International Review of Neurobiology Volume 12

INTERNATIONAL REVIEW OF

Neurobiology VOLUME 12

Associate Editors

w.Ross h E Y

H. J. EYSENCK

D. B o r n

G. HARRI~

Josh DELGAD~

c. W

Sm

JOHN

ECCLES

B

0.ZANGWILL

Consultant Editors

V. AMASSIAN

K. KILLAM

MURRAYB. BORNSTEIN F. TH. BRUCKE

C. KORNETSKY A. LAJTHA

P. DELL

B. LEEEDEV

J. ELKES

Sm AUBREYLEWIS

W. GREYWALTER

VINCENZO LONGO

R. G. HEATH B. HOLMSTEDT

D. M. MAC~AY STEN M & m s

P. A. J.

F. MORRELL

s. KETY

JANSSEN

H. OsMohTD STEPHEN SZARA

INTERNATIONAL REVIEW OF

Neurobiology Edifed by CARL C. PFEIFFER New Jersey Neuropsychiatric lnsfifufe Princeton, New Jersey

JOHN R. SMYTHIES Department of Psychiatry University of Edinburgh, Edinburgh, Scotland

VOLUME 12

1970

ACADEMIC PRESS

New York and London

COPYRIGHT@ 1970, BY ACADEMIC PRESS,INC. ALL RIGHTS RESERVED. NO PART OF THIS BOOK MAY BE REPRODUCED IN ANY FORM, BY PHOTOSTAT, MICROFILM, RETRIEVAL SYSTEM, OR ANY OTHER MEAXUS, WITHOUT WRITTEN PERMISSION FROM THE PUBLISHERS.

ACADEMIC PRESS, INC. 111 Fifth Avenue, New York, New York 10003

United Kingdom Edition published by ACADEMIC PRESS, INC. (LONDON) LTD. Berkeley Square House, London W1X 6BA

Lmmy

OF

CONGRESS CATALOG CARDNUMBER: 59-13822

PRINTED IN THE UNITED STATES OF AMERICA

CONTRIBUTORS Numbers in parentheses refer to the pages on which the authors’ contributions begin.

SAMUELH. BARONDES,~ Departments of Psychiaty and Molecular Biology, Albert Einstein College of Medicine, Bronx, New York (177)

F. BENINGTON, Department of Psychiatry, Uniuersity of Alabama, Birmingham, Alabama (207) M . DAVIS,Laboratory of Clinical Science, and Clinical Research Branch, National Institute of Mental Health, Bethesda, Maryland (145)

JOHN

J. GORDON, Department of Pathology, New York Hospital-Cmell Medical Center, New York, New York (45) A. M . HA-IT, Physiological Institute, University of Basel, Basel, Switzerland (265) PETERLOMAX,Department of Pharmacology, School of Medicine, and the Brain Research Institute, University of California, Los Angeles, California ( 1 ) M . MONNIER,Physiological Institute, University of Basel, Basel, Switzerland (265) R. D. MOF~IN,Department of Psychiatry, University of Alabama, Birmingham, Alabama ( 207)

J. PROKOP,Department of Pathology, New York Hospital-Cornell Medical Center, New York, New York (45) ISAMU SANO,Department of Neurology, The Institute of Higher Nervous Activity, Osuka Uniuersity Medical School, Fukushima-ku, Osaka, Japan (2 3 5 )

R. SAUER,Physiological Institute, University of Basel, Basel, Switzerland (265) * Present address: Department of Psychiatry, The Medical School, University of California at San Diego, La Jolla, California. V

vi

CONTRIBUTORS

J. R. STVIYTHIES, Department of Psychiatry, University of Edinburgh, and The Neurosciences Research Program, Massachusetts Institute of Technology, Boston, Massachusetts (207) M. STERLADE,Labomtoire de Neurophyswlogie, Dbpartement & Physhlogk, Facultb de Me'&ciw, Universitd LaVal, Qdbec, Canada (87)

R. TORACK, * Department

of Pathology, New York Hospital-CorneU Medical Center, New York, New York (45)

M. VAN R o s s v ~ ,Department of Pharmucology, Catholic Uniuersity, Nijmegen, The Netherlands (307)

JACQUES

* Present address: Department of Pathology, Washington University School of Medicine, St. Louis, Missouri.

CONTENTS CONTRIBUTORS

CONTENTS OF

.

.

PRFYIOUS

.

.

.

.

.

.

.

.

.

.

V

VOLUMES

.

.

.

.

.

.

.

.

xi

Drugs and Body Temperature

PETERLOMAX

I. Introduction . . . . . I1. The Regulation of Body Temperature I11. Pharmacological Responses . . IV. Concluding Remarks . . . . References . . . . . .

.

.

.

.

.

.

. . .

. . .

. . .

. . .

. . .

. . .

1 2 7 37

.

.

.

.

.

.

37

45 49 52 70 82

Pathobiology of Acute Triethyltin Intoxication R TORACK. J . GORDON.AND J PROKOP

.

I. Introduction . . I1. Materials and Methods I11. Results . . . IV. Discussion . . . V Conclusions . . References . . .

.

.

.

.

.

.

.

.

.

.

.

. .

. .

. .

. .

. .

. .

. .

. .

. .

.

.

.

.

.

.

.

.

.

. .

. .

. .

. .

. .

. .

. .

. .

. .

83

Ascending Control of Thalamic and Cortical Responsiveness

. .

M . STEWE

I Background . . . . . . . . . . . I1. Testing Responses . . . . . . . . . . I11 Unspecific and Specific Influences on Thalamocortical Complexes . IV. Final Remarks . . . . . . . . . . . References . . . . . . . . . . . .

87 91 97 133 136

Theories of Biological Etiology of Affective Disorders JOHN

M . DAVIS

. Introduction . . . . . . . . The Biogenic Amine Hypothesis . . . . Synthesis and Metabolism of Biogenic Amines . Antidepressants . . . . . . . . Reserpine-Induced Depression . . . . . Electroconvulsive Shock Therapy . . . VII. Lithium . . . . . . . . I I1 I11 IV V VI

Vii

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

. .

. .

. .

. .

145 147 148 149 152 154

.

.

.

.

154

Viii

CONTENTS

VIII . Amphetamine . . . . . . IX Central Receptors and Depression . . X . Catecholamine Metabolism in Depressed and XI. Indole Metabolism in Depressed Patients XI1. Biogenic Amine Levels in Human Brain . XI11. Experimental Drugs and the Biogenic Amine . . . . . . XIV. Electrolytes . . . . XV. Steroids in Depression XVI. Discussion . . . . . . . References . . . . . . .

.

. .

.

.

.

.

.

.

. .

Manic Patients . . . .

.

.

.

Hypothesis

. .

.

.

.

.

.

.

.

.

. .

. .

. .

. .

. . . . . .

.

156 156 157 159 159 160 161 163 165 167

Cerebral Protein Synthesis Inhibitors Block long-Term Memory

SAMUELH . BARONDES I. Introduction . . . . . . . I1. Additional Background . . . . . I11. Inhibitors of Protein and RNA Synthesis . IV. Behavioral Assays . . . . . . V . Consolidation and Redundancy . . . VI Inhibition during Training . . . . VII . Inhibition Shortly after Training . . . VIII . Inhibition Long after Training . . . . IX Inhibition of RNA Synthesis . . . . X Short-Term and Long-Term Memory . . XI Conclusion . . . . . . . . References . . . . . . . .

.

. . .

. .

. .

. .

. .

.

.

.

.

.

.

.

.

. .

. .

. .

. .

.

.

.

.

. . .

. . .

. . .

. .

. .

. .

. .

. .

.

177 178 180 183 185 186 191 192 195 198 201 203

The Mechanism of Action of Hallucinogenic Drugs on a Possible Serotonin Receptor in the Brain

.

J . R. SMITHIES. F. BENINGTON. AND R . D M o m

. I . Introduction: Characteristics of Hallucinogens I1. Specification of a Serotonin Receptor Site . . I11. Specification of the Hallucinogens as Central Serotonin IV. Mitomycin . . . . . . . . . V. Type BO Hallucinogens . . . . . . . . . . VI . Generalization of the Hypothesis VII . A Possible Role of RNA . . . . . . VIII . Testing of the Hypothesis . . . . . . IX. Summary . . . . . . . . . References . . . . . . . . . Note Added in Proof . . . . . . .

. .

. .

. .

.

.

.

.

.

.

. . .

. . .

. . .

207 211 215 223 223 224 225 230 231 232 233

. .

237 245

Antagonists

.

.

.

. .

. .

. .

Simple Peptides in Brain

ISAMU SANO I . y-Glutamyl Peptides . . . . I1. y-Aminobutyryl and b-Alanyl Peptides

.

.

.

.

.

.

ix

CONTENTS

111. N-Acetyl Amino Acids and Peptides IV. Proteases and Peptides . . . References . . . . . .

. .

.

. . .

. .

.

. . .

. .

.

. .

.

252 257 259

The Activating Effect of Histamine on the Central Nervous System

M. MONNIER,R. SAUER,AND A. M. HATT I. Formation, Distribution, and Catabolism of Histamine in the Brain . . . . . . . . . (Bibliographic Report) 11. Effects of Intravascular Administration of Histamine . . . 111. Effects of Intraventricular Administration of Histamine . . IV. Effects of Direct, Intracerebral Administration of Histamine ( Bibliographic Report) . . . . . . . . . V. Release of Histamine in the Hemodialysate of Aroused Animals (Personal Investigations) . . . . . . . . . VI. Conclusions and Summary . . . . . . . . References . . . . . . . . . . . .

266 270 288 293 294 298 303

M o d e of Action of Psychomotor Stimulant Drugs

JACQUES M.

VAN

ROSSUM

I. Psychomotor Stimulant Drugs . . . . . . 11. Effects of Psychomotor Stimulant Drugs in Man . . 111. Effects of Psychomotor Stimulant Drugs in Animals . IV. Kinetics of Absorption, Distribution, and Elimination of . . . . . . . . . Amphetamines V. Antagonism of Amphetamine Action and Interaction with . . . . . . . . , . Drugs . VI. Psychomotor Stimulant Action and Brain Catecholamines VII. Mechanism of Action of Psychomotor Stimulant Drugs . References . . . . . . . . . .

.

.

.

. .

309 323 327

.

.

337

.

Other

.

. . .

.

.

347 356 361 373

.

.

AUTHOR INDEX .

.

.

.

.

.

.

.

.

.

.

.

385

SUBJECTINDEX .

.

.

.

.

.

.

.

.

.

.

.

409

This Page Intentionally Left Blank

CONTENTS OF PREVIOUS VOLUMES Volume 1

Recent Studies of the Rhinencephalon in Relation to Temporal Lobe Epilepsy and Behavioral Disorders W . R. Adey Nature of Electrocortical Potentials and Synaptic Organizations in Cerebral and Cerebellar Cortex Dominick P. Purpura Chemical Agents of the Nervous System Catherine 0. Hebb Parasympathetic Neurohumors; Possible Precursors and Effect on Behavior Carl C . Pfeifler Psychophysiology of Vision G . W. Granger Physiological and Biochemical Studies in Schizophrenia with Particular Emphasis on Mind-Brain Relationships Robert G . Heath Studies on the Role of Ceruloplasmin in Schizophrenia S . Mc?rtens, S . Vallbo, and B . Melunder Investigations in Protein Metabolism in Nervous and Mental Diseases with Special Reference to the Metabolism of Amines F. Gemgi, C. G . Honegger, D . Jordan, H . P. Rieder, and M . Rottenberg AUTHOR INDEX-SUBJECT INDEX

Volume 2

Regeneration of the Optic Nerve in Amphibia R. M . Gaze Experimentally Induced Changes in the Free Selection of Ethanol Jorge Mardones xi

xii

CONTENTS OF PREVIOUS VOLUMES

The Mechanism of Action of the Hemicholiniums

F. w. Schueler The Role of Phosphatidic Acid and Phosphoinositide in Transmembrane Transport Elicited by Acetylcholine and Other Humoral Agents Lowell E . Hokin and Mabel R. Hokin

Brain Neurohonnones and Cortical Epinephrine Pressor Responses as Affected by Schizophrenic Serum Edward J . W a h z e k The Role of Serotonin in Neurobiology Erminio Costa Drugs and the Conditioned Avoidance Response Albert Herz Metabolic and Neurophysiological Roles of ydminobutyric Acid Eugene Roberts and Eduardo Eidelberg Objective Psychological Tests and the Assessment of Drug Effects H. J. Eysenck AUTHOR INDEX-SUBJECT INDEX

Volume 3

Submicroscopic Morphology and Function of Glial Cells Eduurdo De Robert& and H . M . Gerschenfeld Microelectrode Studies of the Cerebral Cortex Vahe E. Amassian Epilepsy Arthur A. Ward, Jr. Functional Organization of Somatic Areas of the Cerebral Cortex Hiroshi Nakahamu Body Fluid Indoles in Mental Illness R. Rodnight Some Aspects of Lipid Metabolism in Nervous Tissue G. R. Webster Convulsive Effect of Hydrazides : Relationship to Pyridoxine Harry L. Williams and James A. Bain

CONTENTS OF PREVIOUS VOLUMES

xiii

The Physiology of the Insect Nervous System D. M . Vowles AUTHOR INDEX-SUBJECT INDEX

Volume 4

The Nature of Spreading Depression in Neural Networks Sidney Ochs Organizational Aspects of Some Subcortical Motor Areas Werner P . Koella Biochemical and Neurophysiological Development of the Brain in the Neonatal Period Williaminu A. Himwich Substance P: A Polypeptide of Possible Physiological SignScance, Especially within the Nervous System F. Lembeck and G. Zelter Anticholinergic Psychotomimetic Agents L. G. A b o d and 1. H . Biel Benzoquinolizine Derivatives : A New Class of Monamine Decreasing Drugs with Psychotropic Action A. Pletscher, A. Brossi, and K . F . Gey The Effect of Adrenochrome and Adrenolutin on the Behavior of Animals and the Psychology of Man A. Hofer AUTHOR INDEX-SUB JECr INDEX

Volume 5

The Behavior of Adult Mammalian Brain Cells in Culture Ruth S . Geiger The Electrical Activity of a Primary Sensory Cortex: Analysis of EEG Waves Walter J. Freeman Mechanisms for the Transfer of Information along the Visual Pathways Koiti Motokawa

XiV

CONTENTS OF PRFIrIOUS VOLUMES

Ion Fluxes in the Central Nervous System F. J. Brinky, Jr. Interrelationships between the Endocrine System and Neuropsychiatry Richard P . Michael and James L. Gibbons Neurological Factors in the Control of the Appetite Andr& Soulairac Some Biosynthetic Activities of Central Nervous Tissue R. V. Coxon Biological Aspects of Electroconvulsive Therapy Gunmr Holmberg AUTHOR INDEX-SUBJECX INDEX

Volume 6

Protein Metabolism of the Nervous System Abel Lajthu Patterns of Muscular Innervation in the Lower Chordates Quentin Bone The Neural Organization of the Visual Pathways in the Cat Thomes H . Meikle, Jr., and James M . Sprague Properties of Afferent Synapses and Sensory Neurons in the Lateral Geniculate Nucleus P, C . Bishop Regeneration in the Vertebrate Central Nervous System Carmine D . Clemmte Neurobiology of Phencyclidine (Sernyl), a Drug with an Unusual Spectrum of Pharmacological Activity Edward F. Domino Free Behavior and Brain Stimulation Jose' M . R. Delgado AUTHOR INDE2-SUBJECr INDEX

CONTENTS OF PREVIOUS VOLUMES

xv

Volume 7

Alteration and Pathology of Cerebral Protein Metabolism Abe 1 Laitha Micro-Iontophoretic Studies on Cortical Neurons K . KrnjeviC Responses from the Visual Cortex of Unanesthetized Monkeys John R. Hughes Recent Developments of the Blood-Brain Barrier Concept Ricardo Edstrom Monoamine Oxidase Inhibitors Gordon R. Pscheidt The Phenothiazine Tranquilizers : Biochemical and Biophysical Actions Paul S. Guth and Morris A. Spirtes Comments on the Selection and Use of Symptom Rating Scales for Research in Pharmacotherapy J . B. Wittenborn Multiple Molecular Forms of Brain Hydrolases Joseph Bemohn and Kevin D. Barron AUTHOR INDEX-SUBJECT

INDEX

Volume 8

A Morphologic Concept of the Limbic Lobe Lowell E . White, Jr. The Anatomophysiological Basis of Somatosensory Discrimination David Bowsher, with the collaboration of Denise Albe-Fessard Drug Action on the Electrical Activity of the Hippocampus Ch. Stumpf Effects of Drugs on Learning and Memory James L. McGaugh and Lewis F . Petrinovich Biogenic Amines in Mental Illness Giinter G. Brune

xvi

CONTENTS OF PREVIOUS VOLUMES

The Evolution of the Butyrophenones, Haloperidol and Tduperidol, from Meperidine-Like 4-Phenylpiperidines Paul A. J . Janssen Amplitude Analysis of the Electroencephalogram (Review of the Information Obtained with the Integrative Method) Leonide Goldstein and Raymond A. Beck AUTHOR INDEX-SUB JECr INDEX

Volume 9

Development of “Organotypic” Bioelectric Activities in Central Nervous Tissues during Maturation in Culture Stanley M . Crain The Unspecific Intralaminary Modulating System of the Thalamus P. Krmpp and M . Monnier The Pharmacology of Imipramine and Related Antidepressants Laszlo Gyermek Membrane Stabilization by Drugs: Tranquilizers, Steroids, and Anesthetics Philip M . Seeman Interrelationships between Phosphates and Calcium in Bioelectric Phenomena L. G. Abood The Periventricular Stratum of the Hypothalamus Jerome Sutin Neural Mechanisms of Facial Sensation 1. Darian-Smith AUTHOR INDEX-SUBJECT INDEX

Volume 10

A Critique of Iontophoretic Studies of Central Nervous System Neurons G. C . Salmoiraghi and C . N . Stefanis Extra-Blood-Brain-Barrier Brain Structures Werner P . KoeUa and Jerome Sutin

CONTENTS OF PREVIOUS VOLUMES

xvii

Cholinesterases of the Central Nervous System with Special Reference to the Cerebellum Ann Silver Nonprimary Sensory Projections on the Cat Neocortex P. Buser and K . E . Bignall Drugs and Retrograde Amnesia Albert Weissman Neurobiological Action of Some Pyrimidine Analogs Harold Koenig

A Comparative Histochemical Mapping of the Distribution of Acetylcholinesterase and Nicotinamide Adenine DinucleotideDiaphorase Activities in the Human Brain T . Ishii and R. L. Fried Behavioral Studies of Animal Vision and Drug Action Hugh Brown The Biochemistry of Dyskinesias G. Curzon AUTHOR INDEX-SUBJECT INDEX

Volume 11

Synaptic Transmission in the Central Nervous System and Its Relevance for Drug Action Philip B. Bradley Exopeptidases of the Nervous System Neville Marks Biochemical Responses to Narcotic Drugs in the Nervous System and in Other Tissues Doris H. Clotlet Periodic Psychoses in the Light of Biological Rhythm Research F . A. Tenner Endocrine and Neurochemical Aspects of Pineal Function Be'k Mess The Biochemical Investigations of Schizophrenia in the USSR D.V. Loxovsky

Wiii

CONTENTS OF PREVIOUS VOLUMES

Results and Trends of Conditioning Studies in Schizophrenia J. S m m Carbohydrate Metabolism in Schizophrenia Per S . Lingjamde The Study of Autoimmune Processes in a Psychiatric Clinic S. F . Semenov Physiological Foundations of Mental Activity N . P. Bechterewa and V. B . Gretchin AUTHOR INDEX-SUB JECr INDEX

Cumulative Topical Index for Volumes 1-10

DRUGS AND BODY TEMPERATURE1 By Peter Lomax Department of Pharmacology, School of Medicine, and fhe Brain Research Institute, Univerrify of California, 10s Angeles, California

I. Introduction . . . . . . . 11. The Regulation of Body Temperature . A. Central Mechanism . . . . B. Peripheral Mechanisms . . . . 111. Pharmacological Responses . . . A. Sites of Action of Drugs . . . B. Methods of Investigation . . . C. Chemical Mediators in the Central Nervous D.Drugs . . . IV. Concluding Remarks . References . . . . . . .

.

.

.

.

.

.

.

.

.

.

. .

. .

.

.

.

.

.

. . . .

. . . .

. . . .

. . . . System

. .

. .

.

.

.

.

.

.

. .

. .

. .

1 2 2 4

7 7 8 11 26 37 37

I. Introduction

The mechanisms invcved in the regu %tionof internal body temperature continue to receive the attention of many investigators, and present concepts are fully examined in the excellent reviews by Bligh (1966a) and Hammel (1968). The effects of drugs on thermoregulation have been extensively studied, and attempts have been made to determine the sites and mechanisms of drug action in the light of this newer knowledge. The papers by von Euler (1961) and Borison and Clark (1967) summarized the pharmacology of temperature regulation up to those dates, and certain aspects of the subject were discussed at a symposium held during the Second International Pharmacological Meeting ( Trabucchi et az., 1964). Any system of the complexity of that concerned in temperature regulation would be expected to be particularly susceptible to The author’s research on this subject has been supported by United States Public Health Service Grant No. B-03007 and by a grant from the American Medical Association Education and Research Foundation. The University of California Brain Information Service contributed bibliographical aid under contract No. 43-6659from the National Institute of Neurological Diseases and Blindness. 1

2

PETER LOMAX

disturbance by the action of drugs. Owing to the well-regulated negative feed-back control of the system, however, any impairment tends to be compensated for, so that, in most cases, deviations from the set point are small in magnitude. Also there are several instances where the action of the drug on peripheral structures involved in temperature regulation is opposite in effect to that exerted on the central thermoceptive structures. Such factors often render it difficult to determine the precise mechanisms by which a druginduced change in core temperature is brought about. In this paper attention will be directed mainly to agents for which there is evidence indicating a direct action on the thermoregulatory centers in the central nervous system. II. The Regulation of Body Temperature

A. CENTRAL MECHANISMS

As early as 1845 Bergmann postulated that temperature-sensitive structures in the brain are responsible for maintaining the balance of heat loss by the cutaneous vessels. Considerable evidence rapidly accrued to support this hypothesis, and in 1938 Magoun et al. closely defined the thermoregulatory centers to the region of the pre- and supraoptic hypothalamic nuclei in the cat. Subsequent studies have confirmed this site as the major controlling center in the dog (Fusco et al., 1961), rat (Satinoff, 1964), rabbit (Downey et al., 1964), goat ( Andersen et al., 1962), ox ( Ingram et d.,1961) , and baboon (Gale and Ruch, 1966). Although other sites, such as the spinal cord (Simon, 1968), have been implicated in the regulation of body temperature the available evidence overwhelmingly confirms the essential nature of the rostra1 hypothalamus. There is basic agreement that temperature itself is the variable which activates the central thermostat. Temperature transducers are located at many sites in the body and these must have connections, directly or indirectly, with the temperature regulator. Although there is controversy as to which temperature the thermostat is ‘looking at” the balance of evidence favors the transduced hypothalamic temperature as the regulated temperature. The temperature of the thermosensitive hypothalamic neurons will follow that of the cerebral arterial blood which in turn is a function of the body core temperature in animals such as the rat, monkey, dog,

DRUGS AND BODY TEMPERATURE

3

and rabbit in which the main cerebral blood supply is via the internal carotid artery ( Abrams and Hammel, 1965; Hayward et al., 1966; Hayward, 1968; Hayward and Baker, 1968). Placing an animal in a cold, neutral, or warm environment produces little change in the temperature of its hypothalamus (Hammel et al., 1963; Lomax et al., 1964). The animal is able, however, to make appropriate adjustments to maintain its body temperature in the face of the imposed thermal stress. Various possible types of control, which are consistent with these observations, have been discussed (Hardy, 1961). There is increasing evidence that the hypothalamic centers behave as a proportional controller for most of the thermoregulatory responses. This means that the regulatory response is proportional to the difference between the actual temperature of the central neurons and a set or threshold temperature maintained in the thermoregulatory centers. If the hypothalamic temperature is higher than the set temperature, responses which increase heat loss are activated; with the hypothalamic temperature below the set temperature heat loss is decreased and/or heat production is increased. Apart from the transduced hypothalamic temperature, there are other inputs to the controlling system from temperature receptors at various locations in the body. These inputs appear to effect a change in the set point of the thermostat and so alter the threshold for the thermoregulatory responses. In the dog it has been shown that the thresholds for the onset of panting (heat loss) and vasoconstriction and shivering (heat gain) can be altered by changing the skin temperature (Hellstr@m and Hammel, 1967; Chattonet et al., 1964) or the core temperature (Hellswm and Hammel, 1967) under conditions in which the hypothalamic temperature is held constant. Raising the skin or core temperature lowers the threshold for the onset of panting and shivering; conversely these thresholds are higher when the core or ambient temperatures are decreased. Similar changes in the set temperature in response to different ambient temperatures have been reported in the rabbit (Downey et al., 1964), ox (Ingrain and Whittow, 1962), and goat (Andersen et al., 1962). In many species the body temperature is lower when the animal is asleep. The fall in temperature at the onset of sleep is due to vasodilation and increased heat dissipation (Hammel et al., 1963, Geschickter et al., 1966). Regulation of body temperature during

4

PETER L o r n

sleep appears to be normal at the lower level, which suggests that the set point has been lowered (see Hammel, 1!368). Studies of temperature regulation during exercise also indicate a downward adjustment of the set point during increased muscular activity. This change is reflected in the considerable increase in sweating seen at the onset of exercise in man (Nielsen, 1966). Possibly an indirect input to the thermoregulatory centers from proprioceptors in the muscles is the mediator of this response. Thus, the temperature controller is believed to integrate the input from the peripheral transducers and readjust the set point to compensate for varying thermal loads. Comparison of the hypothalamic temperature with the set temperature is the direct basis for the activation of the appropriate thermoregulatory responses.

