholinergic transmission but rather the presence of high concentrations of amines which in the periphery are concerned with nervous mechanisms. Summary. If serotonin is a neurohormone related to control of neuronal events, its function differs from that postulated for acetylcholine, and the two systems may coexist in the same nervous structures. This hypothesis is supported by the following information. 1. The morphological differences between the peripheral sites of action of acetylcholine and serotonin. 2. Serotonin is metabolized by MO which is exclusively located in intracellular structures and is not present in cell membranes. On the contrary, cholinesterase is located in external surfaces of neuronal membranes as well as within the neurons. 3. An inhibition of MO by iproniazid does not offer mimicry of the effects of serotonin as shown by the comparison on hippocampal evoked potentials (Revzin and Costa, 1959). In the case of acetylcholine, such a discrepancy would be regarded as exceptional. However, it must be mentioned that inhibitors of MO block transmission in the ganglia (Gertner, 1959) and potentiate the effects of 5-HTP (Udenfriend et aL, 1957b; Costa & al., 1959a), tryptamine (Tedeschi et a,!., 1959b), and dopamine (Goldberg, 1959). Indeed these findings demonstrate the physiological importance of MO and its relationships with biologically active amines normally present in brain structures of mammals.

V. Mechanism of Serotonin Action at Cellular Level A. IMPORTANCE OF MONOAMINE OXLDASE (MO) McIsaac and Page ( 1958), as a result of urine analyses of patients with carcinoid tumors, suggested that N-acetylation or

194

ERMINIO COSTA

conjugation with glycuronic acid are possible catabolic pathways for serotonin. Probably these degradation mechanisms are not important in regard to the neurophysiological actions of serotonin for both Blaschko (1952) and Davison (1958) suggested that indolalkylamines are physiological substrates for MO. This enzymatic activity is a rather generalized property of nerve cells and is unrelated to the cellular serotonin contents (Bogdanski et al., 1958). The presence of MO in neurons may bear a relationship to serotonin and other biologically important amines as well (Davison, 1958). A MO specific for serotonin might be differentiated from the group of enzymes listed as amine oxidase. However this differentiation is not possible on the basis of substrate specificity as serotonin is oxidized by MO and to a lesser extent by diamine oxidases (Blaschko et al., 1959). These two enzymes are differentiated by virtue of their different sensitivity to semicarbazide inhibition. As reported by Zeller ( 1938a, b ) , Werle ( 1940), and Blaschko et al. (1959) this inhibitor differentiates the enzymes on the basis of the presence of the carbonyl residue in the enzymatic reactive site. Other differences, in addition to carbonyl residue, may have a physiological significance. In fact it has been proposed that MO is a complex biological system involving a dehydrogenase linked to a respiratory enzyme chain (Davison, 1958). Thus, numerous different enzymatic sites may characterize the selectivity of enzymatic action. Therefore an extensive study of different MO inhibitors may lead to a better differentiation and classification of MO. Amine oxidases of different biological sources offer a certain degree of specificity toward inhibitors (Corne and Graham, 1957) as well as toward substrates (Zeller et al., 1958). The same is true for MO in regard to (Table I ) iproniazid inhibition (Zeller et al., 1955). The enzyme present in rat brain homogenates is more susceptible than that of the rabbit mesencephalon to the in vitro inhibition by iproniazid phosphate (Table I). Analogous differences have been confirmed with the in vivo experiments on rats (Fig. 5), rabbits (Fig. 6), and dogs (Himwich et al., 1959). Rat brain was analyzed after the cerebellum had been dissected out, using the bioassay technique proposed by Garven ( 1956). The onset of brain serotonin increase in rats chronically injected with iproniazid (Fig. 5) precedes that of spleen, lung, intestine, and kidney.

THE ROLE OF SEROTONIN IN NEUROBIOLOGY

195

However, 72 hours after the beginning of the iproniazid medication the percentile serotonin increases of brain, lung, and spleen are comparable. The increases found in kidney and intestine are significantly smaller throughout the experiment. The results concerning the spleen must be regarded with caution; this tissue does not contain MO (Davison, 1958), and the serotonin of spleen is contained exclusively in the platelets (Humphrey and Jaques, 1954) where it is stored in a diffusible form (Sano et al., 1958). The differences found between tissues other than spleen might derive from specific differences in rate of serotonin turnover (Udenfriend and Weissbach, 1958). TABLE I In Vitro INHIBITION OF RABBITS AND RATSBRAIN AMINEOXIDASE BY IPRONIAZID PHOSPHATE^ Source of the enzyme

ED,"

Fiducial limits (ED,, 1 3.3-9 x 1 0 4 ~ 4-9.2 x 1 0 - 3 ~

7.3 x 1 0 - 4 ~ 7.3 x 10-siv a Homogenates prepared with H,O (1 x 10-3 w/v); serotonin creatinine sulfate 4 x 10-6 gm/ml. Iproniazid was incubated for 20 minutes at 38" C, pH 8 (phosphate buffer according to SZrenson). Enzymatic activity evaluated by measuring serotonin disappearance in 30 minutes at 38". Serotonin analysis carried out according to Garven's biological method (Garven 1956). Statistical analysis according to Stone and Loew ( 1952). Rat brain Rabbit mesencephalon

Increases of serotonin develop more slowly in brain parts of rabbits chronically premedicated with iproniazid ( Fig. 6 ) than in the CNS of rats (Fig. 5 ) . Moreover, on percentile basis, the serotonin increase is smaller in brain parts of rabbits than in rat brain. The serotonin content of different parts of dog brain is also increased by iproniazid medication ( Himwich et al., 1959). This species is less affected by iproniazid than either rats or rabbits. Greater doses (25 mg/kg, twice daily, repeated for at least 4 days) are required to produce a serotonin increase comparable to that found in rats. These species differences may be related to peculiarities in drug distribution, and in the rate of brain serotonin turnover. However, the results of the in vitro experiments (Table I ) weaken the value of these hypotheses and suggest that a different susceptibility to iproniazid distinguishes the MO present in rat and rabbit brain.

SPLEEN

-*

FIG.5. Percentual variations of serotonin contents in tissues of rats chronically injected with iproniazid. A dose of 25 mg/kg iproniazid was injected at 0 time and every 24 hours thereafter. Each group of animals ( n = 6 ) was sacrificed 24 hours after last injection. Ordinates: percentual changes in serotonin contents. Abscissas: hours after injection.

FIG.6. Percentual variations of serotonin contents in brain parts of rabbits chronically injected with iproniazid. A dose of 25 mg/kg of iproniazid was injected at 0 time and every 24 hours thereafter. Each group of animals ( n = 8 ) was sacrificed 24 hours after the last injection of iproniazid. Ordinates: percentual changes in serotonin contents. Abscissas: hours after injection.

198

ERMINIO COSTA

The MO inhibitory action of iproniazid stems from a practically irreversible reaction between the drug and the enzyme. This reaction involves a dehydrogenation of the inhibitor ( Davison, 1957). The irreversibility of the process is documented by the report of Hess 'et al. (1958). In rats injected intraperitoneally with 195 mg/kg of iproniazid 24 hours prior to the analysis of brain MO, the enzyme was found to be inhibited although the isonicotinic moiety was no longer detectable in the tissue. Isopropylhydrazide is probably the active group responsible for the MO inhibitory action of iproniazid and chemically related compounds as well. When this group is attached to molecular carriers other than isonicotinic acid, the antienzymatic effect is not modified provided that the carrier can be easily transported across the cell membranes (Pletscher and Gey, 1958). However, variations in the degradation of the compound might be expected as a result of the properties of the carrier. According to Horita et al. (1959), p-phenylisopropylhydrazine (JB-516) is more selective for brain than is iproniazid. The different distribution might explain why JB-516 is a faster acting cerebral MO inhibitor than iproniazid (Horita, 1958). The time course of the action of JB-516 is illustrated by the experiments reported in Figure 7. With 100mg/kg of JB-516 an almost maximal increase of brain serotonin occurs within 30 minutes. These results were reported, although these doses of JB-516 were lethal within 4 hours because this finding confirms the fast brain serotonin turnover reported by Udenfriend and Weissbach ( 1958). An intraperitoneal dose of lOmg/kg of JB-516 causes the brain serotonin to increase within 6 hours (Fig. 7 ) . The greatest increase was observed at 12 hours and the serotonin level declines thereafter. Spector et al. (1959) report that brain norepinephrine content of cats given JB-516 remains unchanged. This finding is of extreme interest because it lends support to the hypothesis that the enzyme controlling norepinephrine catabolism in cat brain may indeed be different from that which oxidizes serotonin. Trans-2-phenylcyclopropylamine( SKF 385 ) ,which is chemically unrelated to iproniazid, causes a conspicuous increase of brain serotonin content in rats (Fig. 8 ) . This effect has a short latent period and lasts longer than 24 hours. Serotonin levels in other organs are increased but to a lesser extent. Dogs given SKF-trans385 evince a brain serotonin increase as well (Himwich et al., 1959).

199

THE ROLE OF SEROTONIN IN NEUROBIOLOGY

The following findings suggest that this increase in brain serotonin stems from an action on serotonin metabolism: 1. SKF 385 inhibits in vitro the serotonin degradation by rat brain homogenates ( Maas and Nimmo, 1959).

T

300

250 200

I50

0

o

5

10

IS

20

25

30

35

40

45

50

FIG.7. Percentual increase of brain serotonin content of rats injected with betaphenylisopropylhydrazine (JB-516); 04 = JB-516 ( 100 mg/kg intraperitoneally), 0-0 = JB-516 (10 mg/kg intraperitoneally). Each point is a mean of 6 animals. Ordinate: percentual variations of brain serotonin content. Abscissa: hours after injection. Controls ( n = 12) = 100; serotonin = 0.197 & 0.015 pg/gm.

2. SKF 385, like iproniazid and JB-516, increases the toxicity of tryptamine (Tedeschi et al., 195913). 3. Reserpine sedation fails to occur in rabbits prernedicated with SKF 385. Moreover when SKF 385 is injected in rabbits pretreated with reserpine the sedation is promptly relieved (Tedeschi et al., 1959a). It is known that iproniazid and its analogs

200

ERMINIO COSTA

antagonize resperine sedation only if they are given prior to the alkaloid ( Brodie and Shore, 1957). 4. In rats chronically premedicated with SKF 385, both brain serotonin depletion and sedation caused by reserpine are reduced or antagonized (Fig. 9). The sedation was measured 24 hours after 400

\

\ \p‘

m..

200

a

COMROL f N . l Z b I O 0 SEROTONlN..197* ,015

8 /g

04 0

20

10

30

40

I

I

OROINATE- FfRCENTLUL CHANGES IN SEROTONIN CONTENTS. ABSClSSA*HCURS AFTER INJECTION. V-v*AHfWETAMINE

SULFATE f5mg

/kg i.p.1

0 - 0 * 1 r o n r - Z PHENVLCKLOPROPVLAMINE (SnF tranr365J -Cis-.?

HCI.

PHENVLCVCLOPROPYUMINE HCI fSKF C i r 3 8 5 J

EACH POINT IS A MEAN O f 6 ANIUM.5.

FIG.8. Increase of serotonin content of rat brain following SKF 38.5 (2.4 mg/kg intraperitoneally ) .

the second reserpine dose by the degree of palpebral ptosis as suggested by Rubin et al. (1957). In Fig. 9 a fully open eye is rated 0, whereas a maximal ptosis is rated 8. The animals were sacrifhed for brain serotonin analysis 2 hours after the last dose of MO inhibitor. The intensity of the antireserpine effect of SKF 385 is correlated with brain serotonin content. In Fig. 9 the antagonistic

P

300

M

250 -0 -I

200-

-2

150-

-3 -4

100-

-5

RATS G M N ONLY KSEW'INE

1.0

RESERPINE Z.Smg/kg i.p. AT 24 AND 32 HOURS. MAol AT 0 AND EVERY 1 2 HOURS THEREAFTER. EACH POINT IS AN AVERAGE O f 6 ANIMALS.

FIG.9. Brain serotonin content of reserpinized rats prernedicated with MO inhibitors.

202

ERMINIO COSTA

action of SKF 385 toward reserpine is compared to that of iproniazid and other MO inhibitors (Pletscher and Gey, 1958). On a molar basis SKF-trans-385 is at least ten times more antagonistic than other MO inhibitors tested against brain serotonin depletion and palpebral ptosis induced by reserpine. Results of preliminary clinical trials suggest that SKF 385 has amphetaminelike properties and these may be relevant for its antireserpine action. However, as postulated by Brodie et al. (1959), a stimulant action is also related to the degree of MO inhibition. According to our experience, doses of SKF 385, up to 3mg/kg, inhibit brain MO, but do not increase motor activity of dogs, cats, rabbits, rats, and mice. The interrelations between chemical structure and brain serotonin increase for analogs of SKF 385 have been studied. The position of both phenyl and amino group with respect to the cyclopropane ring are important; the cis form is less active than the trans form (Fig. 8). The lack of action of amphetamine sulfate injected in doses greater than those of SKF 385 emphasizes the importance of the cyclopropane ring for this pharmacological effect. Experiments are now in progress to determine the effects of SKF 385 on brain catechol amine concentrations. The cellular location of MO has a definite importance in regard to its neurophysiological function. At present it is believed that this enzyme is exclusively located in mitochondria (Hawkins, 1952; Cotzias and Dole, 1951). Furthermore, histochemical techniques suggest that MO is more abundant in neuronal cell bodies than in nerve fibers ( Azioka and Tanimukai, 1957). The role of MO in the physiological action of serotonin has not been elucidated by pharmacological experiments. In fact, only the toxic effects of serotonin are clearly potentiated by MO inhibition ( Billa and Valzelli, 1958). The pharmacological actions of both serotonin and catechol amines in animals given MO inhibitors are either slightly facilitated or remain unaltered (Bnlzer and Holtz, 1956; Corne and Graham, 1957; Himwich et al., 1959). In contrast, the behavioral effects of biological precursors of both serotonin and catechol amines are enhanced by a premedication with MO inhibitors (Costa et al., 1959a). Thus, the effects of blood-borne serotonin are differentiated from those of the amine synthesized and stabilized in tissues by MO inhibitors. These observations suggest that serotonin enters the cells slowly and support

THE ROLE OF SEROTONIN IN NEUROBIOLOGY

203

the working hypothesis that the locus of action of serotonin is intracellular. B. ACTIONAT RECEPTORSITES The broad variety of physiological systems influenced by serotonin together with the small doses effective both in Vztro and in vivo (Page, 1954, 1958; Erspamer, 1954a) might be an indication of the physiological character of serotonin action. However none seems relevant to the problem of how the compound acts at the synapses, if indeed it does. Valuable indirect evidence regarding the mechanisms of neurotransmitter action can be obtained by studying the effects of their specific inhibitors in various systems. When this approach is applied to serotonin the interpretation of the results becomes difficult. Subsidiary assumptions are necessary in order to offer rational interpretations. Gaddum (1957), as a result of a similar analysis, proposed two types of serotonin receptors; “M” receptors inactivated by morphine and cocaine, and “D” receptors antagonized by lysergic acid diethylamide (LSD) and Dibenzyline. The interpretation of the results obtained with the pharmacological antagonists of serotonin is difficult; in this respect, serotonin resembles epinephrine. Many other analogies can be found between the biological characteristics of epinephrine and serotonin. ( a ) Both are stored in the large granule fraction, which are intracellular structures distinct from mitochondria (Blaschko and Welch, 1953; Blaschko et al., 1957; Baker, 1959) and are displaced from their stores in hypotonic media (Blaschko and Welch, 1953, Hillarp et al., 1953; Baker, 1958). These amine-carrying granules differ from the storage sites of acetylcholine (Hebb and Whittaker, 1958; Baker, 1958). ( b ) Serotonin and catechol amines require the presence of ATP in their cellular binding sites (Carlsson and Hillarp, 1956b; Born et d., 1958b). A typical example is the subcellular fraction of adrenal medulla which stores epinephrine and contains large amounts of ATP (Hillarp et al., 1955). ( c ) Platelets take up both serotonin and epinephrine ( Hardisty and Stacey, 1955; Weil-Malherbe and Bone, 1958; Born et al., 1958a). Epinephrine in adrenal medulla (Blaschko et al., 1956) and

204

ERMINIO COSTA

serotonin in cells may be bound to ATP by ionic bonds (Born and Gillson, 1957). However, it is possible that in analogy with epinephrine (Born et al., 1958a) chemical combination between ATP and serotonin is not uniform for all tissues. This is emphasized by the report of Sano et al. (1958) on the chemical characteristics of serotonin found in platelets. Moreover, Hughes et al. (1958) suggest that the function of ATP may be to furnish the energy required by specific mechanisms concerned with the active transfer of serotonin. The importance of this transfer mechanism is emphasized by in vitro experiments (Vane, 1959; Hughes and Brodie, 1959) . The report of Marrazzi and Hart (1955) that minute intracarotid doses of serotonin inhibit the transcallosal response of pentobarbitalized cats cannot be taken as proof of the synaptic action of serotonin. The interpretation of the surface negative evoked potential is controversial at best (Bishop and Clare, 1953; Purpura and Grundfest, 1956; Von Euler and Ricci, 1958). Dendritic synaptic potentials and cortical surface negative electrical potentials are not identical; therefore, a reduction in the voltage of these potentials is not sufficient proof of an inhibition of synaptic transmission. Furthermore, the peripheral autonomic innervation of the preparation used by Marrazzi and Hart (1955) was intact. Thus, indirect but important effects evoked by serotonin through peripheral stimulation cannot be ruled out (Koella et al., 1959). We found that hippocampal potentials produced by amygdala stimulation are dramatically reduced by intravenous serotonin injections ( Revzin and Costa, 1959). However, this serotonin effect does not take place in bilaterally vagotomized cats. Finally, the importance of Marrazzi’s results would be clarified if the existence of a blood-brain barrier against serotonin could be excluded. On the contrary, the experimental evidence is in favor of a slow diffusion of serotonin from blood stream to nervous tissue (Woolley and Shaw, 1957). Cortical and hippocampal evoked potentials are indeed inhibited in cats given 5-HTP (Lewis, 1958; Revzin and Costa, 1959). This evidence suggests that serotonin synthesized in the brain modifies electrical evoked potentials in cerebral structures, but it does not offer any explanation in regard to the type of influence exerted by serotonin. In contrast to these inconclusive findings the physiological

THE ROLE OF SEROTONIN IN NEUROBIOLOGY

205

importance of serotonin is elucidated by the interrelations between drug action and brain serotonin contents. We have evidence that drugs may determine: (a)an increase (Udenfriend et al., 1957a, b ) or a decrease of serotonin production (Westermann et al., 1958), ( b ) a modification of binding processes of neurohormones (Shore et al., 1955; Pletscher, 1957a), ( c ) a decrease of serotonin destruction ( Spector et al., 1958). Thus, the correlations between changes in brain serotonin induced by drugs and their specific pharmacological actions provide tools for understanding the possible role of serotonin in the nervous system. VI.

Methods for Serotonin Bioassay

Physicochemical ( Weissbach et al., 1958), chemical (Udenfriend et al., 1955), and biological methods (Garven, 1956) have been elaborated for the determination of serotonin present in biological material. Although the specificity of these methods may be considered satisfactory, a differential analysis of bound and free serotonin in tissues has been neglected. While physicochemical methods employ aqueous homogenates of tissues, biological methods ( Garven, 1956; Correale, 1956) use different concentrations of acetone to extract serotonin from tissues. As Correale (1956) points out, extracts made with 80% acetone and tested on the rat uterus are not reliable because of interference by the polypeptide substance P (Amin et al., 1954) which also causes contraction of the rat uterus. Moreover, if the extracts are not purified with petroleum ether another source of error may be found in the presence of a naturally occurring lipid-soluble acid (Vogt, 1958). Correale ( 1958) assays serotonin in aqueous extracts of tissues digested with chymotrypsin in order to destroy substance P. In our experience this method should be modified in order to avoid interference by enzymes catabolizing the seorotonin present in the samples analyzed. Our modification is as follows: The tissues are frozen immediately after removal. They are weighed while frozen and homogenized 1/5 (w/v) in an aqueous solution of JB-516 ( 2 X gm/ml) in order to inhibit the MO (Horita, 1958). One and one-half milliliters of pH 7 phosphate buffer (SZrenson) are added to 1 m l of homogenate and 0.5 ml of chymotrypsin (Worthington) 2.5 x

206

ERMINIO COSTA

gm/ml in saline. The mixture is shaken in a Dubnoff metabolic incubator for 2 hours at 38' C in order to inactivate substance P. The mixture is transferred into a boiling (100" C ) water bath for 5 minutes. Acetone (16ml) is added, and the samples stored overnight in a cold room (4" C ) . The supernatant is filtered through No. 43 Whatman paper previously washed with acetone. The precipitate is re-extracted twice for 1 hour with 10mI of acetone, filtered, and the filtrates are combined. The fluids are evaporated at 30" C in vacuum. Finally the extracts are reconstituted, appropriately diluted with DeJalon fluid, and tested on the isolated uterus of a spayed rat. The bioassay is carried out as follows: The uterus of a spayed rat premedicated with estrogens is suspended in aerated DeJalon fluid at 28-30' C. In this condition the relationship between height of uterine contractions and the logarithm of appropriate serotonin concentrations is linear. Two dilutions of the unknown are always bracketed between four different concentrations of selected serotonin standards. Calculations are made by comparing the biological responses induced by each of the two dilutions of the unknown with the dose response curve exhibited by the rat uterus before and after the assay of each tissue extract. When required, the standards are prepared with brain extracts of suitable control animals in which the natural level of serotonin was inactivated. An example of serotonin recovery with this method is given in Table 11. Note that chymotrypsin does not affect serotonin recovery. The enzyme TABLE I1 RECOVERY OF SEROTONIN~ AFTER INCUBATION OF HOMOGENATES FOR 120 MINUTES^ Brain extract Without chymotrypsin Without chymotrypsin plus 250 ng serotonin With chymotrypsin With chymotrypsin plus 250 ng serotonin a b

c

Serotonin recovered (ngc/gm wet wt.) -~ Mean Range 675

645-707

986 539

877-1109 492-591

809

783-837

Serotonin as creatinine sulfate. Total of six experiments. The abbreviation ng = nanogram ( 10-9 gram),

THE ROLE OF SEROTONIN I N NEUROBIOLOGY

207

destroys the polypeptic material interfering with the bioassay as can be seen by the data given in Table 11. With bioassay procedures, interferences by other substances might be expected. Preparing the extracts with 95% acetone and following Garven’s ( 1956) suggestions concerning the enzymatic destruction of catechol amines by means of a polyphenyl oxidase the tryptamine analogs which could be present in the samples and cause the rat uterus to contract are: 5-meth~xytryptamine~ bufotenine, 5,6-dimethoxytryptamine, tryptamine, and l-methyltryptamine. These compounds are listed according to the order of their respective contractile effect (Erspamer, 1954b). From a quantitative standpoint this interference cannot be important. For example, bufotenine would interfere with the bioassay procedure only if present in amounts of at least 1 pg/gm of rat tissues. One might note that physicochemical analyses has not disclosed such concentrations of bufotenine in mammalian tissues ( Rodnight, 1956). A more serious limitation to the use of bioassay techniques stems from the fact that our knowledge of the possible constituents of brain tissues is unsatisfactory. Hence, theoretically, it is possible that other unknown constituents share the biological properties of serotonin and interfere with the bioassay. This difficulty is minimized by using a widely experimented biological tissue such as the rat uterus (Leitch et al., 1957). Other biological methods (Vane, 1957) suggest the use of the fundus of rat stomach which may be more sensitive than the uterus of spayed rats. The use of several tissues as advocated by Chang and Gaddum (1933) for acetylcholine may also add to the reliability of the bioassay of serotonin. VII.

Importance of Serotonin in Psychopharmacology

A. SEROTOXIN METABOLISMAND RESERPINE EFFECTS The mechanism of reserpine sedation is generally considered as being related to the depletion of cerebral biogenic amines (Carlsson et al., 195713; Brodie and Shore, 1957). However, the precise role played by each known amine and unknown amines has not been defined and many of the findings remain controversial. A direct relationship between behavioral effects and brain serotonin deple-

208

ERMINIO COSTA

tion has been demonstrated for a number of reserpine analogs (Shore and Brodie, 1957a). Yet, catechol amines with central stimulating activity and dopa antagonize resperine sedation ( Carlsson et al., 1957a). These authors, moreover, cite the inefficiency of 5-HTP as a reserpine antagonist as evidence for the importance of catechol amine depletion in explaining the pharmacological action of reserpine. Finally, in dogs reserpine increases the acetylcholine content in the hypothalamus and the temporal and frontal lobes whereas the hippocampus is partially depleted of its acetylcholine content (Malhortra and Pundlik, 1959). The above finding suggests that a number of neurohumoral mechanisms, rather than a single one, are involved in reserpine effects. Electrophysiological experiments demonstrate that thresholds for electrical excitability of brain structures are lowered by reserpine (Killam et al., 1957). Furthermore, reserpinized rabbits exhibit persistent EEG alerting (Rinaldi and Himwich, 1955) and cats given reserpine may exhibit convulsive discharges in limbic cortex (Sigg and Schneider, 1957). The complexity of reserpine action is enhanced when this compound is administered to animals pretreated with iproniazid (Brodie et al., 1956). It is difficult to interpret the excitation observed under these conditions ( Shore and Brodie, 1957b; Chessin et al., 1957) for animals given iproniazid alone do not show as marked changes in behavior. As Muscholl and Vogt (1958) suggested, the depletion of peripheral norepinephrine stores by reserpine might be an important factor in the clinical effects of the alkaloid in that severe impairment of the function of sympathetic postganglionic structures parallels the dramatic norepinephrine depletion. Indeed, it is impossible to consider such a radical alteration of the sympathetic system to be without some effect on behavior. Brodie et al. (1957) demonstrated that brain 5-HTP decarboxylase is not inhibited in reserpinized rabbits ( 5 mg/kg intravenously 16 hours previously). On the contrary, on the basis of indirect evidence, it has been suggested that the synthesis of epinephrine is inhibited by reserpine (Kroneberg and Schumann, 1958). Our attention has been specifically directed toward the 5-HTP decarboxylase. The activity of this enzyme has been evaluated in aqueous homogenates of brain and kidney of rats injected with reserpine. The results of these experiments can be summarized as follows:

THE ROLE OF SEROTONIN IN NEUROBIOLOGY

209

1. One group of rats was injected once a day with 0.25mg/kg intraperitoneally of reserpine for 6 days. The 5-HTP decarboxylase activity of both brain and kidney was not different from that of the control animals. However, the serotonin contents of brain and spleen of these animals were found to be significantly lower than that of the controls. 2. When greater amounts of reserpine were injected (2.5 mg/kg intraperitoneally twice daily for 2 days ) the serotonin synthesis of both kidney and brain homogenates was significantly increased (Fig. 10). In agreement with Brodie’s group, we can exclude the occurrence of decreased serotonin production in brain and kidney of animals chronically injected with reserpine. A further evaluation of 5-HTP decarboxylase activity after reserpine was attempted in experiments in which semicarbazide was used as an inhibitor of the 5-HTP decarboxylase. Semicarbazide ( 50 mg/kg intraperitoneally ) injected 24 and 18 hours before sacrificing the mice consistently inhibited the brain 5-HTP decarboxylase. This inhibition was still present when the mice were injected with both reserpine and semicarbazide; however, the palpebral ptosis induced by reserpine was only slightly antidoted by semicarbazide. Unfortunately, larger doses of the inhibitor could not be used because of lethal effects. This experiment cannot be considered conclusive because: ( 1 ) The semicarbazide inhibits the synthesis of GABA (Killam and Bain, 1957) and probably that of catechol amines and serotonin as well. ( 2 ) In our experiments the 5-HTP decarboxylase activity was dramatically inhibited, only when high concentrations of substrate were used. ( 3 ) In agreement with previous results in guinea pigs ( Weissbach et al., 1957), the brain serotonin of mice given only semicarbazide was not found to be significantly reduced. Brodie et aZ. (1957) demonstrated that the cellular inability to bind serotonin is an important factor in the pharmacological action of reserpine. Brodie and Shore (1957) postulate that this produces excessive serotonin concentrations at receptor sites, for the neurohormone is continuously moving from the sites of production to cellular receptors. The inhibition of enzymes concerned with the active transport of serotonin is the cause for the reduced storage of serotonin, and consequently for the decreased serotonin found in the tissues of reserpinized animals (Hughes et uZ., 1958). This effect stems from an enzymatic inhibition for, among other

BRAIN HOMOGENATE

KIDNEY HOMOGENATE

-

72

-

60

M

w

E

48-

24 -

36

$5

121

-

./

gOF 5-HTP

12

PER

1

1

I

l

l

G. OF TISSUE

Fic. 10. 5-HTP decarboxylase of rat tissues; 0-0 = rats injected with saline (geometrical means of 6 experiments with 18 animals); A-A = rats injected with reserpine (2.5 mg/kg intraperitoneally ) twice daily, 12 hours apart for 2 days (geometrical means of 4 experiments with 12 animals); x= rats injected with reserpine (2.5 mg/kg intraperitoneally ) twice within 12 hours (geometrical means of 4 experiments with 12 animals). All animals were sacrificed 12 hours after the last injection of reserpine.

THE ROLE OF SEROTONIN IN NEUROBIOLOCY

“1

considerations, stoichiometric analyses exclude a competitive antagonism between reserpine and serotonin ( Carlsson et al., 1 9 5 7 ~ ) . As Shore and Brodie ( 1957b) demonstrated, premedication with iproniazid inhibits the depletion of brain serotonin induced by reserpine. If iproniazid is given after reserpine, no change occurs in the usual behavioral and biochemical effects. Carlsson et d. (1957b) demonstrated that a single injection of iproniazid, although inactive per se on brain catechol amine content, protects brain, heart, and adrenal norepinephrine stores from reserpine depletion. Pletscher ( 195713) analyzed critically the interrelations between reserpine and iproniazid in regard to modification of brain serotonin and found that the reserpine effect on the staining characteristics of the enterochromaffin cells was weakened by iproniazid premedication (Zbinden et d.,1957). They conclude that the MO inhibition prevents the release of serotonin after reserpine. Our findings in this regard are summarized in Fig. 11. The brain serotonin depletion of rats given reserpine is prevented by iproniazid premedication. In contrast, iproniazid fails to counteract reserpine effects on spleen serotonin. A few comments are necessary in order to interpret these findings. Udenfriend and Weissbach (1958) find that serotonin, in spleen, is entirely contained in platelets which do not possess enzymes important for serotonin metabolism. The slight MO activity present in spleen extracts is probably contained in smooth muscle cells (Davison, 1958). Thus, the following interpretation is suggested since Brodie and Shore (1957) conclude the essential fact in reserpinized animals is the impairment of serotonin storage. Because in spleen serotonin is neither synthesized nor destroyed, one can block the reserpine effects only by the use of a specific inhibitor. Evidently iproniazid is not a specific inhibitor of reserpine action. The fact that brain serotonin concentrations of rats premedicated with iproniazid and injected with reserpine are greater than control values might be tentatively explained as follows: Several types of serotonin binding sites are present in nerve cells, but not in platelets (Sano et al., 1958); the release of serotonin from the neuronal binding sites is accomplished by physiological stimuli and is accompanied by enzymatic inactivation at the site of action, The concentrations of MO are such that a large safety

212

ERMINIO COSTA

factor is present. If reserpine alone is given, the serotonin, which cannot be bound to storage sites, undergoes enzymatic deamination, and the serotonin level drops. On the contrary, if the enzyme is inhibited and reserpine given, the amine is released but cannot be metabolized and consequently the serotonin concentration increases in proportion to the inhibition of the MO activity of the

SEROTONIN CONTENT n g l g WET

BRAIN

SPLEEN ea.COMR0LS

0.IPRCUlAIlO [3* RESERRNE .+RONIAZID

I RESERPINE

FIG.11. Effect of reserpine on brain and spleen serotonin content in rats premedicated with iproniazid. An intraperitoneal dose of 50 mg/kg iproniazid was injected at 0 time and hour 12 of experiment and a dose of 25mg/kg at hour 48 of experiment. Reserpine ( 2 . 5mg/kg, intraperitoneally ) was injected at hour 24 and 38 of experiment. All animals were sacrificed at hour 50 of experiment. Serotonin analysis was according to Gwen’s method ( 1956). Vertical bars: fiducial limit (S. E . X t ) P = 0.05. Numerals over the bars indicate the number of animals in each group.

tissue. The failure of iproniazid to reverse the reserpine effects when the drug is given to reserpinized animals might be due to the fact that iproniazid is not a rapidly-acting MO inhibitor. With a rapid inhibitor, such as SKF 385, the antagonism can be produced in animals given the antagonist after reserpine (Tedeschi et al., 1959a).

213

THE ROLE OF SEROTONIN I N NEUROBIOLOGY

B. PSYCHOTOMIMETIC DRUGS AND SEROTONIN The psychotomimetic effect of LSD cannot be strictly correlated to its antiserotonin activity as previously supposed (Costa et al., 1957). In Table I11 the correlations between antiserotonin potency, epinephrine potentiation, and psychotomimetic action of various

T T

T

*

1

0

fi W

(20)

(8)

(15)

(8)

(7)

(7)

t fn

m

N

n W

+ fn

0

t

-2 (7)

= Ps .02 COMPARED TO ANIMALS INJECTED WITH SALINE.

1 =

STANOARD ERROR.

( ) = NUMBER OF ANIMALS.

FIG. 12. Adrenal ascorbic acid content of rats. LSD 25 enhances the ascorbic acid depletion caused by epinephrine-BOL fails to do so and differently from LSD is not psychotomimetic.

compounds structurally related to LSD or serotonin are summarized. These data support the tentative conclusion that the intensity of the antiserotonin effect may be independent of the psychotomimetic capacity with the proviso that any comparison between in vivo and in vitro results must be made cautiously. Table I11 and Fig. 12 present data indicating that the psychotomimetic activity of many drugs is correlated with peripheral sensitization to epinephrine. Although the correlation is quite striking, there are exceptions here as in the case of serotonin

TABLE I11 PHARMACOLOGICAL PROPERTIES OF DRUGS CHEMICALLY RELATEXITo INDOLE

Drug

Structurea

LSD-25

A

-N/cn2-cn*

LSM

A

-

ALD-52

A

LAE-32

A

Rl \cn2-cn*

J”’ -z\ \cn+,/O p - c n a

\cnz-cn3 _N/cH2-cHa

AntiEpineph- Psychotomiserotonin rine metic action activityb potentiation in man

RZ

R3

R4

H

H

-

1000

1000

Yes

H

H

-

20

102.0

Yes

H

C,H,

-

2000

780

Yes

H

H

-

120

1020

Yes

Br

H

1030

15

No

0

H

-

?

?

No

H

OH

CH3

0

370

Yes

\n

BOL-148

A

-N/cnz-cna

‘CMa-CN,

- N/cwz-oc.

I

OXY-LSD

A

Bufotenine

B

BAS

B

--CHz--C,HS

CH3

H3C0

H

2

0

No

BAB

B

--CH,--C,H,

CH3

H3C0

CH3

2

0

No

Serotonin

B

H

OH

H

-

20

No

a

’

\cnz-cn,

H

H

b Tested on isolated uterus of spayed rats: LSD activity ( 2 x 10-9) = 1000.

Structure: H.

0

Tested on cat’s nictitating membrane: LSD activity (1.80

pg/kg i.v.) = 1000. I

R.

R,

A

B

6

m

THE ROLE OF SEROTONIN I N NEUROBIOLOGY

215

antagonism (Costa and Zetler, 1958, 1959). Therefore, it is felt that at present none of the above mentioned mechanisms can be the sole interpretation for the psychotomimetic effects of these compounds. However, a prominent role may be played by the catechol amines in view of the correlation between psychotomimetic effects of bufotenine and the release of catechol amines from binding sites (Himwich et al., 1959).

C. ANTIDEPRESSANT DRUGSAND SEROTONIN Mental depression is a symptom frequently found in psychiatric disorders. The classic stimulants do not relieve this condition, but often complicate it with their side effects on behavior, appetite, and sleep cycles. Because of our ignorance in the etiology of psychiatric depression our therapeutic efforts have been directed mainly toward symptomatic relief. The empirical results so obtained may be used in an attempt to understand the physiological mechanisms underlying clinical depression. Imipramine ( Tofranil, from Geigy Chemical Corporation) [N - ( y-dimethylaminopropyl) iminodibenzyl hydrochloride] has been successfully used in clinics as an antidepressant (Kuhn, 1957). Despite the analogy between the chemical structures of imipramine and chlorpromazine, the two drugs diverge with respect to their therapeutic effects. The clinical differentiation between the two compounds is paralleled by the following laboratory results: 1. Chlorpromazine does not modify brain serotonin of rats ( Costa, unpublished observations) whereas imipramine causes an increase of brain serotonin analyzed according to Garven’s (1956) method (Fig. 13). This effect is transient and limited to the spleen and brain. 2. The palpebral ptosis in rats caused by reserpine is reduced by chronic premedication with imipramine (Fig. 14). Chlorpromazine however potentiates this reserpine effect. Imipramine is ineffective if injected after reserpine. 3. In mice the antagonistic effect of iproniazid against the eye closure induced by reserpine was enhanced by 10 mg/kg of imipramine, but was weakened by chlorpromazine. 4. The antiserotonin action of imipramine on the isolated uterus of spayed rats is less than that of chlorpromazine. In Table IV the

400 300

-

400 CWllloL [ * I t ) SemlwlN.

.mf .a5 y/g

-

400

cmma (rr-lz~-coo

om0

-

cQlm)Q (n-12) IOQ

smom"-soot .mt/g

SEmmMIN. M 7 * . 1 3 0 ~ / g

Joo-

300-

-:L I:& 100

CONTROL

3 0

-

200

-

loo-

[N*IZ)~IOO

,

M

400

SEmmNm. .7sof.l44 y/g

300

'O0

aO:

$-& *+: :

-

UlvrRoL ( N = i z ) =too SERITLIYIV. . M f .omt/g

8

v)

*

4

%

t

EYELID

A

CLOSURE

A =0-0.

4 HOURS

AtFTER

B

RESERPINE

INJECTION

lMlPRAMNE.i.p.JOmg/kg IN 3 DOSES, 26h .1gh .ANDIh BEFORE img/kg, OF RESERPINE, i.p..

B = l f A X = I mg/

kg 0 TIME.

c * -=

I mg/kg

RESERPINE, i. p. AT

+

RESERPINE CHLORPROMAZINE. i. p. 3Omg/kg (inj.as inA).

5

TIME

10 IN

15 20 HOURS.

25

FIG.14. Effect of a chronic prernedication with chlorpromazine and imipramine on eyelid closure induced by reserpine.

218

ERMINIO COSTA

antiserotonin activity, in uivo, of imipramine, chlorpromazine, and triflupromazine is compared. This activity was evaluated according to the method proposed by Woolley (1958b). The ED50 of imipramine is significantly greater than that of the two tranquilizers, triflupromazine and chlorpromazine. As reported in Table IV, papaverine is less effective than any of the above mentioned drugs. ED,,a

OF

TABLE IV CHLORPROMAZINE, TRIFLUPROMAZINE~, AND IMIPRAMINEON 5-HTPO INDUCED DIARRHEA IN MICE

Drug injected Triflupromazine Chlorpromazine Imipramine Papaverine a

b

ED,, ( mg/kg 1 1.85 3.64 20.00 > 32.00

Fiducial limit ( P = 0.05) 2.59 to 1.32 8.82 to 1.46 44 to 9.10

-

Calculated according to the method of Litchfield and Wilcoxon (1949). 2-Trifluoremethyl-10-( y-dimethylaminopropyl) phenothiazine hydrochlo-

ride. c

Dose: 100 mg/kg intraperitoneally. Injected intraperitoneally 90 minutes before 5-HTP.

Thus, a nonspecific antispasmodic effect may be ruled out. Furthermore the synthesis of serotonin was not inhibited in the brain of rats premedicated with imipramine and injected with 5-HTP ( Costa, unpublished observations ) . The comparison between the in uiuo antiserotonin action of tranquilizers and imipramine is partially impaired by the rapid catabolism of the latter. These data suggest a similarity between the pharmacological effects of iproniazid and imipramine. This analogy extends to the therapeutic effects but probably does not apply to the mode of action since imipramine is not a MO inhibitor in vitro. A further difference between the mechanisms involved in the therapeutic action of iproniazid and imipramine is also suggested by the following: 1. Imipramine (7.5 mg/kg, intravenously) fails to increase the brain serotonin of rabbits which is elevated by iproniazid (Fig. 6 ) . 2. In pentobarbitalized cats the blood pressure response to norepinephrine is selectively facilitated by imipramine ( Sigg, 1959). On the contrary, iproniazid premedication does not accentuate the

THE ROLE OF SEROTONIN IN NEUROBIOLOGY

219

blood pressure effects of catechol amines (Corne and Graham, 1957). 3. The peripheral effects of 5-HTP are antagonized by imipramine and potentiated by iproniazid. In conclusion, the mechanism by which imipramine improves the symptoms of mental depression is not understood. Nevertheless, the antagonism of that drug toward reserpine sedation and its action on brain neurohormonal levels of rats suggest that its antidepressant action bears some relationship to modifications of brain neurohormonal balance. This possibility is made plausable by the effect of imipramine on the peripheral actions of serotonin and norepinephrine. VIII.

Summary

Available experimental evidence does not encourage the hypothesis that the mechanisms involved in the CNS effects of serotonin and acetylcholine are similar. This suggestion is supported by the following: 1. Unlike specific inhibitors of cholinesterase the specific inhibitors of the enzyme which destroy serotonin neither potentiate the circulatory effects of blood-borne serotonin in vivo nor prolong or accentuate its excitatory action on smooth muscles in vitro. 2. The increase of brain serotonin caused by injection of the biological precursor of serotonin, S-HTP, is accompanied by biphasic behavioral changes in dogs. In rabbits a persistent EEG alert pattern parallels a fourfold rise in serotonin content of mesencephalon-diencephalon whereas features of EEG sleep are associated with two to threefold increases. A mimicry of these effects is not observed when comparable rises in brain serotonin are caused by iproniazid. In contrast, the inhibition of cholinesterase constantly displays a simulation of CNS acetylcholine effects. However, both behavioral and electrophysiological effects (inhibition of evoked electrical brain potential in cats) caused by injections of 5-HTP are potentiated by specific MO inhibitors. Finally, it must be stressed that atropine ( 2 mg/kg, intravenously) reverses the EEG changes produced by increased brain serotonin. The oil-water partition coefficient of serotonin suggests that this amine passes slowly across cellular membranes. This fact em-

220

ERMINIO COSTA

phasizes the importance of possible indirect mechanisms involved in the inhibition of transcallosal potentials caused by blood-borne serotonin in cats (Koella et al., 1959). Moreover, this physicochemical property of serotonin suggests the existence of specific carrier systems for the transport of serotonin across cell membranes. The exclusive intracellular location of the MO needed for serotonin catabolism and the above reported facts do not support a neurotransmitter function for serotonin. It is tentatively proposed that serotonin may play a role in the regulation of neuron excitability, possibly by participating in the control of intracellular ion concentrations. The problem of specificity of 5-HTP decarboxylase is analyzed. Using tissue homogenates a differentiation between dopa and 5-HTP decarboxylase has not been achieved by means of specific inhibitors or by other means. However, the existence of minute differences in the chemical structures of these two enzymes cannot be excluded unless purified enzymatic preparations like those of Clark et al. (1954) are employed. The species variability in the 5-HTP decarboxylase activity of brain homogenates is independent of coenzyme availability. The brain MO activity of different species is not uniform in its susceptibility to iproniazid inhibition. The enzyme of the dog brain is less susceptible than that of rabbit and rat. The latter is in turn more sensitive to this inhibitor both in vitro and in v k o than is the rabbit. Structure-activity relationships are discussed in regard to MO inhibitors. Catron (JB-516) is a faster MO inhibitor than iproniazid. A rapid brain serotonin turnover is demonstrated by use of this inhibitor, for the increase of brain serotonin is observed within 30 minutes after injections of 100 mg/kg. The effect of prernedication with truns-2-phenylcyclopropylamineHCl ( SKF-trans-385) on both brain serotonin content and reserpine sedation is analyzed. This discloses that the cyclopropane ring is essential for the MO inhibitory action of SKF-trans-385. The bioassay methods for serotonin determination are discussed and a new modification is proposed. The depletion of several known biogenic amines is involved in the mechanism of reserpine sedation. None of these amines can be singled out to explain the pharmacological effects of reserpine

THE ROLE OF SEROTONIN IN NEUROBIOLOGY

221

and reserpinelike drugs. Brodie’s theory on the mechanism of brain serotonin depletion caused by reserpine is supported by data on the antagonistic effects of iproniazid. This antagonism is thought to be indirect in nature. The intensity of psychotomimetic effects evoked by indole derivatives is not correlated with their antiserotonin properties. An accentuation of catechol amine effects or a release of catechol amines from body stores is a property more consistently associated with psychotomimetic action. Antidepressant drugs, including imipramine, antagonize several pharmacological actions of reserpine. Both MO inhibitors and imipramine cause a rise of brain serotonin content as revealed by bioassay analysis. The mechanisms involved are different and an exclusive implication of serotonin in this therapeutic activity is improbable. ACKNOWLEDGMENT It is my pleasure to express my indebtedness to Drs. A. Revzin and S. Garattini for allowing me to cite unpublished experimental results obtained in this laboratory with their valuable collaborations. The criticism of Drs. H. E. Himwich, A. M. Revzin, J. R. Smythies, and S. Garattini has been very greatly appreciated. The competent technical assistance of Mrs. M. G. Griffin, Miss G. L. Woods, and Miss D. I. Grant made possible some of the experiments reported in this paper. REFERENCES Abrahams, V. C., Koelle, G. B., and Smart, P. (1957). 1. Physiol. (London) 139, 137. Amin, A. H., Crawford, T. B. B., and Gaddum, J. H. (1954). 1. Physiol. (London) 126, 596. Azioka, I., and Tanimukai, H. (1957). J. Neurochem. 1, 311. Baker, R. V. (1958). J. Physiol. (London) 142, 563. Baker, R. V. (1959). J. Physiol. (London) 146, 473. Balzer, H.,and Holtz, P. (1956). Arch. exptl. Pathol. Pharmakol. NaunynSchmiedoberg’s 227, 547. Barrett, W. E., Plummer, A. J., Rutledge, R. A., and MacPhillamy, H. B. (1958). Abstr. Meeting Am. Soc. Pharmacol. Exptl. Therap. Ann Arbor, Michigan, p. 3. Bertler, A., and Rosengren, E. (1959). Experientia 16, 10. Billa, B., and Valzelli, L. (1958). Boll. soc. ital. biol. sper. 34, 1404. Bishop, G . H. (1956). Physiol. Reus. 36, 376.

222

ERMENIO COSTA

Bishop, G. H., and Clare, M. H. (1953). J. Newophysiol. 16, 1. Blaschko, H. (1952). Pharmucol. Reus. 4, 415. Blaschko, H., Born, G . V. R., D’Iorio, A., and Eade, N. R. (1956). J. Physiol. ( L o n d o n ) 133, 548. Blaschko, H., Friedman, P. J., Hawes, R., and Nilsson, K. (1959). J. Physiol. (London) 146, 384. Blaschko, H., Hagen, J. M., and Hagen, P. ( 1957). 1. Physiol. ( L o n d o n ) 139, 316. Blaschko, H., and Welch, A. D. (1953). Arch. exptl. Pathol. Pharmukol. 219, 17. Bogdanski, D. F., and Udenfriend, S. (1956). 1. Pharmucol. Exptl. Therap. 116, 7. Bogdanski, D. F., Weissbach, H., and Udenfriend, S. (1957). 1. Neurochem. 1, 272. Bogdanski, D. F., Weissbach, H., and Udenfriend, S. (1958). J. P h a m c o l . Exptl. Therap. 122, 182. Born, G. V. R., and Gillson, R. E. (1957). J. Physiol. ( L o n d o n ) 137, 82. Born, G. V. R., Hornykiewicz, O., and Stafford, A. (1958a). Brit. J. Pharmucol. 13,411. Born, G. V. R., Ingram, G. I. C., and Stacey, R. S. (1958b). Brit. J. Phamacol. 13, 62. Brodie, B. B. ( 1958). Modern Med. (Minneapolis) 26, 69. Brodie, B. B., and Shore, P. A. (1957). Ann. N . Y. Acad. Sci. 66, 631. Brodie, B. B., Pletscher, A., and Shore, P. A. (1956). J . Pharmacol. Exptl. Therap. 116, 9. Brodie, B. B., Spector, S., and Shore, P. A. (1959). Pharmacol. Reus. 11, 548. Brodie, B. B., Tomich, E. G., Kuntzman, R., and Shore, P. A. (1957). J. Pharmacol. Exptl. Therap. 119, 461. Brodie, B. B., Finger, K. F., Orleans, F. B., Quinn, G. P., and Sulser, F. (1960). J. Pharmacol. Exptl. Therap. 128,in press. Bulbring, E., and Crema, A. (1958). Brit. J . Pharmacol. 13, 444. Bulbring, E., and Lin, R. C. Y. (1958). 3. PhysioE. (London) 140, 381. Buzard, J. A,, and Nytch, P. D. (1957). J. BWZ. Chem. 227, 225. Carlsson, A., and Hillarp, N. A. (1956a). Kgl. Fysiograf. Sh’lLskap. i L u n d Forh. 26, No. 8. Carlsson, A., and Hillarp, N. A. (1956b). Actu Physiol. Scand. 37, 235. Carlsson, A., Lindqvist, M., and Magnusson, R. (1957a). Nature 180, 1200. Carlsson, A., Rosengren, S., Bertler, A., and Nilsson, J. (1957b). “Psychotropic Drugs,’’ p. 372. Elsevier, Amsterdam. Carlsson, A., Shore, P. A., and Brodie, B. B. ( 1 9 5 7 ~ ) .J. Pharmacol. Exptl. Therap. 120, 334. Chang, H. C., and Gaddum, J. H. (1933). J. Physiol. ( L o n d o n ) 79, 255. Chessin, M., Kramer, E. R., and Scott, C. S. (1957). J. Pharmacol. Exptl. Therap. 119, 453. Clark, C. T., Weissbach, H,, and Udenfriend, S. (1954). J. Biol. Chem. 210, 139. Come, S. J., and Graham, J. D. P. (1957). J. Physiol. ( L o n d o n ) 135, 339.

THE ROLE OF SEROTONIN IN NEUROBIOLOGY

223

Correale, P. (1956). J. Neurochem. 1, 22. Correale, P. (1958). J. Neurochem. 2, 201. Costa, E. (1956). Proc. SOC. Exptl. Biol. Med. 91, 39. Costa, E., and Aprison, M. H. (1958a). J. Nervous Mental Disease 126, 289. Costa, E., and Aprison, M. H. (1958b). Am. J. Physiol. 192, 95. Costa, E., and Himwich, H. E. (1958). IV Inter. Congr. Biochem., Vienna, Symposium 111. Costa, E., and Rinaldi, F. (1958). Am. J. Physiol. 194, 214. Costa, E., and Zetler, G. (1958). Federation Proc. 17, 360. Costa, E., and Zetler, G. (1959). J. Pharmucol. Exptl. Therap. 125, 230. Costa, E., Rinaldi, F., and Himwich, H. E. (1957). “Psychotropic Drugs,” p. 21. Elsevier, Amsterdam. Costa, E., Himwich, W. A,, Goldstein, S. G., Canham, R. G., and Himwich, H. E. ( 1959a). Federation Proc. 18, 379. Costa, E., Pscheidt, G. R., Van Meter, W. G., and Himwich, H. E. (1959b). J. Pharmucol. Exptl. Therup. In press. Cotzias, G. C., and Dole, V. P. (1951). Proc. SOC.Exptl. Biol. Med. 78, 157. Davison, A. N. (1957). Biochem. J. 67, 316. Davison, A. N. (1958). Physiol. Revs. 38, 729. Davison, A. N., and Sandler, M. (1958). Nature 181, 186. Eccles, J. C. (1957). “The Physiology of Nerve Cells.” Johns Hopkins Press, Baltimore, Maryland. Eccles, J. C., Fatt, P., and Koketsu, K. (1954). J. Physiol. (London) 126, 524. Erspamer, V. (1954a). Pharmacol. Revs. 6, 425. Erspamer, V. (195413). Rend. Sci. Farmital. 1, 5. Erspamer, V. ( 1957). “Psychotropic Drugs,” p. 415. Elsevier, Amsterdam. Fatt, P. (1954). Physiol. Revs. 34, 674. Feldberg, W. S. ( 1945). Physiol. Revs. 25, 596. Feldberg, W. S. (1954). Pharmucol. Revs. 6, 85. Florey, E.,and Florey, E. (1953). Nutuwissenschuften 40, 413. Freter, K., Weissbach, H., Redfield, B. G., Udenfriend, S., and Witkop, B. (1958). J. Am. Chem. Soc. 80, 983. Furshpan, E. J., and Potter, D. D. (1957). Nature 180, 342. Gaddum, J. H. (1953). J. Physiol. (London) 121, 15. Gaddum, J. H. (1957). Ann. N . Y. Acad. Sci. 66, 643. Gaddum, J. H., and Giarman, N. J. (1956). Brit. J. Phamuzcol. 11, 88. Gaddum, J. H., and Paasonen, M. K. (1955). Brit. J. P h a m c o l . 10, 474. Garattini, S., and Valzelli, L. (1957). “Psychotropic Drugs,” p. 428. Elsevier, Amsterdam. Garattini, S., Kato, R., and Valzelli, L. (1958). Atti S O C . lomharda sci. med. hiol. 13, 1. Garven, J. D. (1956). Brit. J. Phurmacol. 11, 66. Gertner, S. B. (1959). Federation Proc. 18, 393. Ginzel, K. H., and Kottegoda, S. R. (1954). J. Physiol. (London) 125, 277. Goldberg, L. I. (1959). Federation Proc. 18, 396. Grundfest, H. (1957). Ann. N . Y. Acud. Sci. 66, 537. Gyermek, L. (1955). Lancet ii, 724.

224

ERMINIO COSTA

Hagiwara, S., and Tasaki, I. ( 1958). J. Physiol. (London) 14.9, 114. Hardisty, R. M., and Stacey, R. S. (1955). J. Physiol. (London) 130, 711. Hawkins, J. (1952). Biochem. J. 50, 577. Hebb, C. 0. (1957). Physiol. Reus. 37, 196. Hebb, C. O., and Whittaker, V. P. (1958). J. Physiol. (London) 142, 187. Hess, S., Weissbach, H., Redfield, B. G., and Udenfriend, S. (1958). J. Pharmacol. Exptl. Therap. 124, 189. Hess, S., Redfield, B. G., and Udenfriend, S. (1959). Federation Proc. 18, 402. Hess, W. R. ( 1954). “Diencephalon, Autonomic and Extrapyramidal Functions.” Grune and Stratton, New York. Hillarp, N. A,, Lagerstedt, S., and Nilson, B. (1953). Actu Physiol. scand. 29, 251. Hillarp, N. A,, Hogberg, B., and Nilson, B. (1955). Nature 176, 1032. Himwich, W.A., Costa, E., and Himwich, H. E. (1959). “Biological Psychiatry,” Vol. 2. Grune and Stratton, New York. In press. Hodgkin, A. L. (1951). Biol. Reus. Cambridge Phil. SOC. 26, 338. Holtz, P., and Westermann, E. (1956). Arch. exptl. Pathol. Pharmukol. 227, 538. Holzbauer, M., and Vogt, M. (1956). J. Neurochem. 1, 8. Horita, A. (1958). J. Pharmucol. Exptl. Therap. 122, 178. Horita, A,, McGrath, W. R., and Eltherington, L. G. (1959). Federation Proc. 18, 403. Hughes, F. B., and Brodie, B. B. (1959). J. Pharmucol. Exptl. Therap. 127, 96. Hughes, F. B., Shore, P. A., and Brodie, B. B. (1958). Experientiu 14, 178. Hughes, F. B., Orleans, F. B., Finger, K. F., and Brodie, B. B. (1960). J. Pharmacol. Exptl. Therap. 128, 131. Humphrey, J. H., and Jaques, R. (1954). J. Physiol. (London) 124, 305. Hyden, H. (1958). IV Intern. Congr. Biochem., Vienna, Symposium 111, reprint 4. Jepson, J. B., Udenfriend, S., and Zaltzmann, P. (1959). Federation PTOC. 18, 254. Karki, N. T.,and Paasonen, M. K. (1959). J. Neurochem. 3, 352. Killarn, K. F., and Bain, J. A. (1957). J. P h u m c o l . Exptl. Therap. 119, 263. Killam, E. K., Killam, K. F., and Shaw, T. (1957). Ann. N . Y. Acad. Sci. 66, 784. Kobinger, W. (1958). Arch. ExptZ. Pathol. Phrmukol. 233, 559. Koella, W. P., Smythies, J. R., and Bull, D. M. (1959). Science 129, 1231. Kosterlitz, H.W., and Robinson, J. A. (1957). J. Physiol. (London) 196, 249. Kuhn, R. (1957). Schweiz. med. Wochschr. 87, 1135. Kroneberg, G., and Schiimann, H. J. (1958). Arch. exptl. Pathol. Pharmakol. 234, 133. Leitch, J. L., Maryn, D., Debley, V. G., and Haley, T. J. (1957). J. Phurmucol. Exptl. Therap. 120, 408. Lernbeck, F. (1953). Nature 172, 910. Lennard-Jones, J. E., and Snow, P. J. D. (1956). Gmtroenterologia 85, 169. Lewis, G. P. (1958). J. Pharm. and Pharmacol. 10, 529.

THE ROLE OF SEROTONIN IN NEUROBIOLOGY

225

Litchfield, J. T., Jr., and Wilcoxon, F. (1949). J . Pharmucol. Exptl. Therap. 96, 99. McIsaac, W. M., and Page, I. H. (1958). Science 128, 537. Malhotra, C. L., and Pundlik, P. G. (1959). Brit. J. Pharmucol. Exptl. Therap. 14, 46. Maas, A. R., and Nimmo, J. M. (1959). Nature 184, 547. Marrazzi, A. S., and Hart, E. R. (1955). Science 121, 365. Mott, J. C., and Paintal, A. S. (1953). Brit. J. P h a m c o l . 8, 238. Muscholl, E., and Vogt, M. (1958). J . Physiol. (London) 141, 132. Nachmansohn, D. (1955). Ergeb. Physiol. u. exptl. Pharmakol. 48, 575. Paasonen, M. K., and Dews, P. B. (1958). Brit. J. Pharmacol. 13, 84. Paasonen, M. K., and Giarman, N. J. (1958). Arch. intern. phurmucodynamie 114, 189. Paasonen, M. K., MacLean, P. D., and Giarman, N. J. (1957). J . Neurochem. 1, 326. Page, I. H. (1954). Physiol. Reus. 34, 563. Page, I. H. (1958). Physiol. Revs. 38, 277. Pepler, W. J., and Pearse, A. G. E. (1957). J. Neurochem. 1, 193. Pickles, V. R. (1956). J . Physiol. (London) 194, 484. Pickles, V. R. ( 1957). J. Physiol. (London) 138, 495. Pletscher, A. ( 1957a). Science 126, 507. Pletscher, A. ( 195713). “Psychotropic Drugs,” p. 468. Elsevier, Amsterdam. Pletscher, A., and Gey, K. F. (1958). Science 128, 900. Pletscher, A., Besendorf, H., and Gey, K. F. (1959). Science 129, 844. Plummer, A. J., Barrett, W. E., Maxwell, R. A., Finocchio, D., Lucas, R. A., and Rutledge, R. A. (1958). Abstr. Meeting Am. SOC.P h a m c o l . and Exptl. Therap. Ann Arbor, Michigan, p. 27. Prockop, D. J., Shore, P. A., and Brodie, B. B. (1959). Experknth 16, 145. Pscheidt, G. R., and Himwich, H. E. (1959). Federdion Proc. 18, 123. Purpura, D. P., and Grundfest, H. (1956). J . Neurophysiol. 19, 573. Quinn, G. P., Shore, P. A., and Brodie, B. B. (1959). J. Pharmacol. Exptl. Therap. 127, 103. Revzin, A. M., and Costa, E. (1959). Federation Proc. 18, 436. Rinaldi, F.,and Himwich, H. E. (1955). Ann. N. Y. Acad. Sci. 61, 27. Rodnight, R. (1956). Biochem. J. 64, 621. Rubin, B., Malone, M. H., Waugh, M. H., and Burke, J. C. (1957). J . Pharmacol. Exptl. Therap. 120, 125. Sandler, M., and Snow, P. J. D. (1958). Lancet i, 137. Sano, I., Kakimoto, Y., and Taniguchi, K. (1958). Am. J . Physiol. 196, 495. Schneider, J. A,, and Yonkman, F. F. (1953). Am. J . Physiol. 174, 127. Schneider, J. A., and Yonkman, F. F. (1954). J . Pharmucol. Exptl. Therap. 111, 84. Shanes, A. M. (1958a). Phamnacol. Revs. 10, 59. Shanes, A. M. (195813). Pharmucol. Revs. 10, 165. Shore, P. A., and Brodie, B. B. (1957a). “Psychotropic Drugs,” p. 423. Elsevier, Amsterdam. Shore, P. A., and Brodie, B. B. (195713). Proc. SOC. Exptl. Biol. Med. 94, 433.

226

ERMINIO COSTA

Shore, P. A., Pletscher, A,, Tomich, E. G., Carlsson, A., Kuntzman, R., and Brodie, B. B. (1957). Ann. N.Y. Acad. Sci. 66, 609. Shore, P. A., Silver, S. L., and Brodie, B. B. (1955). Science 122, 284. Sigg, E. B. (1959). Federation PTOC.18, 144. Sigg, E. B., and Schneider, J. A. (1957). Electroencephalog. and Clin. Neurophysiol. 9, 419. Sourkes, T. L. (1954). Arch. Biochem. Biophys. 51, 444. Spector, S., Prockop, D., Shore, P. A., and Bridie, B. B. (1958). Science 127, 704. Spector, S., Maling, H. M., and Shore, P. A. (1959). Federation PTOC.18, 447. Stone, C. A., and Loew, E. R. (1952). 1. Pharmacol. Exptl. Therap. 106, 226. Tedeschi, R. E., Tedeschi, D. H., Cook, L., Mattis, P. A., and Fellows, E. J. (1959a). Federation PTOC.18, 451. Tedeschi, D. H., Tedeschi, R. E., and Fellows, E . J. (1959b). Federution proc. 18, 450. Thompson, J. W. (1958). 3. Physiol. (London) 141, 46. Trendelenburg, U. (1956a). Brit. J. Phamnacol. 11, 74. Trendelenburg, U. (195613). J. Physiol. (London) 132, 529. Trendelenburg, U. (1957). Brit. J. Pharmacol. 12, 79. Trendelenburg, U. (1959). Federation Proc. 18, 1001. Twarog, B. M., and Page, I . H . (1953). Am. J. Physiol. 175, 157. Udenfriend,. S. (. 1958). IV Intern. Congr. Biochem., Vienna, Symposium I l l , reprint No. 5. Udenfriend, S., and Weissbach, H. (1958). Proc. SOC. Exptl. Biol. Med. 97, 748. U,denfriend, S., Weissbach, H., and Clark, C . T. (1955). J . Biol. Chem. 215, 337. Udenfriend, S., Weissbach, H., and Bogdanski, D. F. (1957a). J. Biol. Cheni. 224. 803. Udenfriend, S., Weissbach, H., and Bogdanski, D. F. (1957b). Ann. N . Y . A c d . Sci. 66, 602. Vane, J. R. (1957). Brit. J. Phannacol. 12, 344. Vane, J. R. (1959). Brit. J. Pharmacol. 14, 87. Vogt, M. (1954). J. Physiol. (London) 123, 451. Vogt, W. (1958). Pharmacol. Revs. 10, 407. Von Euler, C., and Ricci, G. F. (1958). J. Neurophysiol. 21, 231. Weil-Malherbe, H.,and Bone, A. D. (1958). Biochem. J. 70, 14. Weissbach, H., Bogdanski, D. F., Redfield, B. G., and Udenfriend, S. (1957). 1. Biol. Chem. 227, 617. Weissbach, H., Waalkes, T. P., and Udenfriend, S. (1958). J. Biol. Chem. 230,865. Werle, E. (1940). Biochem. Z. 304, 201. Welsh, J. H . (1957). Ann. N . Y . Acad. Sci. 66, 618. Westermann, E., Balzer, H., and Knell, J. (1958). Arch. exptl. Pathol. Pharmakol. Naunyn-Schmiedeberg’s 234, 194. Woolley, D. W. (1957). Nature 180, 630. Woolley, D. W. (1958a). PTOC.Assoc. Research Nervous M a t a l Disease 36, 381.

THE ROLE OF SEROTONIN I N NEUROBIOLOGY

227

Woolley, D. W. (195%). Proc. Suc. Exptl. Bwl. LMed. 98, 367. Woolley, D. W., and Shaw, E. (1954a). Science 119, 587. Woolley, D.W., and Shaw, E. (1954b). Proc. Natl. Acud. Sci. U . S. 40, 228. Woolley, D. W., and Shaw, E. (1957). Ann. N. Y. Acad. Sci. 66, 649. Yuwiler, A., Geller, E., and Eiduson, S. (1959). Federation Proc. 18, 606. Zbinden, G., Pletscher, A., and Studer, A. (1957). Kin. Wochschr. 35, 565. Zbinden, G., Pletscher, A., and Studer, A. (1958). 2. ges. exptl. Med. 129, 615. Zeller, E. A. (1938a). Helv. Chim. Actu 21, 880. Zeller, E. A. (193813). Helv. Chim. Actu 21, 1645. Zeller, E. A., Barsky, J., and Berman, E. R. (1955). J. Biol. Chem. 214, 267. Zeller, E. A., Barsky, J., Berman, E. R., Cherkas, M. S., and Fouts, J, R. (1958). J. Phamu~~ol. Exptl. Therap. 124, 282. Zetler, G., and Schlosser, L. (1954). Arch. exptl. Pathol. Pharmakol. 222, 345. Zimmerman, J. H., and Sheppard, H. (1959). Federation Proc. 18, 463. Zupancic, A. 0. (1953). Actu Physiol. Scand. 29, 63.

This Page Intentionally Left Blank

DRUGS AND THE CONDITIONED AVOIDANCE RESPONSE By Albert Herz Department of Pharmacology. University of Munich. Munich. Germany

I. Introduction

.

...........................................

229

11 Methods for Studying the Conditioned Avoidance Response

I11.

IV . V. VI . VII. VIII . IX . X. XI .

XI1.

(CAR) ............................................. A . Methods Involving Horizontal Route of Escape (Shuttle Box) ............................................ B. Vertical Escape Responses ( Pole-Jump) . . . . . . . . . . . . . . . . C . Timed Rope-Climbing Test ........................... D . Leg-Withdrawal Response ............................ E . Avoidance Responses Derived from Free Operant Behavior . . F . Miscellaneous Methods ............................... Tranquilizing Agents .................................... A. Phenothiazine Derivatives ............................ B . Rauwolfia Alkaloids ................................. C . Diphenylmethane Derivatives ......................... D . Interneuron Blocking Agents ..................... Hypnotics and Sedatives ............................ Morphine and Other Analgesics ........................... Adrenergic and Adrenergic-Blocking Substances . . . . . . . . . . . . . Cholinergic and Anticholinergic Drugs ..................... Serotonin and Monoamine Oxidase Inhibitors . . . Psychotomimetic Drugs .................................. Miscellaneous . . . . . . . . . . . . . . General Discussion .......... A. Specificity of Action . . . . . . . . . . . . . . . . . . ........... B. Various Types of CAR Inhibition ...................... C . The Question of Sensory Deficit ....................... D . Significance of Analgesic Activity ...................... E. Fear-Reducing Hypothesis ............................ F. Further Aspects regarding CAR Inhibition ............... Conclusions . . . . . . . . . . . . . . . . . . .................... References . . . . . . . . . . . . . . . . . . ....................

.

1

230 230 231 232 232 233 233 234 234 242 246 249 250 252 254 256 260 262 264 265 265 268 268 269 270 271 272 272

Introduction

The expansion of the pharmacology of the central nervous system (CNS) during the last decade has confronted the phar229

230

ALBERT HERZ

macologist with the problem of developing suitable methods for the investigation of psychic effects in animals. The transference of drug findings from animals to man, which is difficult in the case of peripheral organ systems, presents an even more difficult problem in regard to the CNS. Among the numerous newly developed methods for the evaluation of psychic effects of drugs, those based on conditioned‘ reactions are most commonly used. These methods, although derived from the classic Pavlovian experiments, show little conformity to the original in regard to stimuli presented or responses elicited. The type of conditioning has been adapted to characterize the action of a group of new compounds, the so-called tranquilizers. For this purpose methods employing conditioned avoidance responses (CAR) have been given preference over approach reactions. It was hoped that alterations of the CAR would help to correlate certain properties of this heterogeneous and poorly defined class of drugs and to differentiate these new tranquilizing substances from well-known drugs with similar actions. Aside from its use to test tranquilizing substances, avoidance conditioning is of value in the investigation of theoretical problems concerning psychotropic substances. A review of accumulated data concerning the action of drugs on the CAR may contribute to an understanding of the conditioned avoidance response itself and help to clarify the mode of drug effects. Reviews on drug action and animal behavior have been compiled previously by Witt ( 1956), Brady (1957), Miller (1957), and Riley and Spinks (1958).

II.

Methods for Studying the Conditioned Avoidance Response

(CAR)

A. METHODS INVOLVING HORIZONTAL ROUTEOF ESCAPE (SHUTTLE Box) Warner ( 1932) described a conditioning technique wherein rats were placed in a two-compartment box with a grid floor. Following a warning signal (the conditioned stimulus-CS) an electric 1 For a discussion of the use of the terms “conditioned” versus “conditional” see Brazier ( 1958).

DRUGS AND THE C A R

231

shock (the unconditioned stimulus-US) is applied via the grid; the animals are taught to avoid shock by jumping from one compartment to the other across a low hurdle dividing the box. The barrier or hurdle is usually electrified. There are three types of responses: 1. The animal jumps following the CS; the conditioned avoidance response ( CAR) is performed. Following intensive training, most animals reach a 95-100% performance level of the CAR. 2. The animal does not respond to the CS, but it does respond as soon as the US is applied. This is the unconditioned response ( U R ) also called “escape response” and represents a specific block of the CAR. 3. The animal does not jump at all, but instead takes the shock. This failure of UR, “response failure,” or “take shock,” is often due to motor incapacity of the animal. The shuttle box, sometimes modified, has been used by many workers. Gellhorn et al. (1942) and Kessler and Gellhorn (1943) replaced the hurdle by a wall with an opening and instead of an auditory CS, visual CS were used. In the experiments of Smith et al. (1956, 1957) monkeys had to move from a lighted to a darkened compartment of a shuttle box. The employment of auditory as well as visual CS permits certain differentiations (John et al., 1958a, b ) . Circumstances for the performance of the CAR may be altered by varying the interval between CS and US. A long interval permits the measurement of reaction time or latency (Battig and Grandjean, 1955, 1957; Grandjean and Battig, 1957; Horisberger and Grandjean, 1958; Nieschulz et al., 1958).

B. VERTICAL ESCAPERESPONSES ( POLE-JUMP) By modifying a method of Winter and Flataker (1949, 1951) whereby fasted rats climb a rope to a platform for food (positive reinforcement), Courvoisier et al. (1953) trained rats to climb a rope after an auditory CS to avoid an electric shock US provided by a grid in the bottom of the cage (“rat grimpeur”). This method for eliciting a CAR has been widely used. Cook et al. (1955) introduced a wooden pole with a roughened surface as the safety area. Gatti ( 1958), by lengthening the pole and recording graphically the movements of the rats on the pole, made it possible to study

232

ALBERT HERZ

the motor performance of the animals as well. The pole-jump method also brings about three types of responses: CAR, UR, and failure of UR. In addition, well-trained animals sometimes make a secondary conditioned avoidance response ( SCR) : the animals jump up the pole as soon as they are placed in the cage without waiting for the CS (Cook and Weidley, 1957; Nieschulz et al., 1957). Maffii (1959) considered this an important aspect of behavior, and investigated the action of a series of substances upon the SCR. Prolongation of reaction time or latency-the period between application of the CS and the jump to the pole-may also be a sensitive measure for changes in conditioned behavior (Jenney and Healy, 1959). In contrast to the situation in the shuttle box where there is a double grid, there is in the pole-jump box a position of security; thus, Nieschulz et al. ( 1957) considers emotional factors, such as fear, to be playing a limited role in the CAR elicited in the pole-jump apparatus.

C. TIMEDROPE-CLIMBING TEST This is not a CAR but is widely used for the study of sedative and disoriented behavior in rats after drugs or the injection of blood plasma from schizophrenic patients. Based on methods devised by Macht (1943) and used by Macht and Hoffmaster (1948) and Winter and Flataker (1949, 1951), fasted rats are trained to climb a vertical rope at the top of which is a platform containing food. The climbing time is measured in this positively reinforced conditioned response.

D. LEGWITHDRAWAL RESPONSE In this method, which comes closest to the classic Pavlovian conditioning technique, animals, usually cats, dogs, or monkeys, have to lift a leg in response to a CS to avoid electric shock to the paw (Wikler, 1946; Funderburk and Case, 1947; Hill et al., 1954a, b; Rutledge and Doty, 1955, 1957; and Domino et al., 1958). This method permits the simultaneous study of changes in autonomic functions, e.g., cardiac rhythm, which accompany conditioned motor reactions (Bridger and Gantt, 1955, 1956; Gliedman and Gantt, 1956).

DRUGS AND THE C A R

233

E. AVOIDANCE RESPONSES DERIVED FROM FREE OPERANT BEHAVIOR Modifications of Skinner’s operant behavior method have been developed in which, instead of working for a reward, animals are trained to avoid a shock by pressing a bar or turning a wheel (Dinsmoor and Hughes, 1956; Sidman, 1953). Harrison and Tracy (1955) employed a high frequency auditory stimulus which was constantly present unless lever-pressing produced a period of silence. These techniques have been adapted by the Owen-Verhave group (1957,1958,1959) and by Weiskrantz and Wilson (1955) for pharmacological purposes. Analysis of the action of various drugs has been made by these more complicated and time-consuming methods; however, the results do not differ significantly from those obtained with the three techniques described previously ( Verhave et al., 1 9 5 8 ~ ). In the four conditioned avoidance methods described, the absence of reinforcement is followed by extinction of the CAR. The influence of drugs on this extinction has been investigated by Ader and Clink ( 1957), Miller et al. ( 1957b), Nieschulz et al. ( 1957, 1958), and Taeschler and Cerletti (1958).

F. MISCELLANEOUS METHODS Other methods, showing some resemblance to avoidance conditioning, have been developed and the results will be discussed to provide a more complete picture of drug action. In an approachavoidance conflict method, introduced by Masserman and Yum (1946), cats are trained to secure food by bar-pressing after a bell signal. Intermittently, an aversive air-blast is superimposed, thus causing neurotic behavior of the animals. Modifications of this procedure have been developed by Jacobsen and Skaarup ( 1955a) and Naess and Rassmussen (1958). In the “cat and mouse test” of Sacra et al. (1957) electric shocks prevent cats from attacking mice. Instead of well-trained rats with an almost 100% performance of the CAR, Jacobsen and Sonne (1955, 1956) employed untrained rats. During the acquisition of the CAR the animals were in a stressful situation. Anticipating the sound signal and the following shock, they showed signs of tension in that they sat stiffened with back curved, forelegs high and had piloerection (Jacobsen, 1958).

234

ALBERT HERZ

This stress-induced behavior can be normalized by suitable substances. The conditions for performance of the CAR are, however, different under these circumstances. 111.

Tranquilizing Agents

A. PHENOTHUZINE DERIVATIVES 1. Chloqn-omazirw Investigation of the central activity of antihistaminic drugs with a phenothiazine configuration led to the discovery of chlorpromazine (CPZ) by Courvoisier et al. ( 1953). CPZ may be regarded as the prototype of a number of potent depressants of the CNS. In addition to its central effects CPZ has many other activities, particularly in relation to the autonomic nervous system. The introduction of CPZ into psychiatric therapy, especially in the treatment of schizophrenia, was revolutionary in its effect. CPZ, by its sedative action, induces a state of lethargy and inactivity, diminishes interest in the environment, and facilitates the induction of sleep, but does not cause any-loss of consciousness. At the present time the site of action of CPZ is assumed to be the ascending reticular system (although its efficacy here seems to be less than originally estimated), the diffuse thalamic projection system, and, last but not least, the hypothalamus. Side reactions of antihistaminic drugs, such as the drowsiness and general sedation observed in man, prompted Macht and Hoffmaster ( 1948) to study Benadryl and Pyribenzamine (0.2 mg intraperitoneally/200 gm rat) in rats trained to climb a vertical rope, one end of which was fastened to the floor and the other to a platform on which food was placed. The decreased performance after these drugs was ascribed to CNS depression rather than to muscular weakness. Winter and Flataker ( 1949, 1951) employing this same method found that several antihistaminic drugs, including phenothiazine derivatives, promazine, and diethazine in doses of 5 to 25 mg/kg (intraperitoneally ) , prolonged the climbing time or abolished the response completely. Experiments with partially curarized rats led these authors to conclude that the results were due to confusion and incoordination produced by CNS depression rather than muscular weakness.

DRUGS AND THE C A R

235

In view of these results Courvoisier et al. (1953) carried out a study on the CNS activity of CPZ in animals. To exclude any interference due to metabolic effects, the positive reinforcement of the rope-climbing method was replaced by an aversive stimulus and Pavlov’s rules on conditioning were adapted for the “rat grimpeur” method. [This then became a CAR (pole-jump) test.] CPZ in subcutaneous doses as low as 1mg/kg retarded the response to the auditory stimulus and slowed down the climbing speed as well. Larger doses inhibited the CAR completely. In general, rats treated with CPZ became disinterested in the ringing bell and were no longer incited into “flight from pain” (Courvoisier, 1956). Since these first experiments of Courvoisier, the action of CPZ on the CAR has been investigated from numerous aspects. Using the pole-jump method, Cook and Weidley (1957) determined the dose-response curve for the CAR blocking action of CPZ. Doses of CPZ which inhibited the CAR almost completely (40mg/kg, orally) only slightly affected the UR, and indicated a specific block of the CAR. These results have been corroborated by many authors. It was demonstrated by Guha et al. (1954) that the suppression of the CAR in rats by CPZ was related to the blocking of hypothalamic areas. The experiments of Maffii (1959) showed that the SCR was also blocked in a specific manner. The same effects of CPZ on the CAR of rats have been obtained with the use of the shuttle box, Ambrus et al. (1957) stated that small doses of CPZ abolished the CAR without affecting the pain threshold. CPZ, however, proved to be ineffective in altering the stress-induced behavior of untrained animals. Although doses up to 40 mg/kg (subcutaneously) were employed, the tenseness of the rats was not reduced; the number of conditioned responses was but slightly reduced ( Holten and Sonne, 1955; Jacobsen, 1957). Smith et al. (1957) observed that lmg/kg of CPZ almost completely abolished the CAR of monkeys in a shuttle box. However, no difference in the response of normal monkeys and monkeys with lesions of the temporal and/or frontal cortex could be demonstrated. This is in contrast to the effects of reserpine on the same group of animals. These findings may indicate that the brain cortex is not essential for the action of CPZ. Using leg withdrawal of dogs as an avoidance response, Glied-

236

ALBERT HERZ

man and Gantt (1956) found that the retention of the motor conditioned reflex, an indicator of learned motor reactivity, was interrupted by CPZ, whereas the conditioned cardiac reflex was but slightly influenced. The motor and cardiac orienting reflexes, indicators of unlearned responses, were abolished by even smaller doses of CPZ. In similar experiments, Domino et al. (1958) showed that CPZ, in doses producing minimal or no motor deficit, depressed the avoidance response more markedly than the escape response. CPZ (along with promethazine and some antihistaminic drugs) also decreased the intensity of conditioned, not especially designed nociceptive, reflexes of dogs (Votava and Vanecek, 1956). A modification of the lever-pressing technique to avoid shock (wheel-turning) was applied by Verhave and co-workers (1958a,b, c, 1959). With suitable doses of CPZ avoidance behavior was depressed without abolition of escape behavior, again demonstrating a specific action. Klupp and Kieser (1959) employed a comparable method which also included positive reinforcement. The effects of CPZ were very similar in both kinds of reinforcement and could not be distinguished from the sedative action of phenobarbital. In a discrimination-avoidance situation CPZ also showed depressant activity (Boren, 1957). The neurotic behavior of cats in an approach-avoidance conflict situation could not be normalized by CPZ (Jacobsen and Skaarup, 1955b); however, in the “cat-mouse test” of Sacra et al. (1957) and in the approach-withdrawal reaction of Naess and Rasmussen (1958) CPZ was found to be effective. The use of CPZ during the acquisition and extinction of the CAR has been studied by several investigators. Herz (1960a) observed no significant difference between the blocking activity of CPZ in rats with a recently acquired pole-jump response and well-trained animals. ( See Section VII, anticholinergic drugs. ) Under the influence of CPZ (1.5-3mg/kg) rats were significantly retarded in their acquisition of an avoidance response in a shuttle box. CPZ also facilitated the extinction of the CAR. Assuming that CPZ diminished anxiety and that this is a motivitating factor in the avoidance situation, Ader and Clink (1957) predicted these results. In a similar investigation Miller et al. (195713) also observed an acceleration of extinction of the “fear motivated response” in rats. To exclude motor deficit as the cause of CPZ action, control

DRUGS AND THE C A R

237

experiments with phenobarbital were performed in an escape runway. It is questionable that the authors proved what they set out to demonstrate. Other experiments by this same group have shown that learning proceeds during CPZ treatment (Miller et al., 1957a). In cats, chronically prepared with implanted brain electrodes, Rutledge and Doty (1955, 1957) showed that 7mg/kg CPZ completely blocked a leg flexion CAR. The cats were conditioned to auditory or visual stimuli. In contrast, the same dose only slightly diminished leg flexion produced by cortical stimulation. This differential effect was not observed with morphine or with barbiturates. These results are in agreement with those of other investigators which indicate that CPZ may act somewhere in the afferent or integrating pathways of the brain. 2. Modification of Chloqzwomzine Inhibition of the CAR

The antagonistic effect of CNS stimulants on the CPZ-induced prolongation of the reaction time in rats, conditioned to pole-jumping, was intensively investigated by Jenney and Healy (1959). Amphetamine, ephedrine, and metamphetamine completely and rapidly reversed the CPZ action. Caffeine was only slightly less effective. Methylphenidate, pipradrol, and cocaine decreased the inhibition for about 2 hours, while the antidotal effect of Metrazol and Megimide lasted only a few minutes. Some of the effect may be due to improvement of the muscle tone by these stimulants, especially in the case of methylphenidate, cocaine, and pipradrol. Methylphenidate and pipradrol were also found by Courvoisier et al. (1957) to antagonize the CAR inhibition. These compounds also reversed the catalepsy produced by CPZ (and perphenazine), but not the antiemetic effect. The possibility of a connection between catalepsy and deconditioning will be discussed later. Small doses of LSD significantly antagonized CPZ in the experiments of Cook and Weidley (1957). On the other hand, Winter and Flataker (1957) observed enhancement of the CPZ inhibition of ropeclimbing by higher doses of LSD. The degree of CPZ-induced inhibition of the CAR was increased but the duration shortened by small doses of atropine or scopolamine (Jenney and Healy, 1959).

238

ALBERT HERZ

3. Effect of Other Phenothiazine Derioatizhes The numerous phenothiazine derivatives (see Fig. 1)which have been synthesized and introduced into therapy since CPZ differ mostly quantitatively in the various psychopharmacological tests, including inhibition of the CAR. CPZ, promazine, promethazine, and 2-chloropromethazine were studied comparatively by Fellows PHENOTHIAZINE DERIVATIVES

R

R'

PROMAZINE

&H~-cH~-cH~C b

CHLORPROMAZINE

C H ~ - C H ~ - C H ~ - ~ ~

-GI

TRIFLUPROMAZINE

~H,CH~- CH,NCCY

-CF,

iH2-CH2

-S-CY

cn,

PROMETHAZINE

MEPAZINE

THIORIOAZINE

1H, r\

PERAZINE

CH~-CH,-

CHLORPERAZINE

CH, - CH, -CH,-N n N-CY v

PERPHENAZINE

hH2-GHz-GHe-NvN-CH,-CH,OH

CH,- N

N-GH,

v

A

- CI -a

FIG. 1. Structures of phenothiazine derivatives which have been studied for their effect on the conditioned avoidance response.

DRUGS AND THE C A R

239

and Cook (1957) and Cook et al. (1958) by the CAR (polejump) in rats. The slope of the dose-response curves of the four substances were very similar; the subcutaneous doses producing a 50% block were 2.4, 6.2, 12, and 29.5mg/kg, respectively. The relative potency of these four compounds on the CAR also correlated with the inhibition of spontaneous motor activity in mice. Comparable findings are reported for the relative activities of CPZ and promazine on the CAR by Wirth (1958) and Maffii ( 1959), whereas Courvoisier ( 1956) observed only prolongation of the reaction time in the case of promethazine. Gatti (1958) reports that promazine reduces muscular tone of rats. Triflupromazine was investigated by Burke et al. (1957). In comparison to CPZ it is three times more potent in blocking the CAR, five times more active in taming monkeys, and the toxicity is comparable to that of CPZ. Acylized derivatives of CPZ were investigated by Wirth et al. (1958).

4 . Piperidyl Derivatives of Phenothiazine Mepazine was examined by Nieschulz et al. (1957). Oral doses of 50mg/kg retarded only slightly the CAR (pole-jump) in rats and 25mg/kg did not modify the extinction of this CAR. With the shuttle-box technique, no significant effects on acquisition and extinction of the CAR could be observed; however, in experiments with CPZ these authors were unable to find a specific action on the CAR which could be differentiated from barbiturates. (See Section XI, A; discussion.) Thioridazine, in the experiments of Taeschler and Cerletti (1958), abolished the CAR of rats at subcutaneous doses of lOmg/kg, whereas, extinction of the CAR was facilitated by doses as low as 2mg/kg. The toxicity was found to be less than that of CPZ.

5. Piperazinyl Derivsltiues of Phenothiazine Perazine showed but slight influence on the CAR of rats in the shuttle box, at a dose of 25 mg/kg (Nieschulz et al., 1958). Chlorperazine was 1.6 times more active than CPZ in blocking the CAR (pole-jump) (Fellows and Cook, 1957). Several hours after injection, when the distinct cataleptic action occurred, an

240

ALBERT HERZ

inhibition of the UR was also observed (Courvoisier, et al., 1957). Perphenazine showed a sevenfold greater activity than CPZ after oral doses and a fourteenfold greater activity after subcutaneous injection in blocking the CAR. The LDBOwas only 2/3 to 1/2 that of CPZ. A somewhat better dissociation between inhibition of avoidance and escape behavior was obtained with perphenazine than with CPZ (Irwin and Govier, 1957; Irwin et al., 1959). A muscular component in the inhibitory activity was assumed by Gatti ( 1958). Piperazinyl and morpholino derivatives of phenothiazine with and without acyl substitution were investigated by Wirth et al. (1958).Substances which produced a strong cataleptic effect also frequently blocked the UR. 6. Substances Related to Phenothiazine (See Fig. 2 )

The trans-isomer of 2-chloro-9-( 3'-dimethylaminopropylidene ) thiaxanthene ( chlorprothixene ) was investigated by Mgller-Nielsen and Neuhold (1959). Not only the structure, but also the action showed close resemblance to CPZ. This congener also produced a specific block of the CAR when given in comparable dosage to CPZ. N-Alkyl-piperidyl carbazoles, substances possessing a very complex central activity inchding some stimulant effect, were examined by Nieschulz et al. (1959). The CAR was but slightly affected by these compounds. Hexamid [5,8phenylethyl3-( (3-diethylaminoethyl) -2,4,6-trioxohexahydropyrimidine]produces mainly a depressant effect on the CNS and is used as preanesthetic medication. Stille et al. (1956) observed a distinct but transient inhibition of the CAR in rats climbing for a positive reinforcement. Imipramine (Tofranil) was reported by Sigg ( 1959) to produce no impairment of the CAR in rats. In other tests this compound showed some stimulant effect and is now recommended for the therapy of psychiatric depressions.

7. Summary The numerous studies on CPZ as an inhibitor of the CAR are almost unanimous in regard to the specific effect of this drug. The CAR is inhibited by doses which largely leave the UR unchanged. The reaction time is prolonged; the extinction of the CAR becomes facilitated. Stimulating agents antagonize this specific block.

241

DRUGS AND THE C A R

Stress-induced behavior is but slightly influenced by CPZ. The uniformity in the effect on approach-avoidance con%icts is not as well established. In general, the inhibition of the CAR by SUBSTANCES RELATED TO PHENOTHIAZINE

a ; D C 1 CH-CH,-CH,-W;~

bH2-0;;

CHLORPROTHIXENE (FRUXAL)

N -ALKYL-PIPERIDYL CARBAZOLES

7 -

c 6 H g x H2 C Hp-N:"

c2

" Y

CzHs 0 HEX AMID

IMIPRAMINE (TOFRANIL)

ADRENERGIC BLOCKADE

ETHOXYBUTAMOXANE

F-883

Hz NA-86

RELATED TO ATROPINE

L J BENACTYZINE

FIG.2. Structures of miscellaneous compounds which have been studied for their effect on the conditioned avoidance response.

phenothiazine derivatives seems to be correlated in some manner with their cataleptic or depressive effect. Newer phenothiazines are more potent but the therapeutic index is approximately the same.

242

ALBERT HER2

The available data do not permit a more exact statement. Further aspects of the actions of CPZ will be discussed in Section XI.'

B. Rauwolfia ALKALOIDS

1. Reserpine The tranquilizing effects of reserpine differ from those of most other substances possessing central depressant activity. Characteristic of the action in man is a state of indifference, tendency to sleep, and occasionally psychiatric depression. Contact with the environment is not impaired, but the emotions are considerably flattened. The locus of action may be the sympathetic centers of the hypothalamus, the suppression of which also produces the typical autonomic syndrome: ptosis, miosis, salivation, tremor, bradycardia, and hypotension. Besides the hypothalamus, rhinencephalic structures and the mesodiencephalic activating system may also be affected by reserpine. There is considerable probability that reserpine does not produce the characteristic change in behavior by a direct action of the CNS. It is assumed that certain actions may be mediated by alteration of the brain content of biogenic amines. A synopsis of the actual state of this knowledge is given by Truitt ( 1959) . In the experiments of Cook and Weidley (1957) reserpine, given orally to rats in two doses of 25 mg/kg over a 17-hour period 2 Note added in proof. The central depressive activity (locomotor activity of rats, prolonging effect on barbiturate anesthesia in rats, effect in the climbing test in mice) of CPZ, promazine, perphenazine, chlorperazine, and mepazine was comparatively assayed by Sandberg (1959). The effectiveness of the substances in the three tests showed some parallelism and resembled the previously reviewed potency to block the CAR. Also Taeschler et al. (1960) investigated the activities of a series of phenothiazine derivatives (CPZ, thioridazine, perphenazine, chlorperazine, thiopropazat and others ) in various tests. The inhibition of the CAR (pole-climbing response in rats) caused by these compounds showed many resemblances to their cataleptic power. The potency of the substances to inhibit the emotional defecation was not the same; thioridazine prevented the emotional defecation at doses which did not affect the CAR, whereas in the case of other phenothiazine derivatives, especially perphenazine and thiopropazat, the relation was inverse. The authors consider this inhibition of the emotional behavior as very important regarding the therapeutic-neuroleptic purposes. The inhibition of the CAR is related to extrapyramidal disturbances, produced by these drugs.

DRUGS AND THE C A R

143

prior to test, produced a 65% block of the CAR (pole-jump). In rats, after a subcutaneous dose of 0.5 mg/kg, the inhibition of the CAR began 120 minutes after the injection and lasted more than 10 hours (Gatti, 1957, 1958). Small doses also blocked the SCR specifically (Maffii, 1959). Knoll (1959) found that small subcutaneous doses of reserpine (0.1 mg/kg) given daily tend to cumulate. A specific reserpine-induced block of the CAR in rats, using a shuttle box, was obtained by Slater (1958). Prolongation of the reaction time and partial loss of the CAR following oral reserpine ( 5 mg/kg) was observed in rats in a shuttle box (Horisberger and Grandjean, 1958). The conditioned leg-withdrawal response of dogs was found to be blocked by reserpine in a dosage which did not influence conditioned cardiac responses ( Gliedman and Gantt, 1956). Jacobsen ( 1957), using 5 to 10 mg/kg of reserpine in rats, observed, shortly after subcutaneous administration, a clear-cut decrease in tension and stiffened attitudes, but no effect on the conditioned response occurred before onset of stupor. The short interval between injection of the drug and performance of the test, however, may not have been optimal for studying reserpine effects. Using rhesus monkeys conditioned to move from the lighted to the darkened compartment of a shuttle box Smith et al. (1956) found that reserpine in intramuscular doses of 0.15 to 0.6mg/kg blocked the CAR without loss of the normal motor function; maximal decrease in performance was reached about 4 to 6 hours after injection. Monkeys with lesions in the temporal lobes were affected less by reserpine than were normal animals. Frontal-lobe-lesioned monkeys were intermediate in their response to reserpine. Rescinnamine and CPZ did not show this selectivity of action in these lesioned animals (Smith et al. 1957). These results are in accord with the concept of Schneider et al. (1955) that, for the full effect of reserpine, the presence of the cortex is necessary. Further studies on monkeys by Riopelle and Pfeiffer (1958) showed that tolerance develops towards the inhibition of the CAR after a 2-month administration of reserpine at a daily dose of 0.25 to 0.5 mg/kg per 0s. However, in the performance of tests of “multiple discrimination and delayed responses” tolerance was not manifest. In monkeys having to bar-press to avoid shock and obtain food, Weiskrantz and Wilson ( 1955) found reserpine (0.75 mg/kg)

244

ALBERT HERZ

strongly depressed avoidance and approach behavior. Bilateral amygdalectomy did not alter the effects of reserpine ( Weiskrantz, 1957). These experiments were initiated because the behavior of the reserpinized animals showed many resemblances to amygdalalesioned monkeys. When monkeys were brought into a conditioned fear situation, where they could not avoid shock, Weiskrantz and Wilson (1956) reported that reserpine produced inhibition of both learning and performance. It is assumed that as the drug depresses sensory input, the animals become nonreactive and nonattentive, reacting only to very strong stimulation ( Weiskrantz, 1957). However, in the experiments of Riopelle and Pfeiffer (1958) the learning of discrimination problems in monkeys was not impaired by reserpine. Also in a conditioned fear situation in rats no deterioration of learning after reserpine could be observed by Stein (1956). Klupp and Kieser (1959) investigated the action of reserpine on rats working in a modified Skinner box. Both positively and negatively reinforced responses were inhibited. In the approach-avoidance conflict situation of cats, reserpine simply caused stupor, thus prolonging the feeding cycles, while behavior was not normalized (Jacobsen, 1957). In experiments of John et al. (1958a, b ) , reserpine produced graded responses, when injected into the lateral ventricle of cats in doses from 20 to 200pg. In suitable dosage the avoidance response was affected without abolishing the escape behavior. At sufficiently low doses only the avoidance response to the visual stimulus was blocked, while the response to auditory cues and the approach responses (discrimination for food) remained unaffected. These findings may demonstrate that the loss in avoidance behavior is not due to sensory deficit. The peripherally administered intramuscular dosage of reserpine required for inhibition of the CAR (7.5 to 65 pg/kg) was frequently remarkably smaller than the dosage applied intraventricularly.

2. Alteration of the Effects of Reserpine Small doses of LSD which do not affect the CAR (pole-jump) of rats (0.1 to 0.5 mg/kg, subcutaneously) significantly antagonized the specific reserpine-induced block (Cook and Weidley, 1957). On the other hand, pretreatment with low doses of reserpine (50

DRUGS AND THE C A R

245

pg/kg daily for 5 days) enhanced the effects of LSD (Winter and Flataker, 1957). Since many of the pharmacological reports show reserpine to have parasympathomimetic effects, Pfeiffer ( 1959) investigated atropine as an antagonist to the reserpine-induced inhibition of the CAR. Atropine was ineffective. Pretreatment of rats with iproniazid ( 100 mg/kg) 24 hours prior to reserpine completely prevented the effects of the latter (Horisberger and Grandjean, 1958). This agrees well with the observation that premedication with this monoamine oxidase inhibitor also abolishes sedation induced by reserpine ( Besendorf and Pletscher, 1956). However, the effects of intraventricularly injected reserpine on the CAR in cats as reported by John et al. (1958b) was not influenced by centrally injected serotonin and iproniazid. Norepinephrine, eserine, and atropine were also ineffective, while epinephrine sometimes attenuated the reserpineinduced inhibition of the CAR. Metamphetamine completely reversed the effects of reserpine, when injected intraperitoneally ( 1 to 5 mg/kg), but this stimulant was not as potent when administered centrally (0.5 to 2 m g ) . Therefore, John considers a peripheral mechanism to be responsible for the action of reserpine on the CAR. One might also question whether intraventricular application supplies the drug to the ordinary cite of action in the brain. 1-Piperidinomethyltetralon-2 ( Na-86), a substance with sympathetic blocking activity was synergistic with reserpine when tested on the CAR in rats (Knoll, 1959). 3. Other Reserpinelike Substances

Smith et al. (1957) also investigated the action of other Ruuwolfia serpentina alkaloids with central activity. Rescinnamine and demethoxyreserpine blocked the CAR after intramuscular doses of 0.5 mg/kg in monkeys. A comparison of these compounds with reserpine showed that they are qualitatively indistinguishable from the latter. Demethoxyreserpine proved the most rapid in onset of action, rescinnamine the slowest. In brain-lesioned monkeys, rescinnamine produced the same effect as in intact animals, while demethoxyreserpine was significantly less active in animals with lesions in the frontal cortex. Knoll (1959) found Tetrabenazine also blocked the CAR in a manner very similar to reserpine.

246

ALBERT HERZ

For further results in connection with reserpine see 5-hydroxytryptamine and adrenergic drugs ( Sections VI and VIII ).

4. Summary Reserpine inhibits the CAR in a specific manner. The extended latent period before onset and the long duration of action are characteristic. Prolonged administration of small doses produces a cumulative effect. The inhibitory action of reserpine resembles that of CPZ, but there are some facts demonstrating differences in the mechanisms of action of the two substances. The three active Rauwolfia alkaloids show similar activity. Some aspects of reserpine action will be discussed later in connection with other compounds. C. DIPHENYLMETHANE DERIVATIVES

1. Benactyzine (Benxilic acid diethylaminoethyl ester) Starting with the working hypothesis that acetylcholine is a possible neurotransmitter in certain portions of the brain such as the thalamus and hypothalamus, where emotions seem to be mediated, Jacobsen and co-workers (Holten and Sonne, 1955; Jacobsen and Sonne, 1955, 1956; Jacobsen and Skaarup, 1955a, b ) investigated anticholinergic drugs in a program of behavioral research. For clinical trial they selected benactyzine which had previously been described as a spasmolytic drug resembling atropine. The action of benactyzine differs in many respects from that of tranquilizing drugs like CPZ and reserpine. Alexander (1957) classifies benactyzine as an antiphobicum and clinically it is used for treatment of certain types of psychoneuroses. The central relaxation produced by benactyzine in man is associated with blocking of thought processes and decreased mental concentration (Jacobsen, 1958). Jacobsen and Sonne (1955) examined the action of benactyzine and other compounds upon stress-induced behavior of untrained rats in a modified shuttle box. Benactyzine at a dosage of 1 to 20 mg/kg abolished the stress symptoms: the rats gave up their tense attitudes, moved around and examined the cage, and behaved like normal animals. They also reacted more frequently to the conditioned stimulus. This facilitation of the CAR under the action of benactyzine, however, does not mean that a real acceleration of the

DRUGS AND THE C A R

247

acquisition of the CAR occurs, as withdrawal of the drug reduced the number of conditioned responses again. The reaction time was also shortened. Without further reinforcement, extinguished conditioned responses partially returned after administration of benactyzine. More complicated reactions, however, were inhibited (Jacobsen and Sonne, 1956). Holten and Sonne ( 1955) investigated a series of compounds for their effect in normalizing stress-induced behavior in rats. Several esters of benzilic acid were as active as benactyzine. Trasentine and atropine showed a low degree and scopolamine a very high degree of efficiency. Quaternary ammonium compounds were ineffective, very likely because they sparingly pass the blood-brain barrier. [For further results with Jacobsen’s method see CPZ, reserpine (Section 111, A, B ) , barbiturates, morphine, and alcohol (Sections IV and V ) . ] Benactyzine was also found to be effective in a similar stress situation by Marchal and Schlag ( 1958). Cook et al. (1958) found benactyzine to be ineffective in suppressing the CAR (pole-jump) of rats. Employing the same method, Gatti ( 1957, 1958), however, reported benactyzine (50 mg/kg, subcutaneously) to block the CAR; the UR also was abolished. The animals seemed to be dissociated from their environment. Herz (1960a) found that benactyzine would inhibit the CAR early in the period of training. Later, when the conditioned response was well established, the performance of the CAR was accelerated. There might be a connection between the effect of benactyzine upon conditioned responses during the state of acquisition and blocking of mental processes in man (Jacobsen, 1958). A prolongation of the reaction time in multiple choice tests and an increase in the frequency of errors occurred after trials of benactyzine in man (Hess and Jacobsen, 1957). The SCR was also found to be blocked; however, there was a great variability in responses (Maffii, 1959). Jacobsen and Skaarup ( 1955a) completed their investigations on benactyzine by use of the technique of Masserman and Yum (1946), wherein the behavior of cats was examined in an approachavoidance conflict situation. Among several benzilic acid amino esters benactyzine proved to be most effective in normalizing the animals’ behavior; whereas scopolamine and morphine were ineffective (Jacobsen and Skaarup, 1955b). The same results

248

ALBERT HER2

were obtained by Naess and Rasmussen (1958) and by Sacra et al. ( 1957), using similar techniques. In a discrimination avoidance situation, wherein rats bar-press to postpone a warning stimulus and an electric shock, Boren and Beyer (1957) observed a stimulating effect of benactyzine. In an operant behavior technique, low doses of benactyzine often increased the over-all response rate, whereas higher doses produced a decrease. Scopolamine or atropine had a similar effect (Boren and Navarro, 1958). For further results see Section VII on anticholinergic drugs.

2. Azacyclonol (Frenquel) Azacyclonol is used clinically in confusional and hallucinatory states and has had trials in the schizophrenic psychoses. However, it has but slight activity upon motor excitement. Alexander (1957) terms azacyclonol an ataracticum. Several investigations indicate that this substance is without effect on the CAR in both rats (Cook et al., 1958; Maffii, 1959) and monkeys (Pfeiffer et al., 1957). A specific block of the SCR is reported by Maffii (1959).

3. Hydroryzine (Aturax) Hydroxyzine, an antihistaminic drug, induces drowsiness in man and is recommended in clinical states of anxiety and tension. Employing a shuttle box, Levis et al. (1957) observed inhibition of the CAR in rats after subcutaneous or oral doses of 60 to 80 mg/kg. But Cook et al. (1958) observed no decrement of the pole-jump CAR after hydroxyzine. Gatti ( 1958) reported slowing down of performance of the CAR after 100 to 2OOmg/kg, but hypotonia of the muscles similar to that produced by relaxing substances was noted. Maffii (1959) observed only a block of the SCR.

4. Summary The action of benactyzine differs from that of CPZ and reserpine, The consolidated CAR is not inhibited, the performance time is shortened. Insufficiently consolidated and more complicated conditioned reactions, however, are inhibited. The tension of

DRUGS AND THE C A R

249

animals in stress situations is reduced, conflict-induced behavior becomes normalized also. The clinically noted benefit might be partially correlated with these effects. The action of benactyzine is in many respects similar to that of scopolamine. Azacyclonol and hydroxyzine show little or no influence upon conditioned behavior.

D. INTERNEURON BLOCKING AGENTS Meprobamate is a possible representative of this group of substances, also called “central relaxants.” Meprobamate resulted from Berger’s attempts to prolong the fleeting action of mephenesin (1952). It is now evident that this substance not only inhibits the spinal multineuronal synapses, but also influences higher regions of the CNS. The taming effect of meprobamate in monkeys was one indication that this substance might be a tranquilizing agent (Berger, 1954). Meprobamate, which possesses no action on the autonomic nervous system, is difficult to classify pharmacologically. Pfeiffer et al. (1957) found that this substance had barbituratelike plus some CNS-stimulating effects. Meprobamate did not block the pole-jump CAR in rats (Pfeiffer et al. 1957; Gatti, 1958; Maffii, 1959). Neurotoxic doses (500 mg/kg, orally) produced a nonspecific inhibition; this blocking was found by Cook and Weidley (1957) to be reduced by small doses of LSD. Using the wheel-turning procedure in rats, Verhave et a2. ( 1958c) observed that 500 mg/kg of meprobamate produced only a small increase in latency of the avoidance response. In the experiments of Pfeiffer et al. (1957) the CAR of monkeys in a shuttle box was not inhibited by oral doses of 200mg/kg meprobamate although the animals showed heavy sedation and ataxia. However, Maffii (1959) found inhibition of the SCR. In 7-day-old ducklings the appearance of learned avoidance behavior was not prevented by meprobamate, though it reduced the symptoms of emotional disturbance (Hess, 1957). Jacobsen (1957) reported some slight enhancement of a conditioned response at very high doses (500 to 800mg/kg), which, like benactyzine, also normalized the stress-induced behavior of rats. In dogs the conditioned leg-withdrawal response was not depressed by oral doses of 100 mg/kg, although these doses produced distinct motor deficit (Domino et al. 1958).

250

ALBERT HERZ

In the approach-avoidance conflict behavior of cats meprobamate was found by Sacra st al. (1957) to have some activity. Meprobamate also showed an antineurotic effect in rats in similar tests by Naess. Some subtle effects of meprobamate on conditioned emotional responses were described by Hunt ( 1957); however, it could not be determined whether these reflected muscular relaxation or a real action of the drug on conditioning. Other interneuron blocking agents such as mephenesin, phenaglycodol, and 2-thienyl-5-amino-l,3,4-thiadiazole ( L-1458) were investigated by Maffii (1959). None of these affected the CAR, while the latter two compounds showed some activity in blocking the SCR. The compound 3- [2- ( l-methyl-2-piperidyl) ethyl] indole did not affect the pole-jump CAR in nondepressive doses (Julou et al., 1957). Summary. Interneuron blocking agents do not inhibit the CAR; the occasional retardation of the conditioned responses at very high dosage may be due to muscle weakness. In approach-avoidance conflict situations, meprobamate seems to be an effective barbituratelike sedative.

IV. Hypnotics and Sedatives

The action of substances which were formerly extensively employed where tranquilizers are now used is of more than academic interest. A careful study of sedatives and hypnotics may allow a more detailed delineation of the psychopharmacology of various tranquilizing drugs.