B. PERIPHERAL MECHANISMS In response to signals from the controlling centers in the hypothalamus, several systems can be activated in order to stabilize the body temperature under varying environmental conditions. Many of the physiological thermoregulatory responses are subserved by organ systems which have other primary functions. Examples are seen in the case of vasomotor responses, shivering, or panting. The thermoregulatory activity must thus be coordinated with the animals’ over-all activity, and different systems may be activated in response to the same thermal stress depending on the individual conditions.

1. Behavior

The creation or selection of an environment in which body temperature can be maintained without calling into action other physiological responses appears to be one of the most primitive thermoregulatory capabilities and is one that is present in all vertebrates studied. Such activity has been demonstrated in goldfish (Rozin and Mayer, 1 x 1 ) and lizards (Cowles and Bogert, 1944). Heating or cooling the preoptic region of the lizard can cause the animal to choose between a cool or warm area respectively (Abrams and Caldwell, 1 x 7 ) . Behavioral responses are an important factor in thermal homeostasis in endotherms also, and they show such diverse facets as adjustment of body shape, nest building, and migration. Environmental selection or control under the stimulus of

DRUGS AND BODY TEMPERATURE

thermal discomfort shows its greatest diversity in man-from coal fires of Britain to air conditioning in California.

5

the

2. Vasomotor T m The rate of transfer of heat between the core and the skin of an animal can be regulated by changing the blood flow through the cutaneous vessels. In endotherms, the core temperature is usually higher than the ambient temperature, so that the rate of heat loss can be controlled by the peripheral circulation. These vascular responses are of particular importance in man (Robinson, 1963). In feathered or furred animals, heat exchange may occur at special sites at which the peripheral blood flow is regulated: the horns of the goat (Taylor, 1966), the tail of rodents (Johansen, 196l), the legs of some birds (Kahl, 1963; Steen and Steen, 1965), and the swimming flippers of seals (Irving et al., 1962) are among many examples.

3. Panting and Sweating The evaporation of water from the lungs or the surface of the body is an efficient method to dissipate body heat. A primitive form of this response is the wetting of the fur with saliva as practiced by some mammals. When the evaporation occurs from the respiratory passages almost all of the heat of vaporization is derived from the body core, whereas in the case of evaporation from the body surface part of the heat comes from the external environment. Thus, sweating is a less efficient mechanism for removing heat from the body. The thermal panting rate and the displacement volume in the dog are nicely regulated so that no significant changes occur in the arterial blood gases ( Albers, 1961) . Apocrine sweat glands, associated with the hair follicles, discharge only small quantities of fluid and are of little importance for thermoregulation in most animals ( Bligh, 1967). An exception is the horse where the apocrine glands resemble the human eccrine glands in their role in temperature regulation. The apocrine glands are stimulated by epinephrine in the blood (Lovatt Evans, 1957). The eccrine glands in man secrete much greater volumes of fluid and can be a significant factor in heat dissipation. They are innervated by cholinergic fibers in the sympathetic nerves and can be activated in emotional states, by gustatory reflexes when

6

PETER LOMAX

very spicy foods are eaten, during nausea and vomiting, in fainting attacks and in hypoglycemia as well as in response to thermoregulatory activity.

4. Thermogenesis The ability to markedly increase metabolic heat production by shivering or by nonshivering thermogenesis is an important regulatory mechanism in endotherms. In order to be effective in raising (or preventing a fall in) the core temperature, the increased heat production must be associated with responses which prevent heat loss from the surface of the animal. a. Shivering. Increased heat production from shivering has been described in many species. The activation of this response requires an intact caudal hypothalamus (Hemmingway, 1963) as well as the regulatory signals from the preoptic region. Segmental sensory inflow, especially proprioceptive input, is also an essential factor in shivering (Stuart et al., 1966). b. Nonshivering Themgenesis. Increased heat production in muscle and various visceral organs, apart from that due to shivering, can occur in mammals under conditions of cold stress. The important role of brown adipose tissue as a source of this nonshivering thermogenesis has been demonstrated in the rat (Smith, 1961, 1962) and in several other species including man (Dawkins and Scopes, 1965). Hypertrophy and increased metabolic activity in brown fat is of especial importance in hibernators and in the process of cold acclimation. Hormonal changes in response to heating or cooling the rostra1 hypothalamus have been described in different animals (see Hammel, 1968): in the goat, thyroid activity is depressed or stimulated by heating or cooling respectively, indicating changes in the release of thyroid-stimulating hormone from the pituitary gland; increased release of catecholamines in response to local cooling occurs in the goat; lowering the hypothalamic or environmental temperature increases cortisol levels in the dog; the secretion of antidiuretic hormone can be inhibited by central cooling in the monkey. The exact role of these several endocrine changes in thermoregulation is not entirely clear but it is likely that they play a part in the adaptation of the animal to abnormal environmental temperatures, particularly acclimatization to cold. An attempt has been made in Fig. 1 to summarize current con-

DRUGS AND BODY TEMPERATURE

7

propriDcep!

CONTROLLED

FIG.1. Schematic representation of the controlling and controlled systems for the regulation of body temperature and the general anatomical location of the various components. Upper part of the diagram represents a sagittal section of the brain close to the midline. AC, anterior commissure; F, fornix; M, mammillary body; OC, optic chiasma; T, thalamus.

cepts of the feedback control and effector mechanisms that appear to be of importance in normal temperature regulation. Ill. Pharmacological Responses

A. SITESOF ACTIONOF DRUGS It is clear from a consideration of normal thermoregulation that there are many ways in which drugs could interfere either with the function of the central controlling system or with the responses mediated by the hypothalamic centers. The degree to which such

8

PETER LOMAX

a disturbance can be compensated for by the system will determine the magnitude of any ensuing change in core temperature. Many drugs which act as general depressants or excitants of the central nervous system may cause rather nonspecific changes in body temperature. The depression of thermoregulatory control during general anesthesia would appear to be such a case. Furthermore, as we have seen, many of the peripheral effector responses are mediated as a secondary function of the organs systems concerned and the primary activity of these may be radically effected by drugs with resultant changes in body temperature. A great many of the temperature effects of drugs reported in the literature fall into the latter category and are of limited interest in the present context. Attention will be directed here primarily to agents which lead to changes in body temperature as a result of direct action on the hypothalamic centers. Such compounds may be employed therapeutically because of their effects on body temperature, whereas in others the induced changes in core temperature constitute a significant side effect of the drug. Finally, there are compounds that are of interest because they may act as neurotransmitters in the thermoregulatory centers, and their study may throw light on the function of the central nervous system in general.

B. METHODS OF INVESTIGATION When a drug causes a change in body temperature following systemic administration it is often difficult to decide whether the response is due to a peripheral or central site of action. Indeed, several components of the thermoregulatory control system may be effected and contribute to the total response. In the case of agents which may play a role as neurotransmitters, such as the catecholamines and acetylcholine, an additional complication arises in that they are unable to cross the blood-brain barrier freely and do not enter the brain tissues after systemic injection. In an attempt to unravel these problems some special techniques have been used in the study of drugs effecting temperature regulation.

1. Quutemry Compounds Molecules containing a highly charged quaternary nitrogen atom are unable to enter the central nervous system. Since many

DRUGS AND BODY TEMPERATURE

9

drugs contain an amine group it is frequently possible to synthesize the quaternary analogue. In most instances, the general qualitative actions of the parent compound are not greatly altered by this chemical maneuver so that it is possible to distinguish peripheral from central actions by comparison of the responses to the two compounds when injected systemically. 2. Intraventrkuhr Injection Introduction of drugs directly into the cerebrospinal fluid (CSF) has been used in many species in order to by-pass the blood-brain barrier. Generally, responses can be obtained with doses that would not elicit any effect when injected systemically. Although there is a paucity of information concerning the rate or degree of diffusion of drugs from the ventricles to different brain sites, it appears that the action is restricted to structures lying in proximity to the ventricular wall (Fuxe and Ungerstedt, 1966; Glowinski et al., 1966). 3. Intracerebral Injection

Microinjection of drugs directly into specific nuclei offers a more precise localization of the site of action than is obtainable by intraventricular injection. Using chronically implanted guides, such as those described by Decima and George (1964), the effects of drugs can be studied in conscious unrestrained animals. Provided the total injection volume is limited to approximately l-microliter, control injections are without significant effect and lesions cannot be detected on subsequent histological examination of the brains.

4. lmtqphoretic Methods The solubility of the drug under study and the limitation of the injection volume frequently restrict the total amount of drug that can be applied by direct injection, Also, it is desirable in some instances to deliver the drug slowly over a prolonged period of time rather than as a single dose. The technique of macroiontophoresis has been used in such cases (Jenden and Lomax, 1966). Briefly, the method consists of packing an ion exchange gel into a capillary tube and loading the gel with the drug by electrophoresis. The loaded tube is then inserted through a previously implanted cannula guide to the required location in the animal’s brain. The

10

PETER LOMAX

drug can be delivered at a controlled rate by reversing the direction of the electrical current. Because an ion of the drug is exchanged for another anion or cation in the brain there is no volume effect to consider.

5. Neurochemisty and Histochemisty The development of fluorimetric methods for the estimation of monoamines has led to many studies of the effects of drugs on brain concentrations of these amines. There is evidence that many agents may act indirectly by releasing or depleting monoamines in the central nervous system. Fluorescence techniques in histology have confirmed the presence of norepinephrine (Carlsson et al., 1962) and 5-hydroxytryptamine (And& et al., 1965) in the preoptic region. The estimation of acetylcholine in neural tissue has been carried out mainly by bioassay methods which do not allow sufficient sensitivity to determine regional distribution in detail. Most histological studies have been concerned with the localization of cholinesterase in different regions. Recently a gas chromatographic procedure has been described (Hanin et al., 1968; Hanin and Jenden, 1969) which is specific for acetylcholine and which promises to attain high sensitivity with the use of mass spectrometer detectors. 6 . Measurement

of Body Temperature

A wide variety of techniques has been used for measuring body temperature in experimental animals and man. Temperature measurements at various sites throughout the body of the calf and the rabbit, by insertion of thermocouples directly into the organs, reveal that the liver is at the highest temperature (Walther et al., 1941). The temperature of the liver and the mesentery around the portal vein are the highest temperatures in the abdominal cavity ( Grayson and Mendel, 1956). Although higher temperatures may occur in specialized tissues, such as the hypothalamus (Lomax et d.,1964) or brown adipose tissue (Smith, 1964), it would seem that the liver represents the best approximation to the body “core” temperature. In many studies, temperature-sensing devices are inserted through the rectum, and the recordings reflect the liver temperature and, thus, the transduced hypothalamic temperature (see Section 11,A).

DRUGS AND BODY TEMPERATURE

11

C. CHEMICAL MEDIATORS IN THE CENTRAL NERVOUS SYSTEM

1. Monoumines The role of epinephrine in metabolic heat production and the control of body temperature has long been debated (see review by Griffith, 1951). In 1943 von Euler and his associates showed that small doses of epinephrine injected into the cerebral ventricles or the cistema magna of the dog caused a rise in core temperature. The hypothalamus contains relatively high concentrations of catechol- and indoleamines (Vogt, 1954), and the possibility that norepinephrine and 5-hydroxytryptamine (5-HT) might play a role in normal temperature regulation was raised by Brodie and Shore (1957) and further discussed in the review by von Euler (1961). This suggestion was reiterated by Feldberg and Myers (1963) when they reported that intraventricular injection of epinephrine or norepinephrine caused a fall in temperature in the cat whereas injection of 5-HT had the opposite effect. Similar responses were obtained when the three amines were injected directly into the rostra1 hypothalamus (Feldberg and Myers, 1965a) indicating an effect directly on the thermoregulatory centers. On the basis of these findings, the investigators proposed that normal temperature regulation depends on the balance of release of these amines in the hypothalamus (Feldberg and Myers, 1964a). Subsequent studies revealed that the same responses to intraventricular injection of monoamines occurred during pentobarbital or chloralose anesthesia when temperature regulation is abolished (Feldberg and Myers, 196413). It was concluded from these results that the profound fall in temperature during anesthesia is due to modification of the release of the three amines. There was no consistent correlation, however, between the concentration of 5-HT in the effluent from the perfused ventricles and the rise in temperature during recovery from anesthesia (Feldberg and Myers, 1966). Intraventricular injection of the monoamine oxidase inhibitor tranylcypromine caused a rise in temperature in both anesthetized and conscious cats (Feldberg and Lotti, 1967a) and increased the output of 5-HT on perfusion of the ventricles (El Hawary et UZ., 1967). Experiments on dogs and monkeys revealed that these species gave the same general responses to intraventricular administration of the monoamines and tranylcypromine as those seen in the cat (Feldberg et al., 1967). Systemic administration of pargyline elevated the

12

PETER L O W

brain stem levels of monoamines in the rabbit, with the greatest increase being in 5-HT, and these changes were accompanied by a rise in body temperature. Intravenous reserpine depleted brain amines, and a fall in temperature was seen (Shellenberger and Elder, 1967). It was suggested that these results supported the concept that temperature control depended on the balance of central monoamines. However, the contribution of changes in peripheral amine stores, which must have occurred, could not be determined. Depletion of norepinephrine by reserpine leads to dilation of skin vessels which is a significant factor in the hypothermia which develops. I n the rabbit, intraventricular injection of norepinephrine either caused a rise in temperature or had no effect, and 5-HT led to a fall or was without effect. Neither drug was effective on injection into the rostra1 hypothalamus although leukocyte pyrogen introduced into the same site led to hyperthermia (Cooper et al., 1965). Thus, where responses were obtained they were opposite in effect to those seen in the cat, dog, and monkey. Monoamine oxidase inhibition with tranylcypromine had no effect on the resting temperature or the response to thermal loads (intravenous infusion of hot or cold saline) in the conscious rabbit, although the diencephalic concentractions of norepinephrine and 5-HT were increased ( Cranston and Rosendodf, 1967). In comparative studies made in the cat, rabbit, and sheep, Ruckebusch et al. (1965) found that the direction of the change in core temperature resulting from intraventricular infusion of catecholamines vaned with the species and the prevailing body temperature: normothermic cats developed hypothermia while rabbits and sheep became hyperthermic; during pyrogen-induced fever norepinephrine lowered the body temperature in all these animals. Neither epinephrine nor norepinephrine had any effect on body temperature when injected into the ventricles of the ox. Sedation, vasodilation, increased evaporative heat loss, and a fall in temperature followed injection of 5-HT ( Findlay and Robertshaw, 1967). Large doses of 5-HT administered to oxen maintained in warm environments decreased rectal and hypothalamic temperatures, and the same responses were seen after tranylcypromine. Norepinephrine injected into the ventricles of animals kept in a cold environment ( -l°C) decreased heat production and led to a fall in core and hypothalamic temperatures ( Findlay and Thompson, 1968).

DRUGS AND BODY TEMPERATUFtE

13

These results were not regarded as consistent with the thesis that temperature regulation is mediated by these amines in the ox. Inconsistent and variable responses to intraventricular injection of adrenaline and 5-HT were also reported in the sheep (Bligh, 196613). The occasional rise in temperature seen following 5-HT seemed to be related to excitement and hyperactivity rather than specific thennoregulatory activity. Norepinephrine, epinephrine, and their a-methyl derivatives decreased temperature and oxygen consumption following intravenous injection in chickens up to 23 days old but not in adult fowls. These effects were accompanied by postural changes and sleep. Pentobarbital produced similar effects (D. J. Allen and Marley, 1967). At normal body temperatures and during pyrogen-induced fever, injection of 5-HT into the third ventricle of the goat caused peripheral vasodilation, polypnea, and a marked fall in brain and rectal temperature. Injections of catecholamines were without significant effect ( Andersson et al., 1966). Injection of norepinephrine, epinephrine, dopamine, or 5-HT into the ventricles of conscious mice produced hypothermia in each case ( Brittain and Handley, 1967). Reserpine and p-chlorophenylalanine lowered brain 5-HT in the mouse to the same extent while reserpine and a-methyl-mtyrosine were equally effective in depleting brain norepinephrine. Of the three drugs only reserpine caused a fall in core temperature suggesting a peripheral rather than central site of action (Somerville and Whittle, 1967). Several studies have been made in the rat with inconsistent results. Ingenito and Bonnycastle (1966) estimated the levels of brain 5-HT and norepinephrine after treatment with amine-releasing agents and monoamine oxidase inhibitors in rats exposed to different environmental temperatures. The amine levels in the brain could be increased or decreased pharmacologically without interference with body temperature regulation; conversely drugs which did interfere with thermoregulation did so without changes in whole brain amine levels. There were no significant changes in rat brain concentrations of dopa, dopamine, norepinephrine, or 5-HT during acute exposure to warm or cold environments (Ingenito and Bonnycastle, 1967). Increases in brain norepinephrine levels occurred after cold exposure ( l°C) for 30 days (Ingenito, 1968),but similar increases are found in other organs (heart, liver,

14

PETER LOMAX

spleen, muscle) after cold acclimation (Leduc, 1961). In unanesthetized rats intraventricular injection of 5 H T lowered core temperature while the effect of epinephrine and norepinephrine varied according to the dose administered: small doses (2-6 pg) raised the body temperature, and higher doses (1&100 pg) led to a fall. Tranylcypromine caused hypothermia when injected systemically or centrally; the intraventricular injections were said to be more effective, but no statistical data are given and the results presented are inconsistent on this point (Feldberg and Lotti, 196%). A similar biphasic response was seen in the rat after injection of norepinephrine into the rostra1 hypothalamus (Lomax et al., 1968a); doses of 2 pg or less led to a delayed rise in temperature and higher doses ( 5 fig) caused a slight fall. The contradictory data concerning different species led Kulkarni (1967) to reexamine the responses in conscious cats. He found that 5-HT administered into the lateral cerebral ventricles caused an immediate fall in temperature, followed later by a slight rise. The hypothermia was dose related and seemed to be the major effect. In view of this report, Banerjee, Burks, and Feldberg (1968a) again studied the effect of intraventricular 5-HT in the cat and confirmed the biphasic response. They were of the opinion that the predominant effect is a rise in temperature and attributed the hypothermia to “paralysis of the cells of the anterior hypothalamus” by “excessive” amounts of 5-HTor the use of distilled water, rather than 0.9%NaCl solution, as the solvent. It must be noted, however, that Kulkarni recorded a fall in core temperature with doses (62.5 pg) much smaller than those used by Banerjee and his associates (200 pg) to produce hyperthermia. The electrical activity of neurons in the preoptic area of the dog has been studied and single units identified which increase their firing rates in response to changes in local tissue temperature (Cunningham et aE., 1967). When 5-HT was injected into the ventricles both temperature-sensitive and temperature-insensitive neurons demonstrated a reduction in discharge rate. Administration of epinephrine either into the ventricles or intravenously generally resulted in a reduction of activity of temperature-sensitive neurons with temperature-insensitive cells showing greater variability. These experiments did not provide any evidence for an antagonistic role of 5-HT and epinephrine in the control of body temperature.

DRUGS AND BODY TEMPERATURE

15

Electrical stimulation of the midbrain raphh in anesthetized rats increased the 5-hydroxyindoleacetic acid and lowered the 5-HT concentrations in the forebrain (Aghajanian et al., 1967). These changes indicated that 5-HT can be released by a specific neural pathway. Stimulation of the same site in conscious rats led to a rise in colonic temperature which could be prevented by prior depletion of the stores of 5-HT with reserpine or p-chlorophenolalanine (Sheard and Aghajanian, 1967, 1968). These experiments were interpreted as being consistent with Feldberg’s hypothesis concerning the role of 5-HT in thermoregulation, but the paper must have been in press coincidently with the data of Feldberg and Lotti (1967b) showing that intraventricular injection of 5-HT lowers rectal temperature in the rat. In studies of the circadian rhythms in rat brain amine levels, it was found that alerting behavior coincided with an increase in brain catecholamines ( Friedman and Walker, 1968). The animals’ rectal temperatures were at the highest levels during this same period. Levels of 5-HT in the brain reached a peak at a time when the rats were inactive or asleep and when the core temperature was at its lowest. It cannot be ascertained from these results whether the changes in temperature are related directly to the brain amine levels or whether they are coincidental to the altered activity. Although the relative importance of the various thermoregulatory responses may vary considerably from one species to another, the general pattern of thermoregulation appears to be essentially the same in all homeotherms. Thus, it seems most unlikely that fundamental differences, particularly to the extent of diametrically opposite mechanisms, would have developed in the control of the central thermostat in closely related species. The inconsistencies exhibited by different species appear to form one of the most compelling arguments against acceptance of the ad hoc thesis that the levels of the hypothalamic amines regulate body temperature. It has been argued (Cooper, 1966) that, because the rostra1 and lateral hypothalamus subserves a variety of functions, all of these functions should be examined to determine to what extent species variations occur. Whatever the outcome of such studies, the basic question regarding temperature regulation would remain. The slow recovery from the effects of an intraventricular or intrahypothalamic injection of amines has been noted as negating a physiological role in temperature regulation since responses have been shown to be

16

PETER LOMAX

activated within seconds of a change in hypothalamic temperature (Hammel, 1968). Such long lasting effects are, however, inevitable in these experimental situations where, in effect, an “internal store” of the amine has been created so that the duration of the drug effect becomes a function of the rate of its removal and metabolism. Clearly the amounts injected are several orders of magnitude greater than those that might be released by intrinsic neuronal activity, so that the time scale is similarly distorted. There are many responses involving catecholamines as neurotransmitters which can undergo very rapid changes; vasomotor activity is a case in point. Even if the temperature responses to central administration of the monoamines are regarded merely as pharmacological oddities, the fact of the high concentration of amines in the nerve cells and fibers of the thermoregulatory area remains. The complexity and accuracy of the control system, situated in such a small and inaccessible region, render attempts to unravel its function extremely difficult with the relatively crude tools on hand (Cooper, 1966). Although the balance of evidence renders unlikely the hypothesis that the relative concentrations of amines per se regulate thermoregulatory responses, experiments such as those of Aghajanian and his co-workers and the data derived by Cunningham et al. (1967), in recording unit activity from the preoptic region, are at least suggestive that the monoamines may have some role in the regulation of body temperature.