1. Barbiturates Cook and Weidley (1957) found that barbital (125 mg/kg, intraperitoneally) would block the pole-jump CAR of rats in a nonspecific manner; CAR and UR were simultaneously abolished by doses which caused ataxia and semiprostration. However, a moderate degree of specific block was produced by pentobarbital ( 20 mg/kg, intraperitoneally ) but again only at neurotoxic doses. Similar results were obtained by Courvoisier and Julou (1956) with phenobarbital and butabarbital, Stille et al. (1956) with

DRUGS AND THE C A R

251

phenobarbital, and Knoll and Knoll (1958) also with phenobarbital. A specific block of the SCR, but not of the CAR, was obtained by Maffii ( 1959) using phenobarbital and pentobarbital. In the shuttle-box method, barbital (10 to 80 mg/kg) resulted in prolongation of the reaction time of rats (Biittig and Grandjean, 1955, 1957), but only at a dosage which also produced motor deficit. In the experiments of Miller et al. (1957b), doses of phenobarbital which produced some inhibition of escape behavior of rats did not facilitate extinction of the CAR, while comparable doses of CPZ did. In similar experiments of Nieschulz et al. ( 1958) butabarbital accelerated extinction; but this effect was interpreted as a typical sedative action. Monkeys, drugged with pentobarbital, attempted to jump the barrier in a shuttle box but were prevented by ataxia (Smith et al. 1957). Pentobarbital in cats, though producing marked motor impairment, only slightly affected avoidance behavior. Stressinduced behavior in rats could not be normalized by phenobarbital according to Holten and Sonne (1955). The CAR (leg-withdrawal) of dogs was abolished by phenobarbital only at that dosage which prevented escape from shock (Domino ,et nl. 1958). Weiskrantz and Wilson ( 1955) found pentobarbital to be only a slight inhibitor of the CAR of monkeys, barpressing to avoid shock. In similar experiments in rats, Verhave et al. (1957, 1958b,c) found that the effect of pentobarbital and secobarbital on avoidance behavior followed closely that of escape responses, which also indicated a relatively nonspecific action of barbiturates. In a repetition of these experiments using a cross-over design, the difference between avoidance and escape behavior was slightly more distinct (Owen et al., 1958). In an approach-avoidance conflict situation, amobarbital ( Bailey and Miller, 1952; Naess and Rasmussen, 1958) and pentobarbital (Masserman and Siever, 1944) proved to be effective. Bailey and Miller stated that amobarbital produced a greater reduction in the fear-motivating avoidance than the hunger-motivating approach. 2. Urethun, Methylparafynol, and Glutethimide Urethan showed but slight inhibitory effect on the CAR at doses which were 50% of the narcotic dose (Knoll and Knoll,

252

ALBERT HER2

1958). Methylparafynol affected the CAR at the neurotoxic dose only (Cook and Weidley, 1957; Knoll and Knoll, 1958). Glutethimide was very similar to barbiturates in its action ( Maffii, 1959).

3. Ethanol Only a few reports exist on the action of ethanol on the CAR. Two milliliters of a 10% solution given orally to rats prolonged their reaction time in a shuttle box (Battig and Grandjean, 1955). Stress-induced behavior of rats could not be normalized by alcohol, even at doses which produced ataxia, The CAR was but slightly inhibited (Holten and Sonne, 1955; Jacobsen, 1957). Using the rope-climbing response with positive reinforcement, no significant effect of alcohol was observed on climbing time at doses which produced distinct signs of intoxication (Winter and Flataker, 1951). Several authors have examined the action of alcohol upon approach-avoidance responses. Conger ( 1951) found that hungry rats, discouraged by electric shocks from running down a short alley to get food, resumed running to the goal for the food reward after the administration of alcohol. Masserman and Yum (1946) and Jacobsen and Skaarup (1955a) obtained comparable results on the effect of alcohol in cats under similar experimental conditions. 4. Summary

Barbiturates and similar depressant compounds inhibit the CAR in a nonspecific manner. With some compounds such as pentobarbital and secobarbital a slight difference in the doses influencing the CAR and UR may be found, but even these doses produce ataxia. However, some effectiveness of CNS depressants in approach-avoidance conflict situations is evident. Alcohol is effective in approach-avoidance conflicts, but does not inhibit the CAR. V.

Morphine and Other Analgesics

According to Cook and Weidley (1957) morphine specifically blocks the CAR (pole-jump ) of rats; 16 mg/kg, intraperitoneally, completely abolished the CAR, while the UR remained unaffected. Similar results were obtained by Herz (in press) and by Maffii

DRUGS AND THE C A R

253

(1959), who found that the SCR was also specifically blocked. Prolongation of reaction time was observed by Gatti (1958), but the speed of pole-climbing itself was not found to be retarded. Courvoisier (1956) observed no specific activity of morphine other than its hypnotic action. Morphine, as well as meperidine and methadone, showed but slight activity in normalizing stress-induced behavior of rats (Holten and Sonne, 1955). In the CAR ( leg-withdrawal) of dogs, morphine impaired differentiation or abolished the CAR (Wikler, 1946). In similar experiments by Stephens and Gantt (1956) morphine depressed the motor CAR at a dosage which did not alter cardiac conditioned reflexes (schizokinesis). Domino et al. ( 1958), however, observed no inhibition of the CAR (leg-withdrawal) of dogs by morphine in doses which did not also produce motor deficit. A distinct dichotomy between avoidance and escape behavior was evident when morphine was administered to rats in a wheel-turning procedure (Verhave et al., 1958c; Verhave et al., 1959). Edwards ( 1959), investigating morphine, codeine, meperidine, and ketobemidone, observed that the spinal reflex (flicking the tail in response to a thermal stimulus), unconditioned withdrawal, and conditioned avoidance reflex were impaired at approximately the same dose of these analgesics. The relationship between the analgesic effect and the CAR inhibitory effect of morphine was investigated by Ambrus et al. (1957) and by Verhave et al. (1959). The shuttle-box CAR of rats was blocked by 4mg/kg, a dosage which did not alter pain threshold. On the other hand, a morphine dose, which produced distinct analgesia, only slightly affected the wheel-turning CAR of rats. A state of conditioned fear in rats working in a Skinner box for food was found to be reduced by morphine. Animals, inhibited from bar-pressing by anticipation of shock, resumed their bar-pressing after morphine. Hill et al. (1954a, b ) propose this test for the screening of analgesics. Alteration of the morphine action. Morphine inhibition of the CAR was not antagonized by LSD in contrast to CPZ and reserpine inhibition. Thus, Cook and Weidley (1957) hold that a different mechanism is responsible for the blocking action of morphine on the one hand and of CPZ or reserpine on the other. However, morphine-induced inhibition of the CAR was reversed by nalorphine, which by itself failed to alter the conditioned behavior. The

254

ALBERT HERZ

action of CPZ could not be reversed by this morphine antagonist; on the contrary, a slight enhancement occurred. Herz (1960b) found that the addition of scopolamine caused subeffective doses of morphine to inhibit the CAR (pole-jump) of rats, even though scopolamine itself did not abolish a well established CAR, but instead shortened latency. This action of scopolamine is in accordance with its potentiating depressant effect when used with morphine as a preanesthetic agent. The analgesic effect of morphine was not increased, however. Bulbocapnine, included here only because of its structural similarity to apomorphine, produces a well known state of catelepsy in animals. This catalepsy is enhanced by scopolamine and counteracted by cocaine, amphetamine, and tetrahydro-fi-naphthylamine (Kerman, 1944). In rats, bulbocapnine catatonia resembles that of morphine. A dose of 32 mg/kg in rats produced maximal inhibition of the CAR, although the escape response remained largely unaffected (Verhave and Owen, 1958). Bulbocapnine inhibition of the CAR (pole-jump ) was rapidly reversed by amphetamine ( 5 mg/kg) whereas morphine inhibition was unaltered (Jenney and Healy 1959). Summary. In most instances morphine acts in the same manner as CPZ and reserpine, showing a specific effect on the CAR. The peculiarity of this specific block is evidenced by the antidotal effect of morphine antagonists and the lack of antagonism by LSD. The analgesic effect of morphine seems not to be related to the inhibition of the CAR. Nalorphine alone does not inhibit the CAR or antidote the effect of CPZ. Bulbocapnine inhibits the CAR of rats and is antidoted by amphetamine.

VI. Adrenergic and Adrenergic-Blocking Substances The distinct distribution of norepinephrine in brain structures (Vogt, 1954) and the finding that reserpine depletes the hypothalamus of norepinephrine and serotonin (Holzbauer and Vogt, 1956) directed attention to the possibility that norepinephrine is an ergotropic neurohormone (Brodie and Shore, 1958). Pletscher et al. (1959) postulate that certain central effects of benzoquinolizine derivatives and reserpine (such as sedation and narcosis

DRUGS AND THE C A R

255

potentiation) may be due to altered cerebral norephenephrine levels. Norepinephrine may also be involved in the mode of action of other tranquilizing agents, such as CPZ. Several authors consider the possibility of adrenergic blockade being involved in the central action of CPZ-like substances (Alexander, 1957; Rathbun et al., 1958; Slater and Jones, 1958; Wirth et al., 1958; Mgller-Nielsen and Neuhold, 1959). 1. Epinephrine and Norepinephrine In the experiments of Kosman and Gerard (1955) in which rats had to press a pedal to avoid shock, epinephrine (dissolved in oil) caused a rapid inhibition of the CAR and a partial inhibition of the UR. Since the animals were incapacitated after epinephrine administration and the blockade of peripheral effects by dibenzyline furnished complete protection against loss of CAR, the central origin of this inhibition is questionable. Injections of epinephrine failed to restore the CAR extinguished by lack of reinforcement. Gellhorn ( 1947), therefore, concluded that the reappearance after insulin coma or electrical convulsions of a previously extinguished CAR was not due to liberation of epinephrine from the adrenal medulla. Weil-Malherbe st al. ( 1959) recently found that epinephrine does not pass the blood-brain barrier (with the possible exception of the hypothalamus). This argues against the central action of epinephrine peripherally applied. John et al. (1958b) injected epinephrine and norepinephrine intraventricularly in conditioned cats. Norepinephrine ( 112 pg ) altered neither the CAR nor the reserpine-induced inhibition of the CAR, although it did attenuate the typical reserpine-induced autonomic changes. Epinephrine (25 to 100 pg) slightly increased the latency of the CAR and somewhat diminished reserpine-induced inhibition of the CAR, but its slight antagonistic activity was not consistently demonstrable. John assumes that hypothalamic depletion of norepinephrine is not responsible for the inhibitory effect of reserpine on the CAR. 2. Adrenergic-Blocking Substances ( See Fig. 2 ) Early investigations of Sivadjian (1934, 1935) showed the effects on the CAR of diethylaminomethyl-3-benzodioxane ( 883F)

256

ALBERT HERZ

and piperidinomethylbenzodioxane (933F), substances whose adrenergic-blocking potency had been previously reported by Fourneau and Bovet (1933). Sivadjian observed inhibition of the CAR or prolongation of the reaction time of rats in a shuttle box. Rathbun et aE. (1958) and Slater and Jones (1958) examined similar congeners which also produced general sedation. In the benzodioxane series the secondary amines proved most active. Selective loss of the CAR was observed in rats wheel-turning to avoid shock. Further alkyl substitution increased the central activity of these benzodioxanes. Ethoxybutamoxane and chlorethoxybutamoxane were selected from a series of benzodioxanes by Verhave et al. (1958a). In doses below 1 mg/kg, subcutaneously, these compounds produced a specific inhibition of the CAR, which was antagonized by d-deoxyephedrine. In addition, phenyl substituted cyclic amines derived from known adrenergic-blocking substances demonstrated a strong tranquilizing potency (Owen and Verhave, 1958). l-Piperidinomethyltetralon ( Na-86), a related substance with adrenergic-blocking effects, proved to be an effective inhibitor of the CAR in rats in the experiments of Knoll (1959). It also potentiated the reserpine-induced inhibition,

3. S u m m y Epinephrine and norepinephrine may not have definitive central effects on the CAR when injected peripherally. The few experiments employing the intraventricular route of administration have not delineated, as yet, the effect of these neurohormones on the CAR. Several adrenergic-blocking drugs have tranquilizing properties and also block the CAR in a specific manner. Some of these drugs are more active than CPZ. The significance of antiadrenergic mechanisms in tranquilizing activity remains to be verified. VII. Cholinergic and Anticholinergic Drugs The role of acetylcholine in brain function has been subject to many investigations. Though there is considerable probability that it may act as a neurotransmitter, exact information is relatively meager. Up to now, it can scarcely be determined which of the numerous central actions of cholinergic drugs and cholinasterase in-

DRUGS AND THE C A R

257

hibitors are directly correlated with the action of acetylcholine. Similar uncertainty exists for inhibition of these actions by anticholinergic drugs. Funderburk and Case (1947), impressed by the cerebral convulsant effect of locally applied acetylcholine, investigated the effect of cholinergic drugs on the CAR in cats by the leg-withdrawal or shuttle-box methods. Eserine completely abolished the CAR and atropine or magnesium sulfate antidoted this inhibition. A similar but much weaker action produced by neostigmine was ascribed to the discomfort caused by increased peristalsis and hypotension. Pilocarpine produced only a transitory inhibition of the CAR. The eserine-induced inhibition of the CAR was ascribed to the accumulation of acetylcholine in the brain. Acetylcholinelike effects of reserpine and, to a lesser degree, of CPZ led Pfeiffer and Jenney (1957) to study the action of cholinergic drugs upon the CAR (pole-jump) of rats. The animals were protected from the peripheral effects of the parasympathetic stimulants by atropine methyl nitrate which, being a quaternary ammonium compound, passes the blood-brain barrier with difficulty. Arecoline ( 2 mg/kg, subcutaneously) almost completely inhibited the CAR during a period of 15 minutes. The UR was found to be affected to a lesser degree. At the height of the drug action the animals showed a fine tremor and slight hyperreflexia. The effect of pilocarpine ( 10 mg/kg) and eserine (0.25 mg/kg) resembled that of arecoline, but their duration of action was 60 to 90 minutes. The inhibition of the CAR, as well as the hyperreflexia, could be completely prevented by atropine sulfate. In contrast to the tertiary amines, cholinergic drugs with a quaternary ammonium structure, such as acetylcholine, methacholine, and neostigmine, failed to block the CAR, presumably not being able to penetrate into the brain. Pfeiffer ascribes this typical and specific inhibition of the CAR to a “muscarinic” effect on the brain. To further decide whether the inhibition of the CAR was due to a nicotinicor muscariniclike action of arecoline, Healy and Jenney (1959) performed experiments in trained rats which were made tachyphylactic to nicotine by repeated injections of nicotine. Nevertheless, arecoline remained active. From this and the previous experiments it is concluded that the inhibitory action of arecoline of the CAR is due to “muscarinic stimulation” of the brain.

258

ALBERT HER2

Franks et al. (1958) described the inhibition of a conditioned response following arecoline administration in man. The conditioned stimulus was a pure tone delivered to both ears through a pair of balanced earphones; the unconditioned stimulus consisted of an air puff delivered to the right eye 350 msec after the commencement of the tone. Arecoline plus methyl atropine had a pronounced inhibitory effect on the reflex studied, whereas its effects on other overt aspects of behavior appeared to be negligible. This was in marked contrast to the effects of drugs such as barbiturates which depressed conditioned responses but also produced numerous side effects. Methyl atropine alone had no observable effects on the conditioned response, results being indistinguishable from placebo or no injection. Only slight effects on the CAR were observed by John et al. ( 1958b), following intraventricular injection of eserine in doses of 50 to SOpg, although the animals showed other severe signs of muscarinic stimulation such as twitching and salivation. The authors concluded that severe motor deficit prevented the animals’ CAR. Dimethylaminoethanol (DMAE, deanol), a probable precursor of cerebral acetylcholine (Pfeiffer, 1959), in oral doses up to 1mg/kg did not inhibit the CAR in rats; whereas deanol produced faster performance in young rats learning the CAR. Inhibition of the CAR which was not to be prevented by atropine was produced by the thio analog of deanol [ ( CHB)2NCH2CH2SH)]$mercaptoethylamine ( H2NCH2CHzSH),and by NIN-dimethylethylenediamine [ ( CH3),NCH2CH2NH2] (Jenney, 1958 and Pfeiffer, 1959). Pepeu and Giarman (1959), using the rope-climbing test in rats, observed depression of the drive for food following application of deanol. Anticholinergic drugs. The experiments just described ( Funderburk and Case, 1947; Pfeiffer and Jenney, 1957; Healy and Jenney, 1959) show that atropine will abolish the central effects of eserine, pilocarpine, and arecoline. The reserpine and CPZ-induced inhibition of the CAR was not prevented by atropine, thus demonstrating the specificity of atropine action against muscarinic stimulation. On the contrary, atropine, at doses of 10 mg/kg injected intraperitoneally before or after application of CPZ, enhanced (but shortened) the inhibitory effect of the latter (Jenney and Healy, 1959).

DRUGS AND THE C A R

259

Investigations of Healy and Jenney (1959) on the quantitative relationship of the atropine-arecoline antagonism in regard to conditioned behavior of rats, resulted in the calculation that 1pmole of atropine will block 4 pmoles of arecoline. A competitive mechanism may be responsible for this antagonism. Scopolamine was approximately twenty times more active than atropine. A series of other antitremor drugs also showed significant antiarecoline potency. The order of decreasing activity was scopolamine > atropine > trihexyphenidyl > caramiphen > diphenhydramine. This scale correlates approximately with the recommended initial daily dosage used for antiparkinsonian therapy in man. The order of potency is also very similar to that reported for equivalent antitremor doses in monkeys by Vernier and Unna (1956). Healy and Jenney, therefore, suggested that the arecoline-induced inhibition of the CAR represents a relatively simple method for the assay of antitremor drugs. The authors believe that the fine tremor of the animals is not sufficient to cause the inhibition of the CAR, since these animals are able to walk and jump normally. In contrast to well-established conditioned responses the CAR in the state of acquisition was abolished by anticholinergic drugs as observed by Herz (1959), using the pole-jump CAR of rats. The drugs were injected as soon as a certain number of conditioned responses had been established during the first session of training. Scopolamine in doses of approximately 1mg/kg almost always abolished the CAR; trihexyphenidyl and benactyzine ( 10 mg/kg ) were also effective, while in the case of atropine, high doses (10 to 50 mg/kg ) were necessary. Quaternary compounds like methylscopolamine and methylatropine showed but slight activity, or were completely ineffective, thus demonstrating that tertiary amine anticholinergics act centrally to produce inhibition. As soon as the CAR was consolidated by training, no further inhibition could be observed. On the contrary, a reduction of latency now occurred. In rats, trained to discriminate between two auditory signals ( t o jump at one and not to jump at the other), scopolamine disturbed this discrimination at the beginning of learning only. The deconditioning produced by these anticholinergic drugs may be connected with altered mental processes, such as concentration and memory, as described in man for scopolamine (Ostfeld et al., 1959). ( See Section 111, C; benactyzine. )

260

ALBERT HERZ

In normalizing stress-induced behavior in rats, scopolamine proved twenty times as active as benactyzine. It should be noted that all effective compounds in Jacobsen’s test were anticholinergic (Holten and Sonne, 1955). In the experiments of John et al. ( 1958b) in cats, atropine after intracerebral or intraperitoneal injection did not inhibit the CAR; the action of intracerebrally injected reserpine was also not altered by atropine. In positive reinforcement and maze tests on rats, Michelson et al. (1954) found that atropine and caramiphen impaired performance; this impairment could be partly antagonized by neostigmine. Summary. “Muscarinic stimulation” of the brain causes loss of the CAR in rats and man, protected from the peripheral effects of parasympathetic stimulation by quaternary salts of atropinelike compounds. Tertiary amine anticholinergic drugs restore the CAR and abolish the fine tremor of central origin. The antidotal potency of several atropinelike drugs against this muscarinic effect correlates with their activity in Parkinsonism therapy. In contrast to consolidated conditioned reactions, anticholinergic drugs are able to inhibit the CAR at the stage of acquisition. Scopolamine shows a high activity in normalizing stress-induced behavior in rats, and resembles benactyzine in this respect.

VIII. Serotonin and Monoamine Oxidase Inhibitors

Serotonin or 5-hydroxytryptamine (5-HT ) is presently the center of attention in neuropharmacology. The discovery of its presence in brain structures and the finding that 5-HT is liberated by reserpine have initiated extensive research and discussion on the significance of indolalkylamines in brain function, the pathogenesis of mental disorders, and the mechanisms of tranquilization. We do not know whether 5-HT is selectively sedative or stimulant to various parts of the brain. Since iproniazid and other inhibitors of monoamine oxidase prevent the degradation of 5-HT and also phenylalkylamines, changes in behavior of the animals after iproniazid may be based on cerebral levels of these biogenic amines. (See review by Costa.) In rats, 50mg/kg 5-HT produced marked incapacity but did not specifically inhibit the pole-jump CAR (Pfeiffer and Jenney,

DRUGS AND THE C A R

261

1957). Employing the same technique a 50% block was observed

by Cook and Weidley (1957), following administration of 5-HT ( 10 mg/kg, subcutaneously) in rats. However, increasing doses up to 40mg/kg did not intensify this effect. In a shuttle box, 5-HT ( 5 mg/kg, intraperitoneally) caused a 50% inhibition of the CAR (Grandjean and Battig, 1957). Since Shaw and Woolley (1954) presume 5-HT (in contrast to 5-hydroxytryptophan) to pass the blood-brain barrier with difficulty, it may be questionable whether these results are produced by central mechanisms. In the experiments of Winter and Flataker ( 1956) 5-hydroxytryptophan and 5-HT prolonged climbing time only slightly. John et al. (1958b) gave 5-HT intraventricularly to conditioned cats. Doses of 270pg 5-HT, combined with 135pg iproniazid, resulted in a slight prolongation of the reaction time in some of the animals, but no definite inhibition of the CAR was observed. Also reserpine-induced inhibition was not significantly decreased by the combination of 5-HT and iproniazid. According to John these results argue against the theory that the inhibition of the C.4R by reserpine is due to release of 5-HT from brain structures. Ipronkzzid ( 1-isonicotinyl-2-isopropyl hydrazine ) , in oral doses of 100 mg/kg, produced a slight retardation and limited loss of the CAR of rats in a shuttle box, but 24 hours later, when enzymeinhibition had reached its maximum, the conditioned behavior was again normal. However, pretreatment with iproniazid prevented completely the reserpine-induced inhibition of the CAR ( Horisberger and Grandjean, 1958). (See Section 111, B; reserpine.) Occasionally in a few animals, Maffii (1959) found a block of the CAR after iproniazid, while the SCR was consistently abolished by very high doses. Iproniazid given to mice over a period of several weeks retarded the learning of a CAR (Sui, 1957). Trans-2-phenylcyclopropylamine ( SKF 385), a potent monoamine oxidase inhibitor, blocked conditioned escape behavior, while iproniazid was ineffective (Tedeschi et d.,1959). The antagonism between 5-HT and LSD-25 is reviewed in the next section. S u m m y . Although the studies of 5-HT on the CAR are not in complete agreement, a strong inhibitory action does not obtain. Since, even intraventricular injection is ineffective, this inactivity can not be related to inadequate passage of the blood-brain barrier. The findings following monoamine oxidase inhibitors, e.g. the pre-

262

ALBERT HERZ

vention of reserpine-blocking action, do not involve 5-HT alone, but also catechol amines and perhaps other unknown amines. IX. Psychotomimetic Drugs 1. D-Lysergic Acid Diethylamide (LSD-25) LSD is the most active substance known to produce hallucinations and schizophreniclike states in man. A large volume of literature has now accumulated on relationships between biogenic amines, especially 5-HT and LSD. The speculative hypotheses of Woolley and of Gaddum, whereby LSD may compete in the brain with 5-HT receptors has been partly responsible for this flurry of activity. In spite of this intensive study the situation is complicated and confused. LSD in large doses (1to 2 mg/kg) inhibited the CAR (polejump) of rats in the experiments of Pfeiffer and Jenney (1957). The animals, however, appeared very confused; therefore, any similarity of this effect to CPZ inhibition of the CAR was doubtful. Cook and Weidley (1957) observed blocking of the CAR; at higher doses the UR was also affected, although the animals still appeared to be able to jump for the pole. They also found that small doses of LSD (0.1 to 0.5 mg/kg) which had no measurable effect on the CAR, antagonized the inhibitory action of 5-HT, reserpine, CPZ, and meprobamate, but not that of morphine. These authors suggest that the antagonistic action of LSD may offer a means of differentiating between certain types of specific CAR-blocking agents. Winter and Flataker (1956, 1957) studied the antagonism between LSD and 5-HT by use of the rope-climbing response in rats with positive as well as negative reinforcement. LSD (0.175 to 0.5 mg/kg, intraperitoneally ) caused prolongation of the climbing time-linearly related to the log dose of LSD-accompanied by signs of confusion. 5-HT ( 6 mg/kg) and 5-hydroxytryptophan antagonized these effects, while reserpine (50 pg/kg daily for a period of 5 days), CPZ ( 1 to 5 mg/kg), as well as indole, intensified the LSD-induced inhibition. Meprobamate and benactyzine reduced the LSD effect. The efficacy of 5-HT in these experiments is not consistent with the assumption that 5-HT is unable to pass the blood-brain barrier (see previous section). Mahler et al. ( 1958)

DRUGS AND THE C A R

263

using a similar method also observed prolongation of climbing time and confusion following LSD. Intraperitoneal doses of 0.3 to 0.53 mg/kg, were effective for less than 30 minutes. Using the climbing test in rats with positive reinforcement, Freedman et al. (1958) studied the development of tolerance to LSD (130 pg/kg, intraperitoneally). After 4 days the rats became completely tolerant and the climbing time was normal. Tolerance following repeated injections of LSD was also observed by Winter and Flataker (1956).

2. Mescaline The mental state produced in man by mescaline resembles that of LSD, although the mescaline model psychosis requires a 3- to 5-thousandfold greater dosage. Mescaline (10 to 100 mg/kg, orally) failed to block the CAR (pole-jump) of rats. These doses, however, slightly enhanced the blocking action of CPZ, reserpine, 5-HT, and morphine (Cook and Weidlev, 1957). The SCR was found to be slightly blocked, but not the'CAR (Maffii, 1959). Courvoisier and Julou (1956) observed a state of excitement and apparent disorientation in animals after mescaline when the CS (bell) sounded. It appeared as if the CS produced a hallucinatory crisis. Frequently not only the CAR but also the UR was abolished. Small doses of CPZ, ineffective in altering the CAR, antidoted the disorientation of the animals and restored the CAR and the UR. In the CAR (leg-withdrawal) of dogs, mescaline (35 mg/kg) produced dissociation between the motor and the autonomic conditioned response; only the latter was affected by this dose (Bridger and Gantt, 1955, 1956). As in the case of Courvoisier's rats the dogs reacted to the CS as if it were the US. Bridger and Gantt have an extensive discussion on the possible mechanism of action of mescaline in the CNS which might account for their findings. Tolerance to mescaline developed after daily intraperitoneal injections of 10 mg/kg in about a week (Freedman et al., 1958). Rats showing tolerance to LSD were also tolerant to mescaline; however, crosstolerance was not seen in mescaline-tolerant rats, which, in fact, showed enhanced response to LSD (Freedman and Aghajanian, 1959).

264

ALBERT HERZ

3. Buf otenine Mahler et al. (1958) found that bufotenine (8 to 27 mg/kg) caused a prolongation of climbing time. This effect lasted more than 2 hours and again was probably due to the confusional state of the rats. 4. Summary The rope-climbing response in rats (employing negative or positive reinforcement) may offer distinct possibilities for measuring quantitatively the confusional state produced by LSD. The observed delay in performance has apparently no similarity to the inhibition of the CAR produced by tranquilizing drugs. A series of agents will reverse or enhance the LSD-induced confusional state. The effects of mescaline and bufotenine on CAR show some resemblance to those of LSD, but these drugs are distinctly weaker in action.

X. Miscellaneous The question whether peripheral autonomic responses play some part in the drive of fear was investigated by Auld (1951). Transmission of the afferent impulses through the autonomic ganglia was blocked by tetraethylammonium chloride (TEA) ; the finding that TEA-drugged rats perform slower in a shuttle box than control animals was interpreted as a reduction in the drive of fear. Davitz (1953) observed that the extinction of a conditioned emotional response of rats was retarded by TEA. Similar results, obtained by Arbit (1957), were ascribed to the known curarelike effect of TEA on skeletal musculature. Brady (1953) also investigated this problem and found that TEA increased performance time of a hungerand thirst-motivated running response. The fear-reduction interpretation of the TEA effects, as suggested by Auld, is, therefore, doubtful. Gellhorn (1953) observed, in rats, recovery of extinguished conditioned responses after coma and convulsions which were induced by electroshock, Metrazol, or hypoglycemia. These results were interpreted as an increased excitability of sympathetic autonomic centers in the hypothalamus.

DRUGS A N D THE C A R

265

Behavioral responses following injection of ACTH were studied

by Mirsky et al. ( 1953) and Murphy and Miller ( 1955). ACTH was without effect on the acquisition of the CAR in rats, but significantly prolonged subsequent extinction time. The reversal, by stimulants, of the inhibitory effect of many tranquilizers has been studied. These experiments have been described already. The CAR may not present an ideal method for the investigation of the action of stimulants alone, because frequently the CAR is so rapid that improvement is hard to measure. Some shortening of the reaction time, however, will be obtained. This was observed with caffeine and amphetamine (Battig and Grandjean, 1955, 1957) and with deanol (Jenney, 1958). In “poor performers” these stimulants will also increase the percentage score of avoidance responses in a given test period (Verhave et al., 1 9 5 8 ~ ) . High doses of caffeine or amphetamine also produce inhibition of the CAR or increase latency due to over-stimulation or ataxia (Battig and Grandjean, 1955; Knoll and Knoll, 1958).

XI. General Discussion A. SPECIFICITY OF ACTION A drug-induced specific inhibition obtains when the CAR is abolished completely, while the UR remains intact. Cook and Weidley ( 1957) contrasted specifically active substances, such as CPZ, reserpine, and morphine, with nonspecific inhibitors, such as barbiturates and methylparafynol. The last two drugs failed to produce specific inhibition of the CAR, or showed insignificant effects at neurotoxic doses. This definition for “specific inhibition” is followed by most investigators and is considered an essential characteristic of specific drug action. This specificity may also be expressed numerically as ratios of drug doses which will inhibit the conditioned and unconditioned responses by 50% (1DS;s). By dividing the ID50 of the UR by the IDso of the CR Maffii (1959) obtained the following values : morphine, 3.53, CPZ, 2.84, and reserpine, 1.87. Comparable values are evident in the reports of other authors. Morphine is thus the most specific inhibitor of the CAR, Knoll and Knoll (1958) states that it is characteristic of specif-

266

ALBERT HERZ

ically acting tranquilizers that, after drug action has ended, abolished conditioned reactions return only with reinforcement, whereas after nonspecific compounds, such as barbiturates, reinforcement is unnecessary. This may correlate with the repeatedly observed acceleration of extinction produced by tranquilizing drugs ( Ader and Clink, 1957; Miller et al., 1957a, b ) . According to Maffii, the inhibition of the SCR represents another possibility for the classification of these drugs. Not only do drugs of the CPZ-reserpine type inhibit the SCR, but also drugs such as meprobamate, phenobarbital, hydroxyzine, azacyclonol, which do not cause a specific inhibition of the CAR. This suggests that any nonspecific sedative is effective in producing SCR inhibition. One may further cite the clinical findings which strongly favor the concept of specific inhibition. For example: CPZ, reserpine, deserpidine, rescinnamine, and perhaps morphine are in some instances antipsychotic drugs if not uniformly antischizophrenic, while the sedative type drugs are not antipsychotic, but do allay anxiety. The possibility of a delineation of specific inhibition such as that detailed above in contrast to simple sedative effects has repeatedly been doubted and even disputed. 1. Nieschulz et al. (1957, 1958) were unable to detect any significant difference between the inhibitory effects of several phenothiazine derivatives on the one hand and barbiturates on the other. They could not find a dose which inhibited the CAR without also influencing the UR. Perhaps, this finding may depend partially on the extensive variation in animal reactivity. To eliminate this variable, Owen and co-workers (1958) using the wheel-turning method employed a cross-over design. Klupp and Kieser (1959) compared the effect of CPZ, reserpine, and barbiturates upon positively and negatively reinforced responses in a modified Skinnerbox and could detect no significant difference between the effective dose in either kind of reinforcement. For this reason the authors regarded this action as a common sedative effect. However, one can assume that these results demonstrate that not only negatively but also positively reinforced conditioned behavior may be affected by these substances. (See section XI, F. ) 2. Participation of a motor component, induced perhaps by weakness or reduction of muscle tone, is considered by other

DRUGS AND THE C A R

267

authors to be a possible factor ( Ader and Clink, 1957; Miller et al., 1957b; Jenney and Healy, 1959). Decreased work performance of rats in a revolving cage after CPZ, meprobamate, or perphenazine was observed by Jasmin and Bois (1959). Miller et aZ. (1957b) tried to delimit the motor component of CPZ action in inhibition of the CAR. When compared to barbiturates the motor influences of CPZ ought to be differentiated by alternately testing the animals in an escape runway, with a continuously electrified grid floor, and in the conditioning box. Whether such enforced running is capable of excluding motor deficit, remains to be established by further investigations. Many laboratories use a horizontal, slowly revolving rod or an inclined screen to test for motor deficit in rats and mice at the dose which inhibits the CAR. The observation by Jenney and Healy (1959) that during a series of quickly repeated conditioned responses the reaction time becomes progressively longer may indicate participation of a muscular factor. Irwin (1958) found that tolerance to CPZ which develops with prolonged use, reduces the depression of motor activity but not the inhibition of the CAR. This indicates that depression of the locomotor activity does not contribute to the suppression of avoidance behavior. Recently Irwin (1959) evaluated the suppressant potencies of eight phenothiazine tranquilizers on conditioned behavior ( hurdle-jump ) and locomotor activity (revolving treadwheel) in rats. Equivalent avoidance suppressing doses of all agents were equally effective in suppressing locomotor activity. Acute tolerance development to the avoidance suppressant actions was also noted. Interneuron-blocking agents, which produce a marked muscular weakness, cause only a nonspecific block of the CAR at neurotoxic doses. The question is undecided as to the motor deficit component in inhibition of the CAR by compounds which produced marked cataleptic action such as chlorperazine (Courvoisier et al., 1957; Wirth et al., 1958). Since these compounds produce concomitant inhibition of the UR, one might assume physical insufficiency. However, bulbocapnine, the original prototype of catatonia-producing substances, showed a high specificity in blocking the CAR (Verhave and Owen, 1958). 3. Another argument might be made against the concept of a specific inhibition of conditioned responses. Because of the varying intensity of CS and US, a more powerful influence of the stronger

268

ALBERT HERZ

US is to be expected. This is demonstrated in the experiments of Irwin (1958), where intensification of the CS was accompanied by a reduction of CAR inhibition. Further work should be initiated wherein the auditory or light stimulus is as painful to the species under study as is the electric shock. In spite of these reservations the bulk of evidence justifies the term “specific CAR inhibition” in connection with the CPZ and reserpine type drugs. The majority of studies show (although different methods and various conditions of investigation have been applied) a definite difference between the effects of specific and nonspecific drugs. B. VARIOUS TYPESOF CAR INHIBITION After excluding muscular deficit and sedation there still remain other distinct behavioral states which inhibit the CAR. There appears to be a significant difference in the behavior of animals whose CAR is inhibited by hallucinogens, such as LSD or mescaline, compared to those inhibited by CPZ. The hallucinogens produce apparent confusion or disorientation, whereas CPZ produces indifference or disinterest. Cook and Weidley (1957) found LSD-injected animals to be hyperirritable. Reference has been made to the crisis induced by the CS in mescaline-treated animals (Courvoisier and Julou, 1956; Bridger and Gantt, 1955, 1956). Gatti (1957) describes the confusional state produced in rats by benactyzine. The inhibitory action of scopolamine and other anticholinergic drugs on the acquisition of the CAR by rats may be included in this category (Herz, 1959, 1960a). Increasing doses of CNS stimulants lead to an over-excitation followed by a prolongation of latency (Battig and Grandjean, 1957). With these disorienting drugs simultaneous inhibition of the UR seems to be more or less characteristic, although there are no objective signs of motor deficit. The animals show predominantly an increased psychomotor activity. This provides a third classification of types of drugs which inhibit the CAR.

C. THE QUESTION OF SENSORY DEFICIT The possibility of a drug-induced deficit in afferent impulses leading to inhibition of the CAR is to be considered, although ex-

DRUGS AND THE C A R

269

tensive experiences with CPZ and reserpine therapy in man indicate that such a mechanism is remote. However, this could occur with new drugs. Sensory deficit could be of two types: (1) that in which the CS would not be heard or seen, and ( 2 ) that in which the animal under analgesia would not feel the electrical current of the US. Neurophysiologists have recorded sensory evoked responses by electrodes on the sensory cortex and find that these are not influenced by reserpine ( Schneider, 1954). In addition, there are some experiments which demonstrate this in a more direct way. In the experiments of John et al. (1958a, b ) with intraventricular reserpine, the avoidance response proved to be more intensely inhibited than the approach response ( pattern discrimination for food). The investigations of Gliedman and Gantt ( 1956) on reserpine and CPZ, and of Stephens and Gantt (1956) on morphine, demonstrate that it is more difficult to influence the autonomic conditioned response than the CAR. In the experiments of Riopelle and Pfeiffer (1958) reserpine left the learning of new discrimination problems unimpaired, while the CAR was inhibited. These results preclude the presence of sensory impairment as being responsible for the observed decrease in performance. On the other hand, inhibition of the CAR has been directly employed to measure sensory deficit. Daily dosage of streptomycin for several weeks in rats is followed by some degree of deafness. Courvoisier and Leau (1956) measured the intensity of the acoustic signal which restored the CAR and employed this as a precise measurement of the degree of deafness. Although the correctness of this assumption is unquestionable the example demonstrates the complexities involved in the interpretation of results. For example, with CPZ an increase of the intensity of the CS may also reverse the inhibition of the CAR to some extent (Irwin, 1958).

D. SIGNIFICANCE OF ANALGESIC ACTIVITY Since fear of pain has served frequently as a motive for the CAR (see next section) the analgesic action of some drugs has to be considered in the inhibition of the CAR. In regard to morphine, the findings of Herz (1960b), wherein scopolamine potentiates the CAR inhibitory action of morphine but not the analgesic effect, indicate that the analgesic activity is probably not essential for the

270

ALBERT HER2

inhibitory action. Although CPZ is sometimes effective in clinical pain it does not show an analgesic effect in animals (Lembeck, 1955), and reserpine even reduces morphine-induced analgesia ( Schneider, 1954). Finally it is doubtful whether “pain,” in the common sense, renders the slight electric shock disagreeable for the animals.

E. FEAR-REDUCING HYPOTHESIS In several reports the suppression of the CAR is interpreted as a “fear-reducing” effect of the drug. Undoubtedly fear and anxiety are important motives in the acquisition of the CAR but not so evident is the possibility that suppression of a consolidated CAR by a drug reflects fear-reducing effects of the drug. A series of observations would argue against such a fear-reducing hypothesis. 1. The elegance and ease with which the animals perform the response renders it unlikely that fear and anxiety should be the motivator (Nieschulz et al., 1958). 2. The finding of John et al. (1957a, b ) that avoidance responses cued by a visual stimulus can be abolished by reserpine, while simultaneously the same response to an auditory one remains intact, argues against a reduction of fear by reserpine. 3. In the experiments of Jacobsen and Sonne (1955, 1956) on stress-induced behavior, the reduction of tension is accompanied by an increase in response to the conditioned stimulus. 4. Interneuronblocking agents, such as meprobamate, do not inhibit the CAR, although these drugs are widely used in man for their anxiety- and tension-reducing effects. 5. The CAR is blocked in animals by simple cholinergic stimulants, whereas, in schizophrenic patients these agents produce a lucid interval. 6. The experiments of Brady et al. (1954) in rats with rhinencephalic injury led to the conclusion that in the maintenance or retention of avoidance behavior fear or anxiety do not play a central motivational role. 7 . The fact that the performance of positively reinforced conditioned responses is also decreased argues against the fear hypothesis (see next section).

DRUGS AND THE C A R

271

F. FURTHER ASPECTSREGARDING CAR INHIBITION A series of papers provide confirmation that substances which inhibit the CAR also reduce the performance of positively reinforced tests, even though certain quantitative differences in the affectability of avoidance and approach reactions may occur (John et al., 1958a, b; Riopelle and Pfeiffer, 1958; Klupp and Kieser, 1959). An action on positive as well as negative reinforcement is interpreted by Riopelle and Pfeiffer as “reduction in motivation.” Berger ( 1957) speaks of “isolation from environment,” Courvoisier ( 1956) of “indifference,” and Weiskrantz ( 1957) of “nonreactivity.” These terms all refer to similar behavior, and have the advantage of avoiding close analogies to human emotions. This objective approach appears reasonable, since according to Bradley ( 1959), we are never certain that the effects we see in animals are in any way related to psychological effects in man. Pfeiffer and his team argue for a close connection between inhibition of the CAR and antischizophrenic potency of drugs. They compare the ineffective drugs, azacyclonol, meprobamate, and barbiturates, which do not inhibit the CAR with antischizophrenic drugs, such as CPZ and reserpine (and also the new adrenergic-blocking substances ), which produce marked inhibition of the CAR (Murphree et al., 1959). These authors state “one should note, that of all the pharmacological tests used for the discovery of anti-schizophrenic drugs, the inhibition of the CAR is the most quantitative.” Investigations on the inhibition of the CAR by muscarinic drugs are based on the possible antischizophrenic potency of these compounds, which are indeed capable of reproducing a “lucid interval” in schizophrenic patients (Fulcher et al., 1957). Murphree ,et al. (1959) also consider the possibility that central parasympathetic dominance per se may produce an antischizophrenic milieu. A shift in the cholinergic-adrenergic balance toward a cholinergic predominance may also be produced by tranquilizers which possess adrenergic inhibiting potency. A close relationship between deconditioning activity and usefulness of substances in psychiatry is suggested by Courvoisier (1956). An explanation for such a correlation may perhaps be established by assuming psychoneurotic and psychotic conditions

272

ALBERT HERZ

to be the result of faulty conditioning (Berger, 1957). At present this concept is purely theoretical.

XII.

Conclusions

Among the useful animal methods for characterizing psychotropic substances, avoidance conditioning represents a procedure which is simple to manipulate and specially suited as a standard method for pharmacological investigations of new compounds. This method has proved extremely useful in the testing of tranquilizing agents. In addition, by means of avoidance conditioning a number of pharmacological and experimental psychological questions have been partially answered and many more have been raised. In order to gain a comprehensive impression of the action of a psychotropic substance the method of the CAR should be supplemented by other behavioral techniques. Among these are operant conditioning (Brady, 1956 and Dews, 1956), various motor tests, timed ropeclimbing, and careful neurophysiological studies to delineate the locus of action of a given agent in the central nervous system. (See also sectional summaries.) ACKNOWLEDGMENT

I am indebted to Dr. Pfeiffer for assistance in the translation of this article.

REFERENCES Ader, R., and Clink, D. W. (1957). 3. Pharmucol. Erptl. Therap. 121, 144. Alexander, L. (1957).Ann. N . Y. Acad. Sci. 67, 758. Ambrus, J. L., Ambrus, C. M., Bauer, R. O., and Noell, W. K. ( 1957). 3.

Pharmucol. Exptl. Therap. 119, 129. Arbit, J. (1957).J . Comp. and Physiol. Psychol. 60, 144. Auld, F. (1951).J . Comp. and Physiol. Psychol. 44, 565. Bailey, C. J., and Miller, N. E. (1952).J . Comp. and Physiol. Psychol. 46,

205. Battig, K., and Grandjean, E. (1955).Helo. Physiol. et Pharmacol. Acta 13,

c54. Battig, K., and Grandjean, E. (1957).J . physiol. (Paris) 49, 41. Berger, F. M. (1952).1. Pharmacol. Exptl. Therap. 104, 229.

DRUGS AND THE C A R

273

Berger, F. M. (19%). J. Pharmacol. Exptl. Therap. 112, 413. Berger, F. M. (1957). Ann. N . Y. Acad. Sci. 67, 685. Besendorf, H., and Pletscher, A. (1956). Helo. Physiol. et Pharmacol. Acta 14, 383. Boren, J. J., and Beyer, N. (1957). J. Pharmacol. Exptl. Therap. 119, 134. Boren, J. J., and Navarro, A. P. (1958). Federation PTOC. 17, 351. Bradley, P. B. ( 1959). “Neuro-psychopharmacology.” Elsevier, Amsterdam, Holland. Brady, J. V. (1953). J. Comp. and Physiol. Psychol. 46, 307. Brady, J. V. (1950). Science 123, 1033. Brady, J. V. (1957). Ann. N . Y. Acad. Sci. 66, 719. Brady, J. V., Schreiner, L., Geller, I., and Kling, A. (1954). J. Comp. and Physiol. Psychol. 47, 179. Brazier, M.A. B. (1958). “The Central Nervous System and Behavior.” Josiah Macy, Jr. Foundation, New York. p. 201. Bridger, W. H., and Gantt, W. H. (1955). 1. Pharmacol. Exptl. Therap. 113, 7. Bridger, W.H., and Gantt, W. H. (1956). Am. J. Psychiat. 113, 352. Brodie, B. B., and Shore, P. A. (1958). Federation PTOC.17, 353. Burke, J. C., Hassert, G. L., and High, J. P. (1957). J. Pharmacol. Exptl. Therap. 119, 136. Conger, J. J. (1951). Quart. J. Studies Alc. 12, 1. Cook, L., and Weidley, E. F. (1957). Ann. N . Y. Acad. Sci. 66, 740. Cook, L., Weidley, E. F., Morris, R. W., and Mattis, P. A. (1955). J. Pharmacol. Exptl. Therap. 113, 11. Cook, L., Weidley, E. F., Deegan, J., and Mattis, P. A. (1958). J. Pharmacol. Exptl. Therap. 122, 14A. Courvoisier, S. (1956). J. Clin. Exptl. Psychopathol. dr Quart. Reo. Psychiat. Neurol. 17, 25. Courvoisier, S., and Julou, J. (1956). XX Intern. Physiol. Congr., Brussels, p. 197. Courvoisier, S., and Leau, 0.(1956). Antibiotics dr Chemotherapy 6, 411. Courvoisier, S., Fournel, J., Ducrot, R., Kolsky, M., and Koetschet, P. (1953). Arch. intern. pharmacodynamie 92, 305. Courvoisier, S., Ducrot, R., and Julou, L. (1957). In “Psychotropic Drugs” ( S. Garattini and V. Ghetti, eds.), p. 373. Elsevier, Amsterdam, Holland. Davitz, J. R. (1953). J. Comp. and Physiol. Psychol. 46, 311. Dews, P. B. (1956). Ann. N . Y. Acad. Sci. 66, 268. Dinsmoor, J. A., and Hughes, L. H. (1956). J. Comp. and Physiol. Psychol. 49, 235. Domino, E. F., Karoly, A. J., and Walker, E. L. (1958). J. Pharmacol. Exptl. Therap. 122, 20A. Edwards, R. E. (1959). Federation PTOC.18, 386. Fellows, E. J., and Cook, L. (1957). In “Psychotropic Drugs” ( S . Garattini and V. Ghetti, eds.), p. 397. Elsevier, Amsterdam, Holland. Foumeau, E., and Bovet, D. (1933). Arch. intern. pharmacodynamie 46, 178. Franks, C. M., Trouton, D. S., and Laverty, S. G. (1958). J. Clin. Exptl. Psychopathol. 6 Quart. Rev. Psychiat. Neurol. 19, 226.

274

ALBERT HER2

Freedman, D. X. and Aghajanian, G. K. ( 1959). Federation Proc. 18, 390. Freedman, D. X., Aghajanian, G. K., and Ornitz, E. M. (1958). Science 127, 1173. Fulcher, J. H., Jr., Gallagher, W. J., and Pfeiffer, C. C. (1957). A.M.A. Arch. Neurol. Psychiat. 78, 392. Funderburk, W. H., and Case, T. J. (1947). J. Neurophysiol. 10, 179. Gatti, G. L. ( 1957). In “Psychotropic Drugs” (S. Garattini and V. Ghetti, eds.), p. 125. Elsevier, Amsterdam, Holland. Gatti, G. L. (1958). Rend. ist. super. saiaita 21, 968. Gellhorn, E. (1947). PTOC.SOC. Exptl. Biol. Med. 64, 375. Gellhorn, E. (1953). Ann. N . Y. Acad. Sci. 66, 200. Gellhorn, E., Kessler, M., and Minatoya, H. (1942). Proc. SOC. Exptl. Biol. Med. 60, 260. Gliedman, L. H., and Gantt, W. H. (1956). Southern Med. J. 49, 880. Grandjean, E., and Battig, K. (1957). Helv. Physiol. et Pharmacol. Acta 16, 366. Guha, G., Dasgupta, S. R., and Werner, G. (1954). Bull. Calcutta School Trop. Med. 2, 46. Harrison, J. M., and Tracy, W. H. (1955). Science 121, 373. Healy, S. T., and Jenney, E. H. (1959). Federation Proc. 18, 400 and Personal communication. Herz, A. ( 1959). Arch. exptl. Pathol. Pharmakol. Naunyn-Schmiedeberg’s 236, 110. Herz, A. (1960a). Z. Biol. Herz, A. ( 1960b). To be published. Hess, E. (1956) cited in Hunt (1956). Hess, G., and Jacobsen, E. (1957). Acta Pharmacol. Toxicol. 13, 135. Hill, H. E., Belleville, R. E., and Wikler, A. (1954a). PTOC. SOC. Exptl. Biol. Med. 86, 881. Hill, H. E., Belleville, R. E., and Wikler, A. (195413). Science 120, 153. Holten, C. H., and Sonne, E. (1955). Acta Pharmacol. Toxkol. 11, 148. Holzbauer, M., and Vogt, M. (1956). J. Neurochem. 1, 8. Horisberger, B., and Grandjean, E. ( 1958). Helv. Physiol. et Pharmacol. Actn 16, 146. Hunt, H. F. (1956). Ann. N . Y. Acad. Sci. 66, 258. Hunt, H. F. (1957). Ann. N . Y. Acad. Sci. 67, 712. Irwin, S. (1958). Federation Proc. 17, 380. Irwin, S. (1959). Pharmacologist 1, 51. Irwin, S., and Govier, W. M. (1957). J. Pharmacol. Exptl. Therap. 119, 154. Irwin, S., Slabek, M., Debiase, P. L., and Govier, W. M. (1959). Arch. intern. pharmacodynnmie 118, 358. Jacobsen, E. ( 1957). In “Psychotropic Drugs” (S. Garattini and V. Ghetti, eds. ), p. 119. Elsevier, Amsterdam, Holland. Jacobsen, E. (1958). Antibiotic Med. d7 Clin. Therapy 6, 89. Jacobsen, E., and Skaarup, Y. (1955a). Acta Pharmacol. Toxicol. 11, 117. Jacobsen, E.,and Skaarup, Y. (1955b). Acta Pharmacol. Toxicol. 11, 125. Jacobsen, E., and Sonne, E. (1955). Acta Pharmacol. Toxicol. 11, 135.

DRUGS AND THE C A R

275

Jacobsen, E., and Sonne, E. (1956). Acta Pharmacol. Toxicol. 12, 310. Jasmin, G.,and Bois, P. (1959). Can. J . Biochem. and Physiol. 37, 417. Jenney, E. H. (1958). Federation Proc. 17, 381. Jenney, E. H., and Healy, S. T. (1959). Federation Proc. 18, 407 and Personal communication. John, E. R., Wenzel, B. M., and Tschirgi, R. D. (1958a). Science 127, 25. John, E. R., Wenzel, B. M., and Tschirgi, R. D. (1958b). J. Pharmucol. Exptl. Therap. 123, 193. Julou, L., Courvoisier, S., Bardone, M. C., Ducrot, R., Foumel, J., and Leau, 0 . (1957). Cornpt. rend. SOC. biol. 161, 864. Kerman, E. F. (1944). A.M.A. Arch. Neurol. Psychiat. 62, 61. Kessler, M., and Gellhom, E. (1943). Am. J . Psychiat. 99, 687. Klupp, H., and Kieser, W. (1959). Arch. exptl. Pathol. Pharmakol. NaunynSchmiedeberg’s 236, 97. Knoll, J. ( 1959). Arch. exptl. Pathol. Pharmukol. Naunyn-Schmiedeberg’s 236, 92. Knoll, J., and Knoll, B. (1958). Arzneimittel-Forsch. 8, 330. Kosman, M. E., and Gerard, R. W. (1955). J . Comp. and Physiol. Psychol. 48, 506. Lembeck, F. ( 1955). “Kiinstlicher Winterschlaf.” Urban und Schwarzenberg, Wien-Innsbruck, Austria. Levis, S., Preat, S., Beersaerts, J., Dauby, J., Beelen, L., and Baugniet, V. ( 1957). Arch. intern. pharmacodynamie 109, 127. Macht, D. I. (1943). Exptl. Med. Surg. 1, 260. Macht, D. I., and Hoffmaster, T. (1948). Federation Proc. 7, 242. Maffii, G. (1959). J . Pharm. and Pharmacol. 11, 129. Mahler, D. J., Humoller, F. L., and Dunn, A. L. (1958). Federation PTOC. 17, 103. Marchal, P., and Schlag, J. (1958). Arch. intern. pharmacodynamie 114, 484. Massennan, J. H., and Siever, P. W. (1944). Psychosomat. Med. 6, 7. Masserman, J. H., and Yum, K. S. (1946). Psychosomat. Med. 8, 36. Michelson, M. J., Rozhkova, E. K., and Savateev, N. V. (1954). Biul. eksp. biol. med. 37, 7. Miller, N. E. ( 1957). In “Psychotropic Drugs” (S. Garattini and V. Ghetti, eds. ) , p. 83. Elsevier, Amsterdam, Holland. Miller, R. E., Murphy, J. V., and Mirsky, I. A. (1957a). A.M.A. Arch. Neural. Psychiat. 78, 526. Miller, R. E., Murphy, J. V., and Mirsky, I. A. (195713). J. Pharmacol. Exptl. Therap. 120, 379. Mirsky, I. A., Miller, R. E., and Stein, M. (1953). Psychosomat. Med. 16, 574. Mfiller-Nielsen, I., and Neuhold, K. ( 1959). Acta Pharmacol. Toxicol. 16, 335. Murphree, H. B., Jr., Jenney, E. H., and Pfeiffer, C. C. (1959). Research Publs., Assoc. Research Nervous Mental Disease 37, 204. Murphy, J. V., and Miller, R. E. (1955). J. Cornp. and Physiol. Psychol. 48, 47. Naess, K., and Rasmussen, E. W. (1958). Acta Pharmacol. Toxicol. 16, 99. Nieschulz, O.,Hoffmann, I., Popendiker, K., and Schumacher, H. (1957). Arzneimittel-Forsch. 7 , 259.

276

ALBERT HER2

Nieschulz, O., Hoffmann, I., and Popendiker, K. ( 1958). Arzneimittel-Forsch. 8, 199. Nieschulz, O., Hoffmann, I., and Popendiker, K. ( 1959). Arzneimittel-Forsch. 9, 219. Ostfeld, A. R., Jenkins, R., and Pasnau, R. (1959). Federation Proc. 18, 430. Owen, J. E., Jr., and Verhave, T. (1958). J. Pharmucol. Exptl. Therup. 122, 59A. Owen, J. E.,Jr., Verhave, T., and Robbins, E. B. (1957). J. Pharmucol. Exptl. Therup. 119, 174. Owen, J. E., Jr., Verhave, T., and Robbins, E. B. (1958). Psychol. Rept. 4, 527. Pepeu, G., and Giarman, N. J. (1959). Federation Proc. 18, 433. Pfeiffer, C. C. (1959). Intern. Rev. Neurobiol. 1, 195. Pfeiffer, C. C., and Jenney, E. H. (1957). Ann. N. Y. Acad. Sci. 66, 753. Pfeiffer, C. C., Riopelle, A. J., Smith, R. P., Jenney, E. H., and Williams, H. L. (1957). Ann. N . Y. Acad. Sci. 67, 734. Pletscher, A,, Besendorf, H., and Gey, K. F. (1959). Science 129, 844. Rathbun, R. C., Henderson, J. K., Kattau, R. W., and Keller, C. E. (1958). J. Pharmacol. E#. Therup. 122, 64A. Riley, H., and Spinks, A. (1958). J. Pharm. and Pharmacol. 10, 657, 725. Riopelle, A. J., and Pfeiffer, C. C. (1958). A.M.A. Arch. Neurol. Psychiat. 79, 352. Rutledge, L. T., and Doty, R. W. (1955). Federation Proc. 14, 126. Rutledge, L.T., and Doty, R. W. (1957). Am. J. Physbl. 191, 189. Sacra, P.,Rice, W. B., and McColl, J. D. (1957). Can. J. Biochem. and Physiol. 36, 1151. Sandberg, F. ( 1959). Arzneimittel-Forsch. 9, 203. Schneider, J. A. (1954). Proc. SOC.Exptl. B i d . Med. 87, 614. Schneider, J. A., Plummer, A. J., Earl, A. E., and Gaunt, R. (1955). Ann. N . Y. Acad. Sci. 61, 17. Shaw, E., and Woolley, D. W. (1954). J. Pharmacol. Exptl. Therap. 111, 43. Sidman, M. (1953). Science 118, 157. Sigg, E. B. (1959). Federation Proc. 18, 144. Sivadjian, J. (1934). Compt. rend. acad. sci. 199, 884. Sivadjian, J. (1935). Compt. rend. SOC. bid. 118, 963. Slater, I. H., and Jones, G. T. (1958). J . Pharmacol. Exptl. Therap. 122, 69A. Smith, R. P., Wagman, A. I., and Riopelle, A. J. (1956). J. Pharmacol. Erptl. "hemp. 117, 136. Smith, R. P., Wagman, A. I., Wagman, W., Pfeiffer, C. C., and Riopelle, A. J. (1957). J. Pharmucol. Exptl. Therup. 119, 317. Stein, L. (1956). Science 124, 1082. Stephens, J. H., and Gantt, W. H. (1956). Bull. Johns Hopkins Hosp. 98, 245. Stille, G., Brunkow, I., and Kroger, H. (1956). Arzneimittel-Forsch. 6, 482. Sui, V. M. (1957). Farmukol. i Toksikol. 20, 18; (1959). Ber. ges. Physiol. u. exptl. Pharmakol. 196, 114. Taeschler, M., and Cerletti, A. (1958). Schweiz. med. Wochschr. 88, 1216. Taeschler, M., Fanchamps, A., and Cerletti, A. (1960). Psychiat. et Neurol. 139, 85.

DRUGS AND THE C A R

277

Tedeschi, R. E., Tedeschi, D. H., Cook, L., Mattis, P. A., and Fellows, E. J. (1959). Federation Proc. 18, 451. Truitt, E. B., Jr. (1958). J . Nervous Mental Diseuse 126, 184; (1959) Klin. Wochschr. 37, 577. Verhave, T., and Owen, J. E., Jr. (1958). The Psychol. Record 8, 49. Verhave, T., Owen, J. E., Jr., and Robbins, E. B. (1957). Psychol. Rept. 3, 421. Verhave, T., Owen, J. E., Jr., Fadely, D., and Clark, J. R. (1958a). J , Pharmacol. Exptl. Therap. 122, 78A. Verhave, T., Owen, J. E., Jr., and Robbins, E. B. (195813). Arch. intern. pharmucodynamie 116, 45. Verhave, T., Owen, J. E., Jr., and Slater, I. H. ( 1 9 5 8 ~ )“Psychopharmacology, . Progress in Neurobiology 111,” p. 267. Cassel, London, England. Verhave, T., Owen, J. E., Jr., and Robbins, E. B. (1959). J . Pharmacol. Exptl. Therap. 126, 248. Vernier, V. G., and Unna, K. R. (1956). Ann. N . k’. Acad. Sci. 64, 690. Vogt, M. (1954). J . Physiol. (London) 123, 451. Votava, Z.,and Vanecek, M. (1956). X X Intern. PhySWl. Congr., Brussels. Warner, L. H. (1932). J. genet. Psychol. 41, 57. Weil-Malherbe, H.,kuelrod, J., and Tomchick, R. (1959). Science 129, 1226. Weiskrantz, L. ( 1957). In “Psychotropic Drugs” ( S . Garattini and V. Ghetti, eds.), p. 67. Elsevier, Amsterdam, Holland. Weiskrantz, L., and Wilson, W. A., Jr. (1955). Ann. N . Y. Acad. Sci. 61, 36. Weiskrantz, L.,and Wilson, W. A., Jr. (1956). Science 123, 1116. Wikler, A. (1946). Federation Proc. 6, 213. Winter, C. A., and Flataker, L. (1949). Federation Proc. 8, 169. Winter, C. A., and Flataker, L. ( 1951). J. Pharmacol. Exptl. Therap. 101, 156. Winter, C.A., and Flataker, L. (1956). Proc. SOC. Exptl. BWZ. Med. 92, 285. Winter, C. A., and Flataker, L. (1957). J . Pharmacol. Exptl. Therap. 119, 194. Wirth, W. (1958). Arzneimittel-Forsch. 8, 507. Wirth, W., Gosswald, R., Horlein, U., Risse, K. H., and Kreiskott, H. (1958). Arch. intern. phamzacodynamie 115, 1. Witt, P. N. (1956). Arzneinittel-Forsch. 6, 359.

This Page Intentionally Left Blank

METABOLIC AND NEUROPHYSIOLOGICAL ROLES OF 7-AMINOBUTYRIC ACID' By Eugene Roberts and Eduardo Eidelberg Department of Biochemistry. Medical Research Institute. City of Hope Medical Center. Duarte. California. and Brain Research Program. University of California. 10s Angeles. California

.

I Introduction ........................................... I1. Some Aspects of Distribution of Free Amino Acids in Nervous Tissue ................................................ I11 Detection and Quantitative Determination of y-Aminobutyric Acid (GABA) in Tissue Extracts ......................... IV Distribution of GABA in Animal Tissues .................. V Metabolic Relations of Some of the Free Amino Acids in the Central Nervous System (CNS) .......................... VI Metabolic Relations of GABA in the CNS . . . . . . . . . . . . . . . . . VII . GABA as a Possible Precursor of Other Substances . . . . . . . . . . VIII . Developmental Features of Glutamic Acid Decarboxylase (GAD) and GABA ..................................... IX Factors Which Might be Involved in Regulating the Levels of GABA in the CNS ..................................... A Availability of Substrate ............................. B. Possible Role of Intracellular pH and pC0, ............. C . The Role of Pyridoxal Phosphate in the Function of GAD and GABA-T .................................... D Some Other Considerations Relevant to Vitamin B, and the GABA System .................................... X Lack of Correlation of GABA Levels and Seizure Susceptibility in Areas of Normal Brain ............................... XI . Physiological Mechanism of Action of GABA at the Cellular Level ................................................. A Stretch Receptors ................................... B Heart Ganglia ...................................... XI1 Pharmacological Studies with GABA ...................... A Spinal Cord ........................................ B Brain ............................................. C. Blood Pressure ..................................... XI11 General Comments and Summary ......................... References ............................................

. . . . .

.

.

.

. . . . . .

280 281 281 283 285 288 295 295

300 300 301 305 316 318 319 319 321 322 322 323 325 326 327

1 Supported in part by Grant #1615 from the National Institute for Neurological Diseases and Blindness. United States Public Health Service and the National Association for Mental Health. Inc.

279

280

EUGENE ROBERTS AND EDUARDO EIDELBERG

1.

Introduction

Only recently pharmacology, electrophysiology, and biochemistry have begun to meet at the level of the nerve cell to attempt to explain those properties which are germane to its unique function, the generation and conduction of the nerve impulse. It is not surprising, therefore, that y-aminobutyric acid ( GABA), a substance which has a unique occurrence in the central nervous system (CNS) and which bas inhibitory properties in various neurophysiological test systems, has aroused such great interest. GABA first became of interest in mammalian physiology because it was found in an easily extractable form in large amounts uniquely in brain and spinal cord of various species. Elucidation of the metabolic pathways in which this substance participates has indicated that GABA is a key member of a metabolic shunt around the a-ketoglutarate oxidase step of the tricarboxylic acid cycle in brain, a reaction sequence which includes an oxidative reaction that is coupled with phosphorylation. Physiological experiments have shown that GABA has inhibitory effects in tests with invertebrate and vertebrate nervous systems, and have indicated that GABA or a metabolite thereof may play an important role in the regulation of activity in the nervous system, possibly by rendering neuronal membranes more permeable to potassium and chloride ions. Recently, a number of comprehensive papers have been presented in which the biochemical and neurophysiological findings relating to GABA have been summarized in considerable detail and a symposium has been devoted to this subject2 (Baxter and Roberts, 1960; Elliott and Jasper, 1959; Elliott, 1959; Florey and Florey, 1958; Grundfest, 1958; Jinnai and Mori, 1958; K d e r and Edwards, 1958; Killam, 1958; Roberts, 1956, 1959, 1960; Roberts et at., 1958b; Tower, 1959). Therefore, no attempt will be made in this article to be all-inclusive in covering the literature. Only those aspects of the problem will be discussed in which chemically understandable phenomena might possibly be related to normal or abnormal function in the CNS. 2 Symposium on Inhibition of the Nervous System and y-Aminobutyric Acid, June 22-24, 1959, City of Hope Medical Center, Duarte, California.

ROLES O F y-AMINOBUTYRIC ACID

II.

281

Some Aspects of Distribution of Free Amino Acids in Nervous Tissue

Extensive studies have been made of the distribution of free or loosely bound amino acids and related substances in animal tissues in our laboratories with the aid of two-dimensional paper chromatographic procedures. It was found that in a given species at a particular stage of development each normal tissue, including every type of blood cell, has a distribution of ninhydrin-reactive constituents which is characteristic for that tissue (see Roberts et al., 1957 and 1958a for summaries). In general, the significance of the presence of relatively large quantities of some of the amino acids in the free or easily extractable form in tissues is not known. All of the evidence at present indicates that GABA probably plays an important role in the physiological function of the CNS. Recently reported data indicate that both glutamic and aspartic acids can produce spreading depression in rabbit cortex (Van Harreveld, 1959) and that aspartic and glutamic acids have an excitatory action on cat spinal neurons (Curtis et al., 1959a), while f3-alanine and taurine, as well as GABA, have a depressant action (Curtis et al., 1959b; Curtis and Watkins, 1960). All of the above substances are found in the free or easily extractable form in the CNS. It is probable that other constituents, as well, will be found to have important physiological effects when suitable systems for their demonstration will be found. The presence or absence of a measurable electrical effect upon topical application or injection of a substance may not provide unequivocal evidence about the role which the substance plays at some intracellular site. Therefore, although a great deal of attention is now focused on problems related to GABA, it is important to keep in mind that GABA is only one of the large number of freely extractable substances in nervous tissue and that its activities and functions must always be considered in relation to the total chemical environment in which it is found. 111.

Detection and Quantitative Determination of y-Aminobutyric Acid (GABA) in Tissue Extracts

Most of the determinations of GABA which have been published to date have been performed by the use of standard paper

282

EUGENE ROBERTS AND EDUARDO EIDELBERG

chromatographic procedures (Consden et al., 1944; Dent, 1947, 1948). These are generally adequate for extracts containing relatively large amounts of GABA, since under these circumstances results can be interpreted easily. However, reports of the presence of small amounts of GABA in blood, urine, spinal fluid, etc., based solely on the use of ordinary paper chromatographic methods cannot be accepted without independent verification. The quantitative paper chromatographic procedure, variations of which have been employed most extensively, is cumbersome and requires controls with every set of determinations (Roberts and Frankel, 1950). Column chromatographic procedures have been devised by which GABA can be determined in extracts of tissue (Berl and Waelsch, 1958; Tallan et al., 1954; Okumura et d.,1959; Clouet et al., 1957). The latter methods are most suitable when accurate determinations of a variety of ninhydrin-reactive constituents are desired rather than when determinations of GABA alone are required since the procedures are relatively complicated and timeconsuming. Reports have been made of the determination of GABA by biological assay with crayfish stretch receptor (Florey and Florey, 1958). Since this assay system responds not only to GABA but also to other substances ( McLennan, 1957, 1959; Florey and McLennan, 1959; Brockman and Burson, 1957) the values obtained by this procedure are probably a composite for several substances with different specific activities. However, the crayfish stretch receptor can be used in an important qualitative way, for if no response is elicited by an extract in this extremely sensitive bioassay, it may be presumed that GABA is absent. A newly devised, specific spectrophotometric method for the microdetermination of GABA employing a bacterial enzyme (Scott and Jakoby, 1959; Jakoby and Scott, 1959) already has proven to be of great value. With this method we have failed to detect a significant amount of GABA in the spinal fluid of normal human subjects or in individuals with a variety of pathological states affecting the CNS. However, the enzyme which forms GABA, L-glutamic acid decarboxylase (GAD), has been detected in small amounts in cerebrospinal fluid (Vates et al., 1959).

ROLES OF y-AMINOBUTYRIC ACID

283

IV. Distribution of GABA in Animal Tissues

In the study of tissues and body fluids in various vertebrates which included fish, amphibian, reptilian, avian, and mammalian species, GABA was detected unequivocally only in brain and spinal cord. All ninhydrin-reactive constituents other than GABA which were found on two-dimensional chromatograms of alcoholic extracts of brain and spinal cord also occurred in varying concentrations in the other tissues. Although the extracts of brains of the different species studied showed marked variations in the contents of most of the detectable ninhydrin-reactive constituents, they all showed relatively high concentrations of GABA, glutamic acid, and glutamine. GABA was found to be present in large amounts in the cerebral cortex of dogs but was not detectable in extracts of the brachial plexus or optic nerve (Roberts et al., 1958a). It was detected in brain and spinal cord but not in sciatic nerve of the rabbit (Roberts et al., 1950). GABA has not been detected in extracts of sympathetic ganglia, liver, spleen, or heart muscle (Florey and McLennan, 1955) but has recently been reported to be present in retina and choroid but not in the iris or ciliary body of the eye of the dog and ox (Kojima et al., 1958). The above results emphasize the uniqueness of the occurrence of GABA in tissues of the CNS. Variations have been found in amounts of ninhydrin-reactive constituents in different parts of brain. The most marked differences generally have been noted in the content of GABA. Higher levels of GABA are found in gray than in white matter. In the rat, the corpora quadrigemina and diencephalic regions contained the highest Ievels of GABA, whiIe the lowest levels were found in the whole cerebral hemispheres and the pons and medulla. In the chicken the optic lobes and diencephalon contained larger quantities of GABA than the cerebral hemispheres, hindbrain, or cerebellum (Roberts, 1959, and Sisken et al., 1960). Similar studies employing different parts of rabbit and monkey cortex have been performed in our laboratory. No attempt will be made to catalogue the quantitative data obtained to date on distribution of GABA in various cerebral areas, since the differences from one area to another in the various species are not readily interpretable at the present time in terms of the function of these areas (Baxter et al., 1960a). The amount of GABA present at any time in a particular area

284

EUGENE ROBERTS AND EDWARD0 EIDELBERG

of the CNS is probably a reflection of the balance between the rate of formation and the rate of utilization. Parenterally administered GABA does not appear to enter the brain readily (Roberts et al., 1958a). Intravenous injection of as much as 3 gm of GABA per kilogram into mice produced no change in the GABA levels of whole brain. Similar results were obtained when GABA was injected intraperitoneally into rats. At a time when the blood level was high no GABA was detectable in the cerebrospinal fluid of an unanesthetized monkey which had been given 200 mg/kg of GABA (Van Gelder and Elliott, 1958). These findings suggest that GABA does not pass the blood-brain barrier readily. However, physiologically disturbed or diseased areas of brain might allow passage into cerebral tissue of blood-borne GABA. Electrophysiological evidence indicates that systemically administered GABA ordinarily does not pass the blood-brain barrier in the cat but that following experimental local breakdown of the barrier penetration of intravenously injected GABA does take place (Purpura et al., 1958). After injection of GABA solutions into mice or rats the amino acid was found distributed in liver, kidney, and muscle and was excreted in relatively large amounts in the urine (Roberts et al., 1958a; Van Gelder and Elliott, 1958). Thus, GABA appears to enter quite readily from the bloodstream into tissues other than brain. There is no evidence that the GABA, which is formed in the brain, can leave the brain readily and enter the blood. GABA has not been detected by us in normal serum or in serum of patients with a variety of nonmalignant and malignant diseases at any time, nor has it been found in specimens of cerebrospinal fluid in patients even after massive head injuries. GABA appears to be taken up readily from solution by slices of cerebral cortex but not by liver or kidney slices or diaphragm sectors (Elliott and Van Gelder, 1958). At the present time it is difficult to relate the latter observations to the activity and function of GABA in the intact nervous system. These results might reflect the affinity of nervous tissue for GABA and might be related to the capacity for retention of the amino acid by nervous tissue. GABA levels in extracts of whole brain, as well as the amounts of other amino acids, tend to remain constant under various physiological stresses. For example, no significant changes were produced in GABA levels in whole brains of mice by a large variety of drugs

ROLES OF Y-AMINOBUTYRIC ACID

285

tested in single effective doses or over a period of time. Inanition, dehydration, hypophysectomy, adrenalectomy, thyroidectomy, and tumor growth also did not appear to alter the levels of brain GABA significantly (Roberts et al., 1958a). V.

Metabolic Relations of Some of the Free Amino Acids in the Central Nervous System (CNS)

Five of the amino acids which are found in the easily extractable form in the CNS in relatively large amounts, glutamic and aspartic acids, glutamine, GABA, and alanine, have intimate relationships to the tricarboxylic acid cycle. These are outlined schematically in Fig. 1. The reaction sequences shown in light lines are believed to be common to all tissues, while those in heavy lines appear to be operative only in the CNS. Interestingly, five ways are known GLUCOSE~dvo(lo,

\1

GLvcoKfssur/

FIG. 1. Outline of metabolic relationship of GABA to the tricarboxylic acid cycle.

286

EUGENE ROBERTS AND EDUARDO EIDELBERG

by which reversible interconversion of glutamate and a-ketoglutarate can take place. The latter two substances participate in reversible transaminations with the following keto acid-amino acid couples: oxalacetate-aspartate, pyruvate-alanine, and succinic semialdehyde-GABA. In addition, glutamate and a-ketoglutarate are interconvertible by a newly discovered direct transamination (Albers, 1960) and by the action of glutamic dehydrogenase. It would be expected from the outline shown in Fig. 1 that glucose carbon would be the chief source of the carbon of the above five amino acids and, indeed, it has been demonstrated in experiments with brain slices that the label from g1uc0se-C~~ appears readily in these amino acids ( Beloff-Chain et al., 1955). The most significant changes in free amino acids of brain produced either in vivo or in vitro have resulted from experimental procedures which could directly or indirectly affect the integrity of the carbohydrate metabolism. Hypoglycemia produced by insulin caused a decrease in brain glutamic acid and an approximately isomolar increase in aspartic acid content as well as decreases in GABA (Cravioto et al., 1951; Dawson, 1953). This result could be explained on the basis of the decreased availability of glucose. The consequently lowered level of acetyl-CoA would result in a decrease in the rate of condensation of acetyl-CoA with oxalacetate to form citrate, thus allowing the transamination of glutamic acid with oxalacetate, forming aspartate and a-ketoglutarate, to proceed at a more rapid rate. Fluoroacetate, which blocks the tricarboxylic acid cycle prior to the a-ketoglutarate oxidase step, produced a reduction in the contents of both glutamic and aspartic acids in the brains of rats (Dawson, 1953). It would be expected that the glutamate levels would be reduced if the availability of a-ketoglutarate would be reduced and that a decreased rate of transamination of glutamate with oxalacetate to form aspartate would result in a fall in aspartate level. In experiments with brain slices anaerobic conditions caused a great increase of formation of lactic acid from glucose and completely suppressed the appearance of the glucose carbon in the amino acids (Tsukada et al., 1958). A variety of inhibitors of glucose metabolism was found to inhibit formation of amino acids from glucose although some of them increased glucose utilization and lactic acid formation. Of the brain amino acids, glutamate was formed most rapidly from the carbon of glucose suggesting that the first point of removal of a carbohy-

ROLES OF y-AMINOBUTYRIC ACID

287

drate intermediate as an amino acid occurs at the a-ketoglutarate step under the conditions of the experiments. That there is a remarkably rapid synthesis of glutamine in intact brain was shown by the finding that a large portion of intracerebrally administered ~-glutamic-U-C~~-acid was converted to glutamine in brains of mice even within 1 minute after injection (Roberts et al., 195813). These results also indicate that the incorporation of ammonia into the amide group of glutamine is a very rapid reaction in brain in uiuo. Remarkable increases in the content of glutamine over control levels were found in the brains of hepatectomized dogs (Flock et al., 1953) and during the coma associated with Eck fistula (Bollman et al., 1957) and in rats poisoned with ammonium salts (du Ruisseau et al., 1957). In dogs receiving bilateral intracarotid infusion of lactate-Ringer solution containing 1% of ammonium hydroxide, the only significant change observed in the free amino acids of brain was an increase in the content of glutamine (Eiseman et al., 1959). In the latter instance, it would appear that the withdrawal of glutamate as glutamine and increase of the latter substance in the free amino acid pool was occurring at a rate which was compatible with the maintenance of the normal levels of the other constituents. Although the glutamine synthetase reaction, by which glutamine is made from glutamic acid, ammonia, and adenosine triphosphate (ATP), is not reversible, glutamine can be reconverted to glutamic acid by the action of glutaminase, a hydrolytic enzyme, or by reactions which utilize the amide group, an example of which is the first step in purine biosynthesis, the formation of phosphoribosylamine from glutamine and phosphoribopyrophosphate ( Hartman and Buchanan, 1958). Since glutamine enters brain readily and is present in blood in relatively high concentrations, it could serve as a source of brain glutamate and a-ketoglutarate. It is conceivable that in severe hypoglycemia the intracellular amino acids together with fatty acids could serve briefly as a source of carbon for the maintenance of cellular oxidations. The system of reversible reactions discussed above linked to the irreversible tricarboxylic acid cycle would seem to be well suited for the reciprocal regulation of cellular oxidative processes and the maintenance of relatively constant intracellular levels of the individual free amino acids. Unquestionably, within the above reaction scheme there also will be found feedback mechanisms by which

288

EUGENE ROBERTS AND EDUARDO EIDELBERG

various reaction rates will be found to be regulated by the products of the reactions.

VI. Metabolic Relations of GABA in the CNS GABA is formed in brain to a large extent, if not entirely, from L-glutamic acid by the action of an L-glutamic acid decarboxylase (GAD), an enzyme found in mammalian organisms only in the CNS almost entirely in the gray matter [reaction (l)] (see Roberts, 1956). glutamic decarboxylase

HOOCCH,CH,CH( NH,) COOH b-, L-Glutamic acid

B, coenzyme

+ CO,

HOOCCH,CH,CH,NH, GABA

(1)

The decarboxylase reaction is essentially irreversible, no fixation of radioactive carbon from C1402into glutamic acid being noted under the anaerobic conditions employed for the manometric measurement of the maximal rate of the decarboxylation. GABA-a-ketoglutarate transaminase ( GABA-T) is found in the CNS also chiefly in the gray matter, but in contrast to GAD, it is also found in other tissues (Bessman et at., 1953; Roberts and Bregoff, 1953). It catalyzes the reversible transamination of GABA with a-ketoglutarate [reaction ( 2 ) ] . 0

II

HOOCCH,CH2CH,NH, GABA

, ,

H00CCH CH CH0 Succinic semialdehyde

+ HOOCCH,CH,CCOOH p ** a-Ketoglutaric acid

+

GABA-a-ketoglutarate transaminase

B, coenzyme

HOOCCH,CH,CH( NH, )COOH L-Glutamic acid

(3)

If a continuous metabolic source of succinic semialdehyde were available, GABA could be formed by the reversal of reaction ( 2 ) . However, to date no evidence has been adduced for the formation of significant amounts of GABA by reactions other than the decarboxylation of L-glutamic acid. The best available data on the distribution of GAD and GABA-T are shown in Tables I and 11. The results in Table I are the only ones in which both GAD and

ROLES OF Y-AMINOBUTYRIC ACID

289

GABA-T have been measured in the same laboratory in similar areas of the CNS. The data are shown in separate tables because one set of values was expressed on a protein basis (Table I ) while the other was calculated on the basis of original fresh weight of tissue (Table 11). Both the lipid and water contents of the tissues would have had to be known to express the data of both studies on a comparable basis. The refined microanalytical work of the above two groups (Tables I and 11) has borne out the original observations on grossly dissected portions of cat brain which showed clearly that the levels of GABA and GAD were much higher in gray than in white matter ( Roberts et al., 1951). Extremely low levels of GAD activity were found in the white matter from the monkey, rabbit, or rat brain. Insignificantly low levels of GABA-T were found in the white tracts and peripheral nervous tissue. Wide variations were found in both GAD and GABA-T activity within the gray matter. Relative activities of GAD in whole brains of four species were as follows: monkey, 100; rabbit, 187; rat, 251; and mouse, 340 (Lowe et al., 1958). When individual areas of rabbit and monkey brain were compared, higher values were found in the rabbit brain for each area (Lowe et al., 1958). The histological correlations with the measurements of enzyme activity in the latter two studies would suggest that both GAD and GABA-T are probably associated with neuronal elements rather than with axis cylinders, myelin, or oligodendroglia. A distribution study similar to the two above made in beef brain, but employing the crayfish stretch receptor for the assay of presumptive GABA levels, also has been made (Florey and Florey, 1958). The GABA-T activity was highest in the subcortical structures, where the maximal potential activity in the various areas was generally in excess of the measured GAD activity (Table I). In most cortical structures the rates of the two reactions were much more similar. It is interesting that the variations in GABA-T and GAD from one area to another did not parallel one another. It is of interest to consider whether one or the other of the above two enzymes controls the level of GABA in various cerebral areas. The data presented above show that there is no correlation between GAD and GABA-T levels in various cerebral areas. In Fig. 2 are plotted the GABA levels and GAD activity in various areas of chick brain (Sisken et nl., 1960), in mouse whole brain

290

EUGENE ROBERTS AND EDUARDO EIDELBEXG

DISTRIBUTION OF GAD

TABLE I GABA-T ACTIVITIESIN NERVOUSSYSTEM^, AND

THE

MONKEY GABA-T

Structure Cortical structures Hippocampus Molecularis Radiata Pyramidalis Alveus Occipital cortex Layer 1 Layer 1-2 Layer 213 Layer 3-4a Layer 4 Layer 4b-5 Layer 5 6 Layer 6 Subjacent white

+

GABA-T

GAD

GABA

94 66 68 46

36 25 33 17

2.6 2.6 2.1 2.7

30

-

55

-

40

-

9

2

25 58

1.2c

19

13 0

-

Motor cortex Gross region Layer 1 Layer 2 3 Layer 5 6 Subjacent white

34 65 36 4

Cerebellar cortex Molecular layer Granular laver

150 74

54 46

2.8 1.6

294

6

49.0

276 252 219 209

-

5

55.3

-

-

8

26.1

149

15

9.9

+

+

Subcortical structures Inferior olivary nucleus Superior colliculus (ext. and middle gray) Dentate nucleus Inferior colliculus Abducens nucleus Anterior hypothalamic nucleus

-

62

-

-

-

-

-

Salvador and Albers, 1959; Albers and Brady, 1959. All results expressed in mmoles/kg/hr in terms of protein content. GAD in terms of CO, evolved; GABA-T in terms of succinic semialdehyde formed. 0 Ratio of combined averages. a

b

291

ROLES OF Y-AMINOBUTYRIC ACID

TABLE I (Cont.) GABA-T Structure Subcortical structures-continued Superia colliculus, inner gray Posterior hypothalamus Medial thalamic nucleus Caudate nucleus Reticular formation Putamen Substantia nigra Midbrain central gray Red nucleus Lateral thalamic nucleus Globus pallidus Lateral pyramids Fasciculus gracilis Pituitary stalk Optic tract Supraoptic nucleus Posterior pituitary Pineal Eland

GABA-T

137 124 115 114 113 105 96 94 62 53 52 0 0 0 0

GAD

GABA

68 16 20 8 14

3.5

-

1.4 5.9 3.1 6.6 3.7

-

-

-

23 2 0

Spinal cord (lumbar gray) ( A + D: ventral + dorsal) A B C D

197 202 246 315

7 12 17 26

28.2 16.8 14.5 12.1

Peripheral elements Dorsal roots Ventral roots Adrenal cortex Adrenal medulla Dorsal ganglion Superior cervical ganglion

0 0 0 15 8 20

2 3

-

2 2 1

-

-

-

(Roberts et al., 1951a), and in rabbit brain cortical area 2-3 (Baxter et al., 1960b). There is an essentially linear relationship between the amount of GABA and GAD activity. The fact that the line can be extrapolated to zero lends support to the hypothesis that GAD is the major if not the sole enzyme responsible for the formation of brain GABA. Similar results were obtained when such a plot

292

EUGENE ROBERTS AND EDUARW EIDELBERG

TABLE I1 DISTRIBUTION OF GAD IN WHITEAND GRAYMATTER IN MONKEY AND RABBIT NERVOUS SYSTEMS% b Structure Whole brain Gray matter Globus pallidus Superior colliculus Hypothalamus Visual cortex Cerebral cortex, unspecified Putamen Caudate nucleus, head Uncus and amygdaloid nucleus Motor cortex Thalamus Midbrain reticular substance Ammon’s horn Cerebellum, whole Lateral geniculate body Medullac Cervical spinal cord0 White matter Cerebellum Cerebral peduncle Cerebral ( subcortical) Medullary pyramid Internal supsule Cervical cord, white only Corpus callosum Brachium pontis Optic nerve Dorsal ganglion Sciatic nerve

Monkey

Rabbit

11

20

33 21 21 18

-

16 16 15 13 13 11 10 10 7 5 1 1 1 <1 <1 <1 <1 <1 <1 <1


-

37 27

16 22

-

17 12

-

-

3

L

<1

-

<1

Lowe et al., 1958. All results expressed in mmoles/kg/hr in terms of original wet weight of tissue. GAD in terms of GABA formed. c Areas contain white matter. a

b

was made for different areas of guinea pig brain. These results together with data on action of convulsant hydrazides suggest strongly that the steady state concentrations of GABA in various brain areas normally are governed by the GAD activity and not by the GABA-T.

293

ROLES OF y-AMINOBUTYRIC ACID

Brain also contains a dehydrogenase which catalyzes the rapid irreversible oxidation of succinic semialdehyde to succinic acid [reaction ( 3 ) ] (Albers and Salvador, 1958), which in turn can

I I

0.75

1

1.00 1.25

I

I

I

1.50

1.75

2.W

I

2.25

GAD GUM W fi%??ED/9W/W/1ooIv6) FIG.2. Proportionality of GABA content and GAD activity. Maximal values obtained: 1, chick cerebellum; 2, chick hemisphere; 3, chick hindbrain; 4, chick optic lobe; 5, chick diencephalon; X, whole mouse brain; 0, rabbit brain area 2-3.

be oxidized via the reactions of the tricarboxylic acid cycle (see Fig. 1). HOOCCH,CH,CHO

+ DPN+ + H,O

*-,clehydrogenase

Succinic semialdehyde

HOOCCH,CH,COOH Succinic acid

+ DPNH + H + (3)

As would be expected from the above relationships, glutamic acid, GABA, and succinic semialdehyde can be oxidized by various brain preparations and can support oxidative phosphorylation ( Tsukada et al., 1957; McKhann and Tower, 1959; Sacktor et al., 1960). It has been shown that the oxidation of GABA can take place in isolated washed mitochondria of rat brain, the P:O value for this oxidation approaching 3. Under the conditions in which GABA is the sole added substrate of oxidation of washed mitochondria it appears that $-hydroxy-GABA, rather than succinate, may be the first oxidation product ( Sacktor et al., 1960). These results indicate that brain mitochondria have the potentiality for forming $-hydroxyGABA from GABA. However, all of the results obtained with less "purified" brain preparations such as brain slices or minces or

294

EUGENE ROBERTS AND EDWARD0 EIDELBERG

in vivo are consistent with the major pathway of GABA metabolism in brain going via succinate (Roberts et al., 1958b; Tsukada et al., 1957). The basis for the differences in the observations may be in the lack of sufficient a-ketoglutarate in washed mitochondria to permit the transamination of GABA to take place. GABA also is metabolized in tissues other than the CNS since the GABA-T is not located exclusively in the CNS. A significant proportion of intraperitoneally administered GABA-1-C14 could be isolated from the urine of intact rats as labeled succinate when a relatively large amount of nonradioactive succinate was administered simultaneously with the labeled GABA (Roberts et al., 1958b). The carbon of intraperitoneally injected GABA-4-CI4 was distributed in glutamate, aspartate, alanine, and glycogen in rats (Wilson et al., 1959). The labeling patterns in the above compounds indicated that the succinate pathway can account for all of the metabolism of GABA under these conditions. Because of the complexity of the metabolic interrelationships in which glutamic acid and GABA participate and the many unknown parameters involved, no valid method has yet been devised to get a realistic assessment of the contribution of the GABA pathway to the energy metabolism of brain. However, an estimate of approximately 40% has been made for the contribution of the GABA shunt to oxidative metabolism of brain slices ( McKhann et al., 1960). In summary, the reaction sequences shown by the heavier lines on Fig. 1 are those which appear to be unique to tissue of the CNS. Although GABA can be transaminated by extracts of tissues other than those of the CNS, the GAD appears to be located almost exclusively in the gray matter of the CNS. Therefore, the operation of the reactions involving GABA would not be anticipated to play an important role normally in tissues other than the gray matter of the CNS. Because of the essential irreversibility of GAD and of the succinic semialdehyde dehydrogenase, the entire sequence of reactions dealing with the metabolism of GABA also would be irreversible. Thus, in the CNS, but not in other tissues, there exists an irreversible shunt around the a-ketoglutarate oxidase system, the operation of which depends on the unique occurrence of GAD and GABA in the CNS. The suggestion of the details of this shunt is based upon enzymatic studies and is completely consistent with in vivo observations.

ROLES O F Y-AMINOBUTYRIC ACID

VII.

295

GABA as a Possible Precursor of Other Substances

It has been suggested that GABA might serve as a precursor for a number of chemically related substances (Pisano et al., 1959 have made a critical evaluation of the data; see also Baxter and Roberts, 1960; Elliott and Jasper, 1959 and Roberts et al., 1958b for pertinent references ) . y-Guanidinobutyric acid has been isolated from calf brain (Irrevere and Evans, 1959) and can be formed by the transamidination of GABA with arginine. The enzyme involved is the transamidinase which is known to catalyze the synthesis of guanidinoacetic acid from glycine and arginine (Pisano et al., 1957; Pisano and Udenfriend, 1958; Pisano st al., 1960). To our knowledge this is the only metabolic pathway, other than the previously discussed transamination, demonstrated to date in which GABA is involved as a precursor of another substance which has actually been proven to be present in brain. y-Guanidinobutyric acid has been found to be active in various physiological test systems in which GABA has been shown to have an effect (Purpura et al., 1959; Kuffler and Edwards, 1958). It is considerably less effective than GABA in blocking sensory discharge in the crayfish stretch receptor system (Kuffler and Edwards, 1958). Definitive isolation of (3-hydroxy-GABA or proof of biosynthesis of this substance in brain has not been achieved yet. Homopantothenic acid, y-aminobutyrylcholine, and y-butyrobetaine have still not been proven to be present in brain by unambiguous procedures, and GABA has not been shown to be a precursor of carnitine, a substance of ubiquitous occurrence in the body. Although a number of important trace substances of physiological importance may be formed from GABA in brain, there is no evidence yet for the quantitatively important participation of any pathways in the normal metabolism of GABA other than those discussed in the previous section. VIII.

Developmental Features of Glutamic Acid Decarboxylase (GAD) and GABA

Observations have been reported of the amino acids of developing whole brain in the mouse, chick, and bullfrog (Roberts et al., 1958a). In all three species studied there was a progressive

296

EUGENE ROBERTS AND EDUARDO EIDELBEXG

increase in the content of GABA during development. This was not the case for any of the other constituents detected in chromatograms of extracts of whole brain in these species. Figure 3 shows a replot of previously published data on the changes with age in GAD and GABA levels of whole mouse brain (Roberts et al., 1951a). Up to five days after birth the amount of GABA that could be formed per unit tissue in the optimally constituted homogenate system in 30 minutes was approximately equal to that actually found in the same weight of brain. Thenceforth, up to approximately 20 days of age the rate of increase of GAD activity was much greater than that of the increase in the concentration of GAD

MOUSE ME /AWI 3. Changes with age in GAD activity and GABA levels of whole mouse brain. FIG.

GABA. Similar patterns of change with age have been found in various parts of chick embryo brain (Figs. 4-8) (Sisken et al., 1960) and in the rabbit cerebral cortex (Fig. 9 ) (Baxter et aZ.,

1960b). The above data suggest that the system by which GABA is used by brain also is developing at a rapid rate at these times and that the turnover of GABA might be increasing much more than the actual amount of the substance, itself, during this period. The increased turnover might be related to increasingly important participation of the GABA system in oxidative cerebral metabolism. Preliminary measurements which have been made on the GABA-T activity of whole chick embryo brain at various stages are entirely in keeping with this concept (Sano and Roberts, unpublished). The work on the changes with age in GABA-T activity in separate areas of chick embryo brain is being carried out at the present time.

297

ROLES OF Y-AMINOBUTYRIC ACID

0

5

M 15 HATCH 5 CHICK AGE fDAW FIG.4

10

15 "ADULT

-

2.00

175

-

150 -

loo 075 125

CHICK AGE fMWI

FIG.5 FIGS.4 and 5. Time sequence of changes in GABA levels and GAD activity in chick brain. Fig. 4: optic lobe, Fig. 5: diencephalon; A, appearance of Metrazol-induced seizures; B, appearance of distinct Nissl substance; C, definite PAS-positive reaction.

HEMISPHERE

ABC

t

/

GABA

W 5 L 0

CHICK AGE fDAyzI FIG.6 CEREBELLUM

-

075

OM0.25

I

125

I

r

GAD

Looto'wl

pa\ Y

0.75

y!:

-----___

Al&

a25

0

5

M

15

HATCH

5

10

15

ADULT

CHICK AGE /NUJ FIG.8 FIGS.6-8. Time sequence of changes in GABA levels and GAD activity in chick brain. Fig. 6: hemisphere, Fig. 7: cerebellum, Fig. 8: hindbrain; A, appearance of Metrazol-induced seizures; B, appearance of distinct Nissl substances; C, definite PAS-positive reaction.

ROLES OF Y-AMINOBUTYRIC ACID

299

In both the chick and mouse brains the increase in GAD activity was preceded by the differentiation of the major neuronal elements. As indicated by the arrows in Figs. 4-8, in all chick brain areas studied susceptibility to Metrazol seizures (Peters et al., 1956) and PAS-positive ground substance and distinct Nissl bodies (Sisken et a,?., 1960) all appeared prior to the sharp rise in GAD activity. The maximal increase in activity of GAD in both species corresponds to the period of greatest increment in

0% -
10

RABBIT AGEf4AYJI FIG. 9. Levels of GABA and GAD activity in area 2-3 of the developing rabbit brain cortex.

features which are related to maturation of the nuclear masses and fiber tracts of the CNS. The decrease in GAD from maximal levels observed to lower values in the adult brain of the chick may be attributable to the histologically observable increases in interstitial substance and fiber tracts, which would tend to decrease the amount of active neuronal tissue per unit fresh weight of tissue. However, GABA levels did not decrease in the adult chick brain. Observations have not yet revealed any histologically or cytologically observable variable which can be correlated sufficiently directly with GAD activity so that GAD activity might be attributed to a particular structure.

300

EUGENE ROBERTS AND EDUARDO EIDELBERG

The progressive increase in GABA with age during development in various species and of the activity of the enzyme which forms GABA, as well as their unequal distributions in various areas of brain, may furnish us valuable tools with which to study the role of GABA in the CNS. Any oxidative function ascribed to the GABA system should be low when the rate of turnover of GABA is low and should increase with age in a manner related to the increase in the rate of turnover. If GABA, itself, is an important inhibitory modulator of neuronal activity, then the actual quantity of GABA in a particular cerebral area should be related to physiological measurements which reflect this inhibitory activity. It cannot be predicted whether or not it will be possible to correlate the activity of the GABA system with particular types of neurons or neuronal processes, since the operation of this system may reside within subcellular units common to all neurons. IX. Factors Which Might be involved in Regulating the Levels of GABA in the CNS At least part of the interest in GABA is based on the findings that the reduction in levels of this amino acid in brain below those found normally are correlated with the actual occurrence of seizures or increased susceptibility to experimentally induced seizures in a variety of species (Killam, 1954, 1957, 1958; and Killam and Bain, 1957) and that elevated levels of this substance above normal are associated with greatly increased seizure thresholds in certain brain areas (Baxter and Roberts, 1960; and Eidelberg et al., 1960). It is, therefore, of value to consider those properties of the enzyme systems involved in the metabolism of GABA which might be important in regulating their activity in vim. A. AVAILABILITY OF SUBSTRATE An important factor which could influence the operation of the GABA shunt might be the availability of substrate. It has been mentioned previously that the evidence suggests that under normal conditions GAD might be the rate-limiting reaction of the GABA sequence. From the enzyme experiments performed to date it would appear that glutamic acid has a relatively low affinity for GAD under the conditions that it has been studied ( K , = 6.4

ROLES OF Y-AMINOBUTYRIC ACID

301

x mole/liter). If one assumes an average concentration of 0.01 it4 for glutamate in the CNS and a uniform distribution, then, on the average, the GAD could be operating at a greater than half maximal rate. A decrease in the level of glutamic acid in neuronal tissue could decrease the rate of formation of GABA and thus the rate of operation of all of the enzymes involved in metabolism of GABA. Such a decrease could be produced by insulin hypoglycemia and fluoroacetate poisoning (see Section V ) . Slices of cerebral cortex from human patients operated for cortical epileptogenic foci show decreases in glutamic acid levels during incubation under conditions which allow slices of normal cortex to increase the glutamate levels (Tower, 1959). However, it is useless to speculate further on this matter at this time, since analyses which have been performed to date for levels of the various substrates have not been able to distinguish the material which is present in neurons or in other cell types in the CNS, and it is not possible to assume uniform distribution of materials in the highly heterogeneous systems of nerve cells or even within the individual neurons themselves. Great variations have been found in the content of various ninhydrin-reactive materials, especially GABA, from one area of brain to another and of GAD and GABA-T ( see Section VI ) . B. POSSIBLE ROLEOF INTRACELLULAR PH AND pC02 The pH activity relationships of cerebral GAD and GABA-T are shown in Fig. 10. The results for each of the enzymes are plotted as percentages of maximal activity obtained under optimal conditions of assay for each enzyme. GAD possesses maximal activity at pH 6.5, GABA-T at 8.2. The two curves intersect (possess the same percentage of maximal potential activity) at approximateIy pH 7.5, which is slightly more alkaline than that found for squid axons, pH 7.1 (Caldwell, 1958), a value which might be assumed to be close to the intracellular pH of neurons in the CNS. It is apparent that small shifts in pH in the physiological range could result in changes of the relative activity of the two enzymes and so might alter the content of GABA at the loci where its physiological effects are exerted. If the results of the in vitro enzyme measurements can be translated into the interior of the nerve cell, intrucellular acidosis might be expected to in-

302

EUGENE ROBERTS AND EDWARD0 EIDELBERG

crease the content of GABA and alkalosis to decrease it. Although experiments to test this supposition have not been performed yet, it is of interest to examine at this time the physiological consequences of changes in carbon dioxide ( C 0 2 ) tension which might alter intracellular pH. Carbon dioxide, which is approximately twice as soluble in animal fats as in water (Nichols, 1957), is able to pass through cell membranes extremely rapidly in the gaseous form (Caldwell, 1958; Halpern and Binaghi, 1959) and can cause acidification of cells. The best measurements performed to date show clearly that

FIG.10. The pH activity relationships of brain glutamic acid decarboxylase and GABA-a-ketoglutarate transaminase.

increasing the C 0 2 tension on the outside of nerves decreases the internal pH (Caldwell, 1958), and the results of the latter study indicate that in vivo the internal pH of neurons is probably regulated to a considerable extent by the internal C02-bicarbonate buffer system. The enzyme, carbonic anhydrase, catalyzes the reversible hydration of carbon dioxide, a reaction which also occurs spontaneously, but much more slowly, in the absence of this enzyme. Other and secondary intracellular reactions occur which involve COz, H2C03, HC03-, C03=, H+, and OH-, but are not related directly to the activity of carbonic anhydrase ( Roughton and Clark, 1951) . The relationships may be summarized in the form of equations (see Umbreit, 1957 for thorough discussion) :

ROLES O F Y-AMINOBUTYRIC ACID

303

carbonic anhydrase + H,O

CO, (gas) & CO, (dissolved) HzC03 F1: H +

+ HC0,- + A-

- H,O (other anions) *HA

(4)

and the COz pressure, bicarbonate ion concentration, and pH are related by the following formulation: pH = pK

+ log

(HC03-) ; pK is approximately 6.3. (

co, 1

(5)

The presence of carbonic anhydrase would allow to take place rapidly the reversible interconversion of C 0 2 and bicarbonate ion, to which tissues are highly impermeable (Wallace and Hastings, 1942; Wallace and Lowry, 1942). A momentary increase in intracellular COP either from oxidative processes or from diffusion of CO, from a higher extracellular concentration would result in the combination of H+ ions with other intracellular buffers and formation of bicarbonate ions until an equilibrium would be reached between the CO2-bicarbonate system and the other buffer systems. If a rapid diffusion of CO, took place from the cell into an environment with a lower C 0 2 tension, the reverse process would occur. In the first case there would be a tendency to decrease intracellular pH and in the second to increase it; but carbonic anhydrase, by accelerating the interconversion of C 0 2 and bicarbonate, would decrease the extent and duration of the fluctuations in pH. The modulation of such fluctuations could be important even within a cell, since it might be envisioned that the pH changes caused by COa formation near a mitochondrion would be buffered sufficiently rapidly so that an acid-sensitive protein molecule in the vicinity would not be denatured. From the above discussion it would appear that one of the consequences of the absence of carbonic anhydrase or its inhibition would be a decrease in the ability of cells and tissues to buffer changes in pH with changes in C02. Carbonic anhydrase is widely distributed in the CNS (Ashby et al., 1952) as well as in other tissues. Studies in rats and guinea pigs during postnatal development have shown that a close correlation exists between the level of carbonic anhydrase in brain, which increases greatly, and the grade and threshold of electrically

304

EUGENE ROBERTS AND EDWARD0 EIDELBERG

induced seizures ( Millichap, 1957) . Failure to induce convulsions in the newborn rat was associated with a low level of carbonic anhydrase activity in the brain. A direct relation has been reported between anticonvulsant effect in an electroshock seizure test in mice and the inhibition of carbonic anhydrase by acetazolamide (Diamox) (Millichap d al., 1955) and substantial evidence has been adduced for the hypothesis that the anticonvulsant effects of both methazolamide and acetazolamide are directly attributable to inhibition of brain carbonic anhydrase (Gray et at., 1957). The anticonvulsant effect of COZ is well known. Metabolic acidosis in some instances has been found to decrease seizure susceptibility in humans and experimental animals, while alkalosis has been found to increase it (see Gray et al., 1957 for some pertinent references). It seems clear that the electroencephalographic changes produced by rapid variations of CO, in the blood are related directly to the partial pressure of CO, and not to alterations in over-all acid-base balance (Swanson et al., 1958). Hyperventilation, which results in decreased COz in the blood, slows and increases the voltage of the electroencephalogram, while breathing of a high COz mixture induces a faster rhythm than normal and a lower voltage. Progressive increases in arterial CO, cause somnolence, incoordination, and finally coma. Hypocapnia can cause confusion and even loss of consciousness in normal individuals and can precipitate petit-ma1 seizures in epileptics. Almost immediately after administration, intravenous acetazolamide has been shown in some cases of epilepsy to block the paroxysmal slow-wave discharges provoked by hyperventilation ( Saldias et at., 1957). Recent studies (Dunlop, 1957a) have shown that COz can depress the spontaneous and evoked activity in the fornix-hippocampal system, an area which has a very low threshold for epileptic seizure discharge and appears to be related to epileptiform phenomena. Although CO, also depressed thalamocortical and somatic sensory cortical pathways, the depression was greatest in the fornix-hippocampal system ( Dunlop, 1957b). The above data and observations may be summarized in the following manner: The results of the studies of the properties of GAD and GABA-T suggest that the relative activities of the two enzymes in a particular brain area may be altered by changes in intracellular pH. Acidification would favor GAD and an increase in GABA; alkalinization would favor GABA-T and a decrease in

ROLES OF Y-AMINOBUTYRIC ACID

305

GABA. As will be discussed later, an increase in GABA is associated with a decrease in neuronal excitability and a decrease in GABA with an increase. Increasing extracellular CO, is a physiological method by which an increase in intracellular acidity can be achieved and COz also is known to be a depressant of neuronal excitability. Carbonic anhydrase is an enzyme which enables cells and tissues to minimize pH effects which would result from metabolically formed or externally administered C 0 2 . The anticonvulsant effect of carbonic anhydrase inhibitors is related directly to inhibition of the brain enzyme and is associated with an increased level of C02 in the brain, which would tend to lower the intracellular pH. One of the ways in which C 0 2 may exert its action is by increasing intraneuronal acidity and inducing local increases in levels of GABA. Eventually, if such a hypothesis can be supported by solid experimental data, the results should be related to ionic shifts, the functioning of other pH-sensitive enzyme systems in brain, etc. At the present moment direct links do not exist between changes in intracellular acidity and the GABA system. However, from the above discussion it would now appear feasible to begin the study of the operation of the GABA system in those areas of brain in which susceptibility to excitation can be altered greatly by COZ or carbonic anhydrase inhibitors.

C. THEROLEOF PYRIDOXAL PHOSPHATE IN THE FUNCTION OF GAD AND GABA-T From the physiological point of view probably the most important property of GAD and GABA-T studied to date is the requirement of these enzymes for pyridoxal phosphate ( a form of vitamin Be) as a coenzyme. Prior to the discussion of the specific aspects of interrelationships of pyridoxal phosphate with GAD and GABA-T a brief summary will be given of the relationships which may exist between apoenzyme and the coenzyme of vitamin B6-requiring enzyme systems in general. Below is shown the formula of pyridoxal phosphate: H-CpO

Pyridoxal phosphate

306

EUGENE ROBERTS AND EDUARDO EIDELBERG

The pyridine ring and phenolic, formyl, methyl, and phosphate groups may all be important in attachment of the coenzyme to the protein moiety, the apoenzyme, and the different types of interactions which can occur between coenzyme and the apoenzymes of various Be-enzymes, as well as the structure of the apoenzyme, probably play a role in determining the nature of the reaction catalyzed by the holoenzyme as well as the relative affinity of the apoenzyme for coenzyme (see Snell, 1958 for thorough discussion of many facets of these problems). The spectral properties of crystalline phosphorylase are consistent with the existence of the aldamine structure ( I ) at physiological pH values and of the free Schiff base form at more acid values (11) (Fischer et al., 1958). P#o?.€/!

(1)

PRrn€/ff

AOd71FN

(11)

( 111)

Striking confirmation of the validity of the above formulations was obtained when E-N-pyridoxyllysine was isolated after acid hydrolysis of crystalline phosphorylase which had been reduced with sodium borohydride at an acid pH (Fischer et al., 1958). A similar result was obtained with crystalline cystathionase ( Fischer and Krebs, 1958). It is probable that the amino group of substrate amino acids displaces the protein amino group and interacts with the aldehyde group in the manner shown in (111) (see Snell, 1958 ) . Schematic representations of relations between apoenzyme and coenzyme of vitamin B6 enzymes are shown in Fig. 11. In addition to participating in attachment to the apoenzyme and in reactions with substrates, the aldehyde group of pyridoxal phosphate can react with a wide variety of reagents X making this group unavailable for the attachment of substrate to the enzyme. Such agents may react with the aldehyde group alone, or may also react with other groups of the coenzyme or apoenzyme in addition to the aldehyde group of the B6 moiety. Inactivation of the

307

ROLES OF Y-AMINOBUTYRIC ACID

aldehyde group may occur upon reaction with the commonly employed carbonyl-trapping agents such as hydroxylamine, semicarbazide and hydrazine ( Gale, 1946), penicillamine ( du Vigneaud et al., 1957), histamine, various phenylethylamine derivatives containing a mtu-hydroxy group such as noradrenaline, 3,4-dihydroxyphenylalanine, m-tyrosine, m-tyramine, 3-hydroxyphenylserineY and 3,4-dihydroxyphenylserine ( Schott and Clark, 1952; Holtz and

...

aam?# FIG.11. Vitamin B, enzymes-possible

relationships of coenzyme and

apoenzyme and effects of inhibitors.