2. Acetylcholine and Cholinomimetics The concept that acetylcholine may be a neurotransmitter in the central nervous system was suggested by Dale (1934) shortly after discovery of its role in the transmission of signals across certain peripheral synapses. In a recent review Votava (1967) has considered the evidence that has since accrued to support Dale’s suggestion. It is only recently, however, that unequivocal evidence for the presence of acetylcholine in the brain has been obtained (Hanin and Jenden, 1969). Previously the existence of the amine had been inferred from the occurrence of enzyme systems for its synthesis and breakdown in neural tissue and from the demonstration of cholinomimetic activity in extracts of brain. By all of these criteria it would appear that the rostra1 hypothalamus has a relatively high content of acetylcholine. Somewhat surprisingly, the earliest report of the effect of acetyl-

DRUGS AND BODY TEMPERATURE

17

choline on body temperature concerned experiments in man (Henderson and Wilson, 1937). These authors introduced acetylcholine through cranial burr holes into the lateral ventricles of patients in doses up to 7.5 mg. Marked parasympathetic responses followed the injections including nausea, vomiting, defecation, urination and, in several subjects, profound sweating and a fall in body temperature. Intraventricular injection of atropine blocked these effects. The same doses of acetylcholine injected intravenously did not produce any significant changes. One patient had undergone cervical sympathectomy, with denervation of the head, neck, and left upper limb, prior to the central injection of acetylcholine. In this case sweating occurred in all parts of the body other than the denervated areas, suggesting that it was of central origin and was not due to a direct action of acetylcholine on the sweat glands. The parasympathetic hyperactivity was sometimes followed by sleep. The authors were of the opinion that these responses were mediated by acetylcholine acting on autonomic centers in the brain. Several years earlier Cushing (1931) had noted a fall in temperature after intraventricular administration of pilocarpine in man. In the monkey (Light and Bysshe, 1933) and rabbit (Light et al., 1933) injection of pilocarpine into the lateral ventricles had inconsistent results on the core temperature; sometimes a fall occurred in the monkey and shivering was frequently observed in the rabbit. Slow infusion of small doses of acetylcholine (0.88-13 pg) into the rostra1 hypothalamus of rabbits was without effect on the animal's rectal temperature, although subsequent injection of endogenous pyrogen at the same site caused fever (Cooper et al., 1965). Acute injections of acetylcholine (0.5-2.0pg) into the preoptic nuclei of the rat were also without effect on the core temperature (Lomax and Kirkpatrick, 1968), although injection of carbachol at this location leads to hypothermia (Kirkpatrick et al., 1967a; Hulst and De Wied, 1967). Acetylcholine is hydrolyzed extremely rapidly in the tissues and, because it requires some minutes for any effect the drug might have on the central thermostat to manifest itself as a change in heat content of the body mass, it is hardly surprising that these central injections were apparently ineffective. With carbachol, which is much more slowly metabolized, the action is sufficiently prolonged for the temperature changes to occur. Recently we have applied acetylcholine iontophoretically to the preoptic region of the rat at a delivery rate of 1 pg in 10 seconds

18

PETER LOMAX

(Kirkpatrick and Lomax, 1969). Within a few minutes the body temperature started to fall and reached a steady state about 1°C below the control level. When the delivery current was terminated the temperature rapidly returned to the resting level. There have been a number of studies showing that cholinomimetic agents effect body temperature. In the mouse intraperitoneal injection of tremorine, arecoline, or pilocarpine causes a marked fall in rectal temperature (Everett et al., 1956; Spencer, 1965; Zetler, 1968). The hypothermia can be prevented by prior injection of centrally active anticholinergic agents but quaternary compounds such as methyl atropine are without effect in doses which abolish peripheral parasympathetic responses (Spencer, 1x5;Zetler, 1968). Thus the hypothermia appears to be the result of muscarinic stimulation of the central nervous system. The cholinesterase inhibitor paraoxon similarly lowers body temperature in mice protected from the peripheral effects of the drug by pralidoxime, a reactivator of cholinesterase which does not penetrate the central nervous system (Zetler, 1968). The marked tremorogenic and parasympathomimetic properties of tremorine are due to its biotransformation product oxotremorine (Kocsis and Welch, 1960; Cho et al., 1961).Systemic injection of oxotremorine lowered the temperature of rats in which the peripheral effects of the drug were blocked by pretreatment with methantheline. Intraperitoneal injection of atropine or trihexyphenidyl abolished the hypothermia. Microinjection of oxotremorine directly into the preoptic region, in doses ( 2 pg) too low to have any effect when injected systemically, also caused a fall in core temperature. Injection of carbachol (1pg) into the same sites led to a mean fall in temperature, which was not significantly different from that following oxotremorine ( 2 pg). In a further group of animals injection of atropine or trihexyphenidyl into the rostra1 hypothalamus antagonized the hypothermic action of oxotremorine previously injected systemically (Lomax and Jenden, 1966). The compound N - (4-diethylamino-2-butynyl)-succinimide ( DKJ 21) is an anticholinergic agent with a marked selectivity for the central nervous system (Dahlbom et aZ., 1966). Premedication with DKJ 21 prevented the fall in temperature from oxotremorine, and injection of DKJ 21 into the thermoregulatory centers caused an immediate reversal of the falling temperature due to systemic injection of oxotremorine ( Kirkpatrick et al., 196713). These experiments con-

DRUGS AND BODY TEMPERATURE

19

firmthat the hypothermic effect of oxotremorine in the rat is due to a direct action on the hypothalamic thermoregulatory centers. In a recent study (Hammer et al., 1968a) conducted in Stockholm, it was reported that oxotremorine caused “a drastic hypothermia in the mouse but only a minor drop in body temperature in the rat.” In conjunction with Professor Holmstedt we arranged an exchange of rats with the Swedish group; it was found that in our laboratories the Swedish animals developed a fall in core temperature after oxotremorine which was the same as that in local control animals. It is possible that different experimental conditions could account for the results obtained by Hammer and his colleagues. Administration of tremorine to young chicks produced tremor and hypothermia. A significant increase in brain acetylcholine levels was also seen after administration of tremorine to these animals (Bowman and Osuide, 1967). Intravenous injection of sublethal doses of organophosphorous cholinesterase inhibitors lowered the body temperature of anesthetized rats from 4 to 6°C. Quaternary anticholinesterases were ineffective when given systemically but caused hypothermia when injected into the subarachnoid space. These responses were partially blocked by atropine but not by atropine methyl nitrate. The highest dose of cholinesterase inhibitor that allowed survival was without effect on the rectal temperature of mice, guinea pigs, or rabbits at thermoneutral environmental temperatures ( Meeter and Wolthuis, 1968). In most of these studies there has been agreement that cholinergic (or, more specifically, niuscarinic) agents cause a fall in body temperature, and the available evidence suggests that this is due to a direct action on the thermoregulatory centers in the hypothalamus. These results seem to suggest that the thermoregulatory centers contain muscarinic receptors, the activation of which leads to a fall in core temperature. If such a cholinergic link plays a physiological role in normal thermoregulation it might be expected that centrally active antimuscarinic agents would interfere with temperature control. The classic muscarinic blocking agent atropine, together with other belladonna alkaloids, tends to cause a rise in temperature in man when administered in large doses, especially when the environmental temperature is high. This effect has been attributed mainly to suppression of sweating (Cullumbine and Miles, 1956),

20

PEER LOMAX

although a direct action on the thermoregulatory centers in the brain has also been implicated (Innes and Nickerson, 1965). In the mouse, systemic injection of atropine caused a fall in core temperature and the hypothermia was augmented by adrenalectomy (Dutta, 1948). Intraperitoneal injection of atropine in the rat led to a fall in temperature ( Kirkpatrick and Lomax, 1967). The quaternary analogue atropine methyl nitrate had a similar effect, suggesting that the hypothermia is due to an action at peripheral sites. Microinjection of atropine directly into the rostral hypothalamus caused a rise in body temperature. Both of these responses to atropine have a time course of several hours, which is similar to the duration of blockade of known muscarinic receptors. In view of the fall in temperature associated with centrally active cholinergic drugs, the central hyperthennic effect of atropine could be due to blockade of endogenous cholinergic transmitters, notably acetylcholine. In the case of blocking agents that have only weak activity a t peripheral cholinergic sites, such as DKJ 21 (see above), both systemic and intracerebral injection produce hyperthermia ( Kirkpatrick et az.,

196%).

Although studies with the cholinomimetic amines have been less extensive and have covered only a few species, a consistent pattern of effects has emerged which contrasts sharply with the case of the catechol- and indoleamines. However, there does not appear to be any justification for assigning a unique role to acetylcholine in the regulation of body temperature. The evidence considered above suggests only that acetylcholine may be the neurotransmitter at some synapses in the rostral hypothalamus. It is the neuronal pathways involved which will determine the particular responses, i.e., whether heat loss or heat conservation mechanisms are activated. Since cholinergic stimulation leads to a fall in temperature either only the pathways concerned with lowering body temperature are cholinergic in nature or else the cholinergic synapses are involved in the setting of the thermostat, specifically in depressing the set point. In view of the considerable histochemical evidence for the presence of acetylcholine and catecholamines in the preoptic region and the pharmacological studies showing that they can induce changes in body temperature, the question arises as to whether thermoregulation might not involve an interaction between these

21

DRUGS AND BODY TEMPERATURE

transmitter substances. Some data supporting such a view have arisen somewhat indirectly as a result of studies of the tremorogenic agent oxotremorine and of the interest in the therapy of Parkinson's disease. In the mouse, the tremor and hypothermia induced by tremorine are prevented by treatment with centrally active anticholinergic or sympathomimetic drugs ( Spencer, 1965). In this study sympathomimetics were more active in blocking the fall in temperature than in abolishing the tremor, whereas muscarinic blocking agents were effective against both responses. Imipramine and desmethylimipramine antagonize the effects of oxotremorine in the mouse (Spencer, 1966), particularly the hypothennia which is sometimes reversed so that the animals become hyperthermic ( Morpurgo, 1967). If mice are rendered hyperthyroid by administration of thyroxine, they exhibit only a transient hypothermia in response to oxotremorine (Waite and Spencer, 1968). Peripheral and central adrenergic mechanisms are known to be potentiated in hyperthyroid animals, and the inhibition of the action of oxotremorine might be due to enhanced adrenergic activity in the thermoregulatory centers. TABLE I TEMPERATURE CHANGES FOLLOWING ADMINISTRATION OF DRUGS TO RATS

Dmg

Number of animals

Mean fall in core temperature ("Cf S.E.M.)

Pilocarpine (5 mg/kg ip)

8

2 . 6 f 0.24a

Pilocarpine (40 pg ic)

6

1 . 3 f 0.22

f 0.07

Norepinephrine (2.5 pg ic)

16

0.4

Norepinephrine (2.5 pg ic) Pilocarpine (5 mg/kg ip)

18

0 . 9 f O.15"ab

Tolasoline (2.5 f i g ic) Norepinephrine (2.5 p g ic) Pilocarpine (5 mg/kg ip)

10

2 . 9 f 0.12b

The fall in temperature after intraperitoneal (ip) injection of pilocarpine was inhibited by intracerebral (ic) injection of norepinephrine into the preoptie area. Tolasoline injected centrally blocked the effect of norepinephrine. Significantly different, p - < 0.01.

22

PETER LOMAX

A parallel may exist between the interaction of amines in the thermoregulatory centers and the inhibition of cholinergic transmission in the superior cervical ganglion by norepinephrine. The latter effect appears to be due to hyperpolarization of the efferent neurons by norepinephrine ( DeGroat and Volle, 1966) so that they are less responsive to cholinergic stimulation. The receptors in the superior cervical ganglion are believed to be closely allied to cholinergic receptors in the central nervous system (Perry, 1957). 40-

T

N

7 i

i

39

-

2

2

37-

L

P

5 c

36 T-TOLAZOLINE 2.5~9IC N-NOREPINEPHRINE 2 . 5 ~ 9 IC P -PILOCARPINE 5 rnglkg ip

35-

34

I

I

I

I

I

I

2

3

4

5

A V

I 24

25

I

I

I

I

26

27

28

29

FIG.2. Temperature recordings from two rats. The fall in body temperature following intraperitoneal ( ip ) injection of pilocarpine was markedly inhibited by prior intracerebral ( i c ) injection of norepinephrine. Central injection of tolazoline blocked the inhibitory action of norepinephrine. The central injections were into the preoptic area.

The interaction of adrenergic and cholinergic agents in the thermoregulatory centers has been studied in the rat (Lomax et al., 1968a,b) . Pilocarpine lowers the body temperature of this species when injected systemically or into the preoptic region (Table I ) . When norepinephrine was injected into the rostra1 hypothalamus the fall in body temperature following systemic administration of pilocarpine was markedly attenuated. This inhibitory effect of norepinephrine was prevented by injection of the alpha-blocking agent, tolazoline, into the thermoregulatory centers immediately prior to injection of norepinephrine. The data from these experiments are summarized in Table I, and typical experiments are shown in Fig. 2. Intracerebral injection of tolazoline alone did not

23

DRUGS AND BODY TEMPERATURE

lead to any significant changes in core temperature, but the fall in temperature induced by intraperitoneal injection of pilocarpine was potentiated by pretreatment with tolazoline (Fig. 3 ) . If these responses are indeed analogous to the interaction of the drugs at cholinergic receptors in peripheral ganglia, catecholamines could alter the degree of polarization of cholinergic neurons in the

3

P

1-TOLAZOLINE

2 . 5 , ~i c~

P-PILOCARPINE

5.0 mg/kg ip

I

I

I

A I

2 V 24

I

25

I

A 1

26V 48

I

49

50

A1 51V 72

,

73

I

74

7;

HOURS

FIG.3. Recording of core temperature from a rat. Injection of tolazoline into the rostra1 hypothalamus was without significant effect on the temperature. Hypothermia followed intraperitoneal ( ip ) administration of pilocarpine, and this response was potentiated by prior intracerebral ( i c ) injection of tolazoline.

thermoregulatory centers. Such adjustment of the membrane potentials by endogenous catecholamines could be a factor in the physiological regulation of the set point of the central thermostat. Any interpretation of the results of pharmacological experiments of this nature must, however, remain essentially speculative until the action of endogenous transmitters and thermoregulatory responses can be related to unit activity of the neurons in the preoptic region. 3. Pyrogens

There is a variety of agents that cause pyrogenic responses following injection into experimental animals and man (see review by Atkins, 1960). These include the endotoxin extracted from the cell walls of gram-negative bacteria, gram-positive organisms, staphylococcal enterotoxin, streptococcal exotoxins, viruses, and several specific antigens. The serum of animals injected with these pyrogens is subsequently found to contain a substance with pyro-

24

PETER LOMAX

genic activity which is distinct from the original injected material. Since the polymorphonuclear leukocytes appear to be the source of this endogenous pyrogen (Bennett and Beeson, 1953) it is generally referred to as leukocyte pyrogen. It is likely that the fever that accompanies a wide variety of infections is ultimately due to the pyrogen derived from the white cells of the host. Recent studies have centered on the sites of action of the various pyrogens. Leukocyte pyrogen produced a more rapid and greater febrile response in the rabbit when injected into the carotid arteries than when injected intravenously; bacterial pyrogen had the same latent period, which was longer than that of leukocyte pyrogen, by either route of administration. When leukocyte pyrogen was tagged with lS1I and injected systemically, concentration of the radioactive material was found in the hypothalamus but not in other parts of the brain (Allen, 1965). This evidence suggests a specific hypothalamic effect. Intracisternal injection of leukocytic pyrogen in the rabbit produced fever of short latency and of much greater duration than that evoked by intravenous injection (Adler and Joy, 1965). The site of action of pyrogens in the rabbit brain was studied by Cooper et al. (1967) using bilateral intracerebral injections. Fever developed after injection into the preoptic area of the hypothalamus but not following injection into the caudal hypothalamus, the midbrain, the pons, the cerebellum, or the cerebral cortex. The latency of onset of pyrexia was much shorter after injection of leukocytic pyrogen than following administration of bacterial pyrogen. The amount of leukocyte pyrogen needed to evoke a febrile response was only 1/100 the dose required by intravenous injection whereas bacterial pyrogen had to be injected centrally in approximately the same amount as was required to cause a similar fever on intravenous injection. Essentially similar results were obtained in the cat (Jackson, 1967). Although the findings in these studies do not exclude the possibility of a direct effect of the bacterial endotoxins on the thermoregulatory centers, the longer latencies with these agents suggest that their action may involve mobilization of endogenous pyrogen locally. Single unit activity was recorded in the rostra1 hypothalamus of anesthetized rabbits, and the effects of local temperature changes and intravenous injection of typhoid vaccine investigated ( Cabanac et al., 1968). Thermosensitive units were found which increased their firing rates with increases in local temperature while other

DRUGS AND BODY TEMPERATURE

25

units showed a decreased activity. In response to systemic injection of bacterial pyrogen, the former units were inhibited whereas the latter increased both their spontaneous discharges and their temperature sensitivity. Intracarotid injection of bacterial pyrogen depressed the activity of hypothalamic neurons sensitive to increased body temperature in the cat within 15 to 30 minutes (Wit and Wang, 1968). Calorimetric studies of temperature regulation in man (Palmes and Park, 1965) and rats (Kerpel-Fronius et al., 1966) at different environmental temperatures during pyrogen-induced pyrexia have revealed that several different mechanisms may be used to bring about the rise in body temperature. It appears that temperature regulation is not deranged by the action of the pyrogen, and the changes that occur are consistent with the view that there has been an upward shift in the set point. The behavior of the two populations of neurons in the studies by Cabanac et al. (1968) could also indicate such a resetting effect. The evidence from central injection studies would indicate that, at least in the case of leukocyte pyrogen, there is a direct action of the pyrogenic material on the neurons of the thermoregulatory centers. General acceptance of this simple premise has been delayed, however, by suggestions that intermediate transmitters may be involved. Attempts have been made to incorporate the data from experiments with pyrogens into the general hypothesis that thermoregulatory responses stem from changes in brain monoamine concentrations. The temperature changes following intraventricular injection of 5-HT or catecholamines in normothermic cats (see Section III,C,l) can be elicited also during pyrogen-induced fever (Feldberg and Myers, 1964a). If the effect of the pyrogen is only to raise the set point, leaving normal regulatory mechanisms intact, this finding is to be expected. The fever induced by intravenous injection of leukocyte pyrogen was prolonged slightly in rabbits pretreated with a monoamine oxidase inhibitor ( pargyline). In two animals given higher doses of pyrogen after the monoamine oxidase inhibitor, a fatal hyperpyrexia developed (Cooper and Cranston, 1966). Similar modii3cations of the hyperthennia following leukocyte pyrogen by alterations in the amine stores were demonstrated in the rabbit by Giannan et al. (1968). It is not possible to say from these data whether peripheral or central monoamines, or both, were implicated. While such studies indicate that

26

PETER LOMAX

pharmacological manipulations can markedly change the response to endogenous pyrogens, none of these studies detract from the supposition that the active pyrexial agents act directly on the cells of the rostra1 hypothalamus. In a recent communication Feldberg (1968) was in concurrence with this view.

D. DRUGS 1. Antipyretics

It is generally stated that antipyretics reduce fever but have no effect on animals with normal body temperatures. While this statement holds for salicylates, other antipyretics, including acetophenetidin, antipyrine, quinine, and aminopyrine can cause a fall in rectal temperature in a variety of species if administered in fairly high doses. There is no indication, however, that these responses are due to specific disturbances of thermoregulatory control. Salicylates are most effective in attenuating or eliminating the fever resulting from systemic injection of bacterial or leukocyte pyrogen. In the rabbit, the fever due to leukocyte pyrogen is reduced by salicylate (300 mg iv) to a level equivalent to that seen when half the amount of pyrogen alone is given (Cooper et al., 1968). Prior injection of salicylate did not, in these experiments, prevent the rise in temperature resulting from injection of leukocyte pyrogen directly into the cerebral ventricles. These authors concluded that salicylate does not block the action of the pyrogen on the hypothalamus. It was suggested that possibly salicylates interfere in some way with the entry of the leukocyte pyrogen into the preoptic region. The discharge frequency of some preoptic neurons in the cat was found to increase when the body temperature increased. Systemic injection of pyrogen during this period depressed the activity of these neurons while subsequent injection of salicylate restored the resting firing rate within 30 to 70 minutes. These studies indicated interaction of pyrogen and salicylate on the thermoregulatory neurons (Wit and Wang, 1968). The interesting suggestion has been put forward by Shubert (1960) that the antipyretic action of salicylates and similar drugs might be caused by their ability to chelate copper. During infection the copper levels in the blood are increased due to release from intracellular stores. The complex of salicylate with copper would aid the transport of copper back into the cells. Perhaps

DRUGS AND BODY TEMPERATURE

27

worthy of comment in this respect is the observation that the destruction of catecholamines is catalyzed by copper with which they form oxidizable chelates (Chenoweth, 1956). 2. Morphine

Studies of the effect of morphine on body temperature have been far from consistent; both hyper- and hypothermic responses have been reported (Helfrich, 1934; Reynolds and Randall, 1957). To some extent these inconsistencies have been due not only to marked species variation in response, but also to differences in dosage and route of administration in comparing studies in a single species. Moreover, the effect of the drug may change with repeated injections due to the development of tolerance; the initial hypothermic effect in the monkey can be converted to a hyperthermic response (Eddy and Reid, 1934), and a similar change has been reported in the rat after a single previous injection (Lotti et al., 1966a). In spite of these difhulties the general consensus would indicate that, if an adequate dose is used, an initial injection of the drug produces hypothermia in many species including man, monkeys, dogs, rabbits, guinea pigs, and rats. Administration of morphine to cats, horses, and cattle usually results in excitement, hyperactivity, and a rise in body temperature. Studies in the rat indicated that 5 mglkg, or less, of morphine sulfate injected intravenously caused a slight rise in temperature or was without effect. Increasing the dose to 10 mglkg produced hypothermia and a maximum response occurred with 35 mglkg (Lotti et al., 1965a). The hypothermic effect was localized to the preopticl anterior hypothalamic nuclei by intracerebral microinjection studies. In some animals, injection of morphine (50 pg) into the region of the supramammillary nuclei caused hyperactivity and a rise in temperature. In recent studies, we have noted a rise in temperature following injection of small doses of morphine (15 pg) into the preoptic region (Lomax and Foster, 1968). A pronounced reduction in oxygen consumption occurred immediately after either intravenous administration of morphine or injection of the drug into the rostra1 hypothalamus (Lotti et al., 196613). This decrease in metabolism preceded the fall in temperature, and a direct relationship existed between the magnitude of the depression of oxygen consumption and the rate of fall in core temperature. In the case of the central injections, the hypothermia appeared to

28

PETER LOMAX

be due solely to the decrease in heat production; after systemic injection there was an additional component contributing to the decline in core temperature indicating that a part of the response was mediated at a site other than the thermoregulatory centers. The skin temperature exactly paralleled the core temperature so it would seem that, in the rat, morphine-induced hypothermia is due solely to decreased heat production. If this is in fact the case, then the metabolic response would appear to be a more reliable index of the drug effect than the body temperature since the rate and magnitude of fall in the latter will be a function of the animals' ability to lose heat and will vary with the environmental temperature and the particular experimental conditions. An example of this is seen in the studies of Paolino and Bernard (1968) in which systemic or central injection of morphine decreased the body temperature of rats at low (SOC) but not at high (32°C) environmental temperatures. Further evidence for a direct action of morphine on the thermoregulatory centers was the finding that injection of nalorphine into the preoptic region will reverse the falling body temperature due to prior intravenous injection of morphine ( Lotti et al., 1965b). Injection of nalorphine alone, either centrally or systemically, produced a rise in temperature, of a degree or so, lasting several hours (Lotti et al., 1965b). This late hyperthennia was particularly marked when the action of morphine was antagonized by nalorphine, and it was suggested that it might represent, in part, a stimulatory action of morphine. The development of acute tolerance to the hypothermic effect of morphine can be demonstrated with both systemic and focal injections (Gunne, 1960; Lotti et al., 1966a). Tolerance to the hypothermic action of systemic morphine was prevented by injection of nalorphine into the rostra1 hypothalamus; subsequent systemic injection of morphine led to a fall in rectal temperature, but the animals did not become analgesic (Lomax and Kirkpatrick, 1967). It was concluded that acute tolerance to the temperature lowering effect of morphine is the result of changes occurring at receptors in the thermoregulatory centers. The quaternary N-methyl derivative of morphine was without effect on the core temperature of the rat when injected intravenously, but a fall in temperature occurred after injection into the preoptic region. Surprisingly the hypothermic action of N methyl morphine was not blocked by intraperitoneal injection of nalorphine (Foster et al., 1967).