Westermann, 1957). Unquestionably, many other naturally occurring and synthetic materials will be found to react with the aldehyde group of free or protein-bound pyridoxal phosphate. In the case of bacterial GAD, an enzyme containing tightly-bound pyridoxal phosphate, it actually has been demonstrated that hydroxylamine and glutamic acid compete with each other for attachment to the aldehyde group under the conditions of measurement (Roberts, 1952). The inhibition of the latter enzyme by hydroxylamine, which could be relieved completely by the addition of excess pyridoxal phosphate to the reaction mixture, illustrates the

308

EUGENE ROBERTS AND EDUARDO EIDELBERG

reversible equilibrium shown in reaction 1, Fig. 11. If the attachment of the apoenzyme to coenzyme is strong and the product formed by X is only slightly dissociated under the conditions of testing, reaction 1 would proceed in the forward direction and would give mostly inactive holoenzyme. Examples of this type of reaction are found in the inhibitions of 5-hydroxytryptophan decarboxylase of guinea pig kidney by semicarbazide and hydroxylamine (Clark et al., 1954; and Weissbach et nl., 1957) and of GABA transaminases of Ps,ezrdomonas aeruginosa (Noe and Nickerson, 1958) and brain (Baxter and Roberts, 1958), none of which are readily reversed by pyridoxal phosphate. Free inactive derivatives of the coenzyme could be formed from pyridoxal phosphate by reaction with the bound enzyme and subsequent dissociation from the apoenzyme (reaction 1 followed by reaction 2) or by a combination of X with the free coenzyme (reaction 3 ) . The inactive derivative could compete with pyridoxal phosphate for attachment to the apoenzyme, be converted back to pyridoxal phosphate, or be excreted in the urine. Deoxypyridoxine phosphate can compete with pyridoxal phosphate for the apoenzyme of bacterial tyrosine decarboxylase ( Umbreit and Waddell, 1949) . The hydrazones of semicarbazide and isonicotinic hydrazide and pyridoxal phosphate are formed in vivo and excreted in the urine (Williams and Abdulian, 1956), while the urinary product in the case of penicillamine is probably the thiazolidine derivative (du Vigneaud et nl., 1957; Kuchinskas et nl., 1957). Injection of hydrazides is accompanied by the appearance in tissues of large amounts of the hydrazones of pyridoxal phosphate and pyridoxal (Bain and Williams, 1960). In the cases of some enzymes with relatively low affinity for pyridoxal phosphate the equilibrium shown in reaction 4 may exist between the holoenzyme and apoenzyme and coenzyme. Such enzymes would be expected to be particularly sensitive to the lack of availability of an adequate circulating supply of vitamin B6, showing rapidly reversible decreases in activity during dietary deficiency of the vitamin. Various dietary forms of vitamin B6 can be converted rapidly to pyridoxal phosphate in tissues. GAD has a relatively low affinity for pyridoxal phosphate. A variety of in vitro experiments demonstrated that the coenzyme can be removed from the enzyme protein very easily and that the lost enzymatic activity can be restored simply by the addition

ROLES OF Y-AMINOBUTYRIC ACID

309

of pyridoxal phosphate. Experiments with rats showed that in pyridoxine deficiency there was a decrease in the degree of saturation of the apoenzyme of cerebral GAD with the coenzyme, but no decrease in the content of apoenzyme in the deficient animals. The control level of enzyme activity was restored rapidly to normal upon feeding pyridoxine to the deficient animals. GAD was readily reversibly inhibited by a variety of carbonyl-trapping agents (see Roberts, 1956 and Roberts et al., 1958b for pertinent references). GABA-T also requires pyridoxal phosphate as a coenzyme, but it appears to have a much greater affinity for the pyridoxal phosphate than GAD (Baxter and Roberts, 1958). GAD was found to be inhibited approximately to the same extent by semicarbazide and hydroxylamine, the inhibition by both of these agents being readily reversed by pyridoxal phosphate. However, the inhibition of GABA-T by 1 X lo-* M hydroxylamine was complete and only partially reversible, while 1 x M semicarbazide was not inhibitory at all under the conditions of testing. The above considerations suggested that differential inhibition of GAD and GABA-T might be feasible. If this could be achieved in viuo, then brain levels of GABA could be altered and the physiological consequences of such alterations could be examined. The in vivo effects of carbonyl reagents on the activity of B6 enzymes in brain depend on many factors, among which are the ease of penetration of the blood-brain barrier, rate of detoxication, major ionic forms present at physiological pH values, and the structure of the enzyme itself, which would determine to a considerable extent the probability of the successful collision of the reagent with the aldehyde group of the coenzyme and the rate of dissociation of the complex, once formed. It has now been possible both to lower and raise GABA levels in brains of experimental animals, presumably by interfering with GABA or GABA-T, respectively, in a preferential manner, and the correlative physiological findings suggest strongly that endogenous levels of GABA may be involved in the control of neuronal excitability in the CNS. The first indication of a truly physiological role for the GABA in mammals came from elegant studies on the mechanism of the action of convulsant hydrazides (Killam, 1954, 1957, 1958; Killam and Bain, 1957; Killam (et al., 1960). Administration of a variety of hydrazides to animals uniformly resulted in the production of

310

EUGENE ROBERTS AND EDUARDO EIDELBERG

repetitive seizures following a rather prolonged latent period. The seizure activity began subcortically and spread to the cortex. Since hydrazide-induced seizures could be prevented by the parenteral administration of various forms of vitamin B6 (Jenney et al., 1953; also see Killam, 1954 and Killam et al., 1960) it was conjectured that some enzyme system requiring pyridoxal phosphate was being inhibited and that the decrease in the activity of the enzyme was somehow related to the production of seizures. Attention was directed to GABA and GAD because of their unique occurrence in the CNS, and, indeed, it was found that the seizures were accompanied by substantial decreases in levels of GABA and reductions in the GAD activity of the brain areas studied. It was shown subsequently in living mice that injection of thiosemicarbazide, one of the most potent hydrazides, significantly slowed the rate at which intracerebrally administered glutamicU-C1*-acid was converted to GABA (Roberts et aE., 1958b). However, it is known that pyridoxal phosphate-requiring enzymes are involved in many different types of transformations, among which are reactions involved in the metabolism of the catechol amines and the formation of serotonin. Therefore, although an excellent correlation was found to exist between the inhibition of GAD by the hydrazides, the decrease in GABA levels, and the occurrence of seizures, the results could not constitute direct proof of a causal relationship. A more direct demonstration would be obtained if GABA itself could reverse the action of the convulsant hydrazides. Although this presented experimental difficulties because of the previously mentioned lack of ability of GABA to pass the blood-brain barrier, some experimental evidence has been accumulated (Killam, 1958, Killam et aZ., 1960) which showed that topically applied or intraventricularly administered GABA could effectively counteract the convulsive activity of thiosemicarbazide in cats and that electrically induced seizures in the rhinencephalon and the isocortex were shortened by application of GABA and increased by the injection of thiosemicarbazide. The case for a causal relationship between increases or decreases in seizure susceptibility in specific cerebral areas with corresponding increases or decreases in GABA levels could be strengthened if endogenously raised GABA levels would result in electrophysiologically observable effects which were opposite to those found with convulsant hydrazides or B6 antimetabolites.

ROLES OF Y-AMINOBUTYRIC ACID

311

A large number of drugs failed to elevate GABA levels (Roberts et d., 1958a). The injection of large amounts of GABA into normal mice, rats, and cats failed to elevate brain GABA (Roberts et aZ., 1958a, van Gelder and Elliott, 1958) or to produce changes in electroencephalographic recordings ( Purpura, et al., 1958). Hydroxylamine then was tested because in vitro experiments had shown it to be a more effective inhibitor of GABA-T than of GAD (Roberts and Frankel, 1951a; Roberts, 1952; Baxter and Roberts, 1958) and it was thought that if the preferential inhibition of GABA-T could be achieved in vivo an increase in GABA levels might occur in the CNS. Hydroxylamine administered alone or with methylene blue, to prevent effects of methemoglobinemia which might result, has been found to elevate GABA levels in the brains of rats, cats, and monkeys (Baxter and Roberts, 1960; Eidelberg et al., 1960; and unpublished observations). Two-dimensional paper chromatography of the extracts showed the elevation of GABA to be specific, the other ninhydrin-reactive constituents, with the exception of glutamic acid which showed a slight increase, showing no significant alterations from the controls (Baxter and Roberts, 1960). Typical results are shown in Table 111. Sodium nitrite, an agent that produces methemoglobinemia, did not produce significant elevations of GABA. Electrophysiological measurements showed that, concomitantly with the increase in GABA levels, hydroxylamine produced a remarkable reduction in the duration and spread of electrically induced after-discharges in cats in acute experiments and in conscious monkeys with indwelling electrodes (Eidelberg et al., 1960). The monkeys appeared more placid without being sedated and were much easier to handle for as long as 7 hours after the administration of hydroxylamine. Examples of the striking effects obtained with the monkeys are shown in Figs. 12 and 13. In Fig. 12 are shown typical results obtained with a macaque monkey with chronically implanted cortical electrodes ( Eidelberg et al., unpublished). Seizures were induced by electrical stimulation of the left motor cortex and the stimulation was repeated every 30 minutes with the same stimulus. Hydroxylamine (4.8mg/kg) was injected after the control series of measurements was completed. It is seen clearly that the seizure discharges were markedly reduced in all areas throughout the course of the experiment. A marked reduction was found in the amplitude of the early negative component

312, EUGENE ROBERTS AND EDWARD0 EIDELBERG

w-

I

%.

ROLES OF y-AMINOBUTYRIC ACID

'cr

I

*

.-e,

C .u

313

314

EUGENE ROBERTS AND EDUARDO EIDELBERG

of the direct cortical response in a monkey under local anesthesia and immobilized with gallamine ( Fig. 13). As stated above, hydroxylamine was tested because it is a potent inhibitor of GABA-T. It has not been yet established with certainty, however, that this is the mechanism by which elevation of GABA is achieved in vivo. Hydroxylamine can react with diphosphopyridine nucleotide (Kaplan and Ciotti, 1954) and thus might affect the activity of the succinic semialdehyde dehydrogen4.8 mg/kg NH,OH ( I . V . )

4 \.

0

a‘ 0

0

0

0

0

0

0

0

a

0

0

ae 0

I

2 3 4 5 Hours After Injection

0

20

FIG.13. Influence of injection of hydroxylamine on the direct cortical response in the monkey cortex.

ase. Indirect effects through action of hydroxylamine or its metabolites on cerebral circulation and oxidations cannot be completely excluded. Nonetheless, hydroxylamine produces highly significant increases in GABA levels in brain which are correlated with electrophysiological measurements that show a generally decreased sensitivity to electrical stimulation and with behavioral observations which indicate, at least in monkeys, a more placid response to environmental stimuli. These results are opposite in effect to those found when GABA is reduced and support the hypothesis that GABA must play a role in the regulation of neural excitability.

TABLE I11 EFFECTSOF HYDROXYLAMINE(NH~OH) AND METHYLENE BLUE (MB)

ON

BRAINLEVELSOF GABA

IN

HAT+

GABA Levels found (mg per 100 gm) Experiment Brain Structure

Control

Diencephalon Colliculi Olfactory lobes Cortex Cerebellum Pons Medulla

58 57 42 27 28 -

-

l b

MB( 10 mg) plus MB( 10 mg) NH,OH (4.8 mg) NH,OH (7.2 mg) 38 45 35 22 22 -

71 71 55 47 40

82 89 62 51 39

-

-

Experiment 20 MB( 10 mg) plus Control NH,OH (12 mg) 56 51 40 26 32 25 26

79 86 61 48 44 36 33

Baxter and Roberts, 1960 Each group consisted of 3 Wistar male rats 160 to 190 gm. When both MB and NH,OH were given, the MB (10 mg in 0.5 ml) was injected 20 minutes before the NH,OH (in 0.5 ml, pH 7 ) was administered intraperitoneally. The animals were sacrificed 70-90 minutes after the last injection. c Each group consisted of 2-3 male and female Sprague Dawley rats 200-25Ogm in weight. The animals were sacrificed between 60 and 75 minutes after the injection of NH,OH. a

b

8 0 hi

aE z

34

*

Q

316

EUGENE ROBERTS AND EDUARDO EIDELBERG

D. SOME OTHER CONSIDERATIONS RELEVANTTO VITAMINB6 THE GABA SYSTEM

AND

Observations in many laboratories have established a relationship of convulsive seizures to dietary pyridoxine deficiency in experimental animals and in human beings (see Tower, 1956 and 1958 for reviews of the literature). A variety of pyridoxine antagonists and carbonyl-trapping agents which can react with pyridoxal phosphate in vivo also have been shown to mimic the effects of a dietary vitamin Be deficiency (see preceding discussion). A number of lines of evidence are consistent with the hypothesis that the seizures are related to a decrease in activity of GAD with consequently decreased cerebral levels of GABA, a reduction of which in some brain areas results in failure to regulate or modify excitatory phenomena, allowing indiscriminate spread of electrical impulses with resultant seizures. The GABA levels are at the least a function of the relative rates of formation and utilization by two B6-requiring enzymes, GAD and GABA-T, and of the retention of GABA by the tissue in which it is formed. The measurement of GAD activity in dilute homogenates of brain without the addition of excess pyridoxal phosphate presents technical difficulties, because under these circumstances the activity is a nonlinear function of time (Roberts et al., 1951b) and can decay in an unpredictable manner in the presence of different inhibitors. It would, therefore, appear that the best single measurement that can be made is that of GABA levels when correlations of convulsant activity with effects on the GABA system are being studied. Failure to attain a correlation in some instances between the production of seizures and a decrease in GAD activity as measured in an unsupplemented homogenate cannot be used as evidence against the above hypothesis (Rosen et al., 1960). It must also be kept in mind that while it may well be that seizures in dietary vitamin B6 deficiency and after treatment with some hydrazides can be attributed to the failure to produce GABA at a sufficiently rapid rate to maintain a normal level of neuronal excitability, various Be antimetabolites or other carbonyltrapping agents may have many additional physiological effects which may range from alterations in cerebral blood flow and changes in permeability of the blood-brain barrier to direct or indirect inhibition of many enzymes which do not require pyridoxal phosphate as a coenzyme. Thus, the nature of the effects of

ROLES OF Y-AMINOBUTYRIC ACID

317

exogenously administered chemical agents on seizure thresholds may be extremely complex. Normally, dietary forms of vitamin B6 are absorbed and converted efficiently in tissues to pyridoxal phosphate or pyridoxamine phosphate, the coenzymatically active forms of the vitamin. Pyridoxal phosphate has been shown to be synthesized in brain from ATP and pyridoxal (Roberts and Frankel, 1951b). In addition to dietary deficiency and the effects of carbonyl-trapping agents and antimetabolites, other circumstances may arise by which the tissue levels of pyridoxal phosphate may be kept too low for the optimal functioning of important rate-limiting enzymes such as GAD ( Coursin, 1960). In some pediatric cases showing mental retardation and convulsions there appears to be an abnormally rapid conversion of dietary vitamin Be to pyridoxic acid, which is not utilized by tissues and is excreted in the urine. The forms of vitamin BG which can be converted to pyridoxal phosphate are kept low in blood and presumably in the tissues. Administration of extra vitamin B6 produced only a transitory improvement. The first instance of pyridoxine dependency was described in a child (Hunt et al., 1954) that showed a grossly abnormal electroencepholographic record and convulsions unless given large daily supplements of vitamin B6. During the seizures in this child there was found to be a decrease in cerebral oxygen consumption and a return to a normal cerebral metabolic rate when pyridoxine was administered (Sokoloff et al., 1959). The defect appears to lie in an abnormally high excretion of pyridoxal and pyridoxamine in the urine, possibly because of a failure to reabsorb these substances from the blood by the kidney (Coursin, 1960). Although not identified yet clinically, it is possible that genetic or induced defects in the handling of vitamin B6 also may occur in the absorption of the dietary forms of the vitamin or in the enzymes involved in the conversions of these forms to the coenzymatically active compounds. It also may be anticipated that there may occur genetically conditioned differences in the structure of the apoenzymes of B6 enzymes so that the same enzyme in various individuals (or strains of animals) may possess differing affinities for the coenzyme and could thus achieve saturation with the coenzyme at widely different tissue levels. The finding that a seizure in an infant with a simple dietary deficiency of vitamin B6 was abolished completely almost immediately after intramuscular injection of pyridoxine

318

EUGENE ROBERTS AND EDWARD0 EIDELBERG

(Coursin, 1954) indicates that in a normal individual there can be an extremely rapid rate of conversion of pyridoxine to pyridoxal phosphate and association of the coenzyme with the apoenzyme of a critical cerebral enzyme. Lack of Correlation of GABA Levels and Seizure Susceptibility in Areas of Normal Brain

X.

In the preceding sections it was shown that changes in levels of GABA in the brains of several species are accompanied by profound alterations in susceptibility to seizures. It was of interest to explore whether or not a direct relationship existed between normu2ty found levels of GABA and some measurable criteria of seizure susceptibility. The results in Table IV obtained in 6 cats TABLE IV GABA AND SEIZURE SUSCEPTIBILITY IN CATBRAINCORTEX AND HIPPOCAMPUS@ Brain area

Average

Range

16.8 17.3 16.7 17.4 23.4

12-23 15-21 12-21 11-21 16-30

~

Posterior marginal gyrus Sigmoid gyrus Posterior ectosylvian cortex Suprasylvian gyrus ( 3 areas) Hippocampus

susceptibilityb

+ ++ ++ +++ ++++

Baxter et al., 196Oa. The relative seizure susceptibility corresponded to the following voltage = 9 to 14 volts, ++ = 5 to 8 volts, and = 1 to 4 thresholds: volts. Only mechanical stimulation was required to exceed the seizure threshold of the hippocampus. The same stimulating electrodes and interelectrode distance were used for all cats. In any experiment the current per pulse was consistent with voltage values and thresholds were close to identical in symmetrical loci of both hemispheres, However, because of variability in impedance in the electrode-brain system the current output per pulse varied from one experiment to another. Electrical stimulation was applied locally through a pair of silver ball electrodes coated with silver chloride and resting on the cortical area being tested. Seizure thresholds were defined as the voltage required to elicit consistent and reproducible after-discharges in the local electrocorticogram after the application of a 5-second train of rectangular bidirectional pulses, each lasting 0.1 msec. Threshold voltages and currents were checked on a calibrated oscilloscope at the end of each determination. a

b

+

+++

ROLES OF Y-AMINOBUTYRIC ACID

319

show that virtually identical levels of GABA were found in 3 areas of cat brain cortex which showed a three-fold variation in seizure susceptibility. The highest level of GABA was found in the hippocampus, which has an exceptionally low seizure threshold, requiring only the mechanical stimulation of exploring electrodes to fire convulsive after-discharges. The above results appear to rule out the possibility that there is a direct and simple relationship between seizure susceptibility and the level of GABA normally found in a particular cerebral area. The total tone in the CNS, of which seizure susceptibility is one expression, is the result of a balance of excitatory and inhibitory influences. The GABA system is only one of potentially many inhibitory factors. However, it would appear from the previously discussed results that this system is sufficiently important in the maintenance of this balance so that increases or decreases, respectively, of GABA in a particular cerebral area are closely correlated with increases or decreases in seizure threshold of that area. XI.

Physiological Mechanism of Action of GABA a t the Cellular Level

How is GABA related to the inhibitory process which “in all nerve cells counteracts the depolarizing action of excitatory processes to maintain the polarization of a cell at an equilibrium level near that of its resting value” (Jasper, 1960)? Elegant work with precisely defined invertebrate and vertebrate test systems employing intracellular microelectrodes has indicated that the membrane stabilization process necessary for this inhibition is related to specific increases in the permeability of the postsynaptic membrane to potassium and/or chloride ions. Present opinion favors the idea that postsynaptic membranes generally are electrically unexcitable and that both excitatory and inhibitory effects may be transmitted by chemical mediators having effects on the specialized postsynaptic membrane. However, an example of electrical transmission in a synapse of crayfish cord has been reported recently (Furshpan and Potter, 1957). A. STRETCH RECEPTORS Perhaps the most thoroughly studied preparation has been the crustacean stretch receptor organs ( Alexandrowicz, 1951 and 1952;

320

EUGENE ROBERTS AND EDUARDO EIDELBERG

see Edwards, 1960 for further references on anatomy and histology). The easy accessibility of these organs for manipulation and for insertion of measuring electrodes has led to a good understanding of their function. The body of the sensory neuron lies near the receptor muscle strand and the peripheral portions of the dendrites are imbedded in the strand. An inhibitory axon forms synapses on the dendrites and there is also a motor nerve supply to the muscle strand. The sensory neuron can be activated by passive stretch or by contraction set by the motor nerves (Wiersma et al., 1953) and can be suppressed even in the presence of stretch by stimulation of the inhibitory nerve (Kuffler and Eyzaguirre, 1955). There are some analogies of the above system with the vertebrate muscle spindle (Kuffler, 1958). The sequence of events leading to the firing of the sensory neuron has been summarized as follows (Kuffler, 1958): “Stretch deformation sets up the generator potential in the peripheral portion of dendrites. By electrotonic spread other cell regions are excited and all-or-none conduction generally starts more centrally in the larger portion of dendrites, in the cell body or in the initial axon stretch. The generator potential persists for the duration of the stretch, its magnitude depending on the amount of stretch deformation. This in turn controls the frequency of sensory discharge. The actual site of origin of propagated impulses has been shown to fluctuate according to intensity of depolarization. While the distal dendrite portions are clearly specialized for graded activity, the proximal parts of the dendrites, the cell body, and the axon carry fully conducted impulses.” The impulses in the slowly adapting stretch receptor of the lobster are initiated in the axon about 500 p from the cell body (Edwards and Ottoson, 1958). An analysis of the mechanisms of inhibition of the above system employing intracellular electrodes has suggested that stimulation of the inhibitory nerve causes the liberation of a substance (or substances) which produce a specific effect on the postsynaptic membrane of the sensory cell. The inhibitory process in the stretch receptor is believed to be attributable to increased conductance of potassium and possibly chloride ions and membrane potential changes. Actually, at the equilibrium level of the inhibitory process no significant membrane potential change is set up by inhibitory impulses, although the inhibitory process is fully active (Kuffler, 1958), The association of GABA with the crayfish stretch receptor

ROLES OF ?’-AMINOBUTYRIC ACID

321

occurred as a result of an extensive series of studies on the influence of factors present in extracts of beef brain which had inhibitory effects on the stretch receptor as well as in other physiological test systems (see Florey, 1957, 1960, and Florey and Florey, 1958 for pertinent references). Eventually, GABA was shown to be one of the major factors exerting potent inhibitory action on the stretch receptor (Bazemore et al., 1957) and this opened the way for many investigations of the actions of GABA in a variety of invertebrate preparations. It was found that low concentrations of GABA had effects on the crayfish stretch receptor which were remarkably similar to those found on stimulation of the inhibitory nerve (Kuffler and Edwards, 1958). GABA concentrations 1000 times stronger than those blocking afferent discharge did not block conduction in the sensory axons. GABA action, therefore, appears to be confined to the dendrites and cell body. The action of GABA is not localized necessarily on subsynaptic receptor sites, since it also was found to block the stretch discharge of the median receptor in the seventh thoracic segment, a structure possessing no demonstrable inhibitory synapses. GABA showed the most potent blocking action on the stretch receptor of a wide variety of related compounds tested (Edwards and Kuffler, 1959), but guanidinoacetic and p-guanidinopropionic acids were almost as effective and the activity of y-guanidinobutyric acid was only slightly less. Guanidino compounds, as well as other unidentified active substances, are undoubtedly present in the crude active brain extracts ( McLennan, 1959,19f30). (3-Hydroxy-GABAwas approximately onehalf as active as GABA in the stretch receptor system (Bazemore et al., 1957). All of the active compounds appear to increase the flow of specific ions in the dendrites and cell body and to resemble the natural neural inhibitory transmitter in this respect (Edwards and Kuffler, 1959). However, the evidence at the present time does not warrant assigning to GABA or any of the other compounds the role of the natural inhibitory transmitter substance. B. HEARTGANGLIA In crayfish and lobster heart ganglia (Florey, 1957 and Maynard, 1958) GABA has shown actions resembling those caused by the stimulation of the cardio-inhibitory nerves. In experiments with the neuromuscular junction of the opener muscle of the claw

322

EUGENE ROBERTS AND EDUARDO EIDELBERG

of the crayfish application of GABA and the stimulation of the inhibitory nerve fiber appeared to act similarly in increasing conductance, which depended on an increased permeability to chloride ions (Boistel and Fatt, 1958; see also Van der Kloot and Robbins, 1959). Picrotoxin has been found to block reversibly both the effects of stimulation of inhibitory fibers and the action of GABA on the crayfish stretch receptor (Elliott and Florey, 1956), cardiac ganglion (Florey, 1957), and claw opener muscle (Robbins and Van der Kloot, 1958). In this connection it is of interest that smaller amounts of picrotoxin were required to produce seizures in monkeys wtih experimentally induced epilepsy than in controls (Kopeloff et al., 1957). Although there is now much fundamental information about the nature of the inhibitory process obtained by the use of exquisitely sensitive electrophysiological techniques, the same type of approach will not be able to lead to the ultimate answers in molecular terms of the events which are involved in inhibition. An adequate analysis of the presynaptic and postsynaptic events will require, at the least, knowledge of the chemical nature of the transmitter, of the mechanism of its formation, storage, release, and destruction, of the structure of the postsynaptic membrane, its relation to maintenance of specific ion permeabilities, and the changes produced in it both by excitatory and inhibitory transmitters. The above experiments represent, therefore, only a first step in a long, hard road.

XII.

Pharmacological Studies with GABA

A. SPINAL CORD Results of experiments have been published recently in which intracellular recordings were made from motoneurons, interneurons, and Renshaw cells of the cat spinal cord while various substances were injected ionophoretically inside or outside of the neuronal membrane by means of double-barrel, five-barrel, or multiplecoaxial micropipettes (Curtis et al., 1959a, 1959b and Curtis and Watkins, 1960). GABA, (3-alanine, and taurine were found to be almost equally active in depressing activity of spinal neurons. These substances administered in this manner caused a reduction

323

ROLES OF Y-AMINOBUTYRIC ACID

in amplitude of both excitatory and inhibitory postsynaptic poten-

tials without a change in membrane potential but with an increase in membrane conductance arising from an increase in permeability to chloride. These results are similar to those found for the crayfish stretch receptor (Kuffler and Edwards, 1958) and for the crustacean neuromuscular junction ( Boistel and Fatt, 1958) and point to a common probable mode of action of exogenously administered GABA in both vertebrate and invertebrate systems. The view has been expressed that GABA is more likely a membrane “stabilizer” rather than an inhibitory transmitter ( Curtis and Watkins, 1W).The above point of view is supported by ingenious experiments in which intracellular recordings were made from spinal hamstring motoneurons in the cat while seizurelike discharge was induced by the injection of thiosemicarbazide, which can cause depletion of intracellular GABA levels (Terzuolo et al., 1960). No change from the control was observed in the amplitude and time course of the inhibitory postsynaptic potential in these motoneiirons evoked by stimulating the quadriceps nerve as long as the cells were reacting in a normal fashion to antidromic excitation. These results are incompatible with a transmitter function for GABA, but would be in keeping with the role of a modulator of neuronal excitability which is produced and which exerts its effects intracellularly. It also has been observed that the injection of thiosemicarbazide resulted in enhancement of mono- and polysynaptic reflexes and in the disappearance of the bulbar reticular inhibition of these reflexes (Eidelberg and Buchwald, 1960) I

B. BRAIN Since the first finding that topically applied solutions of GABA and p-hydroxy-GABA exerted inhibitory effects on electrical activity in the brain (Hayashi and Nagai, 1956 and Hayashi and Shuhara, 1956) many experiments have been performed utilizing this technique in an attempt to elucidate the mechanism of action of GABA and to gain new information about the organization of the brain. However, although the observations have been reproducible in various laboratories, the interpretations of the data have differed. Numerous difficulties are involved in the interpretation of evokedpotential data from the highly complex structures in the brain, even if such artifacts as changes in brain volume and direct electro-

324

EUGENE ROBERTS AND EDUARW EIDELBERG

chemical effects of the test substances on the recording electrodes can be ruled out. The possibility also should be kept in mind that GABA normally might be a substance which is entirely intracellular and that application of the substance to the outer surface of neurons may give results which differ from those exerted by the endogenously formed and intracellularly contained GABA. Direct application of GABA solutions on the cortex of cats induced marked changes in the nature of the evoked responses obtained at the point of application, the early negative component of these responses becoming positive (Purpura et al., 1957a, 1957b, and Iwama and Jasper, 1957). When GABA was applied to the cortical surface or given by deep injection the evoked responses recorded in the deeper layers were unchanged. The polarity of spindle bursts also was reversed by GABA to a surface positive form (Iwama and Jasper, 1957). Direct responses in the cerebellar surface were reduced in voltage but showed no change in polarity (Purpura, et al., 1957a and 1957b). These findings were interpreted to mean that GABA selectively blocks the depolarizing, excitatory synapses, which represent the surface-negative postsynaptic potentials in the cerebral cortex, and that the positive potentials observed are hyperpolarizing postsynaptic potentials, ordinarily masked by depolarization (Purpura et al., 1957a and Grundfest, 1958). The failure to observe positivity upon application of GABA to the cerebellar cortex was attributed to the relative lack of inhibitory synapses in the cerebellum, while the inability to find an effect of topically applied GABA at deeper levels in the cerebral cortex was taken to indicate that the action was specific for axodendritic synapses. An alternative interpretation of the reversal from negativity to positivity in the cortical-evoked potentials has been made in which it was suggested that subsequent to the blocking of the surfacenegative waves the deep negative waves, inverted on the surface according to volume conductor theory, were reflected on the surface as positive waves (Elliott and Jasper, 1959). It is very difficult to accept fully any interpretation based on data obtained from such highly complex structures as the cerebral and cerebellar cortices employing gross recording electrodes. Records of the d.c. changes in the cortex induced by GABA and other substances have been obtained (Goldring et al., 1958; Vanasupa et al., 1959; Adey et al., 1959). Again, owing to our lack of

ROLES OF Y-AMINOBUTYRIC ACID

325

understanding of the basic mechanisms underlying these potential changes, very little can be concluded that will clarify the issue of the role of GABA in normal function of the CNS. Intracellular recordings from neuronal structures in the mammalian cortex only recently have been obtained, and no information is as yet available on the effects of GABA and other substances at this level. If these are found to be similar to those obtained in the cat spinal cord (Curtis and Watkins, 1960 and Terzuolo et a]., 1960) it will be reasonable to conclude that GABA does not act on synaptic structures in the cortex either as an inhibitory transmitter or an excitatory blocking agent. In view of the “complexity-within-unity” of the neuron (Bullock, 1959) as well as of the immense variety of interrelations between neurons in the CNS it is best to refrain at present from attempting to construct a unitarian hypothesis for the role of GABA in the CNS.

C . BLOODPRESSURE Parenteral administration of GABA was found to produce a fall in blood pressure in experimental animals (Takahashi et al., 1955, 1958; and Romanowski, 1958) and in man (Elliott and Hobbiger, 1959). In rabbits GABA was the most potent of the “-amino acids tested (Takahashi et al., 1955). The intracisternal injection of GABA caused much more marked and sustained fall of pressure than the intravenous injection. Transection of the brain stem and partial elimination of the medulla revealed that GABA was specifically affecting depressor points in the medial reticular formation and the area postrema (Takahashi et al., 1958). An interesting finding in the cat was that physostigmine greatly enhanced the blood pressure lowering effect of GABA, while atropine completely suppressed the hypotensive effects of GABA (Romanowski, 1958). On the basis of the latter findings together with other pertinent observations it was suggested that GABA may be acting either by causing cholinergic neurons to release acetylcholine or by sensitizing acetylcholine receptors. Not finding the atropine effect in rabbits or dogs nor observing effects on blood pressure with intracisternal injection others have suggested that GABA might exert its effects on blood pressure by affecting vasoconstrictor tone of blood vessels (Elliott and Hobbiger, 1959). It is particularly inter-

326

EUGENE ROBERTS AND EDUARDO EIDELBERG

esting that GABA given in 1-gm doses three times daily has been reported to lower blood pressure and to alleviate symptoms in 51 of 67 randomly selected hypertensive patients (Shibata et al., 1960 ) .

Xlll.

General Comments and Summary

A review of the biochemistry of GABA published in 1956 (Roberts, 1956) was concluded with the following statement: “Perhaps the most difficult question to answer would be whether the presence in the gray matter of the central nervous system of uniquely high concentrations of GABA and the enzyme which forms it from glutamic acid has a direct or indirect connection to conduction of the nerve impulse in this tissue.” Since that time, work in many laboratories, of which some aspects have been discussed in this review, has shown that GABA has inhibitory effects in both invertebrate and vertebrate systems and has indicated that GABA may play an important role in regulation of neuronal activity. Although experimental difficulties existed in studying GABA because of the failure of this substance to penetrate the blood-brain barrier readily, the finding that convulsant hydrazides could lower GABA in brain (Killam, 1954) and that hydroxylamine could increase the content (Baxter and Roberts, 1960) has given us potent tools with which to manipulate the intracellular levels of GABA and to study the electrophysiological correlates of the induced changes. Numerous difficulties are interposed between the finding of correlations and the establishment of causal relationships between chemically and electrically observed events, not the least of which is the failure of the neurobiochemist and neurophysiologist to understand anything but the most superficial aspects of the companion science. It is a rare neurophysiologist who has a good understanding of the tricarboxylic acid cycle and an even rarer chemist who has an appreciation of the details of neuronal structure or the meaning of an oscilloscope recording. Aside from the great theoretical importance of the understanding of the role of an inhibitory substance which seems to be formed and located almost entirely in the gray matter of the CNS, some practical aspects of possibilities in the treatment of human disease are apparent already. Decreases in brain levels of GABA are asso-

ROLES OF Y-AMINOBUTYRIC ACID

327

ciated with the occurrence of paroxysmal discharges, and it has been inferred that, perhaps, some types of convulsive disorders in humans might be related to malfunctioning of the GABA system. Whether or not elevation of GABA in brain will be of value in alleviating any disorders remains to be determined by future clinical research. REFERENCES

Adey, W. R., Dunlop, C. W., Killam, K. F., and Brazier, M. A. B. (1960). “Inhibition of the Nervous System and y-Aminobutyric Acid,” pp. 317-323. Pergamon Press, Oxford. Albers, R. W. (1960). “Inhibition of the Nervous System and y-Aminobutyric Acid,” pp. 196-201. Pergainon Press, Oxford. Albers, R. W., and Brady, R. 0. (1959). J. Biol. Chem. 234, 926. Albers, R. W.,and Salvador, R. A. (1958). Science 128,359. Alexandrowicz, J. S. (1951). Quart. J. Microscop. Sci. 92, 163. Alexandrowicz, J. S. (1952). Quart. J. Microscop. Sci. 93,315. Ashby, W.,Garzoli, R. F., and Schuster, €3. M. (1952). Am. J. Physiol. 170, 116. Bain, J. A., and Williams, H. L. (1960). “Inhibition of the Nervous System and y-Aminobutyric Acid,” pp. 275-293. Pergamon Press, Oxford. Baxter, C. F., and Roberts, E. (1958). J. B i d . Chem. 233, 1135. Baxter, C. F., and Roberts, E. ( 1960). “Neurochemistry of Nucleotides and Amino Acids,” pp. 127-145. Wiley, New York; (1959). Proc. SOC. Exptl. Biol. Med. 101, 811. Baxter, C. F., Roberts, E., and Eidelberg, E. (1960a). J. Neurochem. In Press. Baxter, C. F., Schade, J. P., and Roberts, E. (1960b). “Inhibition of the Nervous System and y-Aminobutyric Acid,” pp. 214-218. Pergamon Press, Oxford. Bazemore, A. W., Elliott, K. A. C., and Florey, E. (1957). J. Neurechem. 1, 334. Beloff-Chain, A., Catanzaro, R., Chain, E. B., Masi, I., and Pocchiari, F., (1955). Proc. Roy. SOC. 144, 22. Berl, S., and Waelsch, H. (1958). J. Ncurochem. 3, 161. Bessman, S. P., Rossen, J., and Layne, E. C. (1953). J. Biol. Chem. 201, 385. Boistel, J., and Fatt, P. (1958). J. Physiol. (London) 144, 176. Bollman, J. L., Flock, E. V., Grindlay, J. H., Bickford, R. C., and Lichtenheld, F. R. ( 1957). A.M.A.Arch. Surg. 76, 405. Brockman, J. A., Jr., and Burson, S. L. (1957). Proc. SOC. Exptl. BioZ. Med. 94, 450. Bullock, T. H. ( 1959). Science 129, 997. Caldwell, P.C. ( 1958). J. Physiol. (London) 142,22. Clark, C. T., Weissbach, H., and Udenfriend, S . (1954). J. BWZ. Chem. 210, 139.

328

EUGENE ROBERTS AND EDUARDO EIDELBERG

Clouet, D. H., Caitonde, M. K., and Richter, D. (1957). J. Neurochem. 1, 228. Consden, R., Gordon, A. H., and Martin, A. J. P. (1944). Biochem. J. 38,224. Coursin, D. B. (1954). J. Am. Med. Assoc. 164,406. Coursin, D. B. (1960). “Inhibition of the Nervous System and y-Aminobutyric Acid,” pp. 294-301. Pergamon Press, Oxford. Cravioto, R. O., Massieu, G., and Izquierdo, J. J. (1951). PTOC.SOC. Exptl. Biol. Med. 78, 856. Curtis, D. R., and Watkins, J. C. (1960). “Inhibition of the Nervous System and y-Aminobutyric Acid,” pp. 424-444. Pergamon Press, Oxford. Curtis, D. R., Phillis, J. W., and Watkins, J. C. (1959a). Nature 183, 611. Curtis, D. R.,Phillis, J. W., and Watkins, J. C. (195913). J . Physiol. (London) 146, 185. Dawson, R. M. C. (1953). Biochim. et Biophys. Acta 11, 548. Dent, C. E. (1947). Biochem. J.41,240. Dent, C. E. (1948). Biochem. J. 43, 169. Dunlop, C. W. ( 1957a). Am. J . Physiol. 190(1 ), 172. Dunlop, C. W. (195713). Am. J. Physiol. 191(I), 200. du Ruisseau, J. P., Greenstein, J. P., Winitz, M., and Birnbaum, S. hl. (1957). Arch. Biochem. Biophys. 68, 161. du Vigneaud, V., Kuchinskas, E. J., and Horvath, A. (1957). Arch. Biochem. Biophys. 69, 130. Edwards, C . (1960). “Inhibition of the Nervous System and y-Aminobutyric Acid,’’ pp. 386-408. Pergamon Press, Oxford. Edwards, C., and Kuffler, S. W. (1959). J. Neurochem. 4,19. Edwards, C., and Ottoson, D. ( 1958). J. Physiol. (London) 143, 138. Eidelberg, E.,and Buchwald, N. A. (1960). Neurology 10, 267. Eidelberg, E., Baxter, C. F., Roberts, E., Saldias, C. A., and French, J. D. (1960). “Inhibition of the Nervous System and y-Aminobutyric Acid,” pp. 365-370. Pergamon Press, Oxford; (1959). Proc. Soc. Exptl. B i d . Med. 101, 815. Eiseman, B., Osofsky, H., Roberts, E., and Jelinek, B. (1959). J. Appl. Physwl. 14, 251. Elliott, K. A. C. (1959). “Biochemistry of the Central Nervous System,” 4th International Congress of Biochemistry, Vienna, p. 251. Pergamon Press, New York. Elliott, K. A. C., and Florey, E. (1956). J. Neurochem. 1, 181. Elliott, K.A. C., and Hobbiger, F. (1959). J. Physiol. (London) 146,70. Elliott, K. A. C., and Jasper, H. H. (1959). Physiol. Reos. 39, 383. Elliott, K.A. C., and van Gelder, N. M. (1958). J. Neurochem. S, 28. Fischer, E. H., and Krebs, E. G. (1958). Abstr. Pacific Slope Biochem. Conf. Univ. of Calif., Los Angeles, California. Fischer, E. H., Kent, A. B., Snyder, E. R., Krebs, E. G . (1958). I. Am. Chem. Soc. 80, 2906. Flock, E. V., Block, M. A., Grindlay, J. H., Mann, F. C., and Bollman, J. L. ( 1953). J . Bwl. Chem. 200, 529.

ROLES OF Y-AMINOBUTYRIC ACID

329

Florey, E. (1957). Naturwissenschaften 44,424. Florey, E. ( 1960). “Inhibition of the Nervous System and y-Aininobutyric Acid,” pp. 72-84. Pergamon Press, Oxford. Florey, E., and Florey, E. ( 1958). J. Physiol. (London) 144, 220. Florey, E., and McLennan, H. ( 1955). 1. Physiol. (London) 129, 384. Florey, E., and McLennan, H. (1959). J. Physiol. (London) 146, 66. Furshpan, E. J., and Potter, D. D. (1957). Nature 180, 342. Gale, E. F. (1946). Aduances in Enzymol. 6, 1. Goldring, S., O’Leary, J. L., and Huang, S. H. (1958). Electroencephalog. and Clin. Neurophysiol. 10, 663. Gray, W.D., Maren, T. H., Sisson, G. M., Smith, F. H. (1957). J. Pharmacol. Exptl. Therap. 122, 160. Grundfest, H. (1958). Federation ROC. 17, 1006. Halpern, B. N., and Binaghi, R. (1959). Nature 183, 1397. Hartman, S. C., and Buchanan, J. M. (1958). J. Biol. Chem. 293,451. Hayashi, T., and Nagai, K. (1956). Abstr. XX Intern. Physwl. Congr., Brussels, p. 410. Hayashi, T., and Shuhara, R. ( 1956). Abstr. XX Intern. Physiol. Congr., Brtissels, p. 410-411. Holtz, P., and Westermann, E. (1957). Arch. exptl. Pathol. u. Pharmakol. 231, 311. Hunt, A. D., Jr., Stokes, J., Jr., McCrory, W. W., and Stroud, H. H. (1954). Pediatrics 13, 140. Irrevere, F., and Evans, R. L. (1959). J. Bwl. Chem. 234, 1438. Iwama, K., and Jasper, H. H. (1957). J. Physiol. (London) 138, 365. Jakoby, W. B., and Scott, E. M. (1959). J. B i d . Chem. 234, 937. Jasper, H. H. (1960). “Inhibition of the Nervous System and y-Aminobutyric Acid,” pp. 12-28. Pergamon Press, Oxford. Jenney, E. H., Smith, R. P., and Pfeiffer, C. C. (1953). Federatbn Proc. 12, 333. Jinnai, D., and Mori, A. ( 1958). Nisshin Igaku 45,1. Kaplan, N. O., and Ciotti, M. M. (1954). J. Biol. Chem. 211, 431. Killam, K. F., 1954, “Enzymatic Bases of the Action of Convulsant Hydrazides,” Ph.D. Thesis Univ. of Illinois, College of Medicine, Chicago, Illinois. Killam, K. F. (1957). J . Pharmacol. Exptl. Therap. 119, 263. Killam, K. F. (1958). Federation Proc. 17, 1018. Killam, K. F., and Bain, J. A. (1957). J. P h a m c o l . Exptl. Therap., 119, 255. Killam, K. F., Dasgupta, S. R., and Killam, E. K. (1960). “Inhibition of the Nervous System and y-Aminobutyric Acid,” pp. 302-316. Pergamon Press, Oxford. Kojima, K., Mizuno, K., and Miyazaki, M. (1958). Nature 181, 1200. Kopeloff, L. M., Chusid, J. G., and Kopeloff, N. (1957). J. Appl. Physiol. 11, 465. Kuchinskas, E. J., Horvath, A,, and du Vigneaud, V. (1957). Arch. Biochem. Biophys. 68, 69. Kuffler, S . W. ( 1958). Exptl. Cell Research Suppl. 6, 493.

330

EUGENE ROBERTS AND EDUARDO EIDELBERG

Kuffler, S. W., and Edwards, C. (1958). J. Neurophysiol. 21, 589. and Eyzaguirre, C. ( 1955). J. Gen. Physiol. 39, 155. Kuffler, S. W., Lowe, I. P., Robins, E., and Eyerman, G. S. (1958). J. Neurochem. 3, 8. McKhann, G. M., and Tower, D. B. (1959). Am. J. Physiol. 196, 36. McKhann, G. M., Albers, R. W., Sokoloff, L., Mickelsen, O., and Tower, D. B. ( 1960). “Inhibition of the Nervous System and y-Aminobutyric Acid,” pp. 169-181. Pergamon Press, Oxford. McLennan, H. ( 1957). J . Physiol. (London) 139,79. McLennan, H. ( 1960). “Inhibition of the Nervous System and y-Aminobutyric Acid,” pp. 413-415. Pergamon Press, Oxford. McLennan, H. (1959). J. Physiol. (London) 146, 358. Maynard, D.M. (1958). Federation Proc. 17, 106. Millichap, J. G. (1957). Proc. SOC. Exptl. Biol. Med. 96, 125. Millichap, J. G., Woodbury, D. M., and Goodman, L. S. (1955). J. Pharmacol. Exptl. Therap. 116, 251. Nichols, G., Jr. (1957). Science 126, 1244. Noe, F. F., and Nickerson, W. J. (1958). J. Bacteriol. 76, 674. Okumura, N.,Otsuki, S., and Aoyama, T. (1959). J. Bwchem. (Tokyo) 46, 207. Peters, J., Vonderahe, A., and Powers, T. (1956). J. Exptl. Zool. 133, 505. Pisano, J., and Udenfriend, S. ( 1958). Federation Proc. 17, 403. Pisano, J., Mitoma, C., and Udenfriend, S. (1957). Nature 180, 1125. Pisano, J., Wilson, J. D., and Udenfriend, S. (1960). “Inhibition in the Nervous System and y-Aminobutyric Acid,” pp. 226-235. Pergamon Press, Oxford. Purpura, D. P., Girado, M., Grundfest, H. (1957a). Science 126, 1201. Purpura, D. P., Girado, M., and Grundfest, H. (1957b). Proc. SOC. Erptl. Biol. Med. 96, 791. Purpura, D. P., Girado, M., Smith, T. G., and Gomez, J. A. ( 1958). Proc. SOC. Exptl. Biol. Med. 97, 348. Purpura, D. P., Girado, M., Smith, T. G., Callan, D. A., and Grundfest, H. (1959). J. Neurochem. 3, 238. Robbins, J., and Van der Kloot, W. G. (1958). J. Physiol. (London) 143, 541. Roberts, E. (1952). J. Biol. Chem. 198,495. Roberts, E. (1956). Progr. in Neurobiol. 1, 11. Roberts, E. (19%). “Inhibition of the Nervous System and y-Aminobutyric Acid,” pp. 144-158. Pergamon Press, Oxford. Roberts, E. (1959). 3rd Intern. Neurochem. Symposium. In Press. Roberts, E., and Baxter, C. F. (1959a). “Biochemistry of the Central Nervous System,” 4th Intern. Congr. of Biochemistry, Vienna, p. 268. Pergamon Press, New York. Roberts, E., and Bregoff, H. M. (1953). J . Biol. Chem. 201, 393. Roberts, E.,and Frankel, S. (1950). J. Biol. Chem. 187,55. Roberts, E.,and Frankel, S. (1951a). J. Biol. Chem. 188, 789. Roberts, E.,and Frankel, S. (1951b). J. Biol. Chem. 190, 505. Roberts, E.,Frankel, S., Harman, P. J. (1950). Proc. SOC. Exptl. Biol. Med. 74, 383.

ROLES OF Y-AhCINOBUTYRIC ACID

331

Roberts, E., Harman, P. J., and Frankel, S. (1951a). Proc. SOC. Exptl. Biol. Med. 78, 799. Roberts, E., Younger, F., and Frankel, S. (1951b). 3. Biol. Chem. 191,277. Roberts, E., Lowe, I. P., Chanin, M., and Jelinek, B. (1957). J. Exptl. Zool. 136, 239. Roberts, E., Lowe, I. P., Guth, L., and Jelinek, B. (1958a). J. Exptl. Zool. 138, 313. Roberts, E., Rothstein, M., and Baxter, C. F. (1958b). Proc. Soc. Exptl. B i d . Med. 97, 796. Romanowski, W. (1959). Bull. acad. polon. sci. 7 , 83. Rosen, F., Milholland, R. J., and Nichol, C. A. (1960). “Inhibition of the Nervous System and y-Aminobutyric Acid,” pp. 338-343. Pergamon Press, Oxford. Roughton, F. J. W., and Clark, A. M. (1951). In “The Enzymes” (J. B. Sumner and K. Myrback, eds.), VoI. 1, Part 2, p. 1250. Academic Press, New York. Sacktor, B., Packer, L., Cummins, J., and Hackley, B. (1960). “Inhibition of the Nervous System and y-Aminobutyric Acid,” pp. 182-188. Perganion Press, Oxford. (1959). Federation PTOC. 18, 314. Saldias, C. A., Cabieses, F., and Eidelberg, E. ( 1957). Electroencephalog. and Clin. Neurophysiol. 9, 333. Salvador, R. A., and Albers, R. W. (1959). J. B i d . Chem. 234, 922. Schott, H. F., and Clark, W. G. (1952). J. B i d . Chem. 196, 449. Scott, E. M., and Jakoby, W. B. (1959). J. B i d . Chem. 234, 932. Shibata, N., Shimizu, S., Kubo, M., Takahashi, H., Yamaguchi, Y., Ezoe, T., and Tsukada, Y. (1960). “Inhibition of the Nervous System and y-Aminobutyric Acid,” pp. 579-581. Pergamon Press, Oxford. Sisken, B., Roberts, E., and Baxter, C. F. (1960). “Inhibition of the Nervous System and y-Aminobutyric Acid,” pp. 219-225. Pergamon Press, Oxford. Snell, E. E. ( 1958). Vitamins and Hormones 16, 77. Sokoloff, L., Lassen, N. A., McKhann, G. M., Tower, D. B., and Albers, R. W. (1959). Nature 183, 752. Swanson, A. G., Stavney, L. S.,and Plum, F. (1958). Neurology 8, 787. Takahashi, H., Tiba, M., Iino, M., and Takayasu, T. (1955). Japan. J. Physiol. 6, 334. Takahashi, H., Tiba, M., Yamazaki, T,, and Noguchi, F. (1958). Japan. J. Physwl. 8, 378. Tallan, H. H., Moore, S., and Stein, W. H. (1954). J. Biol. Chem. 211, 927. Terzuolo, C. A., Sigg, E. B., and Killam, K. (1960). “Inhibition of the Nervous System and y-Aminobutyric Acid,” pp. 336-337. Pergamon Press, Oxford. Tower, D. B. (1956). Am. J. Clin. Nutrition 4, 329. Tower, D. B. (1958). Nutrition Rev. 16, 161. Tower, D. B. (1959). “Biochemistry of the Central Nervous System,” 4th International Congress of Biochemistry, Vienna, p. 213. Pergamon Press, New York.

332

EUGENE ROBERTS AND EDUARDO EIDELBERG

Tsukada, Y., Nagata, Y.,and Takagaki, G. (1957). Proc. Japan Acad. 33, 510. Tsukada, Y., Nagata, Y., Hirano, S., and Takagaki, G. (1958). J. Biochem. (Tokyo) 45, 979. Umbreit, W. W. ( 1957). “Manometric Techniques,” p. 18. Burgess Publ. Co., Minneapolis, Minnesota. Umbreit, W. W., and Waddell, J. G. (1949). Proc. SOC. Exptl. Biol. Med. 70, 293. Vanasupa, P., Goldring, S., and O’Leary, J. L. (1959). Electroencephalog. Neurophysiol. 11, 93. Van der Kloot, W. G., and Robbins, J. ( 1959). Naturwissenschaften 16, 35. Van Gelder, N. M., and Elliott, K. A. C. ( 1958). J . Neurochem. 3, 139. Van Harreveld, A. ( 1959). J. Neurochem. 3, 300. Vates, T., Agranoff, B., and Sokoloff, L. (1959). Federation Proc. 18, 164. Wallace, W. M., and Hastings, A. B. (1942). J . Biol. Chem. 144, 637. Wallace, W. M., and Lowry, 0. H. (1942). J. Biol. Chem. 144,651. Weissbach, H., Bogdanski, D. F., Redfield, B. G., and Udenfriend, S. (1957). 1. Biol. Chem. 227, 617. Wiersma, C. A. G., Furshpan, E. J., and Florey, E. (1953). 1. Exptl. B i d . 30, 136. Williams, H. L., and Abdulian, D. H. (1956). J. Pharmacol. Exptl. Therap. 116, 62. Wilson, W. E., Hill, R. J., and Koeppe, R. E. (1959). J. Biol. Chem. 234, 347. Wyngaarden, J. B., and Ashton, D. M. (1959). J. Biol. Chem. 234, 1492.

OBJECTIVE PSYCHOLOGICAL TESTS AND THE ASSESSMENT OF DRUG EFFECTS' By H. J. Eysenck Institute of Psychiatry, University of London, England

I. Introduction

........................................... .......................

11. A Critique of Current Methodology

111. A Theory of Depressant and Stimulant Drug Action . . . . . . . . . . IV. Experimental Tests in the Assessment of Drug Effects . . . . . . . . V. A Second Paradigm for the Testing of Psychopharmacological Hypotheses ............................................ VI. A Mathematical Model of a Psychopharmacological Hypothesis VII. The Place of Animal Work in a Hypothetico-deductive System VIII. A General Dimensional System of Psychopharmacology . . , . , . . IX . Summary .............................................. References .............................................