DRUGS AND BODY TEhKPERATURE

29

Intraventricular injection of morphine in anesthetized cats produced shivering and a rise in body temperature (Banerjee et al., 1968b). These changes were abolished by prior intraventricular injection of nalorphine, but if the nalorphine was injected after hyperthermia had developed it had little effect. In these experiments the barbiturate anesthesia induced a sharp fall in rectal temperature, and this could be prevented by manipulation of the animals, e.g., by fixing the head in a sterotaxic instrument. The rise in temperature after such maneuvers was similar in degree and duration to that following injection of morphine. Ergotamine injected into the ventricles also led to shivering and a rise in core temperature and abolished the response to morphine. It is somewhat diEcult to draw firm conclusions as to the site or mechanism of action of morphine from these data. The role played by indole- and catecholamines in the central actions of morphine has been the subject of much experimental work. The analgesic effect of morphine is abolished in rabbits and mice by pretreatment with reserpine or a-methyltyrosine (Pacile and Muiioz, 1968; Verri et al., 1968). Pretreatment with a monoamine oxidase inhibitor increased the toxicity and hypothermic effect of morphine in rats and depressed the development of tolerance (Chodera, 1!363). Banerjee and his associates (196813) were unable to determine from their data whether or not monoamines might be implicated in the temperature responses to morphine in the cat. The intercession of a neurohumoral transmitter and an indirect action of morphine on the thermoregulatory centers was postulated by Paolino and Bernard (1968) to explain the 2-hour interval between the injection of the drug and the achievement of the maximum temperature change. The argument in this last paper is not too clear because the author's data indicate that the core temperature in their animals began to fall immediately after injection of the drug into the rostra1 hypothalamus. In preliminary studies in the rat, administration of reserpine (0.25 mglkg twice daily for 2 days) did not modify the hypothermic response to systemic injection of morphine (25 mg/kg ip) (Kirkpatrick and Lomax, 1968). Some attention has been drawn to the role of morphine in depressing cholinesterase activity (Bernheim and Bernheim, 1936) and inhibiting acetylation of choline ( Morris, 1961). The hypothermic action of morphine is uneffected by atropine (Lomax, 1967a), and

30

PETER LOMAX

the available evidence would not seem to implicate acetylcholine in the action of morphine on the thermoregulatory centers. None of the studies considered above provide direct information as to the mechanism by which the temperature responses to morphine are mediated. According to the species, catatonia or excitement may follow systemic administration so that metabolic heat production in muscle may be a factor. Intracranial injection does not always eliminate this factor since shivering and hyperactivity can occur after focal application of the drug. Morphine certainly seems to exert a direct action on the hypothalamic thermoregulatory centers, and changes in the set point could explain some of the changes in body temperature. Further studies are needed to decide these points.

3. General Anesthetics General anesthesia, irrespective of the drug used for its induction, causes homeothermic animals to become poikilothermic. This is usually credited as being the result of nonspecific depression of the central nervous system rather than any selective action on the thermoregulatory centers. Injection of small doses of anesthetic agents into the cerebral ventricles or the rostral hypothalamus can produce changes in core temperature similar to those following systemic administration. Intraventricular injection of pentobarbital or chloralose in cats kept at 22°C led to a fall in rectal temperature and the induction of a soporific state (Feldberg and Myers, 1965b). Direct injection of chloralose into the rostral hypothalamus resulted in a renewed fall in temperature in cats recovering from the effects of barbiturate anesthesia ( Feldberg and Myers, 196513). Pentobarbital lowered the core temperature and reduced oxygen consumption when injected into the preoptic region of unanesthetized cats (Jacobson, 1966). There was no change in the rectal temperature of rats when pentobarbital (30 pg ) was injected into the preopticlanterior hypothalamic nuclei, whereas morphine caused profound hypothermia when injected at the same sites (Lomax, 1966). No correlation was found between the rate of fall of body temperature and the reduction in oxygen consumption in the rabbit (Daudova, 1961) or the rat (Lomax, 1966) during barbiturate anesthesia. In the latter study, the barbiturate induced a marked increase in cutaneous

DRUGS AND BODY TEMPERATURE

31

blood flow, and this appeared to be the predominant factor governing the loss of heat from the body. The importance of peripheral heat dissipation to the fall in temperature was seen in studies of fowls under halothane anesthesia ( Marley and Stephenson, 1968). Larger falls in temperature occurred in young than in adult animals, and the recovery of temperature was more rapid in adults. These differences were attributed to the larger surface area in relation to body mass in chickens and to the relatively poor thermal insulation of down compared with feathers. The fall in body temperature in man during anesthesia with a variety of agents was mainly dictated by the rate of loss of heat from the skin which exceeded the rate of heat production. The decline in metabolic rate in these patients was due to the decreased muscle activity and was depressed further by administration of curare (Goldberg and Roe, 1966). The spontaneous firing rate of preoptic neurons has been studied in the dog after administration of thiopental, ether, or chloraloseurethane ( Murakami et al., 1967). Small amounts of the anesthetics reduced the spontaneous firing rate of neurons having a high activity. The thermal sensitivity of warm-sensitive neurons was decreased and often suppressed by thiopental and chloralose-urethane but not by ether. Thiopental also lowered the spontaneous frequency of firing of temperature insensitive neurons. During anesthesia there is a general decrease in activity and muscle tone, so that if the animal is exposed to normal laboratory temperatures its body temperature tends to fall. There is also a reduction in metabolic heat production in the liver during barbiturate anesthesia. Increased cutaneous blood flow contributes to the loss of heat. Depression of transmission in autonomic ganglion could contribute significantly to the vasodilation after administration of barbiturates. The reduction in central neuronal activity and sensitivity abolishes the normal physiological responses to this thermal stress. The onset of shivering during recovery from anesthesia indicates the functional recovery of temperature regulation, and this may precede the return to consciousness suggesting that the thermoregulatory centers are somewhat less sensitive to the anesthetic agent. However, anesthetics cause such gross disturbances of the neural control of shivering that conclusions based on the threshold levels at which shivering stops and starts have little meaning ( Hemmingway, 1963 ) .

32

PETER LOMAX

It is apparent that the loss of normal thermoregulatory control during general anesthesia is the result of the combined disruption of both central and peripheral mechanisms. There do not seem to be compelling arguments in favor of any predominant site of action. 4. Imidaxolines

The naphthyl substituted alkyl imidazolines were initially studied by Hartmann and Isler (1939) and found to have sympathomimetic properties. Two of these derivatives, naphazoline and tetrahydrozaline, were subsequently introduced into clinical practice as vasoconstrictors for local application to the nasal mucous membranes. Acute toxic reactions to these compounds have been reported in children; in 1961 Mikkelsen found 25 cases in the literature and added two of his own; three further cases were reported by Segagni (1962) the following year. The toxic effects mainly involved the central nervous system: severe drowsiness, sweating, respiratory depression, and shock have been seen. In most of the reports marked hypothermia was a prominent feature. The hypothermic effect of naphazoline and a related sympathomimetic, 4-amyl-N-benzhydryl pyridinium ( B-45), was studied in several species, including man, by Gylfe et al. (1950) and Pfeif€er et al. (1950) who reported a consistent fall in body temperature after systemic administration. Because B-45 contains a quaternary nitrogen atom it will not enter the brain so that it is likely that the fall in core temperature is due to a peripheral action of the drugs. Intraperitoneal injection of naphazoline, tetrahydrozoline, or B-45 in relatively low doses (0.5-1.0 mg/kg) caused a fall in rectal temperature in the rat. Injection of the drugs (5 pg) into the preoptic region caused the body temperature to rise after a latent period of up to 15 minutes. The maximum effect was reached in 90 minutes, then the temperature declined to the resting level over the next 90 to 150 minutes (Lomax, 196713; Lomax and Foster, 1967). These studies indicate that the hypothermia produced by the imidazolines is not due to a direct action on the thermoregulatory centers but is probably mediated peripherally. The compounds may interfere with carbohydrate metabolism (Bain and Kohlenbrenner, 1950), and naphazoline has been shown to cause peripheral vasodilation in the mouse (Richter, 1964). Thus, the fall in

DRUGS AND BODY TEMPERATURE

33

temperature may be due to the combined effects of decreased heat production and increased heat loss. The time course and magnitude of the rise in rectal temperature, when the imidazolines are injected centrally, are similar to those seen after injection of small doses of norepinephrine into the thermoregulatory centers (Lomax et aZ., 1968a) or cerebral ventricles (Feldberg and Lotti, 1967b) of the rat. The imidazolines interfere with cholinergic transmission in the superior cervical ganglion of the cat in the same way as norepinephrine, and this action is abolished by alphu-blocking drugs (Lomax, 196%). Possibly the mode of action of the imidazolines and catecholamines on the hypothalamus is the same in that they hyperpolarize the thermoregulatory neurons and raise the set point of the thermostat. There has been some recent interest in a new imidazoline derivative, 2- (2,6-dichlorphenylamino) -2-imidazoline hydrochloride ( S t 155, Catapresan) (Hoefke and Kobinger, 1966). This agent is a potent inhibitor of spontaneous activity of the brain stem sympathetic centers (Schmitt et al., 1968). Injection of St 155 intravenously reduces or abolishes discharges in the sympathetic nerves, and similar effects can be induced by intracisternal injection of doses which are without effect if given systemically. Evoked potentials in the splanchnic nerves during stimulation of the hypothalamus were also decreased. There do not appear to be any reports concerning the effect of St 155 on body temperature. 5. Psychotropic Drugs a. Chlorpomadne. Systemic administration of chlorpromazine caused a fall in core temperature, in all species examined, when the environmental temperature was lower then the animals’ neutral temperature. At high environmental temperatures the drug may lead to hyperthermia. Rats maintained at 23°C show a fall in rectal temperature, decreased oxygen consumption, and increased cutaneous blood flow in the tail (Kollias and Bullard, 1964; Lotti et al., 1966b). Behavioral thermoregulation was depressed in rats by doses of chlorpromazine which accelerated heat loss in the cold (Weiss and Laties, 1963). It was suggested from these findings that the drug abolishes all mechanisms of temperature regulation for both heat and cold due to a specific action on the thermoregulatory centers (Kollias and Bullard, 1964). A somewhat different conclusion was reached by Ulrich and his co-workers (1967) who

34

PETER LOMAX

found that in guinea pigs and rats the fall in body temperature after injection of chlorpromazine was paralled by a reduction in oxygen consumption. The metabolism then showed a reactive increase, which arrested the falling temperature. This was regarded as evidence that the thermoregulatory centers were functional and were activated by the hypothermia. Microinjection of chlorpromazine into the rostra1 hypothalamus or third ventricle of rats caused a long-lasting rise in body temperature of up to 2°C; injections into other sites in the brain were without effect. If the rats were pretreated with reserpine, intrahypothalamic injection of chlorpromazine had less hyperthermic activity. Control injections of pyrogen-free water into the same central sites caused a prolonged fall in temperature in these reserpinized animals ( Rewerski and Jori, 1967). These experiments were carried out under ether anesthesia; the authors present only the temperature changes occurring in response to administration of chlorpromazine, but in a related study from the same laboratory (Jon and Garattini, 1968) the resting body temperature of the rats was as low as 27°C under these conditions. The phenothiazines have a general depressive action on the central nervous system, especially on subcortical structures including the hypothalamus, and might be expected to cause a nonspecific interference with thermoregulation. There are also peripheral actions of these compounds including alpha-adrenergic blocking activity and effects on oxidative phosphorylation. It is likely that the temperature changes after administration of chlorpromazine are due to both central and peripheral activity. b. Amphetamine. The body temperature is increased by systemic administration of amphetamine to mice, rats, rabbits, and cats. The magnitude and time course of the hyperthennia was closely correlated with behavioral excitation in rats (Morpurgo and Theobald, 1965). Drugs such as chlorpromazine, desimipramine, and other neuroleptic agents that reduce the behavioral effects of amphetamine also decrease the hyperthermic response (Morpurgo and Theobald, 1967). A clear enhancement of the peak hyperthermic effect of amphetamine was seen in animals pretreated with cholinergic blocking agents such as atropine, but atropine methyl nitrate was without effect. Amphetamine decreased the norepinephrine levels in the hypothalamus of the rat (Baird, 1968), and a correlation was found between the lowering of brain norepinephrine and the elevation of body temperature (Beauvallet et al., 1967).

DRUGS AND BODY TEMPERATURE

35

The temperature changes induced by amphetamine appear to be due mainly to the increased activity, resulting from the central effects of the drug, rather than direct stimulation of the thermoregulatory centers. It is possible that central release of catecholamines could also contribute to the hyperthermia. c. Desmethylimipramine ( Desipramine ) . Normothermic animals do not show any changes in body temperature after systemic administration of desipramine. If hypothermia is induced by prior injection of reserpine, then desipramine leads to a sustained increase in body temperature (Garattini and Jori, 1967). A small rise in temperature occurred when desipramine was injected into the hypothalamus of rats under ether anesthesia; the response was not dose related nor was it specific for any particular region of the hypothalamus, thalamus, or third ventricle. The effect of these intracerebral injections was pontentiated by reserpine ( Rewerski and Jori, 1968). The fall in core temperature following oxotremorine in the rat is potentiated and prolonged by pretreatment with desipramine (Sjoqvist et al., 1968). This is due partly to inhibition of the metabolism of oxotremorine. There is some evidence also that desipramine influences the action of oxotremorine on central neurons; combined administration of the drugs causes a greater increase in brain acetylcholine levels than occurs with oxotremorine alone, although desipramine itself has no effect on the levels (Hammer et al., 1968b). The central actions of oxotremorine are essentially cholinergic and may be mediated by release of acetylcholine. There are no data to support the view that desipramine has any important direct action on central cholinergic neurons. 6. Lysergic Acid Diethylamide (LSD 25) There have been many reports that LSD 25 leads to hyperthermia. In rabbits, small doses given by various routes evoked a significant rise in rectal temperature (Horita and Dille, 1955) and similar changes have been reported in rats (Sheard and Aghajanian, 1967). It has been postulated that the hyperthermic effect of LSD 25 is due to centrally induced excitation of the sympathetic nervous system (Rothlin et al., 1956). Decorticated rabbits still became hyperthermic after LSD 25, whereas the effect was abolished by decerebration; it was concluded that the diencephalon was the probable site of action of the drug (Neuhold et al., 1957). Investigations of the mode of action of LSD 25 have suggested it induces central excitation by the same mechanisms as SHT. The

36

F'FXXR LOMAX

responses to high brain levels of 5-HT were similar to those following LSD 25 (Horita and Gogerty, 1958) and Elder and Shellenberger (1962) suggested that LSD 25 and 5-HT acted at the same, or similar, sites to produce hyperthermia. In a study of a series of tryptamine derivatives which raise body temperature, it was concluded that these agents and LSD 25 act on the same central receptors as 5-HT ( Brimblecombe, 1967). As was noted in Section C,1, electrical stimulation of the caudal midbrain raph6 mediates the release of endogenous 5-HT and leads to a rise in rectal temperature in rats (Sheard and Aghajanian, 1967). The spontaneous firing rate of neurons in the caudal midbrain containing 5-HT was reversibly inhibited by parenteral injections of LSD 25 which were too low to cause gross behavioral effects (Aghajanian et al., 1968). It is not clear whether or not LSD 25 exerts any direct action on the thermoregulatory centers. 7. Hormones Several endocrine systems are involved in thermoregulation, particularly during exposure to extremes of environmental temperature and in acclimation to high or low temperatures. These responses may be activated directly by the temperature controlling centers: Local cooling of the preoptic region of the goat caused marked thyroid activation, whereas warming the same region retarded the release of radioactive iodine from the gland; the excretion of epinephrine and norepinephrine was increased during hypothalamic cooling and the normal rise in blood levels of catecholamines, in response to exposure to a low environmental temperature, was prevented by central warming (Andersson et al., 1964). Acute lowering of the environmental or preoptic temperature activated the pituitary-adrenal system in dogs leading to an elevation of plasma cortisol levels (Chowers et al., 1964). In man, but not in animals, the steroid metabolite etiocholanolone is a potent pyrogen (Kappas et al., 1957). The degree and duration of the pyrexia was significantly greater in men than in women (Kimball et al., 1966). The thermogenesis is accompanied by severe chills, suggesting that a central mechanism is operative. Direct injection of cortisol into the rostra1 hypothalamus of rabbits partially inhibited the rise in temperature following intravenous injection of exogenous pyrogen. The same dose and volume

DRUGS AND BODY TEMPEXATURE

37

of cortisol was without effect on pyrogen fever when injected into other loci in the brain or intravenously (Chowers et at., 1968). How far these data reflect the role of cortisol in the physiological control of body temperature is uncertain. IV. Concluding Remarks

Interest in the effects of drugs on temperature regulation has proved to be a common meeting point for investigators traveling along several different roads. Pharmacological agents have been used as tools to study the mechanisms of thermoregulation itself; the effects on the temperature regulating centers have been used as models for the general neurological actions of drugs; temperature changes have been utilized as a convenient method of comparing the potencies of centrally active agents which have other primary effects; the modification by drugs of neurotransmitters in the central nervous system, which in turn lead to changes in core temperature, has been employed in an attempt to unravel the complexities of transmission in the central nervous system. The validity of these several types of study depends essentially upon the demonstration of a specific action of the various agents on the thermoregulatory centers. As we have seen, with the methods of examination available, this can be an extremely difficult point to establish with certainty and in many cases there is even doubt as to whether the predominant site of action of a compound is within the central nervous system at all. It is only when these initial hurdles have been surmounted that we can proceed to determine the detailed mechanisms by which the various responses are mediated. REFERENCES Abrams, R. M., and Caldwell, F. T. (1967). Federation Proc. 26, 556. Abrams, R. M.,and Hammel, H. T. (1965). Am. J . Physiol. 208, 698. Adler, R. D.,and Joy, R. T. (1965). Proc. Soc. Exptl. Biol. Med. 119, 660. Aghajanian, G. K., Rosecrans, J. A., and Sheard, M. H. (1967). Science 156, 402.

Aghajanian, G. K., Foote, W. E., and Sheard, M. H. (1968). Science 161, 706. Albers, C. (1961). Arch. Ges. Physiol. 274, 125. Allen, D.J., and larley, E. (1967). Brit. 1. Phurmacol. 31, 290. Allen, I. V. (1965). Brit. J. Erptl. Puthol. 4.6,25. And&, N. E., Dahlstriim, A., Fuxe, K., and Larsson, K. (1965). Life Sci. 4, 1275.

38

PETER LOMAX

Andersen, H. T., Anderson, B., and Gale, C. (1962). Acta Physiol. Scand. 54, 159. Andersson, B., Gale, C. C., and Hokfelt, B. (1964). In “Major Problems in Neuroendocrinology” (E. Bajusz and G. Jasmin, eds.), p. 42. Karger, Basel. Anderson, B., Jobin, M., and Olsson, K. (1966). Acta Physiol. Scand. 67, 50. Atkins, E. (1960). Physiol. Rm. 40, 580. Bain, J. A., and Kohlenbrenner, R. M. (1950). Federation Proc. 9, 255. Baud, J. R. C. (1968). J. Phumn. P h a m c o l . 20, 235. Banerjee, V., Burks, T. F., and Feldberg, W. (1968a). J. Physiol. (London) 195, 245. Banejee, V., Feldberg, W., and Lotti, V. J. (1968b). Brit. J. Phurmcol. 32, 523. Beauvallet, M., Fugazza, J., and Legrand, M. (1967). Compt. Rend. SOC. Biol. 161, 2161. Bennett, I. L., and Beeson, P. B. (1953). J. Exptl. Med. 98, 493. Bergmann, C. (1845). Arch. Anat. Physiol. p. 300. Bernheim, F., and Bernheim, M. L. C. (1986). J. Phumcol. 57, 427. Bligh, J. (1966a). Biol. Reu. 41, 317. Bligh, J. (1966b). J. Physiol. (London) 185, 46P. Bligh, J. (1967). Enoiron. Res. 1, 28. Borison, H. L., and Clark, W. G. (1967). Aduan. Phumcol. 5, 129. Bowman, W. C., and Osuide, G. (1967). European J . Phannacol. 1, 71. Brimblecombe, R. W. (1967). Intern. 1. Neurophumcol. 6, 423. Brittain, R. T., and Handley, S. L. (1967). J. Physiol. (London) 192, 805. Brodie, B. B., and Shore, P. A. (1957). Ann. N . Y. Acud. S c i . 66, 631. Cabanac, M., Stolwijk, J. A. J., and Hardy, J. D. (1968). J. Appl. Physiol. 2% 645. Carlsson, A,, Falk, B., Hillarp, N. A., and Torp, A. (1962). Actu Physiol. Scand. 54, 385. Chattonet, J., Cabanac, M., and Mottaz, M. (1964). Compt. Rend. SOC. Biol. 58, 1354. Chenoweth, M. B. (1956). Pharmcol. Reu. 8, 57. Cho, A. K., Haslett, W. L., and Jenden, D. J. ( 1961). Biochem. Biophys. Res. Commun. 5, 276. Chodera, A. (1963). Arch. Intern. Phurmucodyn. 144, 362. Chowers, I., Hammel, H. T., Stromme, S. B., and McCann, S. M. (1964). Am. J . Physiol. u)7, 577. Chowers, I., Conforti, N., and Feldman, S. (1968). Am. J. Physiol. 214, 538. Cooper, K. E. (1966). Brit. Med. Bull. 220 238. Cooper, K. E., and Cranston, W. I. (1966). Nature 210, 203. Cooper, K. E., Cranston, W. I., and Honour, A. J. ( 1965). J. Physiol. (London) 181, 852. Cooper, K. E., Cranston, W. I., and Honour, A. J. (1967).J. Physiol. (London) 191,325.

Cooper, K. E., Grundman, M. J., and Honour, A. J. (1968). J. Physiol. (London) 196,56P.

DRUGS AM) BODY TJ3hIPEFtATURJ3

39

Cowles, R. B., and Bogert, C. M. (1944). Bull. Am. Museum Nut. Hkt. 83, 261. Cranston, W. I., and RosendorE, C. ( 1967). J. Physiol. (London) 193, 359. Cullumbine, H., and Miles, S. (1956). Quart. J. Exptl. Physiol. 41, 162. Cunningham, D. J., Stolwijk, J. A. J., Murakami, N., and Hardy, J. D. (1967). Am. J. Physiol. 213, 1570. Cushing, H. (1931). Proc. Natl. Acad. Sci. U . S. 17, 171. Dahlbom, R., Karlen, B., George, R., and Jenden, D. J. (1966). Life Sci. 5, 1625. Dale, H. H. (1934). Brit. Med. J. I, 835. Daudova, G. M. ( 1 W l ) . Byul. Eksperim. Bwl. i Med. 52, 48. Dawkins, M. J. R., and Scopes, J. W. (1965). Nuture 206, 201. Decima, E., and George, R. ( 1964). Ebctroencephulog. Clin. Neurophysiol. 17, 438. DeGroat, W. C., and Volle, R. L. (1966). J. Pharmucol. 154, 20. Downey, J. A., Mottram, R. F., and Pickering, G. W. (1964). J. Physiol. (London) 170, 415. Dutta, N. K. (1948). Brit. J. Pharmucol. 3, 246. Eddy, N. B., and Reid, J. G. (1934). J. Pharmucol. 52, 468. Elder, J. T., and Shellenberger, M. K. (1962). J. Pharmcol. 136, 293. El Hawary, M. B. E., Feldberg, W., and Lotti, V. J. (1967). J. Physiol. (London) 188, 131. Everett, G. M., Blockus, L. E., and Shepperd, I. M. (1956). Science 124, 79. Feldberg, W.(1968). Proc. Intern. Union Physiol. Sci. 6, 65. Feldberg, W.,and Lotti, V. J. (1967a). J. Physiol. (London) 190, 203. Feldberg, W., and Lotti, V. J. (1967b). Brit. J. Phamucol. 31, 152. Feldberg, W., and Myers, R. D. (1963). Nature 200, 1325. Feldberg, W.,and Myers, R. D. (1964a). 1. Physiol. (London) 173, 226. Feldberg, W., and Myers, R. D. (196413). J. Physiol. (London) 175, 464. Feldberg, W.,and Myers, R. D. (1965a). J. Physiol. (London) 177, 239. Feldberg, W., and Myers, R. D. (1965b). J. Physiol. (London) 179, 509. Feldberg, W., and Myers, R. D. (1966). J. Physiol. (London) 184, 837. Feldberg, W., Hellon, R. F., and Lotti, V. J. ( 1967). J. Physiol. (London) 191, 501. Findlay, J. D., and Robertshaw, D. (1967). J. Physiol. (London) 189, 329. Findlay, J. D., and Thompson, G. E. (1968). J. Physiol. (London) 194, 809. Foster, R. S., Jenden, D. J., and Lomax, P. (1967). J. Pharmacol. 157, 185. Friedman, A. H., and Walker, C. A. ( 1968). J. Physiol. (London) 197, 77. Fusco, M. M., Hardy, J. D., and Hammel, H. T. (1961). Am. J. Physiol. 200, 572. Fuxe, K., and Ungerstedt, U. (1966). Life Sci. 5, 1817. Gale, C. C., and Ruch, T. C. (1966). Federation Proc. 25, 123. Garattini, S., and Jori, A. (1967). In “First International Symposium on Antidepressant Drugs” (S. Garattini and M. N. G. Dukes, eds.), Excerpta Med. Found. Intern. Congr. Ser. No. 122, p. 179. Excerpta Med. Found., Amsterdam. Geschickter, E. H., Andrews, P. A., and Bullard, R. W. (1966). J. Appl. Physiol. 21, 623.