I.

333 334 342 348 366 370 375 379 382 383

Introduction

In recent years a good deal of interest has been shown in what has come to be known as psychopharmacology, a term which may be taken to denote the study of the behavioral effects of drugs. While research has been going on in this field ever since Kraepelin tried to put it on a scientific basis, very little of lasting scientific importance has emerged. It is suggested here that this failure to achieve any noteworthy generalizations is due to a lack of theoretical and methodological sophistication, and that what is required in order to improve the situation is a more rigorous theoretical and methodological approach to a subject which, while superficially very simple, is nevertheless bristling with difficulties and complexities not always realized by investigators. The present chapter is devoted to a summary of the theoretical and experimental con1 The work of Micko reported in this paper (and some of the theoretical work also) was supported by the National Institute of Mental Health under grant M-46-59/7590363-93001-01.

333

334

H. J. EYSENCK

siderations and results relating to an attempt to introduce some degree of order into one corner of this chaotic field; no attempt has been made to cover the whole literature on psychopharmacology. Such a summary has been prepared elsewhere (Trouton and Eysenck, 1960), and the interested reader must be referred to this for coverage of the whole field. Interest in this chapter has been rather in certain methodological problems, and experimental studies are quoted more as illustration than for the sake of completeness. I have used one particular hypothesis, and the experimental work relevant to it, to demonstrate how psychopharmacology is related to psychology, how it can derive from it theories and methods for testing such theories, and how without such symbiosis both general psychology and psychopharmacology would be the poorer. Most of the points raised may appear self-evident, but these have seldom been made in print, and in practice they have usually been completely disregarded by experimentalists not brought up in the discipline of psychology. It would appear, therefore, that the contents of this chapter may not be entirely inopportune.

II. A Critique

of Current Methodology

Let us consider the usual practice of leading experimenters in this field. We shall take it for granted that in their work such people will not make any of the obvious and elementary mistakes which rule “out-of-court” immediately the majority of papers published (Trouton and Eysenck, 1960). Such errors arise from the possibility of contamination of results through knowledge of the drug administered on the part of either subject or experimenter, i.e., the failure to use the “double-blind method of administration; the use of unstandardized ratings and other subjective methods of assessment; failure to control and correct dosages and time elements according to the best available knowledge of the drug in question; the use of tests for the measurement of effects which have no theoretical rationale and no evidence of their validity for the particular trait or ability under investigation [the Rorschach test (Eysenck, 1958) may be quoted as an example of this]. Assuming that all the necessary precautions have been observed, we find that in the typical experiment one of two paradigms has been observed,

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

335

the choice depending largely on whether or not strong learning effects are present in the test under investigation. According to the first paradigm, which is suitable for tests not showing strong learning effects, half the sample is tested under placebo conditions, the other half under drug conditions; on the second day the first group is now given the drug, the second group is now given a placebo (P). Changes between placebo and drug conditions are evaluated in terms of appropriate statistical techniques. This method and its various modifications will be referred to as method I. Method 11, which is appropriate for tests showing strong practice effects, also employs a control and an experimental group, suitably matched on relevant variables; the control group performs the test under placebo conditions, the experimental group under drug conditions. Differences between the groups are evaluated again in terms of the appropriate statistical test which, however, would usually be much less powerful because of the inevitable failure to achieve proper matching between the groups. Thus method I1 requires considerably greater numbers of subjects than does method I. Both method I and method I1 are adequate for the purpose for which they are used, but great difficulties arise in interpreting the results which emerge. Let us assume that a given drug, when compared with (P), has produced significant changes in performance on a group of subjects ( S ) on test ( T ) . (Cf. Fig. 1 as an illustration.) How can we interpret such a finding? At first sight there appears to be no difficulty. Supposing the test used was the pursuit rotor, i.e., a test in which the subject attempts to keep a metal stylus in touch with a small metal disc set in a Bakelite turntable rotating at a speed of 60rpm. If a depressant drug is given, performance on this test is lowered. The interpretation usually given would be that: “The drug lowers performance on the pursuit rotor.” This, of course, is a true statement of facts, but it would have meaning psychologically only if pursuit rotor performance were determined in its entirety, in everybody, and at all times, by one identical set of abilities and drives. That this is not so is shown very clearly in the experimental work of Fleishman and Hempel (1954, 1955), Fleishman ( 1956). These authors showed, in brief, that when scores made by different subjects on the pursuit rotor at different points on the learning curve (i.e., after different periods of practice) are intercorrelated with performance scores on a large

336

H. J. EYSENCK

variety of other motor tests, then it is found ( a ) that performance at any one point is a compound of different abilities, i.e., performance on the pursuit rotor has significant loadings on several different motor ability factors; and ( b ) that the spectrum of abilities

300

I

o BEN HYOSCINE

I

3

5

I

3

5 7 9 II 13 15 17 19 21 23 2 5 C Y C L E S OF PRACTICE ( 6 C Y C L E S = I HOUR)

FIG. 1. Effects of administration of stimulant and depressant drugs upon massed practice. Taken from Payne and Hauty (1955).

Y

0

I

3 5 7 9 II TRIALS ON ROTARY PURSUIT

1 3 1 5

FIG.2. Diagram illustrating the different factorial composition of performance on the pursuit rotor at different stages of practice. Taken from Fleishman (1956).

337

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

which goes to make up the factor loadings of pursuit rotor performance changes in the course of practice, i.e., pursuit rotor performance may have a high loading on factor I and a low loading on factor I1 at the beginning of practice, and a low loading on

roo% VARIANCE UNACCOUNTED FOR PSYCHOMOTOR

-

50% MECHANICAL= EXPERIENCE=

i

I

I

I

I

3

5

9 7 TRIALS

I

I

It

13

15

FIG. 3. Diagram illustrating the different factorial composition on discrimination reaction time task. Taken from Fleishman (1956).

factor I and a high loading on factor I1 at the end of practice. Figure 2 shows in diagrammatic form some of Fleishman’s findings. Figure 3 shows the same phenomenon with respect to another widely used test in psychopharmacological work, namely reaction time experiments.

338

H. J. EYSENCK

We may put the matter in a mathematical way by writing: op2 =

+

OF;

+. . a

OF,’

f

68’

+

oE2

in which op2 stands for the total variance on the pursuit rotor test, while F1,Fa . . . F, stands for different abilities; P is Performance, S stands for the specific factors involved in the test, and E refers to the error variance. (The total equation refers to pursuit rotor performance at any one moment of time only.) Now consider the effect of drug on performance. The effect of the drug could be on any one element or any combination of elements in the equation; the simple fact that total performance is lowered tells us nothing whatsoever about the effect of the drug on those elements which are of real concern to us, i.e., the different abilities F1, Fa . . . F, in the formula. The result of any such experiment, therefore, is completely equivocal, and cannot be directly interpreted. The situation is worse, of course, when we consider that the equation given above only holds for one particular level of training, and changes as practice proceeds. Thus, even if we knew the effect of the drug on the different constituents of the equation at one level of training we would not necessarily be able to say anything about other levels of training. A particular example may make this clear. At the beginning of practice on the pursuit rotor the subject spends a lot of time hunting for the disc with his stylus (psychomotor coordination) and hardly any time in simple rotary pursuit movement of the arm. Let us arbitrarily call the abilities involved in these two activities F1 and Fz. When the subject has had a good deal of practice he spends very little time on activities of the kind F1 and more time on activities of the type Fz. Now if the drug happens to affect F1 but not Fz,then we might find systematic changes in pursuit rotor performance in the sense that at the beginning, where F1 played an important part, performance was impeded, whereas later on, when Fz took over, performance was not impeded by the intake of the drug. Thus two different investigators, starting subjects at different levels of practice, or using discs of different size (thus producing different levels of difficulty), might easily obtain contradictory results, and indeed the frequency of contradictory results in this field suggests that the above considerations may be of some importance. It is quite possible that a given drug may have a deleterious effect on Fl; may improve Fz,

TESTS I N THE ASSESSMENT OF DRUG EFFECTS

339

and may leave factor F3 unaffected; under those conditions any experimental result is possible depending on the degree of practice the subjects have had. (Further complications are introduced, of course, by the fact that one ability may often be substituted for another in achieving a given performance score, thus making the analysis even more complex. There is not room in this brief review to deal with complexities of this kind.) The problem suggested by these considerations can be solved by studying the effects of drugs not on an isolated test performance, but rather on factors derived statistically from whole patterns of tests. There are obvious practical difficulties in the way of this solution, and it would certainly be far more expensive, timeconsuming, and complicated than the much simpler type of experimental procedures currently employed, Nevertheless, the writer is firmly convinced that the gain in rigor, and in interpretability and duplicability of results, will eventually enforce some such solution, and will relegate the simple “one test experiment” to the field of preliminary investigation. The difficulty discussed above arises from the complexity of any particular test that may be chosen to measure drug effects. Other difficulties in interpretation arise from the fact that performance on a given test, even if this should be completely unifactorial, is, nonetheless, determined by a number of different parameters which must be specified. Such specification has usually been the task of learning theory. Let us consider again performance on the pursuit rotor, and let us assume that performance of the drug group falls off as compared with the performance of the placebo group. Let us also assume that we are dealing merely with one ability, say F1. Now in terms of learning theory we find that the following elements should be taken into account in assessing the causes of this decrement. (1) There may have been a decrement in what is known in learning theory as S H R or excitatory potential, i.e., the ability of the central nervous system (CNS) to form the new connections which are the neurological basis of learning. ( 2 ) There may have been an increment in what is known in learning theory as IR or reactive inhibition, i.e., neural fatigue affecting those cortical elements underlying the passage of neural impulses during the performance of the task; reactive inhibition is not assumed to be permanent like S H R , but to

340

H.

J. EYSENCK

dissipate in time. ( 3 ) Alternatively there may have been a decrement in drive symbolized in learning theory as D; this is supposed to interact with habit in multiplicative fashion to produce sEIt or performance. (4) Lastly, there may have been an increment in conditioned inhibition or which arises after a long period of massed practice. (As this rather complex variable need play no part in this example or in the remainder of this chapter, I will not try to explain it in detail here. A brief account of the relevant aspects of modern learning theory is given in Eysenck, 1957b.) Expanding the general formula linking these various concepts, we get SEE = ( D X SHE) - (IRX S H R ) , so that if we have no reason to suspect D of being affected either we must conclude that the drug has decreased SHR or that it has decreased IR,or both. Clearly the experiment which leads to the simple report that a drug has decreased performance on pursuit rotor learning, even under the simplified conditions assumed to hold here, is by no means unequivocal; instead of answering a psychological problem, it raises one, and leaves us to wonder which of the several possibilities mentioned above is the correct one. Fortunately the theory in question also enables us to design an experiment to answer this particular question. As mentioned above, IR dissipates with time, whereas sHR does not. This gives rise to so-called reminiscence effect; after a period of massed practice, during which I R has been built up, a rest pause will serve to dissipate this 1,; consequently after the rest pause performance is distinctly superior (because of the elimination of Ia which acts to depress performance) as compared with prerest performance. This sudden improvement after rest is referred to as the reminiscence effect. If now the drug acts on In, then a comparison may be made of two groups, both tested on 2 successive days. The experimental group would be given the drug on the first day and ( P ) on the second day, the control group would be given ( P ) on both days. If the drug affects IR, then a significantly greater reminiscence effect should be found for the experimental group. If, on the other hand, it is SHRwhich is affected, no such reminiscence effects should appear and performance of the experimental group on the second day should remain inferior to that of the control group. As an illustration we may quote the data in Fig. 4 which were collected in an unpublished experiment by Micko. Using the pursuit rotor,

341

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

he tested twenty-five subjects each under 250 mg Doriden ( D ) and twenty-seven subjects under ( P ) conditions during a 5-minute practice period on the first day. Both groups were retested on the second day, again for 5 minutes; this time both groups were under no-drug condition. Scoring was carried out for successive 10second periods, and the time on target during this 10-second period is plotted on the ordinate. It will be seen that during the first day the ( D ) group is inferior to the ( P ) group for almost the whole of the 5-minute period. Because of the great individual differences, however, this difference does not achieve significance. When we 2.5

-

5 2.0 1.5 =

P U R S U I T ROTOR FIRST DAY

-----PLACEBO DORIDEN

I-W 0

P U R S U I T ROTOR SECOND DAY

2.5,-*.%a

*h '\"~\,"~-~,,~,,,",~.~~, u

p - PLACEBO DOR I OE N --.I-*

(a)

(b)

FIG. 4. Performance on pursuit rotor under placebo conditions and after administration of Doriden.

turn to the second day, we find that now the two groups are even more clearly distinguished and the performance of the ( D ) group is found to be significantly lower at the 1%level on the first trial. As this is a cruciaI trial for reminiscence, it can be seen that the results favor the hypothesis that ( D ) acts on reducing the excitatory potential to a significant degree. A similar study is illustrated in Fig. 5 in which are plotted the results from an investigation of eyeblink conditioning, carried out on the same subjects as the experiment just mentioned. It will be seen that during the first day the ( P ) group is superior to the ( D ) group, the mean number of responses being, respectively, 3.32 and 4.15, giving a difference of 0.83. Differences between the two groups on the second day are illustrated in full, and the degree of superiority of the (P) group is indicated by the size

342

H. J. EYSENCK

of the cross-hatched area. On the second day the mean number of responses was 4.96 and 5.89, respectively, giving a difference of 0.93. Again, any individual differences are so large as to make the effect of the drug of doubtful statistical significance, but the inadequate data suggest again that it is S H R or excitatory potential which is most affected by ( D ) . These data are illustrative and are not meant to suggest any definitive conclusions. They are merely meant to show that under the appropriate experimental conditions the problems raised in this section can receive an answer. But again it is essential that we get away from the simple type of paradigm outlined at the

1STDAY

9

10

11

12

13

14

15

16 17

18

FIG.5. Proportion of responses on eyeblink conditioning apparatus under no-drug condition of subjects who on the previous day had been given eight conditioning trials under either placebo or Doriden administration.

beginning of this chapter and resort to a more sophisticated type of experiment, the details of which are dictated by the most advanced and experimentally supported type of theory available in psychology.

111.

A Theory of Depressant and Stimulant Drug Action

We have seen in the preceding section that measures of the behavioral effects of drug actions, in order to be interpretable, should be part of a general psychological theory. It is the main purpose of this chapter to show how such a theory might run, and how support for it may be found in experimental studies which shall not fall foul of the rules suggested in the first section. The

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

343

theory to be discussed is one which has links both with learning theory and with personality theory; these links have been discussed in great detail elsewhere (Eysenck, 1957a,b) and will only briefly be referred to here. We have mentioned in the preceding section the two notions of excitation (leading to SHR)and inhibition (leading to In). Pavlov originally suggested that different organisms differ with respect to the strength of these two variables in such a way that individuals could be arranged along a continuum depending on their excitationinhibition balance; at the one extreme would be subjects strong in excitation and weak in inhibition, at the other extreme those strong in inhibition and weak in excitation. More balanced types would be intermediate. This notion was taken up by the writer (Eysenck, 1957a,b) as a possible basis for the personality dimension of extraversion-introversion, in the sense that the person strong in excitation and weak in inhibition would thereby be predisposed to introverted behavior patterns, while the person strong in inhibition and weak in excitation would be predisposed to extraverted behavior patterns2 It was also proposed that this general rule would hold not only for normals but also for neurotics; it had been shown previously that hysterics and psychopaths tend to lie toward the extraverted end of the continuum, whereas dysthymics (anxiety states, reactive depressives, obsessionals, etc. ) tend to lie toward the introverted end of the continuum. Furthermore, evidence was given to show that brain damage, particularly the operation of leukotomy, has an extraverting effect on personality, and may be assumed to increase inhibitory potential and to decrease excitatory 2 A detailed discussion OF the concepts of extraversion and introversion has been given by the author elsewhere (Eysenck, 1959d). There are many different ways of diagnosing the position of a particular person on the ewtraversion-introversion continuum. I t may be done in terms of ratings, questionnaire responses, psychiatric diagnoses (psychopaths and hysterics tend to be extraverted neurotics; anxiety states, obsessionals, and reactive depressions tend to be introverted neurotics ) ; objective tests of the “miniature situation” kind; physiological tests; and objective performance tests, linked in a rational manner to the general theory of extraversion-introversion(Eysenck, 1957b). It should be noted that these concepts are operationally defined in terms of the instruments used for measuring them and that, while descriptively they are similar in many ways to Jung’s terminology, there is no intention of accepting Jung’s psychoanalytical and somewhat mystical theory regarding the nature and origin of these personality types.

344

H. J. EYSENCK

potential. We thus have a complex of relationships linking the excitation-inhibition balance to personality, both normal and abnormal, and to brain damage. For proof of hypotheses such as those just suggested, we require tests which can be regarded as measures of extraversionintroversion on the one hand, and measures of the excitationinhibition balance on the other. As extraversion and introversion are behavior patterns observable through life, we can obtain estimates of an individual’s position on this continuum either in terms of ratings, psychiatric or otherwise (Eysenck, 1959d), or by means of self-ratings on specially constructed scales, such as the Maudsley Personality Inventory ( MPI ) extraversion scale ( Eysenck, 1959a). As regards the excitation-inhibition balance, we must look for measures of this in the field of general and experimental psychology, particularly in relation to experiments for the explanation of which the concepts of excitation and inhibition have been used. Pavlov himself used salivary conditioning in dogs as a measure of this balance; it is clear that dogs strong in excitation and weak in inhibition should condition well and quickly, and should retain their conditioned responses for a long time, while dogs strong in inhibition and weak in excitation should condition only weakly and with difficulty, and should retain their conditioned responses only for a short while. Many other measures relevant to the excitation-inhibition balance have been proposed (Eysenck, 1957b) , and some of these will be discussed in more detail when we turn to a consideration of the effects of drugs on these measures. We are now ready to define the relationship between groups of drugs and variables considered above. Put briefly, the writer’s drug postulate states that: “Depressant drugs increase cortical inhibition, decrease cortical excitation and thereby produce extraverted behavior patterns. Stimulant drugs decrease cortical inhibition, increase cortical excitation and thereby produce introverted behavior patterns.” We have already discussed the meaning of excitation, inhibition, extraversion, and introversion in this postulate; we must now discuss the meaning of “depressant” and “stimulant.” These terms are often used by pharmacologists to group together drugs such as alcohol, hypnotic and tranquillizing drugs, barbiturates, and so forth, on the one hand, and caffeine and the amphetamines, generally on the other. This pharmacological group-

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

345

ing is accepted as a first rough guide to the selection of drugs which would form the two groups included in the drug postulate, but it should be noted very clearly that ultimately the terms “depressant” and “stimulant” are used in a psychological context and must derive their definition from the behavioral effects of the drugs concerned. We might therefore say that in the long run the postulate may be used as a definition of depressant and stimulant drugs, respectively, in the sense that all drugs which increase cortical inhibition and decrease cortical excitation form the group of depressant drugs, while conversely, all those drugs which decrease cortical inhibition and increase cortical excitation form the group of stimulant drugs. The present pharmacological grouping is merely a suggestive first step toward a proper psychological grouping in terms of this drug postulate. I t might be objected that this is a circular process, in which drugs are defined as belonging to two great groups in terms of their effects on behavior, and that then the behavior effect of these drugs confirms the quoted hypothesis. Such an objection would be true but irrelevant; our interest lies in the possibility of experimentally defining an invariance of the kind outlined in the postulate; the discovery of such an invariance is of considerable interest and importance, although like most scientific advances it does involve a circular argument. If the invariance suggested should in fact be experimentally supported in a number of cases, then we might even be able to invert the argument and say that if a drug known to be a depressant drug according to our psychological definition has a certain effect on a given test, then this test is eo ips0 determined in part by cortical excitation or inhibition and may therefore be regarded as a measure of extraversion-introversion. It would, of course, be necessary to consider very carefully and critically side effects and other possibilities of erroneous inference, but by and large the argument from the invariance obviously can be taken in both directions. There clearly is a vast amount of experimental work waiting to be done before such a postulate can be regarded as firmly established and before we can use it effectively. Before turning to a consideration of the evidence, one example may be given in some detail to suggest the correct method of using the drug postulate. There is a well-known effect which

346

H. J. EYSENCK

emerges in studies using serial learning of nonsense syllables, i.e., studies in which series of nonsense syllables are shown to the subject for 2 seconds at a time, to be learned by him in such a way that he can anticipate the next syllable while the previous one is being shown. Series may be of any length, but usually consist of 8, 12, or 15 syllables, and a variable amount of time, usually 6 seconds, elapses between series. It is usually found that when the position of the syllables in the series is plotted on the abscissa and the number of repetitions necessary before a syllable is learned is plotted on the ordinate, then the resulting curve is bowed upward in the center, suggesting that the central syllables take longer to learn than those at the beginning or the end of the series. This effect is usually explained in terms of the Hull-Lepley hypothesis (Hull et al., 1940) as being due to trace inhibition developing between syllables within the series, and being most extensive at the center, thus producing the bowing there. If this hypothesis were true, then it would follow from our postulate (a)that extraverts would show greater bowing effects than introverts, and ( b ) that depressant drugs would increase and stimulant drugs decrease the bowing effect. Figure 6 shows the results of one test of this hypothesis, in which twelve nonsense syllables were used and four groups of subjects tested under (P), Amytal ( A ) , and Dexedrine (D-A) conditions; testing after (D-A) was begun either relatively soon or relatively late after administration (hence the groups were called “short (D-A)” and “long (D-A)” (Eysenck, 1957b, quoting unpublished work by Dr. Willett). It will be seen that there are no differences in the bowing effect between the groups, and at the time when the results were reported the writer pointed out that: “This failure may mean one of three things: (1) The drug postulate is in error. (2) The Lepley theory, according to which the bowing of the serial position curve is due to inhibition, is in error. ( 3 ) Drug effects are too slight to give demonstrable results in this situation.” As other tests had shown quite large drug effects, the third possibility is not a likely one, and it seems desirable to test the second possibility by designing a direct test of the Hull-Lepley theory. This was done by getting subjects to learn a series of nonsense syllables without interposing any breaks between series; in this way the pattern of trace inhibitions spanning the within-series syllables only would be broken and the bowing

347

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

effect should therefore be largely destroyed. In actual fact no such destruction was observed and the bowing effect was as strong as ever, thus showing effectively that the Lepley hypothesis was wrong and the bowing effect not due to inhibition as postulated (Eysenck, 1959e). It follows that the test is not a proper test of the theory and does not invalidate the drug postulate. It also follows that we would not now expect extraverts to have greater bowing

ul

60

g 55 50 Y L

45

3 40 a 35

2 30 20 25 IS I _

10

vI

I

I

---- SHORT DEXEDRINE DEXEDRINE ..---. AMYTAL

----LONG I

I

I

1

I

1

1

I

2 3 4 5 6 7 8 9 10 II I2 FIG.6. Nonsense syllable learning: serial position curves of groups of subjects administered different drugs. I

effects than introverts, and indeed no differences between extraverts and introverts have been found on this measure. This example shows the kind of reasoning applied in this situation, and it also shows the necessity of being careful in interpreting results, particularly where these are negative. It is always possible that a negative result is due not to a fault in the postulate but rather to an erroneous deduction due in turn to a fault in the general psychological theory of the phenomenon being used as a test. Great care must obviously be taken not to over-interpret data at such an early stage of development of psychological theory (Eysenck, 1960).

348

H. J. EYSENCK

IV. Experimental Tests in the Assessment of Drug Effects

In this section we will deal systematically with experiments carried out in the field of learning and conditioning, motor performance, perceptual performance, and autonomic reactions. This is not a survey of all the available evidence; such a survey has been given by Trouton and Eysenck (1960). We have concentrated very much on work done in the writer’s department with the specific intention of testing the hypothesis outlined above; work of others has only been drawn in occasionally when it seemed particularly relevant. All the work quoted relates to experiments performed in the laboratories; studies dealing with observation of behavior (Laverty and Franks, 1956) have not been dealt with here even though they were carried out in an attempt to test the theory in question. 1. Eyeblink Conditioning

. In view of the fact that Pavlov’s original theory was concerned largely with conditioning it seems appropriate to begin our work with a conditioning measure, more particularly as this had also been used to verify the hypothesis linking extraversion and introversion to the excitation-inhibition balance. The first study was carried out by Franks and Laverty (1955) on sixteen neurotic hospitalized subjects. Four different treatments were used, two of these consisting of different doses of ( A ) , one of ( P ) , and one of no drug or ( P ) , The subjects were given various personality scales including the Guilford R scale, which has been found to be a good measure of extraversion. The experimental test used was one of eyeblink conditioning, carried out in a soundproof conditioning laboratory. The unconditioned stimulus was a puff of air, delivered to one eye at a pressure of appoximately 65mm of mercury from a distance of 2 c m from the eye and lasting 500 msec. The conditioned stimulus was a pure tone of frequency 1,100 cycles per second at an intensity of 65 decibel above the subject’s auditory threshold. The duration of this tone, heard through a pair of padded earphones was 800msec. The air puff was so arranged that it began 350msec after the tone had commenced. Partial conditioning was used, the sequence being such that the reinforce-

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

349

ment ratio was approximately 60% throughout the reinforcement trials. Thirty reinforcement trials were given, interspersed with eighteen test trials, consisting of the conditioned stimulus alone. The intertrial interval varied from 20-30 seconds with a mean of approximately 25 seconds. After the thirty reinforcement trials and eighteen test trials had been given the subjects were given a further series of ten consecutive test trials (called extinction test trials) so that the resistance to extinction could be measured. The results showed the following number of conditioned responses during acquisition: ( A ) , dose of 300mg=2:0; ( A ) , dose of 90mg=4.0; ( P ) = 7.2; ( N ) , control run = 7.9. The number of conditioned responses during extinction for the four groups in the same order was 0.1, 1.9, 3.8, and 3.9, respectively. Extraversion scores under the four treatment conditions were 33.5, 37.8, 22.7, and 23.2. The authors conclude that “The two main hypotheses of this pilot study would seem to be confirmed tentatively. These are (1) that intravenous ( A ) reduces the number of conditioned eyeblink responses during acquisition and increases the rate of extinction; ( 2 ) that intravenous ( A ) increases the extraversion score as measured by Guilford’s R scale, . . . Under the conditions of the present study a dosage difference of 2 grains intravenously produces no significant differences in the measures under investigation. The trends in the data however are consistent with the possibility that eyelid conditionability is some undetermined negative function of the size of dose administered.” In a later study, Franks and Trouton (1958) repeated the experiment using as subjects eighty paid volunteer graduate female students of education, and employing ( A ) and (D-A) as well as ( P ) , as the drugs to be investigated. The results of the experiment are shown in Fig. 7. The (D-A) group shown in this figure had been administered the drug 2 hours beforehand; another group which had been administered the drug 45 minutes beforehand showed no difference from the ( P ) group. The authors conclude: “It was found that the %hour (D-A) group conditioned more readily than the ( P ) group, whereas the ( A ) group conditioned less readily.” The results of the two experiments are in good agreement

350

H. J. EYSENCK

with each other as well as with those of earlier writers from Pavlov onward. They are also in good agreement with the drug postulate in showing that the depressant drug does in fact have an extraverting effect while the stimulant drug has an introverting effect. The data leave unanswered the question of whether the effects are produced by acting on the excitatory or the inhibitory potential, or both; this point has already been discussed in a previous section. DEX.

. .80

,.:

:,*

---*---

-

PLAC. AMYTAL---

I

0

i. .60

0 0 W

.40

0

4

8

I2

ACPU IS I T ION TEST TRIALS

16

0 2 4 6 8 1 0 EXTINCT ION TEST TRIALS

FIG.7. Acquisition and extinction of eyeblink conditioning in groups of subjects administered different drugs. From Franks and Trouton (1958).

2. Nonsense Syllable Learning Several experiments have been carried out in the field of nonsense syllable learning. Some of these were related to the bowing of the serial learning curve, and this point has already been discussed. Two other predictions were tested by Willett (Eysenck, 1957b), namely: (1) that the mean number of anticipatory errors is increased by ( D - A ) and decreased by ( A ) , and ( 2 ) that the number of repetitions needed to reach the criterion of learning is increased by ( A ) and decreased by ( D - A ) . Six subjects in all were used, and tested under (P), ( A ) , and ( D - A ) conditions; testing under ( D - A ) was carried out twice, once with short, the other time with a long interval between the administration of the drug and the carrying out of the test. The mean number of anticipatory errors as a percentage of the

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

351

mean trial number was, as predicted, lowest for ( A ) (38.99%) intermediate for (P) (41.31%) and longest for ( D - A ) (48.13%). These figures refer to the condition under which a long time elapsed between the administration and testing; when (D-A) was administered shortly before testing, the figures were quite out of line with prediction (36.08%). As in our other experiments we have also found this condition to give results less in accordance with prediction, the possibility cannot be ruled out that a lengthy period has to elapse before ( D - A ) becomes fully effective. As regards the number of trials to learn the list of twelve nonsense syllables to the criterion, the following results were obtained. Under ( A ) number of trials was 40.68; under ( P ) conditions it was 37.21; and under ( D - A ) , short and long, respectively, it was 34.00 and 34.07. The results, while in line with prediction, fall short of significance when tested by analysis of variance. More recently Willett (1960) carried out another similar experiment using twenty-four subjects who were tested under ( D ), meprobamate ( M ) , and no-drug conditions. The main purpose of the experiment was to test the rate of learning which it was predicted would be depressed by depressant drugs. The number of trials taken to achieve the criterion just failed to achieve significance, but when the scores of individuals were plotted, it became apparent that the resulting slopes were in the form of a simple exponential function. The score on each trial was then put into log form so that the exponential groups could be plotted as straight lines whose characteristics would be readily calculable. These derived slope values were submitted to an analysis of variance, and it was found that the mean slope of the no-drug group was significantly steeper than that of either of those of the two drug groups. Since the slopes of learning curves represent rates of learning, this result means that both ( D ) and ( M ) have decreased the rate of learning of the nonsense syllables. On the whole then the results of these studies appear to support the general hypothesis that nonsense syllable learning is affected in predictable ways by depressant and stimulant drugs. There is one study by Burstein and Dorfmann (1959) which appears to contradict this statement. They examined the effect of ( M ) on the learning of a list of paired associates of high interitem competition. They paired the performance of some sixty subjects

352

H. J. EYSENCK

divided into a ( P ) group and a group receiving 1200mg of ( M ) , and found at a 5% level of significance that the ( M ) group learned the list more rapidly. They explained this in terms of its action in reducing the anxiety level since, according to the well-known Spence-Taylor hypothesis (Eysenck, 195713; Taylor, 1951), high anxiety will interfere with the learning of complex material. However, Summerfield and Steinberg (1957) have shown in a wellexecuted study that in comparing the retroactive interfering effect of an interpolated task accompanied by, and without the administration of, a depressant drug, this drug reduced the commonly observed decrement when performance of the original task was resumed. This result they put down to the drug inhibiting the process of interference during the interpolated task. It might equally be argued, therefore, in the case of Burstein and Dorfmann, that the action of the ( M ) in their study was rather one of inhibiting interference in a manner similar to that described by Summerfield and Steinberg. A direct test of this explanation would seem to be desirable. 3. Pursuit Rotor Learning

In the experiment now to be reported the pursuit rotor was used in the form described in a previous section. Four groups of subjects took part in this experiment; these were identical with those tested by Franks and Trouton in the experiment on conditioning described above and were given the same four treatments (Eysenck et nZ.,1957a). They performed for a 5-minute period consecutive practice on the pursuit rotor; this was followed by a 10minute rest, a second 5-minute period of continuous practice, a second 10-minute rest, and a third 5-minute practice period. It was predicted that the ( A ) group, due to either an increase in inhibition or a decrease in excitation, would perform worse and worse as time went on, while the (D-A) group for the opposite reasons would be continuously more superior to the ( P ) group, which would be expected to be intermediate between the others. Results are shown in Fig. 8, which on the whole bears out these predictions. The “long (D-A)”group is distinct from the others even in the first 5 minutes of practice. During the second 5 minutes of practice the ( A ) group also is separated out. It is not, however, until the short period of practice that the “short (D-A)” group is

353

TESTS I N THE ASSESSMENT OF DRUG EFFECTS

separated from the ( P ) group. An analysis of variance showed significance for both periods and treatments as well as for the period-treatment interaction. “We may conclude therefore that the drugs used have differentiating effects on performance, that these effects are in line with prediction and that they interact 9 LONG DEX. DEX. AMYTAL PLACEBO

8 7 -

6-

+ W

za

5 -

I-

4 In V W

u-l

3-

I

I

5

10

I

I

I

I

I

I

5

I

I

I

I

10

I

5

10

TRIALS

FIG. 8. Performance on pursuit rotor of groups of subjects administered different drugs. From Eysenck et al. (1957a).

in complex ways with stage of practice and/or time elapsed since administration.” It is interesting to note that not only do the drugs have the predicted effect, but also that it can be shown that extraverts tend to be inferior to introverts as practice goes on. Evidence for this has been shown in the paper by Eysenck and his co-workers (1957a). It may also be of interest to note that the findings of the drug study are in good agreement with the similar work by Hauty and Payne (1955) and Payne and Hauty (1953, 1955).

354

H. J. EYSENCK

4. Reaction Times The action of drugs on reaction times has often been studied (Trouton and Eysenck, 1960), but from the theoretical point of view it is by no means clear what kind of predictions one would make from the theory elaborated in this chapter. There seems no obvious reason why depressant and stimulant drugs per se should affect the simple reaction time with a regular warning signal, provided that the warning signal precedes the stimulus only by a very short amount of time. On the other hand, in experiments without a warning signal, or in experiments where a warning signal is either irregular or precedes a stimulus by a long period of time, we might predict that the impaired vigilance of the extravert or the person subjected to depressant drug treatment would produce longer reaction times and a greater variability of reaction time. The argument is more fully stated in the section on vigilance; the prediction made here will be seen to apply to reaction times only in so far as the simple straightforward reaction time experiment is complicated to take on the character of a vigilance test. An experiment to test this hypothesis was performed by I. Martin (1960) using twenty-four subjects who were given ( M ) and ( D ) , and were also tested under a control condition. Two series of tests were done; in the regular series the reaction times were obtained to twenty tones which were regularly preceded by a warning tone at an interval of 3 seconds. In the irregular series the twenty signal-to-respond tones were preceded by a warning tone at a time interval which varied from 3 seconds to 18 seconds. In both conditions a period of about 15 seconds elapsed between the subject’s response and the presentation of the next preparatory tone. Results were evaluated by means of an analysis of variance and it was found that the only significant effect of the drugs occurred during the irregular presentation, and only under the effect of 250 mg ( D ) . There was a lengthening of reaction times during the regular series under ( D ) but this was not significant. It must be concluded that the hypothesis under investigation is partly confirmed by these data, although it is not clear why 400 mg ( M ) had no apparent effect on the reaction times under either condition.

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

355

5. Kiltesthtic Figural After-Elqects Prolonged bombardment of sensory receptors sets up a form of inhibition which is known as satiation, and which can be measured by means of what Kohler has called figural after-effects ( McEwen, 1958). These may be visual, auditory, or kinesthetic, and they consist in the displacement of contours, or in the changing of sensations following a lengthy inspection period of some object presented to the appropriate sense. It follows directly from the general hypothesis here investigated that satiation should be stronger in extraverts, or after depressant drug administration, while it should be weaker in introverts or after stimulant drug administration. The only study published to date on the effects of drugs is one by Poser (1958) in which he used a measure previously employed by the writer to show that the predicted personality differences between extraverts and introverts did in fact occur (Eysenck, 1957b). Poser tested three groups of ten male volunteers before and 2 hours after administration of ( A ) , ( D-A ), and ( P ) , respectively. “The prediction that ( D - A ) would reduce satiation effects was borne out very significantly. The effect of ( A ) was however indistinguishable from ( P ) effects.” It appears, therefore, that the predicted effects tend to occur, although the failure of ( A ) to have a significant effect requires explanation.

6. Apparent Movement It is well known that if two stimuli are exposed in rapid succession and in close propinquity, then apparent movement may be observed to occur from the one to the other; the apparent movement seen when stationary pictures are projected in the form of film are well-known examples of this. There is no widely accepted hypothesis relating to the explanation of this phenomenon, but it is usually assumed that a certain amount of interaction and irradiation occurs between the two foci in the cortex, which may be presumed to represent the respective receiving centers of the two original images. If we may further assume that the degree of inhibition of the intervening cortical structures determines the speed of transmission from one of these foci to the other, then we can predict that the threshold for apparent movement will be lowered by depressant drugs, in extraverts and in brain damaged

356

H. J. EYSENCK

subjects. This prediction was tested by Costello (1960a) in an experiment in which six subjects were tested twice, once under ( P ) and once under 600 rng of ( M ) , testing on each day being done once before administration of the drug and 1, 2$, 3f/3, and 4% hours after administration. The Wither’s direct exposure tachistoscope and an electronic cycling timer were used. The exposure time was 45 msec, the stimulus figure used was a simple one, consisting of two circles of light 17; cm in diameter and 3% cm apart. After a number of practice trials the subject was instructed to report whenever he saw a change from simultaneity to movement or from movement to simultaneity (this threshold was used because there were difficulties in obtaining a satisfactory lower threshold, i.e. one between succession and movement). The time interval at which the subject reported change was reported in milliseconds. Four trials were given, with alternating ascending and descending trials. The means of the four trials constituted the threshold for that session. The results were treated by analysis of variance and it was found that two significant F ratios emerged, namely those for subjects and treatments. As predicted, ( M ) lowered the threshold as compared with ( P ) conditions, i.e. the change from apparent simultaneity to apparent movement occurred at longer time irtervals under ( M ) . Time of testing apparently had no effect and neither was there any significant subject-treatment interaction. The results on the whole support the hypothesis under investigation.

7. Visml-Field Eflects In spite of the constant use of perimeters and other devices for measuring the visual field, there has been very little study of the effect of drugs in this connection. It has been found that anoxia acts to reduce the extent of the visual field, but this may be due to a generally depressant effect of anoxia on responses. It would appear that the hypothesis with which we are dealing here would be capable of mediating certain predictions because it implies that the excitation and extinction processes which characterize the specific isopters will be changed in such a way as to alter the extent and gradient (slope) of visual sensitivity. An inhibitory drug should reduce and depress the field, whereas a stimulant drug should enlarge and raise it. A specific study to test this prediction was

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

357

carried out by H. Holland (1960) using a Goldman spherical projection perimeter, in which bowl and target lighting are derived from the same source. The bowl lighting is variable over a very wide range and the bowl-target ratio is achieved by filter and photometer screen. In the present investigation the target size was kept constant at 0.25 mm2 and the light at 358 lux which gave, with a bowl albedo of 0.7, a reflected intensity of 250 apostilbs. Only the left temporal field was investigated. Inner and outer thresholds were determined along 14 meridians, every 15" with the exception of the horizontal coordinate which was missed, and readings taken at approximately 83" and 92", respectively. Three conditions of testing were employed, namely, no drug, ( M ) , and ( D ) . The two latter were used as depressant drugs. Twenty-four subjects in all were employed in the investigation. Results differed very much for incoming and outgoing movement of the target. The action of the two drugs appears to reduce the excitatory value of the incoming target but to increase it for an outgoing target. If the field is defined in terms of the inner limen, results determine a clear-cut reduction in sensitivity, as predicted. On the other hand, if they are defined by the outer limen they indicate an increase in sensitivity. Finally, if they are defined by the mean of the two, the target isopter, they indicate no significant change at all. These results are highly significant by analysis of variance, and in particular the difference between inner and outer limen produces very marked differences between ( P ) and drug conditions. Holland points out that "As the inhibitory hypothesis advanced originally would not account for these findings, for it implies that excitation and extinction will co-vary, any explanation must be post hoc and speculative. At least two, however, can reasonably be advanced. The first such explanation suggests that there is some change in the psycho-physical scale, whereas the second postulates a general slowing of response, which, within the framework of the experimental procedure, gives a false picture of changed sensitivity. It would be tempting to accept the second type of explanation, which involves a notion of some kind of mental or perceptual inertia, and to conclude by suggesting methods whereby experiments of a similar nature could control for such a possibility. However, to do so would ignore the fact that several other tests involving rapid responses have been administered to the same sub-

358

H. J. EYSENCK

jects and have failed to lend support to it. . . . Apart from these facts, the concept of mental inertia is a poor one in terms of explanatory value, suggesting little more than that the system is impaired from optimal efficiency, and delineates little but the conditions of its observation." Holland, therefore, prefers the first type of explanation, i.e. some change in the psychophysical scale, but it would take too long to go into the complex details of its discussion. For the purpose of this review we must conclude that the results, while strikingly large in terms of the small amount of drug administered, can only be explained in part by the hypothesis which the experiment was set up to test, and that further work will be required to test the alternative hypotheses advanced ad hoc to account for this failure of prediction.

8. Flicker Perimetry In another experiment using the same subjects and drugs as the preceding one, Holland (Eysenck, 1960) tested the effect of depressant drugs on flicker at the limen of the visual field. The perimeter used was a modified Lister moving-arm type, which has its own illumination source for background and has targets of differing sizes and colors which are changeable. Modification to the instrument took the form of removing the normal target carried and mounting in its place a small neon bulb which subtended an angle of 1.8". Flicker was provided by a pulse generator and the neon activated through a finger key. Owing to the modification the bulb carrier could not be positioned nearer than 4" from the center spot but from this position to 90" it was continuously variable along any meridian. After some practice trials the critical flicker fusion (CFF) thresholds from flicker to fusion and fusion to flicker were assessed in fourteen different positions on the 45" and 135" meridians at lo", 25", 50", and 60" peripheral angle, and along the 90" meridian at lo", 30", 45",60°, 70°, and 80" peripheral angle, all three coordinates being in the left temporal field. The method of limits was employed, and of the successive bursts of light none was longer than 3 seconds, with a 3-second interval between them. Several analyses of variance were carried out all showing the same pattern, i.e. significant F ratios being derived for both in-

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

359

dividual differences and for treatment effects. None of the analyses are complicated by rank order effects. Both ( M ) and ( D ) depressed the CFF threshold, and graphs published by Holland (Eysenck, 1960) demonstrate that the depression is fairly uniform along the sensitivity gradient from the center to the periphery within the positions measured. This is in line with previous work on CFF which has usually demonstrated depression of CFF thresholds by depressant drugs and elevation by stimulant drugs. 9. The Visual After-Image It is well known that stimulation of the eye by either a colored object or a rotating spiral produces certain after-effects once the stimulus is withdrawn. In the case of the colored object this is the well-known after-image; in the case of the rotating spiral it is an after-image which appears to produce effects opposite to those of the rotating spiral used as a stimulus. It is possible to relate the length of these after-effects to the drug postulate somewhat in this fashion. The experiment may be divided into two periods, (1) the period of stimulation during which the spiral or the colored surface is presented to the eye of the subject, and (2), after-effects or reversal period, during which the subject experiences a subjective sensation. In what follows I shall call the stimulation process X and the reversal X;T refers to the time during which either process is in operation. During the period of stimulation ( T ) there is a constant increase in X ; this may be assumed to take the form of a straight-line function, although the actual slope and shape of this line are irrelevant to the argument. This theoretical function, however, is interfered with by a process of satiation which reduces the amount of X to a degree which is proportional to the satiability of the subject, i.e. it will be greater for extraverts than for introverts, and for subjects after a depressant drug as compared with subjects after a stimulant drug. The total amount of stimulation actually received, therefore, is a function not only of T , but also of the state of the organism. We now come to the reverse effect which is assumed to set in the moment the stimulus ceases to be effective, and which is assumed to be proportional in strength to the strength of the original X. Because of the original satiation, X, and stimulus effect, i.e.

x=

360

H. J. EYSENCK

x,

consequently will differ from person to person in terms of their degree of extraversion, dose, and type of drug, etc. However, there is a further reason to predict that there will be differences in X between these groups. It must be assumed that the process underlying the reversal is physiological in nature and consequently subject to satiation; this satiation will again be stronger in extraverts than in introverts, and in subjects after a depressant drug as compared with subjects after a stimulant drug. These effects, leading to a shorter after-effect period for extraverts and depressant drug subjects as compared with introverts and stimulant drug subjects, are assumed to be cumulative; in any case the prediction is clear that after-effects will be shorter for extraverts and depressant drug subjects, and longer for introverts and stimulant drug subjects. The evidence regarding personality correlates, particularly with respect to the spiral, indicates that the predicted relationship to extraversion-introversion is indeed found; both with normal and with neurotic subjects the predicted decrease in after-eff ect with extraversion has been observed (Eysenck, 1960). As regards drug effects, the only study available is one by Costello (1960b) in which he used a red square ( 5 cmz) with a black focusing point in the center as a stimulus. This red square, surrounded by black opaque paper, was exposed in the viewing tube of a tachistoscope and illuminated by two 5-watt neon lamps behind a perspex sheet to which the red square was attached. All testing was done in a dark room. Six subjects were used in all and each subject was given two treatments: ( P ) and 600mg ( M ) . Both the ( P ) and ( M ) were in identically appearing tablet form and taken orally. They were administered on 2 different days, the order of treatments being counterbalanced. On each day the subject was given one session before treatment and four sessions 1, 2 5 , 3 5 , and 4% hours after administration of the drug or ( P ) . The red square was exposed for 30 seconds each time and after-effects noted by the subject on an automatic recording device. Two scores were used, one being the total duration of the after-effect, i.e., the time from the end of stimulation by the red square to the disappearance of the last after-image, the other the latency period, i.e., the time from the end of stimulation by the red square to the appearance of the first after-image. Analyses of variance were performed and significant F ratios

361

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

emerged on both scores for subjects and treatments, and also for a subject/treatment interaction on the total duration scores. Costello concludes that the significant differences observed indicate ( 1) that considerable differences exist between people in the latency and total duration of the after-image, ( 2 ) that ( M ) , as predicted, decreases the latency and total duration, and ( 3 ) that there are differences between people in the effect of ( M ) on the total duration of the after-image. It is noteworthy that although the subjects were tested at intervals up to 4 5 hours after administration of the AMYTAL PLACEBO DEXEDRINE

X---.

A--

Q 0 z

1

I

I

I

I

I

2

3

4

TRIALS FIG. 9. After-effects of rotating spiral on subjects administered different drugs. From Eysenck et al. ( 1 9 5 7 ~ ) .

treatment and treatment/time interactions are not significant. The experiment on the whole decisively supports the hypothesis under investigation. 10. Rotating Spiral After-Image After-effects of the rotating spiral were studied by Eysenck et nl. ( 1 9 5 7 ~ )Six . subjects were tested and were administered in counterbalanced order, in the form of a double Latin square, ( P ) , 300 mg ( A ) , and 10 mg ( D-A). Four trials of 1-minute stimulation each, during two of which the spiral rotated in one direction, two in the other, constituted the testing procedure. Total length of aftereffect provided the score. Results are shown in Fig. 9, and it will be seen that, as predicted, trials under ( A ) showed the shortest

362

H. J. EYSENCK

after-effects, those under (D-A) the longest, with the ( P ) group intermediate. Analysis of variance showed highly significant F ratios for both people and drugs; the effects of replication were barely significant and none of the interactions were significant. Here again, then, results are positive and support the hypothesis under investigation. A further study of the spiral after-effect duration was carried out by J. Easterbrook (unpublished data) the subjects being eight in number and the drugs being again 5 mg (D-A), 90 mg (A), 200 mg ( M ) , and ( P ) . The experimental design, a balanced incomplete block, insured that each drug would be given once after each other drug and in each serial position. The block was completed twice, once for the subjects seen in the morning and once for those seen in the afternoon. The mean duration of after-effects was 24.05 seconds under ( P ) conditions, 29.64 seconds under (D-A), 23.11 seconds under ( A ) , and 24.97 seconds under ( M ) . The over-all analysis of variance just failed to show a significant general effect of treatments, although for the most part these will be seen to have been in line with prediction. However, the greater duration of effects under the (D-A) as compared to the ( P ) treatment is significant by a t test.

11. Suppression of the Primary Visual Stimulus The phenomenon used here is an objective after-image determination first introduced by Bidwell (1897). He found that if a brief stimulation of the eye with red light is followed by prolonged stimulation with white light, then what is seen is not a red flash but a green one. In the experimental arrangements used by Eysenck and Aiba (1957), a light source throws a shaft of white light through a red filter into a viewing tube; the subject holds one eye close to the other end of the viewing tube. The beam of light is interrupted by a rotating disc, into which a small sector has been cut; this sector allows a red beam of light to pass for a period of 20 msec before it is interrupted by the solid disc. The red beam is followed by a piece of white paper pasted on to the disc on the side of the subject and illuminated by a variable light source. Duration of exposure of the white stimulus was 125 msec. Starting with a very bright red, the subject’s task is to reduce the intensity

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

363

of the stimulus until he sees only green and no red at all. Six subjects participated in the experiment, each serving as his own control under ( P ) , ( A ) , and (D-A) conditions. The prediction made was based on a theoretical explanation of this phenomenon. Briefly, the postulated chain of events is this. The brief red stimulus sets up two chains of events. One of these is photochemical, the other neural. Before the neural message reaches the cortex and is transformed into a red sensation, the white light is exposed and sets up preexcitatory inhibition ( Granit, 1955) which destroys the neural effects of the red stimulus. Thus no red is seen, provided its brightness has not been too strong for the inhibitory impulse to overcome. At the same time that the red stimulation ceases, the photochemical activity is reversed and produces the green after-image. The image is now seen simultaneously with the direct sensation from the white stimulation, which thus produces a slight loss in saturation in the after-image. If (D-A) reduces the inhibitory effects postulated while ( A ) raises them, then we would expect the former to have the effect of lowering the threshold of the original red stimulus, while the latter would raise it. The predicted differences were in fact observed, all differences being significant at the 0.01 level or better, and the ( P ) group being intermediate between the other drug groups. The result therefore strongly supports the theory on the basis of which this test was selected. A rather more complex experiment along similar lines was carried out by Aiba (unpublished results). Five subjects were used and tested before and after ( D ) , 10 mg (D-A), and 300 mg ( A ) . The procedure was as follows. After 15 minutes of dark adaptation the subjects were told to look at the stimulus patch and to increase the intensity of the red stimulus by turning the knob of the variac until they could see the first suggestion of red. As soon as they were satisfied with the adjustment, they resumed the dark adaptation until the next trial started. Four of these trials made up one session, and during the next 140 minutes the same procedure was repeated every 20 minutes. Thus there were altogether six sessions, each consisting of four trials. The drugs were administered between the second and the third session each day. Results are shown in Fig. 10. Differences were shown to be significant at a very high level by means of analysis of variance, and it will be seen that as predicted the (D-A) lowered the threshold

364

H. J. EYSENCK

whereas the ( A ) raised it. Of particular interest is the interconnection between drugs and sessions; “this showed that the drugs produced definite effects but that the time at which effects of each drug became manifest was not identical.” This is an important point, because if the wrong time factor is chosen in experiments of this kind, duplication of results may be difficult. m

O----.--<)

I13 -1

3

5

PLACEBO

P

AMYTAL

I2

A * - - - - -d. DEXEDRINE

II

- 10

ill 10

IV 30

V 50

VI SESSION 70 M I N U T E S

FIG. 10. Effects of drugs on the suppression of the primary visual stimulus, as a function of time since administration.

12. Vigilance Vigilance is a term used for the behavior of test subjects who have to watch visual displays, or pay attention to auditory stimulation for lengthy periods at a time, always being ready to respond with a signal (pressing a bell, etc.) to certain relatively rarely occurring changes in the stimulus display. Thus, the subject may be required to watch a pointer going round a dial at a uniform speed; occasionally for a brief moment the pointer doubles its speed and this change in speed has to be noted (Mackworth, 1948). On the theory under investigation one would expect that extraverts would be poorer at tasks of this kind (i.e., would tend to miss more signals, as compared with introverts) and similarly one would expect depressant drugs to cause subjects to miss more signals,

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

365

while stimulant drugs would be expected to have the opposite effect. The argument, of course, is quite a straightforward one depending on the accumulation of reactive inhibition in the neural structures responsible for what the layman calls “attention.” There is ample evidence that the prediction regarding the relationship between vigilance and introversion is in fact as predicted, and there is some evidence from the work of Felsinger and his associates (1953) and of Mackworth (1948), that depressant drugs and stimulant drugs do indeed have the predicted effects. Two experiments were carried out by E. Treadwell (1960) in an attempt to collect further evidence on this point. In the first experiment an auditory task was used. A series of single-digit numbers was recorded on tape and was reproduced at uniform loudness, the subject being instructed to react to any sequence of three odd numbers. The frequency with which the signals occurred was approximately six every 5 minutes, the time intervals between signals ranged from 3 seconds to 1 minute and 41 seconds with a mean interval of 58 seconds. The test lasted for 30 minutes and was preceded by a 2-minute period. Twenty-four subjects were tested after 400 mg ( M ) , 250 mg ( D ) , and no-drug conditions. Mean error scores for the whole test were 4.3 under no-drug conditions, 4.8 under ( M ) , and 6.2 under ( D ) . The results when tested by analysis of variance felI short of significance, although the mean scores are in the predicted sequence. The test was found to be relatively poorly adapted to the needs of a study of this kind, and a new test was designed. In this, singledigit numbers were recorded on tape and played to the subject for 40 minutes, and he had to check these numbers against a printed list. Some of the numbers in the list were erroneous and these had to be indicated by the subject, and the correct number written in. Scores were calculated in two ways, one score consisting of the number of missed errors, the other one of the number of false responses. Eight subjects were tested on this test under conditions of (P), (D-A), and ( A ) administration. The prediction that (D-A) would lower the number of missed and erroneous replies as compared with placebo conditions, while ( A ) would raise the number of missed and erroneous answers was realized. The number of missed responses for (D-A), (P), and ( A ) conditions, respectively, was 3, 7, and 14; the number of false

366

H. J. EYSENCK

responses was 2, 13, and 18. Analysis of variance showed the “missed response” score to be statistically significant; the number of “false signal” scores did not quite achieve significance. On the whole, the results bear out the hypothesis and demonstrate that vigilance is impaired by depressant drugs and improved by stimulant drugs.