40

P h m LOMAX

Giarman, N. J., Tanaka, C., Mooney, J., and Atkins, E. (1968). Advan. Phurmacol. 6 4 307. Glowinski, J., Iversen, L. L., and Axelrod, J. (1966). J. Phamacol. 151, 385. Goldberg, M. J., and Roe, F. (1966). Arch. Surg. 93, 365. Grayson, J., and Mendel, D. (1956). J. Physiol. (London) 133, 334. Griffith, F. R. (1951). Physiol. Reu. 31, 151. Gunne, Lars-M. (1960). Arch. Intern. Pharmucodyn. 129, 416. Gylfe, J., Hendricks, R., Ochs, S., and Williams, H. B. (1950). Federation Proc. 9, 280. Hammel, H. T. (1968). Ann. Rev. Physiol. 30, 641. Hammel, H. T., Jackson, D. C., Stolwijk, J. A. J., Hardy, J. D., and Strpmme, S. B. (1963). J. Appl. Physiol. 18, 1146. Hammer, W., Karl&, B., Rane, A., and Sjoqvist, F. (1968a). Life Sci. 7, 197. Hammer, W., Holmstedt, B., Lundgren, G., and Sjoqvist, F. (196813). Cited by Sjiiqvist et al. (1968). Hanin, I., and Jenden, D. J. (1969). Biochem. Phurmacol. 18, 837. Hanin, I., Jenden, D. J., and Lamb, S. I. (1968). Proc. Western Pharmucol. SOC. 11, 144. Hardy, J. D. (1961). Physiol. Rev. 41, 521. Hartmann, M., and Isler, H. (1939). Arch. Exptl. Pathol. Pharmakol. 192, 141. Hayward, J. N. (1968). Exptl. Neurol. 21, 201. Hayward, J. N., and Baker, M. A. (1968). Am. J. Physiol. 215, 389. Hayward, J. N., Smith, E., and Stuart, D. G. (1966). Proc. SOC. Exptl. Biol. Med. 121, 547. Helfrich, L. S. (1934). Arch. Intern. Pharmacodyn. 49, 259. Hellstr@m,B., and Hammel, H. T. ( 1967). Am. J. Physiol. 213, 547. Hemmingway, A. (1963). Physiol. Rev. 43, 397. Henderson, W. R., and Wilson, W. C. (1937). Quart. J. Exptl. Physiol. 26, 83. Hoefke, W., and Kobinger, W. (1966). Arzneimittel-Forsch. 16, 1038. Horita, A., and Dille, J. M. ( 1955). J. Pharmacol. 113, 29. Horita, A., and Gogerty, J. H. (1958). J. Pharmacol. 122, 195. Hulst, S. G. T., and De Wield, D. (1967). Physiol. Behavior 2, 367. Ingenito, A. J. (1968). Proc. SOC. Exptl. Biol. Med. 127, 74. Ingenito, A. J., and Bonnycastle, D. D. (1966). Can. 1. Physiol. P h a m c o l . 45, 733. Ingenito, A. J., and Bonnycastle, D. D. (1967). Can. J. Physiol. Pharmucol. 45, 733. Ingram, D. L., and Whittow, G. C. (1962). J. Physiol. (London) 163, 200. Ingram, D. L., McLean, J. A., and Whittow, G. C. (1961). Nature 191, 81. Innes, I. R., and Nickerson, M. (1965). In “The Pharmacological Basis of Therapeutics” (L. S. Goodman and A. Gilman, eds.), p. 530, 3rd ed. Macmillan, New York. Irving, L., Peyton, L. J., Bahn, C. H., and Peterson, R. S. (1962). Physiol. Zool. 35, 275. Jackson, D. L. (1967). J. Neurophysiol. 30, 586. Jacobson, F. H. (1966). Federation Proc. 25, 515. Jenden, D. J., and Lomax, P. (1966). Proc. 2nd Intern. Biophys. Congr., Vienna, 1966 p. 426. Gistel and Cie, Vienna.

DRUGS AND BODY TEMPERATURE

41

Johansen, K. (1961). Federation Proc. 20, 110. Jori, A., and Garattini, S. ( 1968). J. Plzarm. Phurmucol. 20, 473. Kahl, M. P. (1963). Physiol. Zool. 36, 141. Kappas, A., Hellman, L., Fukushima, D. K., and Gallagher, T. F. (1957). J . Clin. Endocrinol. 17, 451. Kerpel-Fronius, S., Kiss, A., and Than, G. (1966). Acta Physiol. Acad. Sci. Hung. 29, 267. Kimball, H. R., Wolff, S. M., Vogel, J. M., and Perry, S. (1966). J. C h . Endocrinol. 26, 222. Kirkpatrick, W. E., and Lomax, P. (1967). Life Sci. 6, 2273. Kirkpatrick, W. E., and Lomax, P. (1969). Proc. Western Pharmucol. SOC. 12, 72. Kirkpatrick, W. E., Lomax, P., and Jenden, D. J. (1967a). Proc. Westem Pharmucol. SOC. 10, 51. Kirkpatrick, W. E., Jenden, D. J., and Lomax, P. (196713). Intern. J . NWTOpharmucol. 6, 273. Kocsis, J. J., and Welch, R. M. (1960). Pharmacologist 2, 87. Kollias, J., and Bullard, R. W. (1964). J. Pharmucol. 145, 373. Kulkami, A. S. (1967). J. Pharmacol. 157, 541. Leduc, J. (1961). Acta Physiol. S c a d . 53, Suppl. 183. Light, R. V., and Bysshe, S. M. (1933). J. Pharmucol. 47, 17. Light, R. V., Bishop, C. C., and Kendall, L. G. (1933). J. Pharmacol. 47, 37. Lomax, P. (1966). Brain Res. 1, 296. Lomax, P. (1967a). Arch. Biol. Med. Exptl. 4, 119. Lomax, P. (196713). PTOC.Western Pharmacol. SOC. 10, 67. Lomax, P., and Foster, R. S. (1968). Unpublished data. Lomax, P., and Foster, R. S. (1967). Plmrmacologist 9, 189. Lomax, P., and Jenden, D. J. (1966). Ititern. J. Neuropharmucol. 5, 353. Lomax, P., and Kirkpatrick, W. E. (1967). Med. Pharmacol. Ex&. 16, 165. Lomax, P., and Kirkpatrick, W. E. (1968). Unpublished data. Lomax, P., Malveaux, E., and Smith, R. E. (1964). Am. J. Physiol. 207, 736. Lomax, P., Foster, R. S., and Kirkpatrick, W. E. (1968a). Proc. Western Pharmacol. SOC. 11, 48. Lomax, P., Foster, R. S., and Kirkpatrick, W. E. (1968b). PTOC.Intern. Union Physiol. Sci. 7, 270. Lotti, V. J., Lomax, P., and George, R. (1965a). J. Phurmacol. 150, 135. Lotti, V. J., Lomax, P., and George, R. (1965b). J. Pharmacol. 150, 420. Lotti, V. J., Lomax, P., and George, R. (1966a). Intern. J. Neuropharmucol. 5, 35. Lotti, V. J., Lomax, P., and George, R. (1966b). Intern. J. NeuropharmucoL 5, 75. Lovatt Evans, C . (1957). Brit. Med. Bull. 13, 197. Magoun, H. W., Harrison, F., Bioheck, J. R., and Ranson, S. W. (1938). J . Neurophysiol. 1, 101. Marley, E., and Stephenson, J. D. (1968). Intern. J. Neurophamcol. 7, 165. Meeter, E., and Wolthuis, 0. L. (1968). European J. Pharmacol. 4, 18. Mikkelsen, P. M. (1961). Ugeskr. Laeg. 123, 1064. Morpurgo, C. (1967). Life Sci. 6, 721.

42

PETER LOMAX

Morpurgo, C., and Theobald, W. (1965). Med. Pharmucol. Ezptl. 12, 226. Morpurgo, C., and Theobald, W. (1967). European J. Phurrna~ol.2, 287. Morris, R. W. (1961). Arch. Intern. Phamcodyn. 133, 236. Murakami, N., Stolwijk, J. A. J., and Hardy, J. D. (1967). Am. J. Physiol. 213, 1015. Neuhold, K., Taeschler, M., and Cerletti, A. (1957). Helu. Physiol. Phurmucol. Acta 15, 1. Nieken, B. ( 1966). Acta Physiol. S c a d 68, 215. Paeile, C., and Muiioz, C. (1968). Proc. 3rd Intern. P h a m c o l . Congr., SGO Paulo, 1966 p. 65. Pergamon Press, Oxford. Palmes, E. D., and Park, C. R. (1965). Arch. Enuiron. Health 11, 749. Paolino, R. M., and Bernard, B. K. (1968). Life Sci. 7, 857. Perry, W. L. M. (1957). Brit. Med. Bull. 13, 220. Pfeif€er, C. C., Schlann, L., and Meduna, L. (1950). Federation Proc. 9, 307. Rewerski, W. J., and Jon, A. (1967). Intern. J. Neurophumcol. 7, 359. Rewerski, W. J., and Jon, A. (1968). J. Pharm. P h a m c o l . 2@,293. Reynolds, A. K., and Randall, L. 0. (1957). “Morphine and Allied Drugs,” p. 89. Univ. of Toronto Press, Toronto. Richter, W. (1964). Acta Pharmucol. Toxicol. 21, 91. Robinson, S. (1963). In “Temperature: Its Measurement and Control in Science and Industry” (J. D. Hardy, ed.), p. 287. Reinhold, New York. Rothlin, E., Celetti, A., Konzett, H., Schalch, W. R., and Taeschler, M. (1956). Experientia 12, 154. Rozin, P. N., and Mayer, J. (1961).Science 134, 942. Ruckebusch, Y., Grivel, M. L., and Laplace, J. P. (1965). Compt. Rend. SOC. Biol. 159, 1748. Satinoff, E. (1964). Am. J. Physiol. 206, 1389. Schmidt-Nielsen, K., Dawson, T. J., Hammel, H. T., Hind, D., and Jackson, D. C. (1965). Hualradets Skrifter 48, 125. Schmitt, H., Schmitt, H., Boissier, J. R., Guidicelli, J. F., and Fichelle, J. (1968). European J. Phumnacol. 2, 340. Segagni, E. ( 1962). Minerua Pediat. 14, 702. Sheard, M. H., and Aghajanian, G. K. (1967). Nature 216, 495. Sheard, M. H., and Aghajanian, G. K. (1968). Life Sci. 7, 19. Shellenberger, M. K., and Elder, J. T. (1967). J. P h a m c o l . 158, 219. Shubert, J. (1960). In “Metal Binding in Medicine” (M. J. Seven and L. A. Johnson, eds. ) , p. 325. Lippincott, Philadelphia, Pennsylvania. Simon, E. (1968). Proc. Intern, Union. Physiol. Sci. 6, 289. Sjoqvist, F., Hammer, W., Schumacher, H., and Gillette, J. R. (1968). Biochem. Pharmucol. 17, 915. Smith, R. E. ( 1961). PhysioZogist 4, 113. Smith, R. E. (1962). Federation Proc. 21, 221. Smith, R. E. ( 1964). Science 146, 1686. Somenrille, A. R., and Whittle, B. A. (1967). Brit. J. P h a m c o l . 31, 120. Spencer, P. S. J. (1965). Brit. J. P h a m c o l . 25, 442. Spencer, P. S. J. (1966). Life Sci. 5, 1015. Steen, I., and Steen, J. B. (1965). Acta Physiol. Scand. 63, 285.

DRUGS AND BODY TEMPERATZTRE

43

Stuart, D., Ott, K., Ishikawa, K., and Eldred, E. (1966). Am. J. Phys. Med. 45, 61. Taylor, C. R. (1966). Physiol. Zool. 39, 127. Trabucchi, E., Paoletti, R., and Canal, N., eds. (1964). PTOC.2nd Intern. P h u m c o l . Meeting, Prague, 1963 Vol. 2, p. 135. Pergamon Press, Oxford. Ulrich, W. D., Maess, M., and Wirth, J. (1967). AT&. Intern. Phurmacodyn. 167, 80. Verri, R. A., Graeff, F. G., and Corrado, A. P. (1968). Intern. I . Neuropharmacol. 7, 283. Vogt, M. (1954). J. Physiol. (London) 123, 451. von Euler, C. ( 1961). Pharmacol. Reu. 13, 361. von Euler, C., Linder, E., and Myrin, S.-0. (1943). Acta Physiol. Scand. 5, 85. Votava, Z. (1967). Ann. Rev. Pharmacol. 7, 223. Waite, R., and Spencer, P. S. J. (1968). J. Pharm. Phamtacol. 20, 199. Walther, J., Bishop, F. W., and Warren, S. L. (1941). In “Temperature: Its Measurement and Control in Science and Industry” (Am. Inst. Phys., ed.), p. 474. Reinhold, New York. Webs, B., and Laties, V. G. (1963). J. Pharmacol. 140, 1. Wit, A., and Wang, S. C. (1968). Federation Proc. 27, 634. Zetler, G. (19%)).Intern. J. Neuro’ophurmacol. 7 , 325.

This Page Intentionally Left Blank

PATHOBIOLOGY OF ACUTE TRIETHYLTIN INTOXICATION By R. Torack,l J. Gordon, and J. Prokop Department of Pathology, New York HospitalCornell Medico1 Center, New York

I. Introduction . . . . . . . . . . 11. Materials and Methods . . . . . . . . . A. General Experimental Technique . . . . . . . B. Assay of Triethyltin Effect upon Cell Membranes . . . C. Space Measurements Using Radioisotope Uptake . . . D. Biochemical Assay of Organotin Inhibition of ATPase Activity . 111. Results . . . . . . . . . . . . . A. General Observations and Water Content of the Brain in Alkyltin . . . . . . . . . . Poisoning . . B. Morphological Correlates of Acute Organotin Cerebral Edema . C. Ultrastructure of Choroidal Epithelium after Triethyltin Injection . . . . . . . . into the Cisterna Magna D. Uptake of Sulfate-”S and Sucrose-“C in Triethyltin Poisoning . E. Chemical Assay of ATPase Activity . . . . . . IV. Discussion . . . . . . . . . . . . A. Sequential Morphological Events in Acute Alkyltin Edema . B. Membrane Effect of Triethyltin . . . . . . . C. Space Compartments in Acute Organotin Edema . . . D. The Role of ATPase in the Development of Triethyltin Edema . V. Conclusions . . . . . . . . . . . . References . . . . . . . . . . . .

45 49 49 49

50 50 52 52

53 59 59 65 70 70 75 78 79 82 83

I . Introduction

After the therapeutic use of alkyltin compounds in furunculosis resulted in the death of 110 Frenchmen, a careful investigation of these compounds revealed tliat triethyltin poisoning produced a diffuse fluid accumulation in the white matter of rats which was the first experimental duplication of cerebral edema (Stoner et al., 1955; Magce et al., 1957). Historically, Spatz (1929) called such a diffusely enlarged and wet brain “cerebral edema,” to distinguish it from a focal, firm and dry enlargement of the white matter, Present address: Department of Pathology, Washington University School of Medicine, St. Louis, Missouri. 45

46

R. TORACK, J. GORDON, AND J. PROKOP

typically occurring around neoplasms, which he called “cerebral swelling.” The early investigators believed edema to be distinct from swelling and explained this difference on the basis of an extracellular us. an intracellular localization of the fluid ( Reichardt, 1919; Scheinker, 1941). This distinction was largely derived from light microscopic preparations, which unfortunately resulted in too much morphological distortion of white matter to permit a valid interpretation of this type. The early application of electron microscopic techniques to the study of accumulated fluid in various circumstances revealed the major ultrastructural change to be an enlargement of astrocytes with no change of the small extracellular space (Gerschenfeld et al., 1959; Torack et al., 1959, 1960; Luse, 1960a). Unfortunately at this time, these procedures were only practical for a study of cortex, since methacrylate embedding resulted in too much artifactual splitting of myelin (Luse, 1960b). Following the use of perfusion fixation (Palay et al., 1962) and of Epon as an embedding medium (Luft, 1961),a study of white matter was possible, and differences in the fine structure of the white matter were observed between diffuse edema (Aleu et al., 1963) and perifocal edema (Herzog et al., 1965) which was a new term for localized cerebral swelling. In the former as represented by triethyltin induced edema, the fluid appeared to reside in large myelin clefts, (Aleu et al., 1963; Bakay, 1965a; Lee and Bakay, 1965) whereas perifocal edema involved a significant enlargement of the extracellular space (Herzog et al., 1965; Lee and Bakay, 1966) especially adjacent to the lesion used to induce this condition. This actually represented a complete reversal of the original concepts in which the diffuse edema was considered to be interstitial and the focal swelling was believed to be intracellular. Despite all the controversy over the intra- and extracellular localizations of the fluid, very little information was obtained concerning the mechanisms by which fluid transfer from blood to brain was controlled. The existence of a functional extracellular space was initially questioned (Torack et al., 1960), only to be r e a r m e d at a later date (Brightman, 1965; Torack, 1965a). Astrocytic swelling was originally believed to be the only type of structural change (Torack et al., 1960; Luse, 1960a; Luse and Harris, 1961), but more recently some investigators have claimed this to be an artifact (van Harreveld et al., 1965,1966). At the present time, most investigators

PATHOBIOLOGY OF ACUTE TRIETHYLTIN INTOXICATION

47

have accepted the existence ot ;I distinctive vascular permeability, thc presencv of a limited hut functional extracellular space, and the rapid response of astrocytvs to variations in brain hydration; but the specific cellular malfunction leading to fluid accumulation has remained unclear. The origin of this fluid generally has been conceded to be the plasma (Bakay, 1965b), and at least one aspect of the blood brain barrier must be altered to allow this fluid to enter the brain in greater amounts. Since all perifocal edema has involved a focally destructive lesion the vascular-glial relationships have been artifactually disrupted. This necessarily permits direct access between the plasma compartment and the extracellular space of the brain. Cerebral edema has been shown to involve a low protein fluid ( Stewart-Wallace, 1939), which implied that barrier control to macromolecules was preservcd. Therefore study of changes leading to cerebral edema was not possible in perifocal edema. Experimental models of this type have been limited to a study of brain responses to excess fluid esscntially derived from plasma (Bakay, 196513; Klatzo et al., 1958). In order to study the factors involved in the evolution of edema a nondestructive form of edema was required for which triethyltin intoxication has remained the best experimental model. Triethyltin ( T E T ) toxicity has been of singular value since no apparent structural artifact was produced at the vascular bed ( Aleu et al., 1963; Lee and Bakay, 1965) and the functional impermeability of blood vessels to macromolecules was unchanged (Torack et al., 1960; Bakay, 1965a). Among the various techniques used to induce fluid imbalance, the barrier characteristics were similarly preserved only in water intoxication (Luse and Harris, 1961; Wasterlain and Posner, 1968; Wasterlain and Torack, 1968) and anoxia (Bakay and Lee, 1965). Triethyltin poisoning appeared preferable to water intoxication and to anoxia. Water intoxication was rather difficult to induce and occurred over a relatively short period, creating much more difficulty in control, while anoxia was undesirable because of the poor reproducibility of the lesions. The toxicity of all the alkyl metallic compounds has been carefully studied for almost one hundred years (Harnack, 1878). Although many of these substances affected the nervous system, they appeared to have an effect upon neurons and not upon electrolyte balance ( Magee et al., 1957). More perplexingly, this unique

48

R. TORACK, J. GORDON, AND J. PROKOP

latter effect of triethyltin has not been explained by any similarly specific biochemical effect of this material. Triethyltin has been shown to inhibit oxidative phosphorylation ( Aldridge and Cremer, 1955; Aldridge, 1958 ) ; however, this mitochondrial effect occurred in both cerebral and hepatic mitochondria, with diethyltin and with other trialkyltins. An uncoupling of electron transport was shown to be due to an inhibition of mitochondrial ATPase (Aldridge, 1958) and to be accompanied by a depression of respiration in tissue slices incubated in uitro (Cremer, 1957, 196l), but no alteration of tissue phosphorylation was noted in organotin poisoned brains (Moore and Brody, 1961a,b; Stoner and Threlfall, 1958). Since triethyllead had a similar effect upon oxidative phosphorylation (Cremer, 1962), it has been studied more extensively than most other alkyl metals. However, the effect of triethyllead upon rat behavior was one of tremor and convulsions rather than stupor and coma, and no edema occurred. Therefore, no link between these biochemical alterations, induced by alkyl metals, and brain edema has been established. The fluid imbalance within the central nervous system has been studied widely using electron microscopic, histochemical, biochemical, and radioactive tracer techniques, In the cortex, fluid accumulation was less severe (Katzman et al., 1963) and was characterized by astroglial enlargement (Torack et al., 1960) with an increased sodium uptake (Reed and Woodbury, 1964). The white matter had a much greater degree of fluid involvement, glial swelling did not occur, and the accumulated fluid was within myelin clefts where, apparently, it was not in equilibrium with the extracellular space ( Aleu et al., 1963; Lee and Bakay, 1965; Katzman et al., 1963). Chemically, the fluid had a low protein content, reflecting no doubt, the intact barrier system to particles of this size (Aleu et al., 1963). No inhibition of a “pump” ATPase has ever been observed (Katzman et al., 1963); however, the inhibition of a peculiar membrane bound ATPase in the white matter has been described ( Torack, 1965b ) . Despite all this information, the fundamental questions remained unanswered. How did TET produce edema? Why did the edema affect the white matter so dramatically? How did the fluid get to the myelin cleft? Why did glia apparently react differently in the cortex and white matter? TET poisoning has been induced usually over a period of 5-7 days and, necessarily, the evolution of the morphological and chemical abnormalities was a slow and

PATHOBIOLOGY OF ACUTE TRIETHYLTIN INTOXICATION

49

subtle process. By producing alkyltin poisoning more rapidly, we hoped to accentuate these changes so they would be more easily appreciated, especially with our morphological techniques. II. Materials and Methods

A. GENERALEXPERIMENTAL TECHNIQUE

Young adult albino male rats, weighing 300-350 gm were given a single intraperitoneal injection of triethyltin sulfate equivalent to 4, 7, 8 and 9 mg/kg of body weight. These animals were killed at intervals of 6, 12, 18, and 24 hours by decapitation with a guillotine. The animals given 9 mg/kg did not routinely survive 18 hours, and therefore they were examined only at 12 hours. The brains were divided along the interhemispheral fissure, and one-half was immersed in 2 4 %glutaraldehyde in 0.2 M cacodylate buffer at pH 7.2, while the other half was weighed and then dessicated at 100°C for a minimum of 24 hours to measure water content. One slice of the fixed hemisphere was routinely embedded in paraffin for an hematoxylin and eosin preparation. Frozen sections were cut from the remaining slice and incubated in the Wachstein-Meisel (1957) ATPase medium for 20-45 minutes at room temperature. Those animals selected for electron microscopic study were perfused with 4.5%hydroxyadipaldehyde in Ringer's solution through a cannula inserted into the left ventricle (Torack, 196%). After perfusion for 20 minutes, sections of the corpus callosum were removed and postfixed in 1%OsO, buffered to pH 7.2 with Dalton's chromate for one hour. Following dehydration in a graded series of alcohol the tissues were embedded in Epon. Thin sections were stained with a saturated aqueous solution of uranyl acetate or with lead citrate. Thick sections were mounted on glass slides and examined with a phase microscope.