V.

A Second Paradigm for the Testing of Psychopharmacological Hypotheses

The research design on which the investigations so far quoted have relied is illustrated in Fig. 1 1 ( A ) . It will be seen that this

--

STIMULANT

I

INT ROV ERS 10 N

A

DEPRESSANT

RANDOM ?SAMPLE PLACEBO CONDITIONS

I

EXTRAVERSION

DEPRESSANT

B c

DRUG ACTION DY ST HY M ICS

0 (INTROVERTS) EXC ITAT ION

MIXED NEUROTICS

0 (AMBIVERTS)

HYSTERICS

S EDATlO N THRESHOLD

4 I

(EXTRAVERTS) INHIBITION

FIG.11. Two fundamental research designs for the study of drug effects.

design does not depend in any way on an assessment of the personality of the subjects prior to the experiments. Subjects are allocated at random to control and experimental groups, or all subjects are tested under control and experimental conditions, and what is studied is a general effect of the drugs under investigation on what might be called the standard subject. This paradigm, is, of course, capable of certain improvements, but essentially it represents the great majority of research designs in present day psychopharmacology.

367

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

A rather different design is illustrated in Fig. 11(B). This design makes use of the known position of groups of subjects such as dysthymics and hysterics on the introversion/extraversion continuum, and thus explicitly contravenes the random sampling technique of design ( A ) . This new design may be illustrated by going back to McDougall’s observation that extraverts need less alcohol to reach the point of intoxication than do introverts, who, with the same amount of alcohol simply become more extraverted. Such a research design requires an objective terminus ad quem, i.e., the terminus of “intoxication threshold which would enable us to ascertain the amount of alcohol required by the different groups of subjects to reach the same level of cortical inhibition as defined by this-threshold.

- BIFRONTAL EEC

30

- 0

1.0

c)

I:

2.0

I

I4

$

3.0

4

f-

’

SLUR

-

La

P x

1

4.0

THRESHOLD

0

5.0

6.0

I SEC

-

I

50pv

0

0 0.5 I 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 SODIUM A M Y T A L (MGM/KG)

FIG. 12. Bifrontal EEG recordings illustrating sedation threshold.

The only example of the use of such a technique which the writer has been able to find is a study of Shagass (1954, 1956) and Shagass and Naiman (1955, 1956) using what he calls the “sedation threshold” of ( A ) . The sedation threshold is an objective pharmacological determination, which depends on electroencephalogram (EEG) and speech changes produced by intravenous doses of ( A ) given at the rate of 0.5 mg/kg of body weight every 40 seconds. The patient is tested for slurred speech, and the injection is continued at least 80 seconds after slurred speech is noted. Con-

368

H. J. EYSENCK

tinuous EEGs are recorded from transverse-frontal and sagittalfrontocentral placements. Figure 12 shows that ( A ) produces a rather striking increase of fast frequency ( 1 5 3 0 c/sec activity). The amplitude of this fast frequency is taken by Shagass as a response to the drug and the dosage-response curve plotted. “The typical curve has a sigmoid shape and contains a point of inflexion, preceding which there is a sudden increase in the amplitude of the fast activity, and following which the curve tends to plateau. This inflexion point generally occurs within 40 sec. (0.5 mg/kg) of the time when slurred speech is first noted, and the slur and inflexion point are used together as indicators of the threshold. The threshold is the amount of sodium amytal in mg/kg required to produce an inflexion point in the 1530 c/sec. amplitude curve, which occurs within 80 sec. (1 mg/kg) of the time when the slur is noted. The slur localizes the threshold roughly, the EEG inflexion point does it more precisely. . , . The measurement is highly reliable; its probable error is not greater than 0.5 mg/kg of body weight. Age, sex, and previous intake of sedatives in usual psychiatric dosage have not been found to influence the threshold.” According to the theory outlined in this chapter (which was developed before Shagass’s work was known to the writer), we should be able to make a very definite prediction. (A), being a depressant drug, would be postulated to increase inhibition. An extravert, whose cortex, according to our theory, is already in a relatively inhibited state, should require comparatively little ( A ) before reaching the critical sedation point; such a person should have a low sedation threshold. The introvert, on the other hand, whose cortex is in a state of considerable excitation and low inhibition, would require a considerable amount of ( A ) before reaching the critical sedation point; he would be predicted to have a high sedation threshold. If we express this general hypothesis in terms of neurotic groups and their standing on the extraversionintroversion continuum, then we would expect psychopaths to have the lowest threshold, followed by hysterics. Mixed neurotics would be intermediate and anxiety states, obsessionals, and reactive depressives would have sedation thresholds. An experiment along these lines was carried out by Shagass; his results are given in Fig. 13. It will be seen that these results bear out our prediction in every detail.

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

369

This design too, of course, is capable of certain interesting modifications. If something corresponding to the sedation or intoxication threshold could be found at the introverted end of the continuum, we would predict lower thresholds for dysthymics than for hysterics, and quite generally a reversal of the relations found

I 1 zixg PSYCHO NEU ROSES

DIAGNOSIS

t

CONVERSION HYSTERIA AND HYST ERlCAL PERSONALITY

N0. CASE

26

x x x

2

x x x x x

MIXED NEUROSIS

x x x x x x x x x x x x x x

ANX IETY HYSTERIA OBSESSIVECOMPULSIVE

x x x x x x x x x x x

X

x

x

XXXE X

x x x x x

x x

NEUROTIC DEPRESSION x x x x

I

ANXIETY STATE

1

x x x x x x

_-

x x x x

x x x x x x

x x x x

1.5 2 2.5 3 I5 4 4.5 5 5.5 6 6.5 7+ SEDATION THRESH 0LD (MG M/KG)

FIG.13. Sedation threshold distribution for various neurotic groups.

on Shagass’s research. Thus the same groups of patients at different times might be given different and opposing drugs as well as placebos. However, such developments depend on the discovery of such threshold effects for excitation drugs, which might possibly be looked for in EEG patterns corresponding to wakefulness as opposed to sleep.

370 VI.

H. J. EYSENCK

A Mathematical Model of a Psychopharmacological Hypothesis

The theory discussed in this chapter, it will be remembered, states that depressant drugs have an action on the cortex which increases inhibitory potential and decreases excitatory potential, while the stimulant drugs have an opposite effect. It will also be remembered that the classification of drugs presently offered by pharmacologists was merely accepted as a tentative first hypothesis, and that the ultimate criterion of the appropriateness of classifying a drug in one group or the other was to be its similarity of action on a group of behavioral tests to the action of other drugs in that group. Clearly these terms are too vague to be useful in a rigorous and systematic theory; we require a more clear-cut definition of such terms as “similarity.” The method suggested here derives fundamentally from Fisher’s “Discriminant function analysis,” particularly that generalized form of it which goes by the name of “Canonical variate analysis.” A detailed discussion of this has been given by Eysenck (1957b) and by Patrick Slater (1960), and consequently only a brief verbal description will here be attempted. The first users of this method in psychology have all been concerned with a problem, which in some ways is rather similar to the one posed here. If we have groups of normals, neurotics, and psychotics, and if all these have been submitted to a battery of personality tests which discriminate adequately between them, can we use these scores to answer a very important psychiatric question, namely that of the relative position of these three groups? In other words, do normals, neurotics, and psychotics lie along one single dimension, with neurotics, as it were, less ill psychotics, and psychotics more seriously ill neurotics, or do we require two dimensions, i.e., one of neuroticism and one of psychoticism, to deal with the test performance of these subjects? What canonical variate analysis does in this situation is essentially this. It first of all combines all the scores obtained from our subjects in such a manner as to maximize the differences between the groups; in this way we obtain our first latent vector and its associated latent root. The method then goes on to extract another combination of scores uncorrelated with the first, and giving us the (residual) optimum differentiation between the three groups; this gives us the second latent vector and its associated latent root. If the three groups lie

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

371

along one dimension, then a test of significance applied to the second latent root would show it to be insignificant; in other words, all the differentiating power of the tests is accounted for in terms of the first vector. If, however, two dimensions are involved, then this is shown by the fact that the significance test reveals the second latent root to have a P value smaller than 0.05 ( Eysenck, 1957b). Let us now turn to the application of this principle to the model of a psychopharmacological theory. Let us suppose we are interested in a hypothesis to the effect that two drugs, say ( M ) and ( D ) , are both CNS depressant drugs, i.e., that they differ from each other only in the strength of the effect produced, and not in the direction. We can test this hypothesis by experimentally creating three groups of subjects, i.e., those tested under ( D ) , those tested under ( M ) , and those tested under ( P ) or no-drug conditions. It would be possible to have different people in these three groups, and if tests were used scores on which changed considerably due to practice or learning, this would be the only experimental design open to us. However, in view of the great individual differences on test scores even under no-drug conditions, this design would be extremely wasteful and would require very large numbers of subjects; consequently a design seems preferable in which the three groups are made up of the same people tested under the three conditions in question. To illustrate the method with an actual experiment, Eysenck and Eysenck (1960) selected five tests, namely: nonsense syllable learning, reaction times, level of skin resistance to the passage of an electric current, flicker fusion, and perimeter threshold difference; all these tests were administered to twenty-four subjects under conditions of no-drug, ( D ) , and ( M ) medication. These five tests were chosen from a larger battery, the choice depending upon the total effectiveness of the test separating the groups, and on the necessity of maintaining experimental independance between the tests. Details of some of these tests have already been given on the previous pages. The results of the canonical variate analysis are shown in Table I. It will be seen that the first of the latent roots is significant at the 0.001 level, whereas the second latent root fails to be significant even at the 5% level. The results therefore bear out the hypothesis and show that both ( M ) and ( D ) are depressant drugs, and do not differ from each other in behavioral effects.

372

H. J. EYSENCK

Scores were computed for each subject on both canonical variates in the following manner. The latent vectors gave two sets of weights to apply to the scores obtained from the tests, so that two measures could be calculated for each subject, one for each canonical variate. These scores, Y1 and Yz, were obtained by multiplying the score of the subject on the tests by the appropriate weights, and summing them over the five tests. The means of these scores are given in Table 11. These scores enable us to TABLE I RESULTSOF CANONICAL VARIATEANALYSIS Latent vectors Tests Nonsense syllable learning Reaction times LeveI of skin resistance Flicker fusion Perimeter threshold difference

XI

x2

- 0.040,971

- 0.014,399

-0.013,625 - 0.080,835

-0.161,517 -0.013,827

1.000,000

1.000,ooo 0.147,773

- 0.475,330

Latent roots h , = 0.453,220 = 90.03 k h2 = 0.050,205 = 9.97 %

Diagonal entries of matrix G-1 B = 0.503,425 Significance of roots

<

R13 = k1, X z = 40.446, P 0.001 R,2 = I,,, X 2 = 3.450, not significant

calculate what, by analogy with the psychiatric classifications problems, we may perhaps call the percentage of correct classification, i.e. given that we only knew the canonical variate scores of a person, could we correctly identify him as a person who had been tested under nu-drug condition, under ( M ) , or under ( D ) ? The method used is one originally suggested by Maxwell ( unpublished), and the results are shown in Table 111; it will be seen that more than 70% of accurate classifications could be made under these conditions, which is a remarkable achievement considering ( a ) the very slight amount of drug administered, and ( b ) the fact that both drugs had the same effect, and were only differentiated from the point of view of the strength of the effect.

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

373

It may be useful to discuss briefly the consequences of the statistical results reported so far. In the first place, it is suggested that we have here the beginnings of a truly scientific system of classification of drugs. The experiment reported, of course, is only a very small beginning; only two drugs, five tests, and twenty-four subjects were employed. This, however, does not detract from the general value of the method itself, which could easily be TABLE I1

(- = -)Y Y

MEANGROUPSCORES Y

Subject

Means of scores

-

Y, = 6.030 Y, = 8.622

C

-

M

Y, = 3.873 -

D

Yl= 2.989 -

Y, = 9.047 Y,

= 8.485

TABLE I11 RESULTSOF CALCULATING THE PERCENTAGE OF CORRECTCLASSIFICATION^ C M D

c

M

D

20 1 3

6 14 4

1 6 17

24

a

-

-

24

24

Under these conditions 70.8% correct classification could be made.

applied to much larger groups of drugs, subjects, and tests. In particular, it would be extremely interesting to include so-called stimulant as well as so-called depressant drugs to determine whether the unidimensionality of results could still be maintained. Furthermore, many other tests suggest themselves as being worthy of inclusion, and just as we can assess classificatory properties of the drug in terms of the test results, so we would then be able to assess the psychological significance of each test in terms of the displacement of scores on the test by sets of drugs found to be homogeneous in their action. In the particular case of stimulant

374

H. J. EYSENCK

and depressant drugs, we would also be able to test the classification of drugs in terms of tests by comparing the results with those of testing typically extraverted and introverted groups. In this way we would treat change in test scores as the dependent variable, and drugs on the one hand, and selection of extraverted or introverted subjects on the other, as the independent variable. It is necessary, however, to insert one word of caution at this point. If a latent root derived from the analysis of a given set of test scores is significant, then it is clear that the dimensionality of the universe in question cannot be less then that indicated by the number of significant roots, although it may be greater. (Thus in the psychiatric example mentioned earlier, another dimension additional to that of neuroticism and psychoticism, e.g. extraversion-introversion, might be involved, but could not be discovered by statistical analysis unless the groups were further subdivided, e.g. into hysterics and anxiety states, etc.) In this way the canonical variate analysis is inferior to factor analysis by enabling us only to test the significance of specific hypotheses which are embodied in the choice of groups and tests; factor analysis, although it has other shortcomings, is not so restricted. When, however, one or more of the latent roots are found to be insignificant, it does not necessarily follow that no significant differences exist between groups. A faulty choice of tests may give the impression of lack of difference, when a suitable choice of tests might have given quite different results. There is no possibility of guarding against this, and particularly at the beginning of the study of such a vast field as this it is almost certain that choice of tests will be relatively haphazard, and erroneous conclusions may easily be drawn. In the present case, for instance, it is possible that if reports of subjective feelings had been included as one of the tests, then a significant second root might have been established. (Subjects taking ( M ) reported no after-effects, while those taking ( D ) reported feelings of sleepiness, etc. ) The solution to the problem is twofold. In the first place, growth of knowledge and increase in the number of properly conceived and executed studies in this field will undoubtedly soon help us to cut down drastically the numbers of errors of this kind, and will help us to design batteries of tests which will cover the whole field more completely than is possible at present.

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

375

The second solution to the problem is complementary to the first. What is required more than anything, in our opinion, is the creation of a firm set of theories and hypotheses linking psychopharmacology with general and experimental psychology. It is by testing, confirming, revising, and improving theories of this kind that we must hope that improvement in the present unsatisfactory situation will ultimately come. Such has been the development of most scientific subjects, and it is difficult to see why psychology should be an exception to this rule.

VII.

The Place of Animal Work in a Hypothetico-deductive System

Animal experiments play a large and important part in the study of drugs, but it has always been axiomatic that results from animal work cannot be accepted indiscriminately as applying to human beings as well. Indeed, there has been much argument in the psychopharmacological field as regards the propriety of accepting results from animal work as being in any way relevant to human behavior. While this chapter is concerned in the main with the testing of pharmacological hypotheses in human subjects, it may be relevant to discuss quite briefly the place of animal experimentation within the hypothetico-deductive system under discussion. The difficulties of interpretation of animal work derive precisely from the same sources as do the difficulties of interpretation of work with human subjects discussed in the first section. Behavior, and changes in behavior, can only become meaningful as part of a theoretical system. If that system makes predictions in the animal field, then animal experimentation becomes relevant. The system under discussion here is concerned with such concepts as excitation and inhibition in the central nervous system, and in terms of modern learning theory, these concepts are just as relevant to animal behavior as they are to human behavior; indeed, some critics might claim that they were more relevant to animals than humans! Consequently we may use animals to confirm or refute specific hypotheses, and in turn animal work may give us helpful suggestions as to experiments to be performed with human subjects, and tests to be used in the human field.

376

H. J. EYSENCK

As this is a relatively minor point, only one example will be given of our work in the animal field. The study itself is concerned with alternation behavior in a simple T-maze (Sinha et nl., 1938). Such alternation behavior-that is, turning in the direction opposite to that of previous turns-is conceived of as a consequence of reactive inhibition; in other words, turning right at one point sets up reactive inhibition to further right turns and predisposes the animal to turn left, and vice versa. It is known that the amount of alternation is a function of the number of preceding turns, and it can be

SP2

I

I

I

SP3

SP3

SP2

I

G

----SPI

I

I

I

I

I

I

; i SPD

G

----SPI

.

predicted from the general drug postulate that this function will be increased by a depressant drug and decreased by a stimulant drug. The floor plan of the maze used is shown in Fig. 14. It consists of a rectangular runway communicating with a T-maze only at the lower end of the upright of the T. The alley forming the other side of the rectangular runway, opposite this junction, was permanently divided into two halves. Thus by suitable manipulation of a series of temporary blocks within this outer runway, subjects could be run to a food reward either from the foot of the T (starting position 0; SPO) or from either of the lower corners of the outer rectangle, left or right (starting position 1; SPl), or from either of the upper corners (starting position 2; SP2), or from either side of the permanent block on the upper side of the rectangle (starting posi-

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

377

tion 3; SP3). An animal would thus have made 0, 1, 2, or 3 turns, respectively, before entering the final T and these turns could be either to the right or to the left. Sixty-four rats in all were run through the maze, the drugs used being ( A ) , 15 mg/kg of body weight, ( P ) , and Meratran, 10 mg/kg of body weight. Each animal was tested under each of the three conditions. All rats were given preliminary adaptation training on the apparatus, and throughout the training they were forced to one or the other of the goal arms of the T by blocking the opposite side at the junction according to a Gellerman series, in order to guard against the development of position habits. An analysis showed that the training procedures, designed to establish equal asymptotic habit strength in running to either right or left goal from the various starting positions, were in general successful. All subjects were then given the same alternation test lasting two days, which consisted in allowing the rat a free choice in the T, both goal arms now being left open. This was done on eight trials on two successive days, and each such free-choice trial was followed by a forced trial with one or other of the arms blocked. Thus, a total of eight trials, four free alternating with four forced, was given each day with a constant intertrial interval, despite the differences in starting position. The starting position side was selected as before-for rats other than those in group &by use of a modified Gellerman series, but the goal side which was blocked on the forced trials was in each case determined by rat’s own choice on the previous free-choice trail, so that each rat was forced away from the side previously chosen. This was a further precaution against the development of position habits.3 During each free-choice trial an alternation response was scored only when the rat turned to the goal arm of the T in the direction opposite to that of its previous turns in the maze on that same trial. A rat with a position habit would thus not vitiate the results because it would, by always turning to one side and being started from each side four times, score only four out of the possible eight alternations-that is, chance level. Alternation responses for rats from starting position 0, without previous turns, were scored in a comparable way, the justification for which is the assumption that 3

Data in this paragraph from Sinha et al. (1958).

378

H. J. EYSENCK

IR may persist from one trial to the next in this simple situation (Zeaman and House, 1951). In this case, an alternation response was defined as a turn on free-choice trial to the goal arm of the T in the direction opposite to that to which it was forced on the previous trial. For the first (free-choice) trial on any day, the scoring convention adopted was to regard the last (forced) trial on the previous day as influential. The possibility of artifacts in

3 t f

0

$

I

I 2 STARTING POSITION

1

3

FIG.15. Effect of drugs on alternation behavior in rats. Taken from Sinha et al. (1958).

the data caused by position habits among these rats is more serious as it is not possible to reduce their effect to chance level in the manner described above. Moreover, by our scoring, a rat which always turned to the same side would receive the maximum alternation score of eight, since it would in each case be forced to the opposite side on the intermediate trials. Examination of the data, however, showed that no rat in this group responded consistently in this way.4 Results are shown in Fig. 15. It will be seen that in general 4

Data in this paragraph from Sinha et al. ( 1958).

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

379

there is an increase in alternation behavior as a function of the number of preceding turns, and it will also be seen that alternation behavior is increased under ( A ) and abolished under Meratran. Analysis of variance showed these effects to be highly significant. It may be concluded that the prediction of the drug postulate is verified. Many fairly obvious experiments suggest themselves for use with human subjects as a consequence of the demonstration that alternation behavior is predictably affected by stimulant and depressant drugs. Almost any situation in which a choice has to be made between two or more stimuli, and in which such choice is repeated fairly regularly, might be used in this connection. The choice might be physical, as in the choice of one of different paths, or in the choice of one of different foods at meal time; or it might be perceptual, as in the choice of one of several pictures to look at, or one of several films to see. There has been a good deal of controversy about the perceptual and motor theories of alternation behavior (House, 1956), but it would seem that the generalization mediating our prediction would equally well embrace both types of behavior.

VIII. A

General Dimensional System of Psychopharmacology

In the preceding sections, an attempt has been made to set forth and to illustrate one particular hypothesis of psychopharmacological action, and to work out the paradigms appropriate to the use of the hypothetico-deductive method in this context. It should not be thought, however, that the general principles and methods here suggested are restricted to the dimension of extraversion-introversion, or the effects of stimulant and depressant drugs. The method is a perfectly general one, and it applies in principle to all dimensions of personality (Eysenck, 1947), and to those groups of drugs affecting a person’s position on these dimensions. The only limitation at present appears to be that a proper dimensional analysis of the relevant personality traits must have been carried out before work in the psychopharmacological field can be commenced; this of necessity means that only a relatively small number of dimensions can at present be studied in this manner ( Eysenck, 1959d) .

380

H. J. EYSENCK

Apart from intelligence and the general field of cognitive abilities, the only two dimensions which have been identified sufficiently well to furnish the necessary substratum for work of this kind are neuroticism or general emotional instability, and psychoticism. The evidence regarding these factors has been discussed elsewhere by the writer (Eysenck, 1959d, 1960), and we need only state here that these two factors appear to be orthogonal to each other and to extraversion-introversion ( Eysenck, 1957b; Eysenck, S. B. G., 1956); intellectual ability is largely independent of all three personality factors. It is likely that drugs which stimulate the sympathetic division of the autonomic nervous system, and which inhibit the parasympathetic division, increase neuroticism, while drugs which inhibit the sympathetic division and stimulate the parasympathetic division are likely to decrease neuroticism ( Trouton and Eysenck, 1960); there is too little systematic evidence to make it possible at the moment to arrive at any clear answer. The possibility does exist, however, of finding such an answer by applying the methods outlined in this chapter to the large group of tests known to measure neuroticism (Eysenck, 1959d; Eysenck, 1952b; Eysenck et al., 1957b). An attempt to carry out such a project would be likely to advance our knowledge of both neuroticism and of drug action to a considerable extent. As regards psychoticism, the position is even less satisfactory than with respect to neuroticism. A discussion can perhaps begin by quoting some comments made by Freyhan (1959), apropos of suggestions made by various speakers at a psychopharmacological conference, that certain drugs had “antischizophrenic and/or antipsychotic effects.” Freyhan said : “I question the legitimacy of these terms. It does not make clinical sense to say ‘antischizophrenic’ when schizophrenic must include great varieties of syndromes such as muteness, explosive aggressiveness, grandiose omnipotence or cosmic despair. Are we to assume that all these typical syndromes are to be acted upon by the same drug in the same manner with the same effectiveness? And I wonder about the applicability of the formulation ‘more psychotic’ in reference to a patient who, under the influence of a stimulant drug, changed from inconspicuous to disturbed or disturbing behavior. Was this patient ‘less psychotic’ when he sat quietly in a corner, surrendering to bizarre fantasies

TESTS IX THE ASSESSMENT OF DRUG EFFECTS

381

which he kept to himself? Neither psychosis nor schizophrenia denote molecular units which can be quantitatively analyzed or, indeed, be construed as therapeutic targets per se.” While recognizing the truth of the criticism as applied to much current work, the conclusion expressed in the final sentence cannot be supported by available evidence. The question of whether there is one particular general factor running through all psychotic disorders is not one which permits of being answered by means of simple clinical observation; it requires an answer in terms of properly planned experiments. Such evidence as there is suggests that there does exist a continuum from normality to psychotic disorder, and that this may properly be called psychoticism; it appears furthermore that objective measures of this continuum can be constructed and used with sufficient precision to make possible the assessment of “antipsychotic” effects of any particular drug (Eysenck, 1952a,b). It also appears that there does exist some functional unity in the concept of schizophrenia, apart from the general notion of psychotiaism, and that this also is measurable by means of objective tests (Payne and Hewlett, 1960). While these results are by no means definitive, they do suggest that experiments aimed at the demonstration of antipsychotic or antischizophrenic effects of certain drugs may be by no means premature. Let us apply these considerations to another point made by Freyhan. He maintains that “drug therapy must aim at target symptoms. These symptoms, regardless of diagnostic entities, can be fairly specific and are often mutually exclusive.” This statement is true in its second part, but leaves out the important fact that each patient has a position not only on one dimension ( psychoticism, say) but also on all other dimensions of personality; it is his position on these other dimensions which may determine the visible symptom. However, drug treatment is concerned with changes in the underlying causes of his psychoticism; the specific symptoms are merely incidental and may be presumed to disappear once the correct general antipsychotic drug has been applied. Such considerations would of course not apply to drugs which are only meant to treat the symptom itself; they might “cure”’ specific symptoms in one psychotic patient and exacerbate them in another. This would indicate that we were dealing with a drug affecting a dimension of personality orthogonal to psychoticism, but interacting with it in producing the symptom in question.

382

H. J. EYSENCK

These points have been made in some detail because they are vital to a proper understanding of the consequences of introducing the dimensional system of personality description into the psychopharmacological field. Symptoms are often determined by several dimensions at once; they are therefore subject to all the difficulties discussed in the second section of this paper. Anxiety is perhaps the outstanding example of such a symptom; it has loadings both on the factors of neuroticism and of introversion (Eysenck, 1957b, 1959d). Changes in anxiety due to the administration of a drug cannot therefore be interpreted in themselves; the drug may affect introversion, neuroticism, or both. A factorial design alone can give us an answer to this question. IX. Summary

In this review the writer has tried to summarize recent work carried out on a specific theory of psychopharmacological action, and to indicate how the methodology developed in order to deal with the problems arising in this attempt could be generalized to solve similar problems in other areas. It was pointed out that most present-day studies, however perfect they might be technically, yet fall short of making possible a psychological interpretation of the results because they do not take into account ( a ) the multifactorial nature of the score on any given test, or ( b ) the determination of performance by several different and distinct factors (drive, habit, inhibition, etc.). It was then pointed out that these difficulties could only be overcome by means of the formulation of theories and hypotheses linking drug action with established personality dimensions, and by the use of certain advanced statistical techniques (factor analysis, discriminant function analysis, canonical variate analysis. ) A specific hypothesis was stated, to the effect that depressant drugs increase cortical inhibition and decrease cortical excitation, while stimulant drugs have an opposite effect. Twelve deductions from this hypothesis were made, and experiments confirming or infirming these deductions discussed. It was concluded that on the whole the results were in line with the hypothesis. Results from an alternative experimental paradigm were also found to support the hypothesis.

TESTS IN THE ASSESSMENT OF DRUG EFFECTS

383

In the final section, an attempt was made to extend the general approach to other dimensions of personality, and to show that psychiatric symptoms require to be dealt with in the same manner as do scores on psychological tests, and that they are subject to the same difficulties and complexities. It was concluded that only by introducing the concepts of dimensional analysis from the field of personality research into psychopharmacology could the many discrepancies and contradictions so rife at present be reconciled. REFERENCES Bidwell, S. (1897). Proc. Roy. SOC.,B61, 268. Burstein, E., and Dorfmann, D. (1959). J. Psychol. (London)47, 81. Costello, C. G. (1960a). J. Mental Sci. 322. Costello, C. G. (196Ob). J. Mental Sci. 326. Eysenck, H. J. (1947). “Dimensions of Personality.” Routledge & Kegan Paul, London. Eysenck, H. J. (1952a). J. Pem. 20, 345. Eysenck, H. J. (1952b). “The Scientific Study of Personality.’’ Routledge and Kegan Paul, London. Eysenck, H. J. (1957a). J. Mental Sci. 103, 119. Eysenck, H. J. (1957b). “Dynamics of Anxiety and Hysteria.” Routledge & Kegan Paul, London. Eysenck, H. J. (1958). Personality tests: 1950-55. In “Recent Progress in Psychiatry” (Fleming, G. W. T. H., ed.), pp. 118-159. J. A. Churchill, London. Eysenck, H. J. ( 1959a). “Manual, Maudsley Personality Inventory.” Univ. of London Press, London. Eysenck, H. J. (195913). “Behaviour Therapy and the Neuroses.” Pergamon Press, New York. Eysenck, H. J. ( 1 9 5 9 ~ ) .In “Drugs and Behavior” (J. G. Miller, ed.). Wiley, New York. In Press. Eysenck, H. J. (1959d). “The Structure of Human Personality” (2nd ed.). Methuen, London. Eysenck, H. J. (1959e). Brit. J. Psychol. 360. Eysenck, H. J. (ed.). (1960). “Experiments in Personality.” Routledge & Kegan Paul, London. Eysenck, H. J. and Aiba, S. (1957). J. Mental Sci. 103, 661. Eysenck, H. J., and Eysenck, S. B. G . (1960). In “Experiments in Personality” (H. J. Eysenck, ed. ). Routledge & Kegan Paul, London. Eysenck, H. J., Casey, S., and Trouton, D. S. (1957a). J. Mental Sci. 103, 645. Eysenck, H. J., Granger, G., and Brengelmann, J. (1957b). “Perceptual Processes and Mental Illness.” Chapman & Hall, London. Eysenck, H. J., Holland, H. and Trouton, D. S. ( 1 9 5 7 ~ ) .J. Mental Sci. 103, 650. Eysenck, S. B. G . (1956). J. Mental Sci. 102, 517.

384

H. J. EYSENCK

Felsinger, J. H., Lasagna, L., and Beecher, H. K. (1953). J , Phurmacol. Exptl. Therap. 109, 284. Fleishman, E. A. (1956). Air Force H u m n Eng., Personnel, Training Research, Tech. Rept. 66-8,142. Fleishman, E. A., and Hempel, W. E. (1954). Psychometrika, 19, 239. Fleishman, E. A., and Hempel, W. E. (1955). J . Exptl. Psychol. 49, 301. Franks, C. M., and Laverty, S. G. (1955). J . Mental Sci. 101, 654. Franks, C. M., and Trouton, D. (1958). J . Compand PhysioE. Psychol. 51, 220. Freyhan, F. A. (1959). “Biological Psychiatry,” p. 323. Grune and Stratton, New York. Granit, R. ( 1955). “Receptors and Sensory Perception.” Yale Univ. Press, New Haven, Connecticut. Hauty, G. T., and Payne, R. B. (1955). J . Exptl. Psychol. 49, 60. Holland, H. ( 1960). In “Experiments in Personality” (H. J. Eysenck, ed. ) , Routledge and Kegan Paul, London. House, J. B. (1956). J . Genet. Psychol. 89, 183. Hull, C. L., Hovland, C. I., Ross, R. T., Hall, M., Perkins, D. T., and Fitch, F. B. ( 1940). “Mathematico-deductive Theory of Rote Learning.” Yale Univ. Press, New Haven, Connecticut. Laverty, S . G., and Franks, C. M. (1956). J . Neurol. Nsurowrg. Psychiat. 19, 137. McEwen, P. (1958). Brit. J . Psychol. Monogr. Suppl. 31, 1-92. Mackworth, H. ( 1948). M.R.C. Special Report No. 268. Martin, I. ( 1960). In “Experiments in Personality.” (H. J. Eysenck, ed. ), Routledge & Kegan Paul, London. Payne, R. B., and Hauty, G. T. (1953). U.S.A.F. Sch. Aviat. Med. Progr. Rept. No. 21. Payne, R. B. and Hauty, G. T. (1955). J . Exptl. Psychol. SO, 343. Payne, R. W., and Hewlett, J. H. G. (1960). In “Experiments in Personality.” (H. J. Eysenck, ed.), Routledge & Kegan Paul, London. Poser, E. G. (1958). Am. Psychol. 13, 334. Shagass, C . ( 1954). Ebctroencephabg. and Clin. Neurophysiol. 6, 221. Shagass, C. (1956). Psychosomat. Med. 18, 410. Shagass, C., and Naiman, J. (1955). A.M.A. Arch. Neurol. Psychiat. 34, 397. Shagass, C., and Naiman, J. (1956). I. Psychosom. Research 1, 49. Sinha, S. N., Franks, C. M., and Broadhurst, P. L. (1958). J. Exptl. Psychol. 66, 349. Slater, P. ( 1960). “Experiments in Personality” (H. J. Eysenck, ed.) . Routledge & Kegan Paul, London. Summerfield, A., and Steinberg, H. (1957). Quart. J . Psychol. 9, 146. Taylor, J. (1951). J . Erptl. Psychol. 41, 81. Treadwell, E. (1960). In “Experiments in Personality” (H. J. Eysenck, ed.). Routledge & Kegan Paul, London. Trouton, D. S., and Eysenck, H. J. (1960). In “Handbook of Abnormal Psychology” (H. J. Eysenck, ed. ), Pitmans, London. Willett, R. A. ( 1960). In “Experiments in Personality” (H. J. Eysenck, ed. ) , Routledge & Kegan Paul, London. Zeaman, D. and House, B. J. (1951). J. Exptl. Psychol. 41, 177.

AUTHOR INDEX Numbers in italic indicate the pages on which the references are listed.

A Abbie, A. A., 9, 14, 38 Abdulian, D. H., 308, 332 Abrahams, V. C., 193, 221 Adams, R. D., 30, 40 Ader, R., 233, 236, 266, 267, 272 Adey, W. R., 9, 14, 38, 324, 327 Adkins, F. J., 139, 172 Aghajanian, G. K., 263, 274 Agranoff, B., 282, 332 Aiba, S., 362, 383 Aird, R. B., 101, 134 Akert, K., 10, 14, 38 Albe-Fessard, D., 139, 141, 173 Albers, R. W., 286, 290, 293, 317, 327, 330, 331 Alcaino, F., 50, 55, 56, 57, 59, 60, 70, 75 Alexander, L., 246, 248, 255, 272 Alexandrowicz, J. S., 319, 327 Ambrus, C. M., 235, 253, 272 Ambrus, J. L., 235, 253, 272 Amin, A. H., 156, 164, 165, 171, 176, 177, 185, 187, 205, 221 Anderson, J., 127, 135 Andrew, A. M., 9, 38 Aoyama, T., 282, 330 Aprison, M. H., 177, 185, 223 Apter, J. T., 10, 14, 38 Arbit, J., 264, 272 Arey, L. B., 5, 6, 38 Artom, C., 101, 133 Ashby, W., 303, 327 Ashton, D. M., 332 Auld, F., 264, 272 Axelrod, J., 169, 171, 173, 255, 277 Azioka, I., 191, 202, 221

B Bachtold, H. P., 145, 173 Battig, K., 231, 251, 252, 265, 268, 272, 274

BaiIey, C. J., 251, 272 Bain, J. A., 209, 224, 300, 308, 309, 327, 329 Bakay, L., 101, 133 Baker, R. V., 203, 221 BaIzer, H., 186, 187, 202, 205, 221, 226 Bardone, M. C., 250, 275 Barelare, B., Jr., 42, 76 Barlow, H. B., 5, 9, 38 Bamard, G. L., 139, 172 Barrett, W. E., 181, 221, 225 Barsky, J., 194, 227 Barsoum, G. S., 159, 171 Bauer, R. O., 235, 253, 272 Baugniet, V., 248, 275 Baxter, C. F., 280, 283, 287, 289, 291, 294, 295, 296, 299, 300, 308, 309, 310, 311, 315, 318, 326, 327, 328, 330, 331 Bazemore, A. W., 321, 327 Beckett, S., 139, 171 Beecher, H. K., 365, 384 Beelen, L., 248, 275 Beers, D. N., 13, 38, 40 Beersaerts, J., 248, 275 Beerstecher, E., Jr., 43, 45, 52, 56, 74, 76 Belleville, R. E., 72, 75, 232, 253, 274 Beloff-Chain, A,, 286, 327 Benfey, B. G., 110, 135 Bennett, H. S., 131, 133 Berger, F. M., 249, 271, 272, 272, 273 Berl, S., 282, 327 Berman, E. R., 194, 227 Bemstein, J., 147, 173 Berry, H., 56, 74 Berry, J. F., 102, 116, 135 Berry, L. J.. 43, 45, 52, 5B, 74, 76 Bertler, A., 181, 187, 207, 211, 221, 222

385

386

AUTHOR INDEX

Besendorf, H., 181, 225, 245, 254, 273, 276 Besendorf, J., 145, 173 Bessman, S. P., 288, 327 Beveridge, J. M. R., 101, 133 Beyer, N., 248, 273 Bickford, R. C., 287, 327 Bidwell, S., 362, 383 Biederman, S., 3, 38 Billa, B., 202, 221 Binaghi, R., 302, 329 Birks, R. I., 85, 86, 89, 92, 96 Bimbaum, S. M., 287, 328 Bishop, G. H., 5, 8, 38, 175, 192, 204, 221, 222 Blaschko, H., 194, 203, 222 Block, M. A., 287, 328 Bloor, W. R., 101, 134 Boell, E. J., 31, 36, 38, 39 Boelter, M. D. D., 101, 134 Bogdanski, D. F., 164, 171, 176, 177, 178, 180, 185, 188, 193, 194, 205, 209, 222, 226, 308, 332 Bois, P., 267, 275 Boistel, J., 322, 323, 327 Bollman, J, L., 287, 327, 328 Bone, A. D., 203, 226 Boren, J. J., 236, 248, 273 Borison, H. L., 80, 83, 91, 96 Born, G. V. R., 203, 204, 222 Bovet, D., 82, 84, 95, 96, 97, 256, 273 Bovet-Nitti, F., 95, 96 Boyle, P. I., 122, 126, 134 Bradley, P. B., 271, 273 B d y , J. V., 230, 264, 270, 272,273 Brady, R. A., 52, 56, 61, 74 Brady, R. O., 290, 327 Brazier, M. A. B., 230, 273, 324, 327 Bregoff, H. M., 288, 330 Brengelmann, J., 380, 383 Brenner, C., 138, 139, 172 Breusch, R. S., 5, 38 Bridger, W. H., 232, 263, 268, 273 Broadhurst,J'. L., 376, 377, 378, 384 Brockman, J. A., Jr., 282, 327 Brodie, B. B., 145, 164, 171, 172, 173,

176, 181, 182, 200, 202, 204, 205, 207, 208, 209, 211, 222, 224, 225, 226, 254, 273 Brooks, V. G., 87, 88, 96 Brown, R. G., 67, 70, 74 Brown, W. D., 44, 52, 56, 74 Brunkow, I., 240, 250, 276 Buchanan, J. M., 287, 329 Buchwald, N. A., 323, 328 Bucy, P. C., 138, 172 Biilbring, E., 178, 180, 222 Bull, D. M., 204, 220, 224 Bullock, T. H., 325, 327 Burke, J. C., 200, 225, 239, 273 Burson, S. L., 282, 327 Burstein, E., 351, 383 Buser, P., 8, 38, 139, 141, 173 Buzard, J. A., 189, 222

C Cabieses, F., 304, 331 Caldwell, P. C., 126, 134, 301, 327 Callan, D. A., 295, 330 Campbell, K. M., 42, 76 Campbell, W. W., 101, 134 Canham, R. G., 180, 181, 193, 223 Carlsson, A., 161, 172, 181, 203, 208, 211, 222, 226 Carver, J. W., 67, 74 Case, T. J., 232, 257, 258, 274 Casey, S., 352, 353, 383 Catanzaro, R., 286, 327 Cavallini, G., 95, 96 Cerletti, A., 149, 172, 233, 239, 276 Chace, R. R., 11, 13, 40 Chaikoff, I. L., 101, 134, 136 Chain, E. B., 286, 327 Chamorro, A., 140, 141, 142, 144, 150, 172 Chang, H. C., 207, 222 Chanin, M., 281, 331 Channon, H. J., 118, 134 Cherkas, M. S., 194, 227 Chessin, M., 208, 222

302,

202, 207,

242,

143,

387

AUTHOR INDEX

Chibnall, A. C., 118, 134 Chusid, J. G., 322, 329 Ciotti, hl. M., 314, 329 Clare, M. H., 204, 222 Clark, A. M., 302, 331 Clark, C. T., 187, 189, 205, 220, 222, 226, 308, 327 Clark, J. R., 236, 256, 277 Clark, W. G., 331 Clink, D. W., 233, 236, 266, 267, 272 Clouet, D. H., 282, 328 Code, C. F., 165, 172 Cohn, W. E., 101, 134 Coleman, E. W., 93, 96 Conger, J. J., 252, 273 Consden, R., 282, 328 Conway, E. J., 122, 126, 134 Cook, L., 199, 212, 226, 231, 232, 235, 237, 239, 242, 244, 247, 248, 249, 250, 252, 253, 261, 262, 263, 265, 268, 273, 277 Corne, S. J., 194, 202, 219, 222 Correale, P., 177, 205, 223 Costa, E., 95, 96, 177, 178, 180, 181, 182, 184, 185, 193, 194, 195, 198, 204, 213, 215, 223, 225 Costello, C. G., 356, 360, 383 Cotzias, G. C., 202, 223 Coursin, D. B., 317, 318, 328 Courvoisier, S., 231, 234, 235, 237, 239, 240, 250, 253, 263, 267, 268, 269, 271, 273, 275 Cravioto, R. O., 286, 328 Crawford, T. B. B., 156, 164, 165, 171, 176, 177, 185, 187, 205, 221 Crelin, E. S., 33, 38 Crema, A., 180, 222 Cummins, J., 293, 331 Curran, P., 130, 136 Curtis, D. R., 281, 322, 323, 325, 328

D Dasgupta, S. R., 235, 274, 309, 310, 329 Dauby, J., 248, 275 Davies, R. E., 112, 134 Davison, A. N., 187, 194, 195, 198, 211, 223

Davitz, J. R., 264, 273 Dawson, R. M. C., 100, 102, 105, 113, 114, 129, 134, 286, 328 Debiase, P. L., 240, 274 Debley, V. G., 207, 224 Deegan, J., 239, 247, 248, 273 del Castillo, J., 88, 96 Delore, P., 56, 74 Dember, W. N., 49, 73, 74 Dent, C. E., 282, 328 Desmedt, J. E., 90, 96 Detwiler, S. R., 25, 39 Deuel, H. J., 100, 134 Deutsch, W., 112, 134 Dews, P. B., 181, 225, 272, 273 Dinsmoor, J. A., 233, 273 D’Iorio, A., 203, 222 Dobric, V., 165, 173 Dole, V. P., 202, 223 Domeij, B., 148, 172 Domer, F. R., 93, 96 Domino, E. F., 232, 236, 249, 251, 253, 273 Dorfmann, D., 351, 383 Doty, R. W., 232, 237, 276 Doyle, W. L., 120, 134 Ducrot, R., 231, 234, 235, 237, 240, 250, 267, 273, 275 Duke-Elder, S., 4, 38 Dunlop, C. W., 304, 324, 327, 328 Dunn, A. L., 262, 264, 275 du Ruisseau, J. P., 287, 328 Dusser, De Barenne, J. G., 138, 172 du Vigneaud, V., 307, 308, 328, 329

E Eade, N. R., 203, 222 Earl, A. E., 243, 276 Eccles, J. C., 30, 39, 108, 115, 134, 176, 191, 192, 223 Eccles, R. M., 30, 39 Edwards, C., 280, 295, 320, 321, 323, 328, 330 Edwards, R. E., 253, 273 Eggleston, L. V., 111, 135 Eggman, L. D., 110, 134 Ehrenthiel, 0. F., 166, 172

388

AUTHOR INDEX

Eidelberg, E., 283, 300, 304, 311, 318, 323, 327, 328, 331 Eiduson, S., 187, 227 Eiseman, B., 287, 328 Eisenman, A. J., 72, 75 Elliott, K. A. C., 280, 284, 295, 311, 321,322, 324, 325,327,328, 332 Ellison, F. S., 34, 40 Eltherington, L. G., 198, 224 Emerson, G. A., 44, 66, 67, 68, 70, 74, 75, 76 Erspamer, V., 176, 177, 189, 203, 207, 223 Essig, C. F., 139, 172 Euler, U. S . v., 148, 158, 159, 172 Evans, R. L., 329 Eyerman, G. S., 289, 292, 330 Eysenck, H. J., 334, 340, 343, 344, 346, 347, 348, 350, 352, 353, 354, 355, 358, 359, 360, 361, 362, 370, 371, 379, 380, 381, 382, 383 Eysenck, S. B. G., 371, 380, 383 Eyzaguirre, C., 320, 330 Ezoe, T., 326, 331

F Faber, J., 31, 39 Fadely, D., 236, 256, 277 Fanchamps, A., 242, 276 Farah, A., 86, 96 Farrell, M. A., 94, 96 Farthing, T. E., 13, 40 Fatt, P., 176, 190, 223, 322, 323, 327 Faure, M., 106, 134 Feldberg, W. S., 176, 223 Fellows, E. J., 193, 199, 212, 226, 239, 261, 273, 277 Felsing, B., 57, 76 Felsinger, J. H., 365, 384 Ferrari, W., 95, 96 Ferreira-Berrutti, P., 24, 39 Findlay, M., 101, 134 Finger, K. F., 181, 182, 222, 224 Finocchio, D., 181, 225 Fischer, E. H., 306, 328 Fitch, F. B., 346, 384 Flataker, L., 231, 232, 234, 237, 245, 252, 261, 262, 263, 277

Fleishman, E. A., 335, 336, 337, 384 Flock, E. V., 287, 327, 328 Florey, E., 177, 223, 280, 282, 289, 321, 322,327, 328, 329, 332 Florey, E., 177, 223, 280, 282, 283, 289, 320, 321, 322, 327, 328, 329, 332 Folch, J,, 100, 106, 134 Forsander, O., 45, 66, 73, 75 Fourneau, E., 256, 273 Foumel, J., 231, 234, 235, 250, 273, 275 Fouts, J. R., 194, 227 Frankel, S., 282, 283, 289, 291, 296, 311, 316, 317, 330, 331 Franks, C . M., 258, 273, 348, 349, 350, 376, 377, 378, 384 Frazer, F., 72, 75 Freedman, D. X., 263, 274 Freinkel, N., 110, 134 French, J. D., 300, 311, 328 Freter, K., 223 Freyhan, F. A., 380, 384 Friedkin, M., 101, 134 Friedman, P. J., 194, 222 Fries, B. A., 101, 134, 136 Fulcher, J. H., Jr., 271, 274 Funderburk, W. H., 232, 257, 258, 274 Furshpan, E. J., 176, 223, 319, 320, 329, 332

G Gaddum, J. H., 156, 159, 164, 165, 171, 178, 177, 182, 185, 186, 187, 188, 190, 203,205,207, 221,223 Gaitonde, M. K., 282, 328 Gale, E. F., 307, 329 Gallagher, W. J., 271, 274 Gantt, W. H., 232, 236, 243, 253, 283, 268, 269, 273, 274, 276 Garattini, S., 177, 182, 223 Gardiner, J. E., 87, 89, 96 Garven, I., 164, 172 Garven, J. D., 194, 195, 205, 207, 212, 215, 223 Garzoli, R. F., 303, 327

389

AUTHOR INDEX

Gatti, G. L., 231, 239, 240, 243, 247, 248, 249, 253, 268, 274 Gaunt, R., 243, 276 Gaze, R. M., 5, 7, 8, 9, 22, 25, 27, 31, 35, 39 Geller, E., 187, 227 Geller, I., 270, 273 Gellhom, E., 139, 171, 231, 255, 264, 274, 275 Gerard, R. W., 112, 134, 255, 275 Gertner, S. B., 193, 223 Gesler, R. M., 94, 96 Gey, K. F., 181, 198, 202, 225, 254, 276 Giarman, N. J., 177, 182, 185, 186, 188, 189, 190, 223,225,258, 276 Gillson, R. E., 204, 222 Ginzel, K. H., 180, 223 Giovinco, J. F., 86, 87, 89, 92, 96 Girardo, M., 284, 295, 311, 324, 330 Gliedman, L. H., 232, 236, 243, 269, 274 Gosswald, R., 239, 240, 255, 267, 277 Goldberg, L. I., 193, 223 Goldring, S., 324, 329, 332 Goldstein, L., 141, 172 Goldstein, S. G., 180, 181, 193, 202, 223 Gomez, J. A., 284, 311, 330 Goodman, L. S., 304, 330 Gordon, A. H., 282, 328 Govier, W. M., 240, 274 Graham, J. D. P., 149, 172, 194, 202, 219, 222 Grandjean, E., 231, 243, 245, 251, 252, 261, 265, 268, 272, 274 Granger, G., 380, 383 Granit, R., 6, 39, 363, 384 Grant, G. A., 93, 96 Gray, W. D., 304, 329 Greenberg, D. M., 101, 134 Greenberg, L. A., 51, 59, 60, 75 Greenfield, P., 31, 36, 38, 39 Greenstein, J . P., 287, 328 Grindlay, J. H., 287, 327, 328 Grundfest, H., 175, 204, 223, 225, 280, 295, 324, 329, 330

Guha, G., 235, 274 Guth, L., 281, 283, 284, 285,295, 311, 331 Guyenot, E., 34, 39 Gyermek, L., 182, 223

H Hagen, J. M., 127, 134, 203, 222 Hagen, P., 203, 222 Hagiwara, S., 176, 224 Hahn, L., 101, 134 Haley, T. J., 207, 224 Hall, M., 346, 384 Halpern, B. N., 302, 329 Hame, B. J., 384 Hanahan, D. J., 106, 134 Hanzon, V., 121, 134 Hardisty, R. M . , 203, 224 Harkness, W. D., 69, 75 Harman, P. J,, 283, 330, 331 Harper, A. A., 109, 112, 134 Harris, G. W., 166, 172 Harrison, J. M., 233, 274 Hart, E. R., 204, 225 Hartline, H. K., 9, 34, 39 Hartman, S. C., 287, 329 Hassert, G. L., 239, 273 Hastings, A. B., 303, 332 Hauty, G. T., 336, 353, 384 Hawes, R., 194, 222 Hawkes, C. D., 42, 76 Hawkins, J., 202, 224 Hayashi, T., 323, 329 Healy, S. T., 232, 237, 254, 257, 258, 259, 267, 274, 275 Hebb, C . O., 87, 96, 107, 134, 176, 203, 224 Hederra, A., 43, 45, 47, 50, 54, 55, 56, 57, 59, 60, 62, 70, 75 Hegsted, D. M., 44, 48, 76 Hempel, W. E., 335, 384 Hems, R., 111, 135 Henderson, J. K., 255, 256, 276 Herrick, C. J., 5, 6, 23, 33, 37, 39 Herz, A,, 236, 247, 254, 259, 268, 269, 274

390

AUTHOR INDEX

Hess, E., 274 Hess, G., 247, 249, 274 Hess, S., 176, 189, 198, 224 Hevesy, G., 101, 134 Hewlett, J. H. G., 381, 384 Hiestand, W. A., 66, 75 High, J. P., 239, 273 Hill, H. E., 232, 253, 274 Hill, R. J., 294, 332 Hillarp, N. A,, 181, 203, 222, 224 Himwich, H. E., 180, 181, 182, 184, 193, 194, 195, 198, 202, 208, 213, 215, 223, 224, 225 Himwich, W. A., 180, 181, 193, 194, 195, 198,202, 215, 223,224 Hirano, S . , 286, 332 Hobbiger, F., 325, 328 Hodgkin, A. L., 122, 126, 131, 132, 134, 135, 175, 224 Horlein, U., 239, 240, 255, 267, 277 Hoffmann, I., 231, 232, 233, 239, 240, 251, 266, 270, 275, 276 Hoffmaster, T., 232, 234, 275 Hogben, C. A., 122, 135 Hogberg, B., 203, 224 Hokin, L. E., 101, 102, 103, 105, 106, 107, 108, 109, 110, 112, 113, 114, 116, 118, 119, 121, 124, 125, 127, 128, 129, 131,134, 135, 136 Hokin, M. R., 101, 102, 103, 105, 106, 107, 108, 109, 110, 112, 113, 116, 118, 119, 121, 124, 125, 127, 128, 129, 131, 135 Holland, H., 357, 361, 383 Holt, L. E., 42, 76 Holten, C. H., 235, 246, 247, 251, 252, 254, 260, 274 Holtz, P., 186, 203, 221, 224, 307, 329 Holzbauer, M., 181, 224, 254, 274 Hooker, D., 25, 39 Hoppe, J. O., 94, 96 Horisberger, B., 231, 243, 245, 261, 274 Horita, A,, 198, 205, 224 Homykiewicz, O., 203, 204, 222

Horvath, A., 307, 308, 328, 329 House, J. B., 379, 384 House, 378, 384 Hovland, C. I., 346, 384 Huang, S. H., 324, 329 Hughes, F. B., 181, 204,209, 224 Hughes, L. H., 233, 273 Hull, C. L., 346, 384 Humoller, F. L., 262, 264, 275 Humphrey, J. H., 195, 224 Hunt, A. D., Jr., 317, 329 Hunt, H. F., 250, 274 Hyden, H., 185, 224

I Iino, M., 325, 331 Ingram, G. I. C., 203, 222 Inscoe, J. K., 169, 171 Irrevere, F., 329 Irwin, S., 240, 267, 268, 269, 274 Isbell, H., 72, 75 Iwama, K., 324, 329 Izquierdo, J. J., 286, 328

J Jacobsen, E., 233, 235, 236, 243, 244, 246, 247, 249,252,270, 274,275 Jacobsohn, D., 166, 172 Jakoby, W. B., 282, 329, 331 Jaques, R., 195, 224 Jasmin, G., 267, 275 Jasper, H. H., 280, 295, 319, 324, 329 Jelinek, B., 281, 283, 284, 285, 287, 295, 311, 328, 331 Jenkins, R., 259, 276 Jenney, E. H., 104, 115, 136, 146, 173, 232, 237, 248, 249, 254, 257, 258, 259, 260, 262, 265, 267, 271, 274, 275, 276, 310, 329 Jepson, J. B., 189, 224 Jinnai, D., 280, 329 John, E. R., 231, 244, 245, 255, 258: 260,261, 269, 270, 271, 275 Johnson, M., 111, 135 Johnston, C. D., 69, 75

391

AUTHOR INDEX

Jones, G. T., 255, 256, 276 Jorgensen, C. B., 110, 136 Jorpes, E., 109, 135 Julou, J., 237, 240, 250, 263, 267, 268, 273, 275

K Kaada, B. R., 138, 172 Karki, N. T., 159, 182, 224 Kahlson, G., 166, 172 Kakimoto, T.,195, 211, 225 Kaplan, N. O., 314, 329 Karoly, A. J., 232, 236, 249, 251, 253, 273 KasB, Y.,80, 83, 91, 96 Kates, M., 122, 135 Kato, R., 177, 223 Kattau, R. W., 255, 256, 276 Katz, B., 88, 96 Keller, C. E., 255, 256, 276 Kennard, M. A., 138, 172 Kennedy, E. P., 100, 119, 135, 136 Kent, A. B., 306, 328 Kerman, E. F., 254, 275 Kessler, M., 231, 274, 275 Keynes, R. D., 122, 126, 131, 132, 134, 135 Khalidi, A. I., 149, 172 Kieser, W., 236, 244, 266, 271, 275 Killam, E. K., 208, 224, 309, 310, 329 Killam, K., 323, 325, 331 Killam, K. F., 208, 209, 224, 280, 300, 309, 310, 324, 326, 327,329 Kirschner, L. B., 130, 135 Kiyohara, K., 112, 135 Kleinknecht, F., 138, 172 Kling, A., 270, 273 Klupp, H., 236, 244, 266, 271, 275 Knell, J., 186, 187, 205, 226 Knoll, B., 251, 252, 265, 275 Knoll, J., 243, 245, 251, 252, 256, 265, 275 Kobinger, W., 182, 224 Koella, W. P., 204, 220, 224 KoeIle, G. B., 85, 96, 193, 221 Koeppe, R. E., 294, 332

Koetschet, P., 231, 234, 235, 273 Kohonen, J., 45, 66, 73, 75 Kojima, K., 283, 329 Koketsu, K., 176, 223 Kollros, J. J., 6, 7, 31, 39 Kolsky, M., 231, 234, 235, 273 Kopeloff, L. M., 322, 329 Kopeloff, N., 322, 329 Kornberg, A., 116, 120, 135 Kosman, M. E., 255, 275 Kosterlitz, H. W., 178, 224 Kottegoda, S. R., 180, 223 Kramer, E. R., 208, 222 Krebs, E. G., 306, 328 Krebs, H. A., 111, 135 Kreiskott, H., 239, 240, 255, 267, 277 Kremer, W. F., 139, 172 Kristofferson, A. B., 49, 73, 74 Kroger, H., 240, 250, 276 Kroneberg, G., 208, 224 Kubo, M., 326, 331 Kuchinskas, E. J., 307, 308, 328, 329 Kuffler, S. W., 280, 295, 320, 321, 323, 328, 329, 330 Kuhn, R., 215, 224 Kuntzman, R., 181,208,209,222,225

L Lagerstedt, S., 203, 224 Landsford, E. M., Jr., 57, 76 Langford, H. G., 138, 172 Larsell, O., 6, 39 Lasagna, L., 365, 384 Lasher, A. V., 94, 96 Lassen, N. A., 317, 331 Laverty, S. G., 258, 273, 348, 384 Lawrow, J., 58, 76 Layne, E. C., 288, 327 Leaf, A., 126, 127, 135 Leau, O., 250, 269, 273, 275 Lehninger, A. L., 101, 134 Leitch, J. L., 207, 224 Lembeck, F., 165, 172, 185, 224, 270, 275 Lennard-Jones, J. E., 185, 224 Lesser, E. J., 127, 135 Lester, D., 51, 59, 60, 75

392

AUTHOR INDEX

Lettvin, G. Y., 6, 35, 39 Levis, S., 248, 275 Lewis, G. P., 184, 204, 224 Lichtenheld, F. R., 287, 327 Lin, R. C. Y., 222 Lindberg, O., 101, 133 Lindqvist, M., 161, 172, 208, 222 Lionetti, F., 130, 136 Litchfield, J. T., Jr., 218, 225 Little, P. E., 93, 96 Loew, E. R., 195, 226 Loiseleur, J., 44, 75 Long, J. P., 78, 80, 82, 84, 86, 87, 89, 91, 92, 94, 96, 97 Longo, V. G., 82, 84, 96, 97 Lowe, 1. P., 281, 283, 284, 285, 289, 292, 295, 311, 330, 331 Lowry, 0. H., 303, 332 Lucas, R. A., 181, 225 Lucy, J. D., 166, 172

M Maas, A. R., 199, 225 McColl, J. D., 233, 248, 250, 276 hlccrory, W. W., 317, 329 McCulloch, W. S., 6, 35, 39, 139, 173 McEwen, P., 355, 384 McGrath, W. R., 198, 224 Macht, D. I., 232, 234, 275 MacIntosh, F. C., 85, 86, 89, 90, 92, 96 McIsaac, W. M., 193, 225 McIvor, R. A., 93, 96 McKay, I. F. S., 112, 134 McKhann, G. M., 294,317, 330,331 McKibbin, J. M., 106, 135 Mackworth, H., 364, 365, 384 MacLean, P. D., 177, 225 McLennan, H., 282, 283, 321, 329, 330 McMurray, V. M., 24, 39 McMurray, W. C., 102, 116, 135 McNair, M. B., 170, 173 McPhail, M. K., 80, 97 MacPhillamy, H. B., 181, 221 Madison, J. B., 92, 93, 97 Mafii, G., 232, 235, 239, 243, 247,

248, 249, 250, 251, 252, 261, 263, 265, 275 Magni, F., 30, 39 Magnusson, R., 208, 222 Magnusson, T., 161, 172 Mahler, D. J., 262, 264, 275 Malhotra, C. L., 208, 225 Maling, H. M., 180, 198, 226 Malone, M. H., 200, 225 Mann, F. C., 287, 328 Mann, P. J. G., 115, 136 Mansi, W., 30, 39 Marchal, P., 247, 275 Mardones, J., 43, 44, 45, 47, 48, 50, 52, 53, 54, 55, 56, 57, 59, 60, 62, 64, 65, 66, 67, 70, 73, 75, 76 Maren, T. H., 304, 329 Marinetti, G. V., 102, 106, 136 Marini-Bettolo, G. B., 95, 96 Marrazzi, A. S., 204, 225 Marshall, E. K., Jr., 44, 75 Martin, A. J. P., 282, 328 Martin, A. R., 88, 97 Martin, I., 354, 384 Maryn, D., 207, 224 Masi, I., 286, 327 Masserani, E., 95, 96 Masserman, J. H., 233, 247, 251, 252, 275 Massieu, G., 286, 328 Matthey, R., 10, 12, 39 Mattis, P. A., 199, 212, 226, 231, 239, 247, 248, 261, 273, 277 Maturana, H. R., 5, 6, 12, 24, 35, 39 Maxwell, R. A., 181, 225 May, R. M., 25, 39 Maynard, D. M., 321, 330 Meckelnburg, K. L., 148, 173 Meier, R., 124, 136 Mellanby, J., 109, 136 Memtt, H. H., 138, 139, 172 Michelson, M. J., 260, 275 Mickelsen, O., 330 Milholland, R. J., 316, 331 Miller, F. R., 139, 172 Miller, N. E., 230, 251, 272, 275

393

AUTHOR INDEX

Miller, R. E., 233, 236, 237, 251, 265, 266, 267, 275 Millichap, J. G., 304, 330 Minatoya, H., 231, 274 Miner, N., 29, 39, 40 Minkowski, M., 138, 172 hlinz, B., 139, 140, 141, 142, 143, 144, 145, 147, 151, 153, 154, 155, 159, 162, 172, 173 Mirone, L., 45, 52, 56, 64, 65, 75 Mirsky, I. A., 233, 236, 237, 251, 265, 266, 267, 275 hfitoma, C., 295, 330 Miyazaki, M., 283, 329 Mizuno, K., 283, 329 Moller-Nielsen, I., 240, 255, 275 Moore, B. M., 66, 67, 68, 75, 76 Moore, S . , 282, 331 Moore, W. T., 44, 66, 67, 68, 70, 74, 76 Morelec-Coulon, J., 106, 134 Mori, A., 280, 329 Morris, R. W., 231, 273 Morton, M. E., 110, 136 Mott, J. C., 180, 225 Murayama, M. M., 101, 134 Murphree, H. B., Jr., 271, 275 Murphy, J. V., 233, 236, 237, 251, 266, 267, 275 Muscholl, E., 208, 225 Mutt, V., 109, 135

N Nachmansohn, D., 175, 225 Naess, K., 236, 248, 251, 275 Nagai, K., 323, 329 Nagata, Y., 286, 294, 332 Naiman, J., 367, 384 Nash, J. B., 44, 67, 68, 70, 74 Navarro, A. P., 248, 273 Neuhold, K., 240, 255, 275 Nichol, C. A., 293, 316, 331 Nichols, G., Jr., 302, 330 Nickerson, W. J., 308, 330 Nieschulz, O., 231, 233, 239, 240, 251, 266, 270, 275, 276 Nieuwkoop, P. D., 31, 39

Nilson, B., 203, 224 Nilsson, J., 181, 207, 211, 222 Nilsson, K., 194, 222 Nimmo, J. M., 199, 225 Noble, G. K., 3, 39 Noe, F. F., 308, 330 Noell, W. K., 235, 253, 272 Noguchi, F., 325, 331 Nytch, P. D., 189, 222

0 Okumura, N., 282, 330 OLeary, J. L., 324, 329, 332 Olley, J., 106, 134 Onfray, E., 43, 52, 53, 75 Orkand, R. K., 88, 97 Orleans, F. B., 181, 182, 222, 224 Ornitz, E . M., 263, 274 Osaki, H., 110, 136 Osmond, H., 169, 173 Osofsky, H., 287, 328 Ostfeld, A. R., 259, 276 Otsuki, S., 282, 330 Ottoson, D., 320, 328 Owen, J. E., Jr., 233, 236, 249, 251, 253, 254, 256, 265, 266, 267, 276, 277 Owens, A. H., Jr., 44, 75

P Paasonen, M. K., 159, 173, 177, 181, 182, 186, 189, 223, 225 Packer, L., 293, 331 Page, I. H., 83, 97, 176, 193, 203, 225, 226 Page, L. B., 127, 135 Paintal, A. S., 180, 225 Palade, G. E., 115, 117, 121, 136 Pasnau, R., 259, 276 Patterson, J. L., 138, 172 Payne, R. B., 336, 353, 381, 384 Pearse, A. G.E., 193, 225 Pelton, R. B., 57, 58, 68, 69, 70, 76 Pepeu, G., 258, 276 Pepler, W. J., 193, 225 Perkins, D. T., 346, 384 Perlman, I., 101, 136 Peters, J,, 299, 330

394

AUTHOR INDEX

Petit, M., 44, 75 Pfeiffer, C. C., 87, 92, 97, 104, 115, 136, 146, 173, 231, 235, 243, 244, 245, 248, 249, 251, 257, 258, 260, 262, 269, 271, 274, 275, 276, 310, 329 Phillis, J. W., 281, 322, 328 Piatt, J., 25, 39 Pickles, V. R., 184, 192, 225 Pinkston, J. O., 138, 173 Pisano, J., 295, 330 Pitts, W. H., 6, 35, 39 Pletscher, A., 145, 164, 171, 173, 181, 189, 198, 202, 205, 208, 211, 222, 225, 226, 245,254, 273,276 Plum, F., 304, 331 Plummer, A. J., 181, 221, 225, 243, 276 Pocchiari, F., 286, 327 Popendiker, K., 231, 232, 233, 239, 240, 251,266, 270, 275,276 Popper, L., 138, 173 Porter, R. R., 138, 172 Poser, E. G., 355, 384 Potter, D. D., 176, 223, 319, 329 Powers, M., 94, 97 Powers, T., 299, 330 Preat, S., 248, 275 Preston, J. B., 88, 97 Pricer, W. E., Jr., 116, 120, 135 Prieto, R., 62, 65, 75, 76 Prockop, D. J., 145, 173, 182, 205, 225, 226 Pscheidt, G. R., 182, 184, 223, 225 Pundlik, P. G., 208, 225 Purpura, D. P., 204, 225, 284, 295, 311, 324, 330

Q Quastel, J. H., 115, 136 Quinn, G. P., 181, 182, 222, 225

R Raper, H. S., 109, 112, 134 Rasmussen, E. W., 236, 248, 251, 275 Rathbun, R. C., 255, 256, 276 Ravel, J. M., 57, 76

Redfield, B. G., 188, 189, 198, 209, 223, 224, 226 Redfield, G. B., 308, 332 Redman, C. M., 110, 114, 121, 129, 131, 136 Reed, J. G., 52, 56, 74 Rehm, W. S., 122, 136 Reitzel, N. L., 80, 84, 86, 87, 89, 91, 97 Remond, A., 139, 173 Renshaw, A., 127, 135 Revzin, A. M., 193, 204, 225 Ricci, G. F., 204, 226 Rice, W. B., 233, 248,250,276 Richter, C. P., 42, 62, 65, 76 Richter, D., 282, 328 Riley, H., 230, 276 Rinaldi, F., 178, 180, 184, 208, 213, 223, 225 Rioch, D. Mck., 138, 173 Riopelle, A. J., 231, 235, 243, 244, 245, 248,249, 251, 269,271,276 Risse, K. H., 239, 240, 255, 267, 277 Robbins, E. B., 233, 236, 251, 253, 266, 276, 277 Robbins, J., 322, 330, 332 Roberts, E., 280, 281, 282, 283, 284, 285, 287, 288, 289, 291, 294, 295, 296, 299, 300, 307, 308, 309, 310, 315, 318, 317, 318, 326, 327, 328, 330, 331 Robins, E., 289, 292, 330 Robinson, J. A., 178, 224 Rochon-Duvigneaud, A., 4, 39 Rockhold, W. T., 66, 75 Rodnight, R., 207, 225 Rogers, L. L., 57, 58, 68, 69, 70, 76 Romanowski, W., 325, 331 Rosa, E. C., 170, 173 Rosen, F., 316, 331 Rosenberg, B., 64, 76 Rosenberg, T., 121, 136 Rosengren, E., 187, 221 Rosengren, S., 181, 207, 211, 222 Ross, R. T., 346, 384 Rossen, J., 288, 327 Rossiter, R. J., 102, 116, 135

AUTHOR INDEX

Rothstein, A,, 124, 136 Rothstein, M., 280, 287, 294, 295, 309, 310, 331 Roughton, F. J. W., 302, 331 Rozhkova, E. K., 260, 275 Rubin, B., 200, 225 Ruffo, A., 111, 135 Rutledge, L. T., 232, 237, 276 Rutledge, R. A,, 181, 221, 225

S Sacktor, B., 293, 331 Sacra, P., 233, 248, 250, 276 Saffran, M., 124, 135 Saldias, C. A., 300, 304, 311, 328, 331 Salvador, R. A., 290, 293, 327, 331 Sandberg, F., 242, 276 Sandler, M., 187, 189, 223, 225 Sano, I., 195, 204, 211, 225 Sarzana, G., 101, 133 Sastry, P. B., 85, 86, 89, 92, 96 Savateev, N. V., 260, 275 Schachner, H., 101, 134, 136 Schade, J. P., 291, 296, 327 Schally, A. V., 124, 135 Schlag, J., 247, 275 Schlosser, L., 165, 173, 176, 227 Schmid, K., 170, 173 Schmidt, E. C . H., Jr., 42, 76 Schmidt-Nielsen, K., 110, 136 Schneider, J. A., 180, 208, 225, 226, 243, 269, 270, 276 Schott, H. F., 307, 331 Schreiner, L., 270, 273 Schueler, F. W., 77, 78, 80, 82, 83, 84, 86, 89, 91, 92, 94, 96, 97 Schumann, H. J., 208, 224 Schumacher, H., 232, 233, 239, 266, 275 Schuster, E. M., 303, 327 Schwartz, J. R., 110, 136 Scott, C . S., 208, 222 Scott, E. M., 282, 329, 331 Segovia, N., 43, 45, 47, 53, 54, 55, 56, 62, 64, 65, 66, 67, 70, 75 Segovia-Riquelme, N., 44, 48, 50, 57, 59, 60, 75, 76

395

Segre, E., 101, 133 Senoh, S., 169, 171 Shadewald, M., 66, 67, 76 Shagass, C., 367, 384 Shamblin, J. R., 92, 97 Shanes, A. M., 190, 225 Shaw, E., 177, 185,227,261,276 Shaw, T., 204, 208, 224 Shelp, W. D., 106, 107, 108, 135 Shen, S. C., 31, 36, 38, 39 Sheppard, H., 182, 227 Sherwin, A. L., 110, 135 Shibata, N., 326, 331 Shimizu, S., 326, 331 Shive, W., 57, 76 Shore,.P. A., 145, 172, 173, 176, 180, 181, 182, 198, 200, 202, 204, 205, 207, 208, 209, 211, 222, 225, 226, 254, 273 Shuhara, R., 323, 329 Sidman, M., 233, 276 Siekevitz, P., 115, 117, 121, 136 Siever, P. W., 251, 275 Sigg, E. B., 208, 218, 226, 240, 276, 323, 325, 331 Silver, A., 107, 134 Silver, S. L., 181, 205, 226 Sinha, S. N., 376, 377, 378, 384 Sirnes, T. B., 69, 70, 72, 76 Sisken, B., 283, 289, 296, 299, 331 Sisson, G. M., 304, 329 Sivadjian, J., 255, 276 Skaarup, Y., 233, 236, 246, 247, 252, 274 Slabek, M., 240, 274 Sladen, W. J. L., 110, 136 Slater, I. H., 233, 236, 243, 249, 251, 253, 255, 256, 265, 276, 277 Slater, P., 370, 384 Sloane-Stanley, G. H., 100, 136 Smallman, B. N., 87, 96 Smart, P., 193, 221 Smith, C. M., 147, 162, 173 Smith, F. H., 304, 329 Smith, R. P., 231, 235, 243, 245, 248, 249, 251, 276, 310, 329

396

AUTHOR INDEX

Smith, S. W., 119, 136 Smith, T. G., 284, 295, 311, 330 Smythies, J. R., 204, 220, 224 Snell, E. E., 306, 331 Snow, P. J . D., 185, 189,224,225 Snyder, E. R., 306, 328 Sokoloff, L., 282, 317, 330, 331, 332 Solomon, A,, 130, 136 Sonne, E., 233, 235, 246, 247, 251, 252,253, 260, 270, 274, 275 Soumalainen, H., 45, 66, 73, 75 Sourkes, T. L., 186, 226 Spector, S., 145, 173, 180, 198, 202, 205, 226 Sperry, R. W., 3, 7, 11, 12, 15, 16, 17, 18, 20, 21, 24, 25, 26, 29, 33, 36, 39, 40 Spinks, A., 230, 276 Stacy, R. S., 203, 222, 224 Stafford, A., 203, 204, 222 Stavney, L. S., 304, 331 Steck, E. A., 94, 96 Stein, L., 276 Stein, M., 244, 265, 275 Stein, W. H., 282, 331 Steinberg, H., 352, 384 Steiner, E. C., 85, 96' Stemler, F. W., 66, 75 Stephens, J. H., 253, 269, 276 Stem, P., 165, 173 Still, E. U., 112, 134 Stille, G., 240, 250, 276 Stokes, J., Jr., 317, 329 Stone, C. A., 148,173, 195, 226 Stone, L. S., 12, 13, 17, 24, 33, 34, 40 Stotz, E., 102, 106, 136 Strickland, K. P., 101, 102, 116, 134, 135 Stroer, W. F. H., 7, 23, 40 Stroud, H. H., 317, 329 Studer, A,, 189, 211, 227 Sui, V. M., 261, 276 Sulser, F., 182, 222 Summerfield, A., 352, 384 Swanson, A. G., 304, 331 SzAkely, G., 29, 33, 40

T Taeschler, M., 233, 239, 242, 276 Takagaki, G., 286, 293, 294, 332 Takahashi, H., 325, 326, 331 Takayasu, T., 325, 331 Tallan, H. H., 282, 331 Taniguchi, K., 195, 211, 225 Tanimukai, H., 191, 202, 221 Tasaki, I., 176, 224 Taurog, A., 101, 136 Taylor, J., 352, 384 Taylor, R. D., 83, 97 Tedeschi, D. H., 193, 199, 212, 226, 261, 277 Tedeschi, R. E., 193, 199, 212, 226, 261, 277 Tennenbaum, M., 115, 136 Terzuolo, C. A., 323, 325, 331 Thies, R. E., 88, 96 Thompson, J. W., 180, 226 Tiba, M., 325, 331 Tinbergen, N., 3, 40 Tomchick, R., 173, 255, 277 Tomich, E. G., 181, 208, 209, 222, 226 Torchiana, M. L., 148, 173 Toschi, G., 121, 134 Tower, D. B., 280, 294, 301, 316, 317, 330, 331 Tracy, W. H., 233, 274 Treadwell, E., 365, 384 Trendelenburg, U., 180, 185, 192, 226 Trouton, D. S., 258, 273, 334, 348, 349, 350, 352, 353, 354, 361, 380, 383, 384 Trubey, R. H., 57, 76 Truitt, E. B., Jr., 242, 277 Tschirgi, R. D., 231, 244, 245, 255, 258,260,261,269,270, 271, 275 Tsukada, Y., 286, 293, 294, 326, 331, 332 Twarog, B. M., 176, 226

U Udenfriend, S., 164, 171, 176, 177, 178, 180, 185, 187, 188, 189, 193, 194, 195, 198, 205, 209, 211, 220,

AUTHOR INDEX

222, 223, 224, 226, 295, 308, 327, 330, 332 Umbreit, W. W., 302, 308, 332 Underwood, J., 149, 173 Unna, K. R., 259, 277 Ussher, N. T., 13, 40 Ussing, H. H., 122, 126, 136

V Valzelli, L., 177, 182,202, 221, 223 Vanasupa, P., 324, 332 Van de Kloot, W. G., 322, 330, 332 Vane, J. R., 189, 190, 192, 204, 207, 226 Vanecek, M., 236, 277 van Gelder, N. M., 284, 311, 328, 332 Van Harreveld, A., 281, 332 Van Maanen, E. F., 88, 97 Van Meter, W. G., 184, 223 Varela, A., 62, 65, 73, 75, 76 Vass, C. C. N., 109, 134 Vates, T., 282, 332 Verhave, T., 233, 236, 249, 251, 253, 254,256,265, 266, 267, 276,277 Vernier, V. G., 259, 277 Vitale, J. J., 44, 48, 76 Vogt, M., 162, 173, 181, 187, 208, 224,225, 226,254,274, 277 Vogt, W., 205, 226 Vonderahe, A., 299, 330 Von Euler, C., 204, 226 Votava, Z., 236, 277

W Waallces, T. P., 205, 226 Waddell, J. G., 308, 332 Waelsch, H., 282, 327 Wagman, A. I., 231, 235, 243, 245, 251, 276 Wagman, W., 231, 235, 243, 245, 251, 276 Waksman, B. H., 30, 40 Walaszek, E. J., 145, 147, 149, 151, 153, 154, 155, 159, 162, 172, 173 Waldeck, B., 161, 172 Walker, E. L., 232, 236, 249, 251, 253, 273

397

Wallace, W. M., 303, 332 Walls, G. L., 2, 3, 4, 40 Warner, L. H., 230, 277 Watkins, J. C., 281, 322, 323, 324, 328 Waton, N. G., 84, 92, 97 Waugh, M. H., 200, 225 Weidley, E. F., 231, 232, 235, 237, 239, 242, 244, 247, 248, 249, 250, 252, 253, 261, 262, 263, 265, 268, 273 Weil-Malherbe, H., 173, 203, 226, 255, 277 Weiskrantz, L., 233, 243, 244, -351, 271, 277 Weiss, P., 15, 24, 25, 29, 37, 40 Weiss, S. B., 119, 136 Weissbach, H., 176, 177, 178, 180, 185, 187, 188, 189, 193, 194, 195, 198, 205, 209, 211, 220, 222, 223, 224, 226, 308, 327, 332 Welsch, A. D., 222 Welsh, J. H., 177, 178, 203, 226 Wenzel, B. M., 231, 244, 245, 255, 258,260,261, 269, 270, 271,275 Werle, E., 194, 226 Werner, G., 235, 274 Westerfeld, W. W., 52, 56, 58, 61, 74, 76 Westermann, E., 186, 187, 205, 224, 226, 307, 329 Whittaker, V. P., 203, 224 Whittam, R., 126, 136 Wiebers, J. E., 66, 75 Wiersma, C. A. G., 320, 332 Wikler, A., 72, 75, 232, 253, 274, 277 Wilcoxon, F., 218, 225 Willett, R. A., 351, 384 Williams, H. L., 248, 249, 276, 308, 327, 332 Williams, R. J., 43, 45, 52, 56, 57, 58, 70, 76 Wilson, J. D., 295, 330 Wilson, W. A,, Jr., 233, 243, 244, 251, 271, 277 Wilson, W. E., 2!34, 332 Winitz, M., 287, 328

398

AUTHOR INDEX

Winter, C. A., 231, 232, 234, 237, 245,252,261,262,263,277 Wirth, W., 239,240, 255,267, 277 Witcoff, H., 100, 136 Witkop, B., 169, 171, 223 Witt, P. N., 230, 277 Witter, F. F., 102, 136 Wlassak, R., 6, 40 Woodard, G., 69, 75 Woodbury, D. M., 304,330 Woolley, D. W., 177, 178, 185, 192, 204, 218, 226,227, 261,276 Wyngaarden, J. B., 332

Y Yamaguchi, Y., 326, 331 Yamazaki, T., 325, 331 Yonkman, F. F., 180, 225

Younger, F., 316, 331 Yum, K. S., 233, 247, 252, 275 Yuwiler, A., 187, 227

Z Zaltmann, P., 189, 224 Zarrow, M. X., 64, 76 Zaur, I. S., 13, 40 Zbinden, G., 189, 211, 227 Zeaman, D., 378, 384 Zeller, E. A., 194, 227 Zerahn, K., 122, 126, 127, 136 Zetler, G., 165, 173, 176, 180, 182, 215, 223, 227 Zilversmit, D. B., 100, 101, 134, 136 Zimmerman, J. H., 182, 227 Zimmermann, B. U., 124, 135 Zupancic, A. O., 190, 227

SUBJECT INDEX y-Aminobutyric acid a-Ketoglutarate A transaminase Acetylcholine ( ACh) distribution in monkey nervous sysin brain, 176 tem, 288-94 CAR, 257 pH activity relationships of cerebral effect on enzyme secretion of panGAD and GABA-T, 301-05 creas, 103 pyridoxal phosphate as a coenzyme effect of reserpine on content, in for, 309 various areas of dog brain, 208 effect of topical application, 138-40 Amobarbital in approach-avoidance conflict, 251 effects on phospholipid metabolism, Amphetamine ( Dexedrine ) ( D-A ) 103-10 action of chlorpromazine, 237 enzymes in ACh stimulated exbulbocapnine catalepsy, 254 change of phosphate in phosphaeffect of, on ethanol intake, 68-69 tidic acid, 118-19 learning under amphetamine, 346, inhibition of, by HC-3, 85-89 350-51 mechanism of action of, 132 results of in stimulation of P32 incorporation eyeblink conditioning, 349-50 into phospholipids, 103-104, loS, kinesthetic figural after-effects, 113-18 355 synaptic transmission, 85 nonsense syllable, 346 transmission and function, 190-93 pursuit rotor, 352-53 Acidosis, intracellular, GABA content, rotating spiral after-image, 361301-05 62 Adrenalectomy suppression of primary visual brain CABA levels, 285 stimulus, 363-64 pressor response in adrenalectovigilance, 365 mized animals, 142 salt intake in rats, 42, 67 Amytal (A) &Adrenochrome, potentiates CEPR, alternation behavior in simple T149 maze, 377-79 results of Alanine, tricarboxylic acid cycle, 28588 eyeblink conditioning, 348-50 p-Alanine kinesthetic figural after-effects, activity of spinal neurons, 322 355 effect in vertebrate nervous sysnonsense syllable learning, 346, tems, 281 350-51 Alcoholism, human pursuit rotor learning, 352-53 theories of etiology, 71-74 rotating spiral after-image, 361withdrawal symptoms, 72 62 Alkalosis, GABA content, 301-05 suppression of primary visual Alternation behavior, in simple Tstimulus, 363-64 maze, 376-79 sedation threshold, 367-68 y-Aminobutyric acid (GABA) vigilance, 365 metabolic and neurophysioIogica1 Analgesia, effect of drugs and CAR, roles, 280-332 269-70 399

400

SUBJECT INDEX

Anesthetics, ethyl ether and pressor response, 151 Anticholinesterase agents Hemicholiniums as, 78, 93 Apoenzyme relationships between apoenzyme and coenzyme of vitamin B, requiring enzyme systems, 305-14 Apparent movement, in assessment of drug effects, 355-56 Arecoline CAR, 257-59 effect on CEPR, 147-49 Asparagine, alcohol intake of rats, 58 Aspartic acid effect in rabbit cortex and cat spinal neurons, 281 tricarboxylic acid cycle, 285-88 ATP (Adenosine triphosphate) chemical combination between serotonin and ATP, 204 effects of ACh on incorporation of P32 from a-Glycerophosphate P32 and ATP into phosphatidic acid, 116-19 as energy source for osmosis, 12628 turnover of phosphatidic acid and ATP specific activity, 111 Atropine abolition of central effects of eserine, pilocarpine, and arecoline, 258 abolition of effect of ACh, 104 cholinergic agents on enzyme secretion in pancreas, 103 action of chlorpromazine, 237 reserpine, 245 as antagonist to HC-3, 94 atropine arecoline antagonist, 25859 blocking pressor response, 147 blood pressure lowering effect of GABA, 325 CAR, 257, 260

effect of GABA, 325 stimulating effect of, on phospholipid turnover, 104-05 stress-induced behavior in mts, 247 Azacyclonol (Frenquel) CAR, 248 potentiating effect on CEPR, 145 voluntary alcohol intake in man, 6869 Azaserine effect of, on free choice of alcohol by rats, 70

B Barbital, pole-jump CAR, 250-51 Barbiturates, CAR, 250-51 Benactyzine in animals, 246-48, 259-60 in man, 246 Biotin, deficiency and free choice of ethanol, 56 Bufotenine CAR, 264 potentiates CEPR, 149 Bulbocapnine, effect in animals, 254 Butabarbital, pole-jump CAR, 250

C Caffeine, action of Chlorpromazine, 237 Canonical variate analysis, 370-75 Caramiphen, CAR, 259-60 Carbamylcholine, effect on enzyme secretion of pancreas, 103 Carbon tetrachloride, cirrhosis induced by, 69 Catechol m i n e s content in hypothalamus and CEPR, 166-69 decrease of, after reserpine, 181-82, 208 Central nervous system (CNS) regeneration amphibia, 1 mammals, 1 Chlorethoxybutamoxane, effect on CAR, 256

401

SUBJECT INDEX

2-Chloropromethazine, 50% block of pole-jump with, 239 Chlorperazine central depressive activity of, 242 influence on CAR, 239 Chlorpromazine ( CPZ ) antagonistic effect of CNS stimulants on CPZ, 237 clinical and laboratory differentiation of imipramine and CPZ, 21519 CNS activity of, in animals, 235-37, 240, 242, 243 effect of, on CEPR, 144, 151 on schizophrenia, 234 other phenothiazine derivatives, 238-39, 241 piperazinyl derivatives of phenothiazine, 239-41 serotonin effects, 182 voluntary alcohol intake, 68439 Chlorprothixene (Fruxal), CAR, 240 Choline active transport mechanism of, 87 antidote for HC-3 poisoning, 86-87, 94 deficiency and free choice of ethanol, 56 effect on enzyme secretion of pancreas, 103 not an antidote for HC-3, 88-89 substances similar to, as antidote for HC-3, 87 Choline acetylase, in brain structures, 176 Cholinergic transmission, effect on, by hemicholiniums, 77 Cholinesterase activity, development of, in optic lobe, 31 Cirrhosis free choice of ethanol, 69-70 induced in rats with carbon tetrachloride, 69 Cocaine action of CPZ, 237 bulbocapnine catalepsy, 254

Codeine, CAR, 253 Coenzyme relationships between apoenzyme and coenzyme of vitamin B, requiring enzyme systems, 305-14 Conditioned avoidance-response (CAR) drugs, 229-72 Conditioned stimulus (CS ), 230-31 secondary CS, 232 Cyanocobalamine (Vitamin B-12), free choice of ethanol, 56

D Deanol ( 2-dimethylaminoethanol) CAR, 258 effect of, on alcohol intake, 68-69 not an antidote to HC-3, 87 thio analog of, and CAR, 258 Depressant drugs alternation behavior in simple Tmaze, 376-79 apparent movement, 355-56 in an application of canonical variate analysis, 371-74 definition and theory, 344-45, 36669 eyeblink conditioning, 350 flicker perimetry, 358-59 influence on cortical inhibition and excitation, 344 kinesthetic figural after-effects, 355 nonsense syllable learning, 351-52 reaction time, 354 rotating spiral after-image, 361-62 “sedation threshold,” 368 suppression of primary visual stimulus, 363-64 vigilance, 365-66 visual after-image, 360-61 visual-field effects, 356-58 Diethazine, effect of and vertical rope climbing, 234 Diethylaminomethyl-3-benzodioxane, effects on CAR, 255-56 Diethylstilbestrol, alcohol intake in gonadectomized rats, 67 Dimethoxyreserpine, CAR, 245

402

SUBJEXT INDEX

Diphenhydramine ( Benadryl ) CAR, 259 effect on vertical rope climbing, 234 Diphenylhydantoin ( Dilantin ), after topical application of ACh, 139 Disulfiram ( tetraethylthiuram disulfide, antabuse) effect of drugs related to, 69 effect on free choice of alcohol in rats, 69 use in treatment of alcoholism, 69 t-dopa, effect of on CEPR, 167 t Edrophonium, effect on hemicholiniums, 78, 80 Electroencephalogram ( EEG) effect of HC-3 upon, 84 rabbit, after 5-HTP, 184 “sedation threshold,” 367-69 Electrogenesis, conductile vs. synaptic, 175-77, 190 Electroshock therapy, brain serotonin content, 182 Emmetropic frogs, 2, 4 newts, 2 Endocrines influence of, on free choice of alcohol adrenals, 67 gonads, 66-67 pancreas, 65-66 thyroid, 64-65 Endoplasmic reticulum, site of phospholipid turnover in salt gland, 120-21 Ephedrine, action of CPZ, 237 Epileptogenic foci, cerebral cortex shows decreased glutamic acid levels, 301 Epinephrine analogies with serotonin, 203-04 CAR, 255 cortical epinephrine pressor response (CEPR), 138-71

in hypothalamus, 158-63 reserpine, 245 Eserine (see Physostigmine) action of reserpine, 245 CAR, 257-58 P32 incorporation into phospholipids, 104 Estradiol benzoate, alcohol consumption in deer mice, 67 Ethanol animal differences in intake, 44-45 CAR, 252 diet and free choice of ethanol, 5058 forced administration of, @2-64 free choice by rats, 42ff heredity and ethanol preference in rats, 45-49 induced changes in free selection, 42-73 3rd choice and free choice of ethanol, 59-61 Ethoxybutamoxane, effects on CAR, 256 Excitation, cortical, 339-40, 343, 344 Excitation-inhibition balance, 343-47 Experimental conditions animal vs. human experimentation, 375-79 appropriate, 335-42 errors in, 334-35 scheme of appropriate experimental conditions, 342-47, 366-69 Experimental tests in assessment of drug effects alternation behavior in simple Tmaze, 376-79 apparent movement, 355-56 eyeblink conditioning, 341-42, 34850 flicker perimetry, 358-59 kinesthetic figural after-effects, 355 nonsense syllable learning, 346, 350-52, 371-75 pursuit rotor learning, 335-42, 35253

SUBJECT INDEX

reaction times, 354, 371-75 rotating spiral after-image, 361-62 suppression of primary visual stimulus, 362-64 vigilance, 364-66 visual after-image, 359-61 visual-field effects, 356-58 Extraversion-introversion apparent movement, 355-56 definition and theory, 343-47, 36669 kinesthetic figural after-effects, 355 pursuit rotor learning, 353 visual after-image, 360-61 Eye, amphibia autograft, 12, 26 color vision, 3 homograft, 12, 13, 24 magnification factor, 5 orientation, 4 polarity, 33 rod-cone rates, 4 rotation, 16ff and reimplantation, 33 shape differentiation, 3 visual elements, 5 Eyeblink conditioning in assessment of drug effects, 34850 Doriden, 341-42 Eye vesicle, transplanted to position of ear vesicle, 25 I-

Factor N, free choice of ethanol, 43, 53, 54 Factor N,, free choice of ethanol, 43, 54, 55, 62, 70 Flicker perimetry, in assessment of drug effects, 358-59

G Genetotropic disease, alcoholism as a, 43 Glutamic acid alcohol intake of rats, 58

403

distribution in animal tissues, 283 effect in rabbit cortex and cat spinal neurons, 281 tricarboxylic acid cycle, 285-88 1,-Glutamic acid decarboxylase (GAD) in cerebrospinal fluid, 282 changes with animal age, 295-99 distribution in monkey and rabbit, 288-94 inhibition by hydrazides, 310 pH activity relationships of cerebral GAD and GABA-T, 301-05 pyridoxal phosphate as a coenzyme for, 308-09 as rate-limiting reaction of GABA sequence, 300-01 Glutamine alcohol intake of rats, 57 distribution in animal tissues, 283 effect of antagonists on free choice of ethanol by rats, 70 tricarboxylic acid cycle, 285-88 Glutethimide ( Doriden) in an application of canonical variate analysis, 371-74 CAR, 252 effects on eyeblink conditioning, 341-42 flicker perimetry, 358 nonsense syllable learning, 351 pursuit rotor test, 340-42 reaction times, 354 suppression of primary visual stimulus, 363 vigilance, 365 visual field, 357 Glycine, alcohol intake of rats, 58 Gonadectomy, alcohol consumption in rats, 66-67 Gonads, sex difference and alcohol intake, 66 “Gosh-type’’ agent, HC-3 as antagonist, 93 y-Guanidinobutyric acid GABA precursor for, effect of, 295

404

SUBJECT INDEX

H Hemicholiniums, 77ff central site of action, 84-90 postsynaptic, 87-90 presynaptic, 84-87 distribution and excretion studies on, 93 other actions, 92-94 possible inhibitor of polio virus, 94 rate of onset of action, 89-90, 93 Hexamid, effect on CNS, 240 Hexobarbital, effect on pressor response, 150 Hippocampus potentials from, influence of iproniazid on serotonin content, MO inhibitors on 5-HTP, 191 reduction of hippocampal potentials by serotonin, 204 Histamine effect of systemic application on pressor response, 151 topical application on systemic blood pressure, 140 estimation of, in hypothalamus, 165-66 Hull-Lepley hypothesis, 346-47 Hydrazides, hydrazide induced seizures and B,, 309-10 Hydroxylamine, GABA levels, 31 1-14 5-Hydroxytryptophan ( 5-HTP) behavioral changes in animals after, 180-81 brain serotonin concentration after I.V. 5-HTP, 184 decarboxylation of, in sympathetic ganglia, pheochromocytomata, 186 depressed cortical potentials after, 184, 208 inhibition of hippocampal potentials in cats given, 204 rate of metabolism of, in brain homogenates, 188-87 5-HTP decarboxylase activity of, in homogenates of brain

and kidney injected with reserpine, 208-09 not inhibited in reserpinized animals, 208 specificity of, 185-89 Hydroxyzine ( Atarax) in animals, 248 in man, 248 Hypermetropic frogs, 2 newts, 2 Hypophysectomy brain GABA levels, 285 pressor response in hypophysectomized animals, 142 Hypothalamus CEPR, 142 neurohumoral substances in extracts of rabbit hypothalamus, 157-66 neurovegetative divisions of, 176 primary amine content of, 187

1 Imipramine ( Tofranil) CAR, 240 clinical and laboratory differentiation of CPZ, 215-19 symptomatic relief of mental depression, 215-19 Inhibition, cortical, 339-40, 343, 344 Insulin coma, 5-HTP carboxylase, 182 Insulin, effect of, on alcohol intake, 66 Iproniazid action of reserpine, 245 CAR, 261 CEPR, 145-46 depression by 5-HTP of cortical potentials, 184 effect of chronic iproniazid on serotonin in rats, dogs, rabbits, 19495 excitation in animals given reserpine pretreated with iproniazid, 208, 211 monoamine oxidase inhibitor action of, 198

SUBFCT INDEX

Isoproterenol, effect of topical application on systemic BP, 140

J JB-156 (Catron), fast acting cerebral MO inhibitors, 198

K Ketobemidone, CAR, 253 Kinesthetic figural after-effects, in assessment of drug effects, 355

L Leukotomy, effect on personality, 343 LSD-25 (d-lysergic acid diethylamide ) action with chlorpromazine, 237 morphine, 253 reserpine, 244-45 CAR, 262-83 effect of, on alcohol intake, 69 potentiates CEPR, 149 psychotomimetic effect of, not strictly correlated to its antiserotonin activity, 213-15

M Mecamylamine, prevents effect of arecoline and nicotine on CEPR, 148 Megimide, and action of CPZ, 237 Mental depression, symptomatic relief with Imipramine, 215-19 Mepazine effect on CAR ( pole-jump), 239 central depressive activity, 242 Meperidine, stress-induced behavior in rats, 253 Mephenesin, CAR, 250 Meprobamate ( Miltown, Equanil) (MI action in animals, 24950 in an application of canonical variate analysis, 371-74

405

effects on apparent movement, 356 flicker perimetry, 358 nonsense syllable learning, 35152 reaction times, 354 rotating spiral after-image, 362 vigilance, 365 visual after-image, 360-61 visual-field effects, 357 voluntary alcohol intake, 68-69 no effect on CEPR, 145 Meratran, alternation behavior in simple T-maze, 377-79 Mescaline CAR, 263 effect on CEPR of pretreatment with mescaline, 149-50 Metamphetamine action with chlorpromazine, 237 reserpine, 245 Methacholine CAR, 257 effect on CEPR, 147 Methadone, stress-induced behavior in rats, 253 Methionine sulfoximine, alcohol consumption in rats, 70 Methyl atropine, CAR, 259 Methylparafynol, CAR, 252 Methylphenidate (Ritalin) action of chlorpromazine, 237 antagonizes CAR inhibition, 237 effect of, on ethanol intake, 68-69 Methyl-scopolamine, CAR, 259 Metrazol, action of CPZ, 237 Mitochondria, MO in, 191. MO (monoamine oxidase) importance, 193-203 location, 191 physiological substrates of, 194 specificity, 194 MO Inhibitors iproniazid as, 198, 211-12 JB-516 (Catron), 198

406

SUBJECT INDEX

pretreatment with, and electrical potentials in hippocampus, 191 SKF 385, 198-202 Monosodium glutamate, alcohol intake of rats, 58 Monosynaptic reflexes, enhancement of, with thiosemicarbazide, 323 Morphine alteration of action of, 253-54 CAR, 252-53 effect on CEPR, 151 free choice by rats, 42 Multineuronal synapses inhibition of spinal, with meprobamate, 249 Multisynaptic pathways reduced threshold of, after increased brain serotonin, 184 Muscarinic stimulation of brain, 25860 Myasthenia gravis, result of presynaptic biochemical defect, 90

N Na-$6 1-Piperidinomethyltetralon inhibits CAR, 256 synergistic with reserpine, 245 Nalorphine reversal of morphine-induced inhibition of the CAR, 253 Naphazoline ( Privine ) effect on pressor response, 140 Neostigmine CAR, 257 effect on CEPR, 147 hemicholinium, 78, 80 sensitization with, after topical application of acetylcholine, 139 Neurohumoral substances, difficulty in estimation of, 157, 159 Neuromuscular blocking agents hemicholinium as, 78, 80, 84-85, 89 Neuron, sensory events leading to the firing of, 320

Neuronal membrane effect of GABA, 6-alanine and taurine on, of cat spinal cord, 322-23 Neuronal specificity Sperry’s hypothesis, 26 advantages, 30 obstacles, 26-29 Neurotransmitters mechanism of action, 190-93 nature of changes brought about by, in postsynaptic membrane, 190 specific membrane receptors for specific, 190 Nicotine, CEPR, 148 NMN ( N-methylnicotinamide) HC-3 as inhibitor of NMN uptake, 86 Nonsense syllable learning in assessment of drug effects, 35052 results on, with subjects under placebo, amytal and dexedrine conditions, 346 Norepinephrine action of reserpine, 245 brain content depressed by certain Rauwolfia alkaloids, 181 CAR, 255 depletion of peripheral stores by reserpine, 208 in hypothalamus, 158-63 pressor response after topical application, 140

0 Optic lobe, amphibia anatomy, 6 electrical activity, response to stimulation, 8-9 embryology, 7 sensorimotor coordinating area, 9, 14 Optic nerve fibers, 5, 6 growth in abnormal direction, 24-25

SUBFCT INDEX

operative procedures, 10, 11, 24, 26, 33 regeneration amphibia, 1-38 teleost fishes, 5, 6 selective restitution of nerve fibers, 22ff visual recovery after optic nerve regeneration, explanation, 15ff Optokinetic response frogs, 4 return of, after section of optic nerve amphibia, 11-12 teleost fish, 12 Orientation response (taxis), frogs, 3 Oxytocin, release of after CEPR, 142

P Pain, fear of, motive for CAR, 270 Pancreas effect of antidiabetic sulfa (Nadisan) on alcohol intake, 66 effect of cholinergic agents on enzyme secretion of, 103 Pancreatectomy alcohol consumption, 65 carbohydrate consumption in rats, 42 Pancreozymin, stimulates phospholipid turnover, 109 Pantothenate deprivation and free choice of ethanol, 56 Pentobarbital (Nembutal) effect on pressor response, 150 pole-jump CAR, 250-51 Perazine, influence on CAR, 239 Perphenazine CAR, 240, 242 central depressive activity of, 242 Phenaglycodol, CAR and SCR, 250 Phenobarbital, and pole-jump CAR, 250-51 Phosphatides, diacyl glyceroas carriers in membranes, 99-133

407

Phosphoinositide turnover significance of stimulation of, 12830 stimulation in pancreas, 109 Phospholipid turnover concerned with sodium transport, 110-12, 121-26 effect of atropine on, 104-05 effect of cholinergic agents on, 103 in endocrine and exocrine gland, 109-10 individual, in cerebral cortex slices, 106-07 in various parts of brain, 107-08 Physostigmine (see Eserine) antidoting effect of, to HC-3, 89, 94 blood pressure lowering effect of GABA, 325 effect on CEPR, 147 stimulates cortical suppressor areas, 139 Pilocarpine CAR, 257 effect on CEPR, 147 Pinocytosis, and phospholipid effect, 131 Piperidinomethylbenzodioxane (933F), effects on CAR, 256 Pipradrol (Meratran) action of chlorpromazine, 237 antagonizes CAR inhibition, 237 voluntary alcohol consumption, 6869 Polysynaptic reflexes, enhancement of, with thiosemicarbazide, 323 Postsynaptic membrane changes brought about by the transmitter in, 190 excitatory and inhibitory effects transmitted by chemical mediators effecting, 319-25 Procaine effect of systemic application, 151 effect of topical application on blood pressure, 138

408

SUBJECT INDEX

Progesterone, and alcohol intake in deer mice, 67 Promazine 50% block of pole-jump with, 239 vertical rope climbing, 234, 242 voluntary alcohol intake, 68-69 Promethazine, 50% block of polejump with, 239 Propylthiouracil, and alcoholic intake of rats, 64 Psychiatric symptoms how to deal with them, difficulties and complexities, 380-82 Psychological tests, and assessment of drug effects, 333-83 Psychopharmacology definition, 333 methodological problems, 333-83 Pursuit rotor test in assessment of drug effects, 35253 interpretation of results, 335, 340 performance under placebo and Doriden conditions, 340-42 Pyribenzamine, effect of, and vertical rope climbing, 234 Pyridoxal (vitamin B-6) as a coenzyme, 305-14 convulsive seizures and dietary pyridoxine deficiency, 316-18 5-HTP decarboxylase activity, 18889 relationships between apoenzyme and coenzyme of vitamin B, requiring enzyme systems, 305-14 Pyridoxine deprivation, and free choice of ethanol, 56

R Reaction times, in assessment of drug effects, 354 Rescinnamine CAR, 245 in frontal lobe lesioned monkeys, 243, 245 Reserpine action in animals, 242-44, 246

action in man, 242 altering effects of, 244-49 antagonism by iproniazicl, 145 effect on norepinephrine and serotonin content of brain, 145, 156 potentiation of CEPR, 144, 146, 151 reserpine-like substances, 245-46 serotonin metabolism and, 207-12 tranquilization by, and brain serotonin content, 181-82, 184 voluntary alcohol intake, 68-69 Respiration, depressant action on of HC-3, 80-84, 91-92 Retina local sign, 14, 15 regeneration ( amphibia), 12, 14 rotation of retinal ganglion cell “field,” 33 Riboflavin deprivation and free choice of ethanol, 56 Rotating spiral after-image in assessment of drug effects, 361-62 3 Salt gland endoplasmic reticulum as site of phospholipid turnover, 130-21 role of phosphatides in ion transport, 110-12 Schizophrenia abnormal factors in serum, 169-70 effect of serum from, on CEPR, 151-53, 167 Scopolamine CAR, 259-60 chlorpromazine, 237 enhances bulbocapnine catalepsy, 254 morphine and CAR, 254 “Sedation threshold,” test for, 367-69 Semicarbazide, used as a 5-HTP carboxylase inhibitor, 309 Serotonin bioassay of, 156, 164 CAR, 260-62 concentration of, in brain, 156

409

SUBJECT INDEX

effect of reserpine on serotonin content of brain, 145, 156 topical application on systemic bloocl pressure, 140 effect on CEPR after pretreatment with serotonin, 149 reserpine, 245 role of in neurobiology, 175-227 Serum, from patients with organic diseases and CEPR, 154 pregnant women and CEPR, 153-54 schizophrenics and CEPR, 151-56 SKF 385 CAR, 261 as a monamine oxidase inhibitor, 198-202 Sodium chloride, secretion of albatross, 110-12, 121-22 Sodium nitrite, and GABA levels, 311 Sodium pump, phosphatidic acid as a carrier, 121-26, 128-30 Specific inhibition of CAR definition, concepts, 265-68 Stimulant drugs apparent movement, 355-56 in an application of canonical variate analysis, 371-74 definition and theory, 344-45, 36669 eyeblink conditioning, 350 flicker perimetry, 358-59 influence on cortical inhibition and cortical excitation, 344 kinesthetic figural after-effects, 355 nonsense syllable learning, 351-52 reaction time, 354 rotating spiral after-image, 361-62 sedation thresholds, 368 suppression of primary visual stimulus, 363-64 vigilance, 365-66 visual after-image, 360-61 visual-field effects, 356-69 Strychnine, free choice by rats, 42

Substance P, estimation of in hypothalamus, 165 SU 3118, and brain biogenic amine depletion, 181 SU 5171, and brain biogenic amine depletion, 181 Sulfa drugs Na sulfadiazine and alcohol consumption, 70 sulfasuxidine and alcohol intake, 70 Sympathetic ganglia, phospholipid effect in, 108-09 Synaptic transmission, and phosphatidic acid cycle, 131-32

T Taurine activity of spinal neurons, 322 effect in vertebrate nervous systems, 281 Testosterone, and alcohol intake in gonadectomized rats, 67 Tetrabenazine augmentation of CEPR, 144-45 brain norepinephrine content, 181 CAR, 245 serotonin content, 181 Tetraethylammonium chloride ( TEA ) blocks afferent impulses through autonomic ganglia, 2'64 Thiamine, deprivation and alcohol preference by rats, 43, 52 Thioctic acid, and free choice of ethanol by rats, 55 Thiopropazat (Dartal), and CAR, 242 Thioridazine, effect on CAR, 239, 242 Thiosemicarbazide GABA production, 310 intracellular GABA levels, 323 Third choice (sugar, fat or saccharin solution) effect of, on alcohol consumption of rats, 59-61 Thyroidectomy alcoholic intake of rats, 65 brain GABA levels, 285

410

SUBJECT INDEX

Thyroxine alcoholic intake of rats, 64 effect of systemic administration on pressor response, 151 Trasentine, effect in stress-induced behavior in rats, 247 Triflupromazine, compared to CPZ, 239 Tryptamine serotonin, 189 increased toxicity of, 199 d-tubocurarine post synaptic blockade by, and the effects of HC-3, 90 Tumor growth and brain GABA levels, 285

U Unconditioned response ( U R ) , 231 Unconditioned stimulus ( U S ) , 231 Urethane, effect on CAR, 251

V Vasopressin, CEPR, 143 Vigilance, in assessment of drug effects, 364-66 Visual after-image, in assessment of drug effects, 359-61 Visual environment, response to, frogs, 3

Visual-field effects, in assessment of drug effects, 356-58 Visual field, frogs anterior, 7 in motion, 2 posterior, 7 stationary, 2 Visual stimulus, suppression of primary in assessment of drug effects, 362-64 Vitamin A, deficiency and free choice of ethanol, 56 Vitamin B complex deprivation and carbohydrate intake in rats, 42 Vitamin B-6 as a coenzyme, 305-14 convulsive seizures and dietary pyridoxine deficiency, 316-18 5-HTP decarboxylase activity, 18889 relationships between apoenzyme and coenzyme of vitamin Bo-requiring enzyme systems, 305-14 Vitamins multiple vitamin deficiencies and free choice of ethanol, 56 single vitamin deficiency and free choice of ethanol, 56