B. ASSAYOF TRIETHYLTIN EFFECTUPON CELLMEMBRANES In three similar normal rats, 0.1 ml of 1 x M TET in isotonic saline was injected into the cisterna magna, after a cut down to the foramen magnum. Four minutes later, 1%osmium tetroxide in phosphate buffer at pH 7.2 was injected into the cisterna magna. After removal of the choroid plexus from the 4th ventricle, it was cut into 1 mm lengths, dehydrated and embedded in Epon as described above.

50

R.

TORACK,

J. GORDON, AND J. PROKOP

C. SPACEMEASUREMENT USINGRADIOISOTOPE UPTAKE

The radioisotope uptake studies were carried out on SpragueDawley rats, weighing 375 to 425 mg. These animals were anesthetized with intraperitoneal sodium pentathol. Then using an open cut down technique, a plastic cannula was inserted into the femoral vein and firmly tied in place. The cannula was connected to a constant perfusion pump, which was an electrically driven piston designed to push in the plunger of a syringe at an even rate delivering 0.45 ml of fluid per hour. The radioactive substances used for perfusion were sucrose-14C and sulfate-:%. Sucrose-14C was perfused at a rate of 4.5 pcurieslhour, sulfate-.Y3 at a rate of 4.8 pcurieslhour. The rats were injected with 7 mg/kg of triethyltin given intraperitoneally. At 4, 12, and 22 hours the animals were lightly anesthetized with ether and sacrificed by decapitation. The brain was rapidly removed from the skull, separated from the meninges, and blotted. Blood samples were also taken. Normal animals were used as controls. In addition one group of rats was poisoned in the usual slow manner with a daily htraperitoneal injection of 1 mglkg of TET. After 7-10 days these animals became sick and they were perfused with sucrose-"C, exactly as previously described. One hemisphere from each brain was weighed immediately upon removal from the skull, and homogenized with a Tri-R stirrer in 4 ml of distilled H,O containing a wetting agent. Following homogenization, 1 ml samples were pipetted onto planchets. The planchets were slowly evaporated to dryness (about 18 hours at room temperature) and the radioactivity counted with a Geiger counter and a scaler ( Nuclear-Chicago Model #18lB). The opposite hemisphere was also weighed immediately and then dessicated at 100°C for at least 24 hours. The plasma was separated from the red blood cells by centrifugation. Then 0.5 ml samples were pipetted onto planchets and slowly dried. The radioactivity was counted in the same way as the honiogenates. All radioactive counts were corrected for background, decay, and absorption.

D. BIOCHEMICAL ASSAYOF ORGANOTIN INHIBITIONOF ATPASE AcnvrrY Adult male Sprague-Dawley rats weighing 350400 gm were perfused through the left ventricle with cold isotonic sucrose while under deep ether anesthesia. Following removal, the cerebral hemi-

PATHOBIOLOGY OF ACUTE TRIETHYLTIN INTOXICATION

51

spheres were washed in cold 0.3 M sucrose and homogenized with eight up and down strokes of a motor-driven teflon pestle in a glass tube, cooled by an ice bath. In this way a 20% homogenate was prepared in 0.3 M sucrose and R 1OX homogenate was prepared in 2% and 10.5%glutaraldehydc,. Preliininary studies indicated that there was no difference in the amount of enzyme activity between the 10%and 20% homogenates of both media. Both media were diluted to the desired concentration with 0.2 M imidazole and the pH was adjusted to 7.2 with HCl. After preparation, the homogenates were transferred to dialyzing tubing having a 4.8 m p pore diameter and dialyzed against 6 litcm of deionized distilled water for 48 hours at 2.0"C. The assay procedure was similar to that used by Gordon and Torack (1967) to demonstrate Mg++-ATPase in aldehyde-treated tissue. Each assay tube, containing a final volume of 5.0 ml, was prepared from a basic reaction media containing 0.3 mM MgNO?, sufficient trisaminoniethane and maleic acid to adjust the p H to 7.2, and an appropriate amount of sucrose to achieve iso-osmolality. The cffect of TET concentrations of lo-?32 through M was measured. A stock solution of TET, prepared by dissolving a weighed amount of inhibitor in deionized distilled water, was adjusted with sufficient solid imidazole to achieve a pH of 7.2. After adding 0.4 ml of the dialyzed homogenate to 3.95 ml of basic reaction media, the required amount of inhibitor stock was added with sufficient distilled deionized water to bring the volume to 4.85 ml. Quadruplicate tubes of each inhibitor concentration and a control were preincubated for 20 minutes in a water bath set at 37"C, after which 1.5 ml of cold tris ATP was added and the incubation continued for another 15 minutes. The reaction was stopped by immersion in a Dry Ice and acetone bath. To each tube of frozen material, 5 ml of 10%trichloroacetic acid (TCA) was added. The tubes were allowed to thaw, without agitation, at 4°C. After brief centrifugation, triplicate 1 ml aliquots of supernatant were analyzed for inorganic phosphorus ( P , ) by thc method of Taussky and Shorr ( 1953). Reagent and tissue blanks were run simultaneously to adjust for spontaneous hydrolysis of ATP and endogenous P,. The enzyme activity was expressed in micrograms of P, released per milligram of dried retentate per hour (pg P,/mg/hr). The dry weight of the retentate was obtained by drying quadruplicate 1 ml samples of the nondialyzable material at 98°C for a minimum of 48 hours and weighing on a Mettler balance.

52

R . TORACK, J . GORDON, AND J. PROKOP

Ill. Results

A. GENERAL OBSERVATIONS AND WATER CONTENT OF ALKYLTIN POISONING

THE

BRAINIK

Two distinctive effects upon the behavior of these rats were noted with these higher doses of TET (Fig. 1 ) . Initially, after a period of 15-30 minutes, the animals became lethargic or stuporous. This was best seen in rats receiving 8-9 mg TET/kg and not observed in the 4 mglkg rats. All the animals tended to revive after about 2-3 hours, but at about 12 hours following injection,

HOURS

FIG.1. A diagrammatic representation of the behavior of rats during acute triethyltin intoxication. The vertical ordinate refers to degrees of stupor.

they again became lethargic, developed weakness of the hind limbs, and finally lapsed into stupor. This biphasic involvement was best noted in the 7-8 mg/kg animals. The rats receiving 9 mg/kg never fully recovered from the initial stupor and lingered semistuporously for about 12 hours. Then they became less easily aroused, developed labored respirations, and died between 12 and 18 hours. A few rats, given 10 mglkg TET, remained deeply stuporous and expired between 2 and 8 hours after injection. The 4 mglkg dosage group did not manifest any abnormal behavior even at 24 hours. After 24 hours the animals that had received 7-8 mg TET had a degree of stupor and lower limb weakness which was comparable to that noted after 7-10 days in those animals receiving a daily dosage of 1 mg/kg TET. The only difference between these two groups of animals was the weight loss which chronically poisoned animals invariably develop. No statistically significant increase in water content ever oc-

PATHOBIOLOGY OF ACUTE 'TRIETHYLTIN INTOXICATION

53

curred before 18 hours evcn in those animals receiving 9 mg/kg T E T (Table I ) . After 24 hours, a single dose of 7 and 8 mg/kg T E T resulted in a 1.19%and 1.14%increase in brain water respectively. No increase in wet weight was noted in the brains of those rats receiving 4 mg/kg TET.

8 Hours 1'3 Hours 18 Hoiirs 24 Hours

Control

79.05 79.45 79.1 79,35 79.1

f 0.75 f 0.55 f0.7 & 0.45

f0.6

79. (15 f 0 :&I 7 9 . 4 %f o :%I i0.7 f 0.3 XO.3 f 0 . 5 79. 1 f 0 . 7 80 '3 f 0 . 3 80 0:2 k 0 . 4 8 79.00 f 0 38 79.3 f 0.4

-

77 77 f 0.48 -

77 58 f 0 40

B. MORPHOLOGICAL CORRELATEX OF ACUTE ORGAXOTIN CEREBRAL EDEhlA The paraffin-embedded tissue stained with H & E was practically useless in this study. To be sure, severely edematous brains were accompanied by a loosening of the white matter, but this could be predicted even in the gross specimen by the enlarged and soft appearance of the corpus callosum. The light microscopical evaluation of questionably cdematons tissue usually had poor correlation with the actual water content of that tissue. This difficulty occurred in tissue fixed in glutaraldehyde, hydroxyadipaldehyde, or formaldehyde and was probably more related to paraffin embedding than to fixation itself. Thick sections of hydroxyndipaldehyde-perfused and Eponembedded tissue, when examined with a phase microscope, were quite informative in sevcral ways. These evaluations revealed whether the tissue represented the dense white matter of the corpus callosum or the less compact myelinated axons of the corona radiata (Fig. 2 ) . The latter arras were always admixed with neurons and plentiful astrocytrs. Astrocytic swelling was occasionally more easily and more reliably c\xluated in this tissue than in thin sections which were studied witli electron microscope (Fig. 3 ) .

FIG. 2. A phase Iiiicrograph of norninl rat h i r i at the junction of the corpus callosum and the cortex. Urrisely inyeliiioted ‘ireas ( DWhI ) and nstrocytes ( A s ) are clearly identified. X2000. FIG.3. A phase micrograph of dense white iiiatter at 12 hours after injec-

PATHOBIOLOGY OF ACUTE TRlETHYLTIN INTOXICATION

55

The recognition of the cell swelling was facilitated by the greater volume of the cell which was present in a thick section and by the increased numbers of cells included in sectioning the whole block. Myelin cleft formation also could be accurately evaluated in these sections (Fig. 4); however, axonal enlargement with or without the aggregation of cytoplasmic organelles could not be reliably determined from these thick sections. Therefore, these phase microscopic studies were invaluable for tissue orientation, for a morphological evaluation of some cellular changes in a greater mass of tissue, and for a correlation of micro- and ultrastructure. 1. Myelin Cleft Formation This most distinctive feature of white matter in TET intoxication was present in the corpus callosum of every rat 24 hours following an injection of 7 or 8 mg/kg. The clefts were also observed in the compact myelinated fibers of the internal capsule, the looser fibers of the corona radiata, and the scattered fibers of the neuropil. They occurred along the interperiod line exactly as described previously by Aleu et al. (1963) (Fig. 5 ) . These clefts were noted infrequently at 12 hours with any dosage (Fig. 6 ) , whereas at 18 hours 75%of all the blocks of corpus callosum demonstrated this finding. No clefts occurred in 6 hours even when a dose of 10 mg/ kg was used. 2. Axons Axons demonstrated two patterns of involvement. At 12 hours after injection, a focal dilation of axons could be observed (Figs. 7 and 8 ) . This was inconsistently seen in the corpus callosum of animals poisoned with 7 or 8 mglkg, but it was regularly seen with a dose of 9 mglkg. The myelin sheath was usually intact around such a dilatation, and no predilection for nodal areas could be distinguished. Within such enlargements, the neurofilaments in the center of the axon were more closely approximated than those near the axon cell membrane (Fig. 8 ) . Thus the extra fluid did not appear to diffuse equally throughout the axoplasm. No increase tion of 9 mg/kg TET. Enlargement of astrocytic cell bodies ( A s ) and perivascular foot processes (AsF) is clearly present. X1000. FIG.4. A phase micrograph of dense white matter at 24 hours after injection of 8 mg/kg TET. Severe vacuolization around intact axons ( A ) can be seen. An astrocyte ( A s ) can be identified and is not enlarged. X1000.

FIG.5. An electron inicrograph of a large iiiyelin cleft occurring at 24 hours after the injection of 7 mg/kg TET. X8000. FIG. 6. These niyelinated nerve fibers in compact white matter appear

PATHOBIOLOGY OF ACUTE TRIKIHYLTIN INTOXICATION

57

in filamentous or tubular organelles was visible. Some macrotubules appeared dilated, but mitochondria were normal in number and appearance. These dilatations were seen infrequently at 18 hours after injection and rarely at 24 hours; indeed, their presence was inversely proportional to thc presence of myelin clefts. At 24 hours, other dilated axons were noted; however, in contrast to the previously described axons, these were filled with a varying mixture of irregular dense bodies, mitochondria, dilated macrotubules, microtubules, and neurofilaments (Figs. 9 and 10). These organelle-laden axons were occasionally seen at 18 hours but not at 12 hours after TET injection. They were not visible by phase microscopy. 3. Astrocytic Smelling

Astrocytes with an enlarged cytoplasmic volume unaccompanied by any significant increase of cytoplasmic organelles were observed in the neuropil from 12 to 24 hours after injection. At 12 hours they were more prominent in the higher doses of organotin, especially in the 9 nigfkg dose. In the white matter, however, astrocytic swelling was only apparent at 12 hours after injection, and this was most consistently seen in those animals receiving 9 inglkg of TET (Figs. 11 and 12). At 18 and 24 hours these cells were never enlarged; in fact, the astrocytic foot processes at the surface of blood vessels frequently appeared shrunken (Fig. 13). These shrunken perivascular foot processes often appeared separated both from each other and from the vascular basement membrane. All organelle structures within the astrocytcs were normal in appearance whether these cells were enlarged, normal or shrunken. 4. Blood Vessels, Extracellular Space, Oligodendrocytes All these structures were essentially intact in all animals. Occasionally at 24 hours some widening of the vascular basement membrane was suggested. Irregularly the extracellular space reached a width of 700-900A. This was never consistently present but was consistently absent in the areas of prominent cleft formation. The oligodendrocytes were never observed to be abnormal. normal at 12 hours after the injection of 9 mg/kg TET. No cleft formation can be seen. X8000.

FIG.7. This myeliiiated nerve fiber in the corona radiata shows two focal

PATHOBIOLOGY OF .4CUIX TRIFXHYLTIN INTOXICATION

c.

ULTRASTRUCTURE OF

59

CHOROIDAL EPITHELIUM AFTER TFUETHYLTIN CISTERNA MAGNA

INJECTION INTO THE The ultrastructiire of normal choroidal epithelium has been characterized by numerous villous processes on the ventricular surface (Fig. 14). These processcxs were usually of fairly uniform length and width. They were covered by an extension of the cell membrane and contained cytoplasmic ground substance identical to that found elsewhere in the cell. Four minutes after injection of TET, the villi had lost their normal appearance. They were markedly irregular in size and shape, with some being considerably dilated (Fig. 15). The enclosing extension of the cell membrane was also changed from a single linear electron dense structure to one having many discontinuities frequently represented by a series of small vesicles (Fig. 16). Beneath the villi there was a linear zone in which the cytoplasmic ground substance had a decreased density (Figs. 15 and 16). This decreased density also appeared to involve the villi. The mitochondria were normal as were all the other cytoplasmic organelles. A similar morphological change in the villi was observed when Amphoteiicin B was injected into the cistern (Fig. 17). The villi were also irregular in size and shape with decreased density of the ground substance and a lack of a uniform linear limiting cell membrane. D. UPTAKE OF sULFATE-d's POISONING

AND

SUCROSE-'*C

IN

TRIETHYLTIN

The sucrose and sulfate spaces of the brain were calculated by the following formula: Brain space

=

counts per gram of wet brain counts per milliliter of plasma

x

100

The sulfate space appeared to reach an equilibrium at 4 hours and did not change throughout, being essentially the same at 22 hours. There was also no difference in values between the control and TET animals (Table 11). enlargements ( F E ) of axoplasni 12 hours after the injection of 9 mg/kg TET. At these sites and elsewhere, the inyelin appears normal. X8000. FIG.8. A higher magnification of a focally swollen nerve (insert of Fig. 7 ) fiber reveals the neurofibrils ( N F ) to remain cslosely packed in the center but to be ahsent in the periphery of the enlargement. X28,OOO.

60

R. TORACK, J. GORDON, AND J. PROKOP

FIG.9. At 24 hours after injection of 8 mg/kg TET, this dilated axon contains nunierous mitocliondria ( M ) and dense bodies ( DB). The inyelin remains intact. X17,OOO. FIG.10. Another dilated axon 24 hours after injection has an increased density of the axoplasm in addition to the nuinerous dense bodies ( D B ) . X 17,000.

PATHOBIOLOGY OF .%CU'IY 'I'RIETHYLI'IN INTOXICATLOX

61

FIG. 1L A niassively enlai ged mtrovyte ( As ) in roinpact white niattci 12 honrs after the injection ot 9 rng/kg TET. The oligodendrocyte (01) remains normal. X 3000.

FIG. 12. At 12 hours after injection of 9 mg/kg TET the perivascular nstrocytic foot processes ( A s F ) are greatly enlarged. X3000. FIG. 13. At 24 hours after injection of 7 nig/ky TET the perivascular .istrocytic foot processes (As F ) are small. A localized enlargement of extracellular space ( E C S ) is visible. X26,OOO. 62

FIG.14. An electron micrograph of normal choroid plexus in which the choroidal villi are thin and regular. X8500. FIG. 15. Four minutes after the injectioii of TET in the cisterna inagna the villi appear enlarged and niniieroiis vcsicles ( V ) are seen in these strnctures. An electron-lucent zone ( E L ) of the cytoplasni beneath the villi can be seen. X9000. 6:3

64

R. TORACK, J. GORDON, AND J. PROKOP

FIG. 16. A higher magiiification of choroidal villi after TET reveals the

PATHOBIOLOGY OF ACUI'E TRIETHYLTIN INTOXICATION

65

The normal sucrose space slowly increased during the interval from 12 to 22 hours. Although there was no significant difference in I4C values between the control and TET groups (Table 11), there was a definite tendency for the T E T space to be less than the control animals. This trend became more definite at 22 hours and in chronic poisoning, but it never reached the point of statistical distinction. The control group had a mean brain water content of 79.09%k 0.38 (Table I ) . The mean brain water content of the 22-hour acute group was 80.03%% 0.48 while that of the chronic T E T group was 80.24%-C 0.46. This represented an increase of 1.19%for the acutely poisoned rats and an increase of 1.15%for the chronic group. There was no statistical difference in the mean brain water content between the 4- or 12-hour TET animals and the control groups.

E. CHEMICAL ASSAYOF ATPASEACTIVITY 1. Histochemistry The histochemical demonstration of ATPase activity in the glutaraldehyde-fixed tissue following incubation in the Wachsteinpresence of numerous vesicular ( V ) and tubular profiles within the enlarged villi. X20,OOO. FIG. 17. Four minutes after the injection of Amphotericin B into the cisterna magna, prominent enlargement of villi and increased numbers of vesicles ( V ) are visible. An electron-lucent zone ( E L ) beneath the villi is also present. X 13,000.

FIG. 18. Glutaraldeliycle-resistal7t ATPase activity is present in normal rat corpus callosum and a t the ependymal lining ( E ) after 45 minutes incubation a t room temperature. X250. FIG. 19. Glutaraldehyde-resistant ATPase activity is observed only at the ependymal lining ( E ) at 24 hours after injection of 8 nig/kg TET. X250. FIG.20. An electron micrograph of normal rat corpus callosum, fixed in

PATHORIOLOGY OF -4CU'TE 1'RIETHYLTIN INTOXICATION

67

Meisel medium revealed a significant decrease in the intensity of the reaction product only in the 24-hour animals receiving 7 or 8 mg/kg T E T (Fig. 18). The reaction product after glutaraldehyde perfusion or prolonged immersion (24 hr) was found in a fibrillar pattern and was present only in the white matter (Fig. 19). Previous electron microscopic studies ( Torack, 1965c) had revealed a localization of this precipitate to the surface of glial cells and to macrotubular structures in the axoplasm ( Fig. 20). 2. Biochemistry A plot of the residual ATPase of brain homogenates prepared in sucrose against the concentration of TET in the incubation media gave a curve that could be divided into stimulatory and inhibitory components ( Fig. 21 ) . At T E T concentration of 1 x 10 ' and 1 x lo-' 11.1, the ATPase activity exceeded the control Iewl of 103.8 7.2 pg/nig/hr. This control level was also lower than tlic 119.0 pg/mg/hr level that was reported previously in homogc nates which were incubated with TET but which had not been preincubated (Torack et al., 1967). Therefore the possibility existed that TET at low concentrations protected the ATPase from denaturation during prcincubation. However, experiments in which ATPase was measured in 0, 1 x 1eS, 1 x lo-', and 1 x lo-', A! TET without preincubation gave essentially the same results, indicating a true stimulation. By extrapolating the stimulatory and inhibitory curves to their point of intersection, an approximate maximal ATPase activity of 116.5 p,g/mg/hr in the presence of 2 x lo-' hl TET was obtained. At higher concentrations of TET ( 1 x 10-',-1 x 10 hi) there was an exponential fall of ATPase activity. An analysis of the inhibitory segment of the curve revealed that it could be further resolved into three components, two different TET sensitive components and a TET insensitive component. A plot of the log of P, released against the concentration of organotin demonstrated the existence of a plateau of maximal effect at 65.3 pg/mg/hr. This TET-insensitive activity was subtracted from the total, and the log of the difference, or TET-sensitive component, was replotted (Fig. glutaraldehyde and incubated in the Wachstein-Meisel medium, reveals a reaction product in large tubules of axoplasm ( A T ) and on the surface of a glial cell (GC) resembling an oligodendrocyte. X14,OOO.

68

R . TORACK, J. GORDON, AND J. PROKOP

22). The resulting graph demonstrated the existence of a biphasic TET inhibition of phosphate releasc. Initially the log of the TETsensitive activity fell off curvilinearly with increasing TET concentration. However, at a concentration of 1 x M TET, there was a break where thc decrease in activity appeared linear. This was better demonstrated in Fig. 23, in which this component was

120F 110

s-2,-

::I, 20

I 10

, 100

,

zoo

100

300

:I, Moles of TET/liter(x108) ---_--o--- -

400

-0

,

,

200 300 400 Moles of TET/liter ( x I O - ~ )

, 500

FIG. 21. The effect of triethyltin on ATPase activity in rat cerebral homogenates. The main graph is a plot of the Pg-releasing activity of brain homogenates, which are incubated with TET in concentrations varying from lo-' to 5 X M , showing a plateau of TET inhibition. The insert shows the effect of low TET concentrations ( 1 0 - 8 4X lo-' M ) including the apparent stimulation of activity at very low concentration. The dotted line represents the extrapolation to maximal enzyme activity.

plotted on a condensed concentration scale. This linear component was extrapolated back to zero concentration of TET. The amount of Pi released was determined for this component at TET concentration of 1x 10-5-1 x lo-' M , and these values were subtracted from the total TET-sensitive activity. A plot of the log of this difference against inhibitor concentration was another straight line. Cerebral Pi releasing activity in the presence of ATP thus had two TET-sensitive components. One component was sensitive to low concentrations of T E T with K,,,,, a value which was determined by

69

PATHOBIOLOGY OF ACUTE T R I E T H Y L T I N I N T O X I C A T I O N

' 1 mathematical analysis ot t l i c l i n t . ot 4.3 x 10 as cstimated from the graph. The other componrnt \vas scwsitive to high concentration and also had K , which WAS detcwnined by a mathematical analysis of the line of 2 x 10 I . The TET sensitivity ot tlics P,-rclcasing activity in cerebral tissue, homogenized in glutaraltlehydc, did not coincide with either component of inhibition in untreated tissue. Homogenates prepared '8,

0

3.

1

120 10

50

80 100

250

400

Moles of TET/liter ( x lo6)

FIG.22. The components of TE'I inhibition of cerebral ATPase. The log of the enzyme activity minus the plateau v;ilue (component IV ) plotted against TET. The linear coniponeiit (111) is extrapolated to M . The insert shows a plot of the differences of the experimenta1:y determined points and the activity of coinponent 111 at cwresponding TET concentrations, yield linear component 11.

in 2%glutaraldehyde showed a sensitivity to TET, but P,-releasing activity at each concentration of inhibitor tested exceeded that in tissue which was not exposed to organotin. At a concentration of 10.5%glutaraldehyde, there was further inhibition of the activity in the absence of T E T as previously reported (Gordon and Torack, 1967). At this glutaraldehyde concentration the addition of T E T caused no additional effect. This level of TET-resistant activity was similar to that of fresh tissue homogenized in sucrose alone (Fig.

23).

70

R . TORACK, J. GORDON, AND J . PROKOP

IV. Discussion

A. SEQUENTIAL MORPHOLOGICAL EVENTSIN ACUTE ALKYLTIXEDEMA The sequential study of rats during a 24-hour period, in which an acute TET-induced brain edema develops, reveals a striking and changing pattern of cellular involvement. At the end of this period the water content and morphological alterations are virtually identical to those described at the end of an edema which develops slowly over a period of 7-10 days. Previous studies involving this o Sucrose A \

0

I

2.0% Glutaraldehyde

A 10.0% Glutaraldehyde

\.

2

3

4

5

6

7

8

9

10

Moles of T E T / l i t e r ( x 10')

FIG.23. The effect of TET on the cerebral ATPase activity in glutaraldehyde-treated brains. The enzyme activity of the sucrose control is plotted from a T E T concentration of lo-' M , while the activity of glutaraldehyde-treated brain is plotted from zero concentration of TET.

latter type of edema induction have not revealed any similar transitory cell changes ( Alcu et ul., 1963; Lee and Bakay, 1!365). However, since the end result is identical, it appears reasonable to assume that similar changes are occurring in animals of the latter group, but that the cellular alterations are too subtle to be appreciated by our present techniques. Indeed, even in this study, the striking cellular enlargements occur at a specific time and dosage (i.e., 12 hours with 9 mg/kg T E T ) , whereas a decreased dosage does not evoke a responsc of similar magnitude at any time. A description of the sequential cellular changes occurring in acute

PATIIOBIOLOGY OF ACUI E 'I'HIETHYI.TIN INTOXICATION

71

alkyltin edema appears to be important because it reveals a morphological pattern of involvemcnt lcadiiig to edema of the brain. The absence of any distinctivc structural changes in the brain until 12 hours after injection of TET suggests that the etiology of the early stupor in these aiiitnals is different from that which develops slowly between 12 and 24 hours. The absence of any significant fluid accumulation prior to 12 hours indicates that this is a metabolic effect of organotin which has no direct relationship to fluid imbalance. Various specific nic~tnboliceffects of T E T have I~eenwell identified in uitro, siich a s reduced oxygen consumption and an interruption of oxidativc phosphorylation. However, these findings have been considercd to be of dubious significance in explaining cerebral edema Iwcausc they have been largely unconfirmed in vim and becausc they were not limited to the brain. Most of these in t h o studies in\~olvcdanimals in a terminal stage of a more slowly evolving c d e n i a . Perhaps these metabolic effects are more properly relatctl to this c d y stupor which is never scen with low dosage (less than 7 mg/kg) and which is unassociated with any change of brain architt.cture. The clinical, structural, and ch(*micalevcnts leading to the full development of cerebral edeinii s c ~ w ito have their onset at about 12 hours. Progressive hind limb wc~nkness and stupor, various cellular alterations, and incrwsc>clwatrr content are observed only in the 12-24 hour interval after injcction. Therefore, any changes occurring at about 12 hours would appear to represent the earliest sites of TET toxicity. Astrocytic swclling and axonal swelling are the striking cellular changes at this time. The astrocytic enlargement occurs in both gray aiid whitc matter simultaneously; however, in the white matter this is a transitory change. Therefore at 12 hours, the astrocytes in both places appear to be rcactiiig similarly, and a functional distinction between these cells is not justified. Apart from their incrcwcd volume and a dilution of cytoplasmic organelles, thew cells appear normal. Mitochondria, in particular, appear quitc compact and unlike those enlarged or distorted mitochondria in which rcspiratory function has been specifically altered ( Tandler et oZ., 1968; Wilson and Leduc, 1963). These morphological findings suggest that the basic defect involves fluid control h y these cells, and they would support the concept proposed by Reed et al. (1964) that increased permeability of the cell membrane is responsibl(, for the, cc.11 enlargement.

72

R. TORACK, J. GORDON, AND J. I’ROKOP

A similar swelling exists within axons, except the involvement is focal rather than diffuse. Again, no change in organelle fine structure can be observed other than a dilation of macrotubules and n separation of neurofibrils. The fact that the neurofibrils are more widely separated near the axon membrane suggests that the fluid enters at this point and supports the existence of a leaky cell menibrane. The focal nature of this swelling implies that fluid is not entering the axon uniformly, and the compact appearance of thc myelin around thew enlargements indicates an occurrence in the internodal segment. This is somewhat perplexing since myelin is usually considered to act as a diffusion barrier to electrolytes (Hodgkin, 1951; Caldwell, 1968) while the nodes of Ranvier are the regions where ionic currents pass and free exchange occurs between the axoplasm and the extracellular space. However, Treherne ( 1961) has shown that exchanges of ions appear to proceed rapidly between plasma and nerves, and Villegas et al. (1961) have observed no measurable diffusion barrier to small particles across sheaths and cells surrounding the squid axon. Morc recently Freeman et al. (1966) have demonstrated swelling of fibers to occur as a fast osmometer with changes in external NaCl concentration and accompanied by only minor changes in the sheaths themselves. Singer and Salpeter ( 1967) h a w even found radioactive histidine and leucine passing through Schwann cells and myelin into the axoplasm. Therefore, although inyclin itself has few enzymes (Adams, 1965) and a low metabolic rate ( O’Brien, 1965; O’Brien and Sampson, 1965), particles and fluid may pass through to reach the axoplasm. Perhaps this is occurring also in carly TET poisoning. The paranodal thickening of the axon membrane recently observed by Peters (1966) may be a factor in keeping the fluid localized. The next obvious question concerns the origin of the fluid itself. A priori, we assume this origin to be the circulating plasma. In this case we would also have to assume that the vascular wall is leaky, permitting sodium and water to leave the vascular lumen and pass into the c>xtracellular space. The functional capacity of even the small extracellular space as an effective site for fluid and electrolyte exchange no longer appears questionable (Kuffler et al., 1966; Lasansky and Wald, 1962; Nicholls and Kuffler, 1964). However, although an increase in cercbral fluid and sodium has been well documentcd, no one has ever dernoiistrated an increased extra-

PATHOBIOLOGY OF ACUTE 1'RIETHYLTIN INTOXICATIOX

73

cellular space, morphologicallv (Torack et al., 1960; Aleu et al., 1963; Lee and Bakay, 1965) or physiologically (Streicher, 1962: Katzman et al., 1963; Reed et nl., 1964). and we do not find anv consistent enlargement in thcse studies. If all these cell membranes are leaky perhaps the fluid and electrolytes quickly pass from the extracellular space into cells, thus preventing any extracellular accumulation. The latent tendency for most cells to acquire fluid and to swell has been documciited by several investigators (Luck6 and McCutcheon, 1930; Opie, 1948). Swelling would be quite acceptable under these circumstances. One perplexing problem remains; why don't oligodendrocytes swell? The progressive accumulation of fluid is accompanied by a development of myelin clefts. There is no reason to doubt that bulk of fluid is located within thcse clefts. This cleft formation clearly occurs at a stage later than the swelling of axons and astrocytes. Yet we cannot see any morphological indication as to origin of cleft formation. Since the split is at the interperiod line (Aleu et al.. 1963), the possibility of simple hydration of myelin arises (Robertson, 1955; Hirano et al., 1966). However, other forms of fluid accumulation do not result in this degree of splitting (Lee and Bakay, 1966; Herzog et al., 1965). Probably the lack of protein in the edema fluid would facilitate this degree of splitting, but on thc other hand since myelin is essentially a membrane structure we cannot eliminate a similar membrane effect of alkyltin on myelin as upon other membranes. An alteration of myelin secondary to axonal involvement does not appear likely since the myelin around such axons appears largely normal. That these clefts, once formed, act as a sink for the fluid and clcctrolytes does seem quite evident. This has already been proposed by Katzman et al. (1963) from the finding that the increased sodium is not readily exchanged with "'Na in the plasma. The present study clearly demonstrates the inverse relationship which exists between the cell swelling and the clefts. This also can be produced by such a sink action. The accumulation of sodiuni witliin these clefts may be partly explained by the absence of an ion pump in this site, whereas the sodium exchange at cell surfaces is maintained. In other words. sodium continues to be turned o \ w in swollen cells, but, in myelin clefts, it apparently stagnates. Another consequence of sodium movement into clefts should be a reversal of any concentration gradient of water and electro-

74

R. TORACK, J. GORDON, AND J. PROKOP

lytes in the extracellular space resulting from increased vascular permeability. Thus the movement of fluid and electrolytes from the extracellular space into clefts should leave less than a normal content of sodium and fluid in this space. At this point sodium, potassium, and intracellular fluid should tend to pass through hyperpermeable membranes of swollen cells back into the extracellular space causing a reduction of the cell swelling. Finally, since this entire process is dependent upon cleft formation, the

00

WHITE MATTER

No

CORTEX

+

00 NORMAL

SWELLING

( 12 hr-9rng/kg)

FIG. 24. A diagrarnrnatic representation of sequential changes in cell volume and in sodium localization diiring the evolution of acute triethyltin intoxication.

persistence of swollen astrocytes in the neuropil now becomes quite understandable since there are few myelinated nerve fibers in this area. This seems to be the final stage in organotin edema (Fig. 24). The last distinctive morphological finding in this series is the axonal dilatation associated with the accumulation of various cytoplasmic organelles. This type of dense body and mitochondria1 aggregation has been described in various types of nerve injury (Hirano et al., 1967; Lampert, 1967a; Gonatas, 1967; Nevin, 1967), and some more recent evidence indicates this may be due to impaired axon flow ( Weiss, 1967; Lampert, 196713). The later sequential occurrence of this finding suggests that the focal accumulation

P.4THOBIOLOGY OF ACWl E TRIETHYLTIN INTOXICATIOS

75

ot fluid results in a structural or functional abnormality leading to the aggregation of these organelles. Both axonal changes appear to be associated with myelin s h c d i s which are not significantly involved in cleft formation and may indicate a distinction of these myelinated nerve fibers. Inciclcntly, the axons passing through the majority of the split myelin \heaths do not appear abnormal in any way. Thus neither the early axonal swelling nor the organelle laden dilatation seems to be directly rvlated to myelin cleft formation. €3. MEMBRANEEFFECTOF TRIETIIVLTIN

From the foregoing discnssion, the concept of an organotininduced leaky membrane becoincs the crucial factor in explaining astrocytic swelling and shrinkagc, axonal dilatation and organelle aggregation, endothclial hyl"rI)"mral~ility, and perhaps even the myelin cleft itself. Therefore, somc' understanding of the physiological effects of hyperpermeablc membranes is necessary, especially to understand how they would result in variations of cell size and electrolyte content. The regulation of fluid and electrolytes appears to be chiefly controlled by the physical character of the cell membrane and thc prcwnccl of ion pumps capable of moving diffusible ions agaimt a concentration gradient (Robinson, 1966; Conway, 1954; Kavanau, 1966; Eiseninan et al., 1967). Fluid transfer across cell membranw can be altered by changing the osmotic properties of cells ( i.c., intracellular protein), by significantly altering the character ot thc cbxtracellular fluid, by poisoning ion pumps located on cell sarfaces, or by transforming the molecular arrangement of the cell membrane to permit more rapid exchange of diffusible ions (Dick, 1966). In organotin edema, the \ize and character of the extracellular compartment has never been known to be significantly altercd (Streicher, 19622; Katzman et al., 1963; Reed et al., 1964), Na", K', and Mg'+ ATPase activity has been found to be unaffected ( KatTman et al., 1963), and no alteration of cellular proteins has been dcmonstrated. Therefore, if cellular fluid regulation is changed b y TET, the likely abnormality would be in the composition of the cell membrane itself. Although the ion kinetics involved in hyperpermeable membranes is exceedingly complex, a simplified and superficial review of some of the consequences affecting cell volume appears necessary a t this point. Previous studies of the mechanisms regulating cell water have revealed that, although diffusible ions do pass

76

H. TORACK, J. GORDON, AND J. PROKOP

through cell membranes, this ionic transfer is not conipletely free ( Conway, 1954; Ling, 1966; Lichtcnstein, 1966). This limitation of ion movement coupled with the activity of ion pumps has the effect of rendering these membranes impermeable to these diffusible ions so that high concentrations of potassium and sodium can be maintained inside and outside these cc4s respectively ( Skou, 1965). Any reduction of the inherent impediment to ion flow would initially require increased activity of the ion pump to maintain normal intra- and extracellular ionic concentration. Membrane permeability can also be increased to a degree in which the ion pumps no longer could maintain these ionic gradients ( Hempling). At this point, potassium would leave the cell, and the cellular concentration of sodium would rise. With a physiological extracellular fluid these ions would exchange proportionally, and no cell volume change need occur since the total ion content would remain constant. However, if an abnormally high amount of one ion, particularly Na', was located extracellularly, then sodium would enter the cell faster than potassium would leave. This would result in an increase in the total ion content causing the cell to swell. When the ion kinetics in TET-poisoned rat cortex have been studied by Reed et (11. (1964), a decrease in potassium and an increase in sodium has been described which they believe is intracellular. Quite logically they have proposed a leaky membrane as the reason for these abnornial ion concentrations. However, Katzman et al. (1963) could not find similar values in the white matter, which is more important since this is where most of the fluid and sodium occurs. They have found a decrease in potassium and an increase in sodium, but the sodium is not readily exchanged. So the leaky membrane theory has not gained acceptance. The results of this study do reveal that cell swelling, similar to the cortex, does occur in the white matter before significant cleft formation. This would support a membrane effect as the basis for fluid accumulation in white matter as well as in gray matter. As previously mentioned, the final evolution of normal ccll size could also be explained by a freely permeable membrane. Obviously, any additional evidence that this actually occurs would be of immense support for the entire thesis. Recently we have been studying various factors that influence cell swelling, by observing the changes in choroidal epithelium which are in-

PATHOBIOLOCY OF ACUTE TRIETHYLTIN INTOXICATION

77

duced by various siibstancrns injected into the cisterna magna. In this way, the effect of hypotonic and hypertonic saline, ouabain, sucrose, deoxycholate, and Amphotericin B upon choroid epithelial cell size and ultrastructure has been investigated. This system affords a new opportunity to study osmotic effects on mammalian cells in uiuo, particularly on neuroectoderni. Of these various compounds, Amphotericin B is of greatest interest since it is believed to punch small holes in cell membranes, which allow increased transfer of sodium and potassium while not altering the sodium pump (Lichtenstein and Leaf, 1966; Sharp et al., 1966; Fanestil, 1968). Amphotericin B is believed to interact with sterol components in membranes, especially cholesterol (van Zutphen et al., 1966; Demel et a!., 1965), to cause a reorientation of molecular structure. The effect upon red cells is a loss of K+ within 3 minutes (Butler et al., 1965) and a pitting of the cell membrane when it is negatively stained and viewed with an electron microscope ( Kinsky et al., 1966). This is the type of membrane effect which is being proposed for TET. A comparison of the effects of TET and Amphotericin B upon choroidal epithelium indicates that both alter the size and shape of the villi, both transform the electron microscope representation of the cell membrane from a linear bilaminar structure to a series of discontinuous vesicles, and both produce an electron-lucent zone in the subvillous ground substance. In each case the remainder of the cytoplasmic organelles appear normal. The swelling of the villi and dilution of the ground substance could be the result of increased sodium and fluid transfer into the cell. The vesicular pattern of the cell membrane could be related to an alteration of membrane structure, so that the regular diester linkage, which osmium is believed to produce in membranes (Korn, 1967), does not occur. Changes in choroidal epithelial cells, similar to these, also have been noted when deoxycholate is injected into the cistern especially when sucrose is substituted for saline as the vehicle for this compound. However, deoxycholate also appears to alter membranes of the endoplasmic reticulum as well. Deoxycholate is believed to produce bigger holes in cell membranes than Amphotericin B. We do not know how specific these membrane changes are; however, they are not observed after the injection of ouabain, hypotonic or hypertonic saline, or hypertonic sucrose. None of these latter com-

78

R. TORACK, J. GORDON, A N D J. PROKOP

pounds are believed to alter membrane structure. Therefore, the electron microscopical changes of choroidal epithelium induced by TET are practically identical with those resulting from Amphotericin B. These findings support the contention that TET alters membrane structure in a manner similar to Amphotericin B.

C. SPACECOMPARTMENTS I ~ YACUTEORCANOTIN EDEMA The sulfate space is generally accepted as representative of the extracellular space (Walser et al., 1953; Barlow et d.,1961), and no significant change could be demonstrated at 4, 12, or 22 hours in TET animals as compared to thc controls. The significance of the sucrose space is somewhat controversial, for, although originally conceived as another measure of extracellular space, values taken after more than 4 hours of uptake are always much higher than thiocyanate or sulfate (Raskin and Fishman, 1966). This slow compartment has been considered to be glial (Reed and Woodbury, 1963) and neuronal (Katzman et al., 1969) in origin. Although not statistically significant, there was a suggestion that the 22-hour acute TET group of rats and the chronic TET group have somc reduction in the size of the sucrose space. Certainly no increase in this space occurs. In this reduction is real, it would imply a decrease of an intracellular space since the sulfate space is unchanged. A reduction of cell size is suggested late in acute alkyltin poisoning at which time electron micrographs reveal the astrocytes to be somewhat shrunken. Since these cells account for a relatively small number of cells in the white matter, the lack of a more distinctive alteration in sucrose space is not surprising. The absence of any increase in sulfate space particularly early in the evolution of organotin edema is consistent with the morphological data revealing normal sized spaces. Physiologically this remains somewhat surprising, but it emphasizes the necessity of a widespread alteration of nieinbrane permeability involving endothelial cells, glial cells, and axons. Then fluid and sodium entering the brain could be quickly transferred into cells and would not produce significant dilatation or concentration in the extracellular space. In the final stage of acute edema, some perivascular enlargement of the extracellular space is noted in electron micrographs. However, in this tissue the extracellular space around myelin clefts is practically absent which probably renders the measurement of the total space unchanged.

PATHOBIOLOGY OF ACUTE TRIETILYLTIN ISTOXICATIOS

D. THE ROLE EDEMA

OF

ATPASEI N

THE

79

DEVELOPMENT OF TRIETHYLIIX

1. ~listoc.hemicnlly Deniotistrcrhlc Activity The demonstration of ;I inc,tnbrRnc.-l)ound ATPase in the white matter of rats, which is inhi1)itcd during the evolution of TET edema (Torack, 1965b), has h > n considered to be a good indication that this enzyme is involvcd in fluid regulation. This type of enzymatic activity must be tlistinguislied from the mitochondria] ATPase, which several biochemists a l s o have demonstrated to be inhibited by T E T ( Aldridgc m c l Cremer, 1955; Aldridge, 1958; Cremer, 1967a,b; Moore and Brody, 1961a,b). These latter enzymes are not easily demonstrated in fixed tissue because the enzyme protein is too readily inactivated by aldehyde fixation which is used to stabilize cell structure. Howcwr, the sturdier niernbrane-bound Mg" ATPase survives tissue fixation, and a final reaction product can be consistently localizcd to ;I variety of cellular structures in both the whitc, and gray matter (Torack and Barrnett, 1963, 1964; Torack, 1 9 6 5 ~ ) .The distribation of this end product can be altered by the use of different aldehyde fixatives prior to the enzymatic incubation ( Torack, 198%). Since these aldehydes affect proteins in rather specific \ w y s ( Fravnkel-Conrat and Mecham, 1949; Milch, 1965), this difference in distribution of surviving enzyme activity can be consickred to indicate different enzyme proteins all capable of hydrolyzing ATP. The ATPase activity which srirvivcd glutaraldehyde fixation has evoked particular interest, not only bccause it has been located in the white matter but also because certain glial cells in the white matter have been shown to possess this activity (Torack, 196%). The location of this somewhat spcicific ATPase on these cells has appeared to indicate a chemical distinction of these cells in the white matter. Glial cclls, especially astrocytes, have been considered to partake in fluid transport ( Ile Kobertis and Gershenfeld, 1961; Luse and Harris, 1961) and ion transport (Hertz, 1965). The inhibition of a cell membranc enzyme of these cells in the white matter has been exciting sinccx it could explain how organotin produces edema and why edema occiirs i n the white matter. Additional data, however, have rendered this concept less definite. To begin with, the accurate localization of a metallic end product in heavily myelinated tissuc is extremely difficult because

80

R . TORACK, J. GORDON, AND J. I'ROKOP

altered myelin inevitably precipitates hravy metals, such as lead, causing them to form deposits not only on myelin but also on adjacent cells. A variety of celliilar localizations can be demonstrated in axons and glial cells, which may vary even in the same block of tissue. In tissues that are optimally preserved, the final product is present on large tubules of the axoplasm and on cell membranes of glial cells sometimes resembling oligodendrocytes. These may represent the real sites of glutaraldehyde resistant Mg++ ATPase. However, a more serious objection arises from the sequential analysis of cell swelling and ATPase inhibition in acute T E T poisoning. At 12 hours after injection no consistent change in enzymatic activity is present, although the swelling of cells is most prominent. On the other hand, an almost complete loss of staining in the white matter at 24 hours is unaccompanied by any distinctive morphological change in glial cells. Since myelin is generally believed to be enzymatically inert (Adams, 1965), it is dificult to ascribe the cleft formation to this enzyme inhibition. There could be a relationship between the ATPase inhibition and the aggregation of organelles in axons, but this latter finding has not been considered to be a factor in fluid accumulation. Although the ATPase inactivation appears to be a terminal event, a degree of enzymatic inhibition could be present at the onset of fluid accumulation but not be detectable by the low sensitivity of the histochemical technique.

2. Biochemicd Assuy of Triethyltin Efects on ATPase The Pi-releasing activity of cerebral homogenates can be separated into four components on the basis of T E T sensitivity. At low concentrations of and lo-; A4, TET stimulates the amount of Pi released. Higher concentrations cause inhibition of Pi-releasing activity. The inhibition can be divided into two components, one of which has an approximate K , , , , of 4.3 x lo-'; h-l; the other has a K,,, of 2 x lO-.'M. The existence of a maximal inhibitory effect of T E T of 55.9% of the control activity indicates that the remaining activity is not sensitive to this alkyltin compound. The biological significance of these results require cautious interpretation. The high energy of the phosphodiester bond of ATP is probably utilized in cell processes by the transfer of the terminal phosphate group to an intermediate compound (Skou, 1960). This should be followed by the replacement of the phosphate group on

PATHOBIOLOGY OF ACUTE T R I E T H Y L T I N INTOXICATION

81

the intermediate by another compound as seen in muscle contraction, protein and nucleic acid synthesis, or as postulated in the active transport of Na+ (Badcr et nl., 1968; Heinz, 1967; Skou, 1965). The over-all effect of such a system would be the release of inorganic phosphate from ATP. Thus, it is erroneous to speak of an ATPase without the demonstration of a single protein moiety capable of splitting ATP. If all that is bcing measured is the release of inorganic phosphoros it is bcttcr to dcfine the assay as one for ATPase activity. Furthermore, thc intervening reactions, if any, in the hydrolysis of ATP are not cwident, making interpretation of any such experiments speculative. With this precaution in mind, it can be tentatively postulated that the existence of four components of ATPase activity in cerebral homogenates with respect to TET scnsitivity reflects the existence of more than one site of action for TET on the energy metabolism of the brain. These sites may h a w different tissue, cellular, or macromolecular localizations. Triethyltin has been shown to effect the energy metabolism of brain slices (Cremer, 1957, 1961, 1967a,b; Moore and Brody, 1961a ) and to uncouple oxidative phosphorylation in liver mitochondria ( Aldridgc, 1958). Though mitochondria normally synthesize ATP, they ciin he induced to hydrolyze it as a result of dinitrophenol ( DNP ) treatment or nonspecific damage. The resultant ATPase activity has been shown to be sensitive to the presence of TET ( Aldridge, 1958), and the kinetics of this TET sensitivity shows a striking similarity to part of that for brain ATPase. The stimulation of DNP ATPase at low concentrations with a half maximal concentration of 1 x M T E T is close to that for the stimulation of cerebral ATPase. In addition the K,,, of the first component of T E T inhibition of cerebral ATPase approximates those found for liver mitochondria of 2.5 x lo-: and 1x by Aldridge (1958) and Rrody and Moore (1962), respectively. Homogenization of brain without special precautions as has been done in this study is likely to result in damaged mitochondria with a high Mgt7 ATPase activity (Aldridge, 1958). Therefore, it is possible that cerebral mitochondria are the sites sensitive to the low concentration of TET. We have been chiefly concerned with a glutaraldehyde-resistant ATPase in the histochemical procedures. An enzyme capable of developing ion gradients in mitochondria also has been demonstrated to survive glutaraldrh ydc fixation ( Utsumi and Packer,

82

R. TORACK, J . GORDOh-, AND J . PROKOP

1967). Accordingly we have attempted to demonstrate a similarity between the kinetics of TET inhibition of P, release in glutaraldehyde-treated material and a component found in untreated tissue. A TET-sensitive activity is found in cerebral tissue homogenized in 2% glutaraldchyde, in agreement with earlier data (Gordon and Torack, 1967). Though glutaraldehyde inhibits cerebral ATPase activity, the activity of 2%glutaraldehyde-treated material is always higher than that of untreated brain in the presence of T E T with the concentrations studied. Thus, glutaraldehyde appears to alter the sensitivity of the ATPase, in cerebral homogenates to TET, preventing a direct comparison between unfixed and glutaraldehyde-treated brains. This altered sensitivity could also mask an early inhibition of enzyme by TET and at least partly explain the poor correlation between the histochemical studies and the onset of fluid accumulation. Although an inhibition of glutaraldehyde-resistant ATPase activity by T E T can be demonstrated histochemically and biochernically certain questions remain unanswered. What is the precise cellular location of this enzyme? Is this a specific ATPase? Is some inactivation occurring at 12 hours in acute TET poisoning? Until these difficult problems can be resolved, a role of TET-induced ATPase inhibition in the evolution of cerebral edema must remain possible but not proven. V. Conclusions

The rapid induction of triethyltin intoxication during a 24-hour period appears to emphasize the sequential morphological, chemical, and physiological responses of the nervous system which lead to cerebral edema. In these animals two distinct episodes of stupor and coma are identified. One occurs shortly after the administration of the organotin, is not present following a dose of 4 mg/kg or less, and does not appear to be related to fluid imbalance. The second episode is slowly progressive and appears to be temporally related to the accumulation of fluid in the brain. The sequential study of altered morphology in the white matter reveals that a period of cell swelling, involving astrocytes and axons, occurs at about the time of earliest increase of fluid in the brain. At this time there is similar involvement of astrocytes in the gray matter, so astrocytes in both the gray and white matter appear to be reacting similarly. In the white matter this period of cell

PATHOBIOLOGP OF ACUTE TRIETHYLTIN INTOXICATION

53

swelling is limited by the dtw~lopmciitof myelin clefts. The excess fluid in thr white matter appc’ars to accumulate in these clefts, and the cytoplasmic volume of w l l s in thr white matter returns to normal or slightly less than norm;il. Thcsr, morphological findiiigs, prwioris elcctrolyte determinations, and the measurement of normal-sizcd extracellular spaces, as determined by ;{‘S uptake, :ir(’ consistent with the concept that the cell membranes in trirthvltin poisoning arc hyperpermcable. A membrane effect could also Iic~lpto explain the massive splitting of myelin sheaths within which the cxcc’ss fluid stagnates. Additional evidcnce for a direct c4lcc.t upon cc.11 membranes is obtained from the morphologically similar effcct upon choroidal epithelia which appears after the injection of tricthyltin and Amphotericin B. Amphotericin B is believed to alter the molecular pattern of cell membranes to increase permoaliility without affecting ion pumps. The induction of like changes in choroidal epithelia supports the contention that a similar c1i;uigc. in membrane structure occurs with triethyltin. The possibility remains thiit tlic mcmbrane alteration is due to an inhibition of a mcml)rniic-bouncl ATPase which has been shown to be produced by organotin lioth in histochemical and biochemical assays. The relativvly late involvement of this enzyme inhibition during acute alkyltin poisoning suggests that this enzyme effect is less specific than a dircct c>ffectupon membrane structure. The focal swelling of some sons without significant myclin cleft formation appears to indicate a specificity of these nerve fibers. Two types of swelling a r c obsc~rvcd.The first type occurs simultaneously with astrocytic, s u ~ l l i n gand appears to be due to Iiyperpermeable mcmbranvs. The second swelling is only observed late in this study and appears to 11c due to an aggregation of various cytoplasmic organelles. Siniiliir aggrcgates have been observed previously following impaired ax011 flow. Therefore, the early swelling owing to fluid imbalance is believed to impair axon flow resiilting in the focal aggregation o f organelles.

Adains, C. W. M. ( 1965). “IIistoclrc.inistl.? ant1 Cytocheinistvy of the Nervous System.” Elsevier, Ainstwdaii~. Aldridge, W. N. ( 1958). Bioclioii. 1. 69, 367-376. Aldridge, W. N., and Creiner, J. E. ( 1 9 5 5 ) . Biochem. J. 861, 406-418.

84

R. TORACK, J. GORDON, AND J. PROKOP

A h , F. P., Katzman, R., a n d Terry, R. D. ( 1963). J . Necwopathol. Exptl. Neuml. 22, 4OS3-413. Hader, H., Post, R. L., :and Bond. C,. 11. ( 1968). Biochim. Rioph!is. Arta 150, 41-46. Bakay, L. ( 1965a). J. Neurol. Sci. 2, 52. Bakay, L. (1965b). Progr. Brain Res. 15, 155. Bakay, L., and Lee, J. C. ( 1965). “Cerebral Edema.” Thomas, Springfield, Illinois. Barlow, C. A., 1)omek, h’. S., Goldberg, hl., and Roth, L. J. ( 1961). Arch. Neurol. 5, 102-110. Brightman, M. (1965). J. Cell B i d . 26, 99-123. Brody, T. M., and Moore, K. E. (1962). Federation Proc. 21, 1103-1106. Butler, M7. T., Alling, D. W., and Cotlove, E. (1965). Proc. Soc. Expt!. B i d Med. 118, 297-300. Caldwell, P. C. (1968). Physiol. Reo. 48, 1-64. Conway, E. J. (1954). Symp. Soc. Exptl. B i d . 8, 297-324. Cremer, J. E. (1957). Biochem. J. 67, 87-96. Cremer, J. E. ( 1961). Biochem. Pharmacol. 6, 153-160. Cremer, J. E. (1962). J. Neurocheni. 9, 289-298. Cremer, J. E. (1967a). Biochem. J. 104, 212-223. Cremer, J. E. (196713). Biochem. J. 104, 22-3-228. Dnnel, R. A,, van Deenen, L. L. M., and Kinsky, S. C. (1965). 1. Biol. Cliein. 240, 2749-2753. De Robertis, E. D. P., and Gershenfeld, H. M. (1961). Intern. Reu. Neurobiol. 3, 1-65. Dick, D. A. T. (1966). “Cell Water.” Britterworth, London and Washington, D.C. Eisenman, G., Sandbloni, J. P., and Wdker, J. L., Jr. (1967). Science 155, 965-973. Fanestil, D. D. (1968). J. Lab. CIiti. Med. 71, 548-554. Fraenkel-Conrat, H., and Mecham, D. K. (1949). J. B i d . Chem. 177, 477. Freeman, A. R., Reuben, J. P., Brandt, P. W,, and Grundfest, H. (1966). J. Gen. Physiol. 50, 423. Gerschenfeld, H. M., Wald, F., Zadunaisky, J. S., and De Robertis, E. 1). P. ( 1959). Neurology 9, 412425. Gonatas, N. K. (1967). Nature 214, 352-355. Cordon, J. S., and Torack, R. M. (1967). J. Nerrrochem. 14, 1155-1160. Harnack, E. (1878). Arch. Erptl. Path. Pharmakol. 114, 39. Heinz, E. (1967). Ann. Reo. PhlJsioZ. 29, 21. Hempling, H. ( 1968 ). Personal communication. Hertz, L. (1965). Nature 206, 1091-1094. Hirano, A., Zimnierman, H. M., and Levine, S. ( 1967). 1. Neuropathol. Exptl. Neurol. 24, 244-255. Hirano, A,, Zimnierman, H. M., and Levine, S. (1966). J. Cell B i d . 3, 397411. Hirano, A., Zimnierman, H. M., and Levine, S. (1967). J. Neuropathol. E x p t l . Neurol. 26, 200-213. Hodgkin, A. L. ( 1951). Biol. Rez;. 26, 339-409.

PATHOBIOLOGY OF ACUTF 1'RlE1'HTLTIN INTOXICATIOlV

85

Katzman, R., Aleu, F., and Wilson, ( z . (1963). Arch. Neurol. 9, 178-187. Katzman, R., Lehrer, G . M., and Wilson, C. E. (1969). J. NeuropathoL Exptl. Neurol. 28, 167. Kavanau, J. L. ( 1966). Federatiori Proc. 25, 10901-1107. Kinsky, S. C., Luse, S. A., and van Ileenen, L. L. (1966). Federation Proc. 25, 1503-1510. Klatzo, I., Piraux, A., and Laskowski, E. J. (1958). J. Neuropathol. Exj)tl. Neurol. 17, 548-564. Korn, E. D. ( 1967). J. Cell Biol. 34, 627-639. Kidler, S. W., Nicholls, J. C., and Orknnd, R, K. (1966). J. Neurophysiol. 29, 769-787. Lampert, P. W. (1967a). Actu Nenropathol. 9, 99-126. Lampert, P. W. (1967b). J. Neurojxitliol. Exptl. Neurol. 26, 345-368. Lasansky, A., and Wald, F. (1962). J. Cell Hiol. 15, 463479. Lee, J. C., and Bakay, L. (1965). Arch. Nerrrol. 13, 48-57. Lee, J. C., and Bakay, L. (1966). Arch. Nerrrol. 14, 36-49. Lichtenstein, N. S. (1966). Anti. N.Y. Accid. Sci. 137, 441472. Lichtenstein, N. S., and Leaf, A. (1966). Anti. N.Y. Acad. Sci. 137, 556-565. Ling, C.. N. (1966). Ann. N.Y. Acad. Sri. 137, 830-836. Lnckb, B., and hlcCutcheon, h4. (19SO). A.M.A. Arch. Puthol. 10, 662-670. Lnft, J. hl. (1961). J. Biophys. Riochenz. Cyfol. 9, 409414. Luse, S. A. (1960a). J. Neurosurg. 17, 439-446. Lnse, S. A. (1960b). J. Ultrustrurt. Rev. 4, 108-112. Luse, S. A., and Harris, B. (1961 ). Arcli. N ~ I I K4,J ~139-152. . Xlagee, P. N., Stoner, H. B., and Barnrs, J . hl. ( 1957). J. Pathol. Bacteriol. 73, 107-124. ?\lilch, R. A. (1965). Gerontologici 10, 117-136. Moore, K. E., and Brody, T. hl. ( 196la). Bioclieiti. PIiarmucol. 6, 134-142. Sloore, K. E., and Brody, T. hl. (19611)). J . Pltarmacol. Exptl. Therap. 132, 6-12. Nevin, N. E. ( 1967 ). J. Netrro/xitliol. E x p t l . Weurol. 26, 77-84. Nicholls, J. C., and Kuffler, S. It. (1964). J. Nwrophysiol. 27, 645-671. O'Brien, J. S. (1965). Science 147, 1099-1107. OBrien, J. S., and Sampson, E. L. (1965). Science 150, 1613-1614. Opie, E. L. (1948). J. Exptl. Metl. 87, 425-444. Palay, S. L., McGee-Russel, S. M . . (;ordon, S., and Grillo, hl. A . ( 1962). J. Cell B i d . 12, 385-410. Peters, A. (1966). Quart. J. Exlit/. Pltysiol. 51, 229. Raskin, N. H., and Fishinan, n. A. (1966). Arch. Neurol. 14, 21-30. Reed, D. J., and Woodbury, D. hl. (1963). 1. Physiol. (Loridon) 169, 8101850. Reed, D. J., Woodbury, 1).M., and Holtzer, R. L. (1964). Arch. Neicrol. 10, 604-6 16. Reicliardt, M. (1919). Allgeni. 2. Pqcliiut. 75, 34-103. Robertson, J. I). (1955). J. B i o p h y ~ .Biochem. Cytol. 4, 349-364, Hobinson, J. R. ( 1966). Federation Proc:. 25, 1109-1 111. Scheinker, I. ( 1941 ). A.M.A. Arcslt. Nvrrrol. Ps!/cliiaf. 45, 117-129.

86

R . TORACK, J . GORDON, AND J. PROKOP

Sharp, G. W., Coggins, C. H., Lichtenstein, N. S., and Leaf, A . 1966 . J. Clin. I m e s t . 45, 1640-1647. Singer, M., and Salpeter, M. M. (1967). Nature 210, 1225-1227. Skou, J. C . (1960). Biochim. Biophys. Acta 42, C23. Skou, J. C . (1965). Physiol. Reo. 45, 59G-617. Spatz, H. (1929). Arch. Psychiat. Neroenkrankh. 88, 79C794. Stewart-Wallace, A. M. (1939). Brain 62, 42fj-438. Stoner, H. B., and Threlfall, C . J. (1958). Biochem. J. 69, 376-385. Stoner, H. B., Barnes, J. M., and Duff, J. L. (1955). Brit. J. Pharmacol. 10, 16-25, Streicher, E. (1962). J. Neuropathol. Exptl. Neurol. 21, 437. Tandler, B., Erlandson, R. A., and Wynder, E. L. (1968). A m . J . Puthol. 52, 69-95. Taussky, H. H., and Sliorr, E. (19ij3). J . B i d . Chem. 202, 675-685. Torack, R. M. (1965a). Z . Zellforsch. Mikroskop. Anat. 6’6, 352-364. Torack, R. M . (1965b). Am. J. Pathol. 46, 245-261. Torack, R. M . ( 1 9 6 5 ~ )J. . Histochem. Cytochem. 13, 191-205. Torack, R. M., and Barrnett, R. J. (1963). J . Histochem. C!ytocheni. 11, 763-772. Torack, R. M., and Barrnett, R. J. (1964). J . Nrnropafhol. E x p t l . NettroZ. 23, 46-59. Torack, R . hl., Terry, R. D., and Ziminerman, H. M. ( 1959). Am. J. Pathol. 35, 1135-1147. Torack, R. M., Terry, R. D., and Zimmerman, €I. M. (1960). A m . J. Puthol. 36, 273-287. Torack, R. hl., Dufty, hl. L., and Gordon, J. S. ( 1967). I n “Brain Edema” ( I . Klatzo and F. Seitelberger, eds.), p. 491. Springer, Berlin. Utsumi, K., and Packer, L. ( 1967). Arch. Biochem. Biophys. 121, 633-640. van Harreveld, A , , Crowell, J., and Malhotra, S . K. (1965). J . Cell Biol. 25, 117-137. van Harreveld, A,, Collewyn, H., and Malhotra, S . K. (1966). Am. J. Physiol. 2110, 251-256. van Zutplien, H., van Deenen, L. L. M., and Kinsky, S. C. (1966). Biochem. Biophys. Res. Commzm. 240, 2749. Villegas, R., Capnto, C., and Villegas, L. (1961 ). J. Gen. Phy.sio1. 46, 245. Wachstein, M., and Meisel, E. (1957). Am. J. Clin. Pathol. 27, 13-23. Walser, M., Seldin, D. W., and Grollman, A. ( 1953). J. Clin. Znoest. 32, 299311. Wasterlain, C. G., and Posner, J. (1968). Arch. Neurol. 19, 71-79. Wasterlain, C. C . , and Torack, R. M. (1968). Arch. Nerrrol. 19, 79-87. Weiss, P. A. ( 1967). Neurosci. Res. Program Bull. 5, 371-400. Wilson, J. W., and Leduc, E. H. (1963). J. Cell B i d . 16, 281-296.

ASCENDING CONTROL OF THALAMIC AND CORTICAL RESPONSIVENESS By M. Steriade Loboratoire de Neurophysiologie, Departement de Phyriologie, Facult; de Mkdecine, Univerrite Lovol, Quebec, Canada

I. Background . . . . . . . . . . . . 11. Testing Responses . . . . . . . . . . A. Sensory Thalamocortical Systcnis . . . . . . . B. Motor Cortex Responses and Pyramirlal Tract Evoked Discharges 111. Unspecific and Specific 1nflucwcc.s o n Thalamocortical Coiiiplexcs . A. Visual Relays . . . . . . . . . . . B. Auditory Relays . . . . . . . . . . C. Somesthetic Relays . . . . . . . . . . . . . . . . . . . . D. Motor System . IV. Final Remarks . . . . . . . . . . References . . . . . . . . . . . .

87 91 91 94 97 97 113 116 119 133 136

I. Background

The concept of unspecific and specific ascending influences exerted on sensory and inotor systems is largely an outgrowth of experiments and theories concerning sleep and wakefulness. Thus. in order to analyze the factors affecting the synaptic transmission of afferent messages through tlialamic and cortical neuronal chains, the relay of messages through thc motor cortex to the efferent paths, and to correlate some of these physiological activities with perceptual integration at differcwt levcls of awareness, it is necessary to present a brief preliminary srirvey of cerebral systems regulating wakefulness and sleep. Indrcd, wakefulness or awareness and their neurophysiological correlate, the “tonus central” of the cerebrum ( Bremer, 1935), represent “ a l>rercqiiisite for the performance of the highest functions of the central nervous system” ( Moruzzi. 1958, p. 23). The modern version of thc hypothesis that sleep is a consequence of deafferentation (Kleitman, 1929) found its crucial experimental proof in Bremcr’s expcrimcnts ( 1935) in which electro87

88

.

hZ. STERIADE

cncephalographic ( EEG ) and behavioral sleep were induced by interruption of the corticopetal paths by means of complete mesencephalic transection ( cerveau isole‘ preparation), Although lethargy was produced in the monkey by lesions of the lateral hypothalamic areas (Ranson, 1939), that effect was ascribed to elimination of a downward discharge rathcr than to interruption of an ascending tonic influence (Ranson and Magoun, 1939). In 1949, Moruzzi and Magoun described the EEG patterns of arousal (low voltage and fast rhythms) obtained by stimulating the brainstem reticular core with high-frequency electrical pulses. These patterns could be reproduced even after section of lemniscal pathways, thus leading to the hypothesis stating that the reticular formation ( R F ) is the wakefulness center and that sleep occurs by elimination of the waking influence of the ascending reticular system rather than by reduction of classic sensory inflow (Moruzzi and Magoun, 1949; see also Lindsley et al., 1950). Sleep can no longer be regarded as a purely passive process. The Pavlovian type of sleep, actively precipitated by monotonous sensory stimulation (see details in Moruzzi, 1960), was confirmed in studies by Roitbak (1960) and Pompeiano and Swett (1962) showing that low-frequency stimulation of the skin or of low threshold cutaneous fibers induced EEG synchronization, sometimes associated with behavioral sleep symptoms. Similarly, stiniulation of the large diameter fibers of cardiovascular origin in the vagal trunks caused closing of the pupil as well as EEG spindles and slow waves (Dell and Padel, 1965). Other experimental data have shown that behavioral sleep and its common correlate, the synchronization of the EEG, can be induced by stimulating with lowfrequency pulses near the massa intermedia (Hess, 1949), the intralaminar thalamic nuclei (Akert et ul., 1952), the caudate nucleus (Heath and Hodes, 1952; Buchwald et al., 1961), the region of the solitary tract (Magnes et ul., 1961), and the preoptic basal forebrain area ( Sterman and Clemente, 19624 Behavioral sleep and/or EEG synchronizing reactions were obtained in unanesthetized acute preparations and behaving animals also by ‘The hypnogenic effect of such stimulations could be ascribed to the parameters of electrical stimulation, since synchronizing ( inhibitory-like) reactions obtained by low-frequency stimulation of the caudate nucleus, midline thalamic, or other unspecific structures change into activating reactions on fast repetitive stimulation of the same points. Even the most “activating” structure,

ASCENDIXG REGU1,ATING CEREBRAL SYSTEMS

89

chemical‘ or high-frequency el(~trica1stimulation of the preoptic basal forebrain areas ( Sterman and Clemente, 1962b; HernhndezPe6n et al., 1963), the ventral part of the basolateral amygdala ( Kreindler and Steriade, 1964), the ventral pontine RF (Demetrescu and Demetrescu, 1962b ) , and well-defined districts of the medulla ( Steriade, 196913). On thc other hand, high-frequency stimulation of the R F led to deeper (“activated or “paradoxical”) sleep (Dement, 1958; Jouvet, 1962) when it was performed toward the end of a period of intermediate (slow wave) sleep (Jouvet and Michel, 1960; Rossi et aZ., 1961). Conversely, a long series of experiments showed that when synchronizing mechanisms were cut off from superior levels by complete transections, discrete lesions, or reversible inactivation, persistent EEG desynchronization and behavioral arousal could be obtained. Thus, Nauta (1946) suggcsted a sleep center in the anterior hypothalamus (the “trophotropic” area of Hess), transection of this region having induced marked hyperactivity in rats. The Pisa group obtained enduring EEC, and behavioral patterns of wakefulness following complc~te ( Batini et al., 1959) or ipsilateral (Cordeau and Mancia, 1959) midpontine pretrigeniinal transection, inactivation of synchronizing centers of the lower brainstem by intravertebral administration of barbiturate ( Magni et al., 1959; Rosina and Mancia, 1966) or cooling of the floor of the fourth ventricle (Berlucchi et al., 1964). Similarly, Bonvallet and her associates (Bonvallet and Bloch, 1961; Bonvallet and Allen, 1963) found that interruption of the pathways between the medulla and rostral RF or localized lesions of the nucleus of the solitary tract were followed by an increase in the duration of cortical arousal and suggested a decrease in ncgative feed-back control of rostral activating reticular system. Thc above mentioned data permitted the mesencephalic HF, niay produre on low-frequency stimulation a widespread synchronization of the electrocortical activity in the awake animal, changing into the well-known arorisal rraction b y increasing frequency of electrical pulses ( Kaada et uZ., 1067 ). Therefor?, the same parameters ( namely high-frequency electrical pulses whose effects reproduce those elicited by chemical stimulation) should be used for differentiation between “activating” and “inhibitory” systems. ’As pointed out by Cordeau et al. ( 1963), synchronization of the EEG accompanied by drowsiness and sleep may be induced by injections of aL-tylcholine (even in the rostral 1)rainstem H F ) , while injections of adrenaline produce behavioral arousal and cles).nclironization of the EEC.

90

Xi. STERIADE

the inference of a deactivating or inhibitory influence exerted by medullary and lower pontine structures upon the mesencephalic activating system or more rostra1 structures (Dell et al., 1961; Moruzzi, 1964). It seems that the tonic, synchronizing influence of the cerebellar fastigial nucleus on the electrocortical rhythms, recently described by Manzoni et al. (1968), is not different from that of other rhombencephalic structures. Evidence was also presented for a deactivation of the mesencephalic reticular system “from in front” ( Jouvet, 1967a) by experiments showing the crucial role of the orbital cortex in regulation of thalamocortical synchronization (Jouvet, 1962; Velasco and Lindsley, 1965). Opposite alterations of cortical enzymic activity was found to occur in association with opposite patterns of EEG activity in the midpontine and rostropontine preparations ( Steriade et ul., 1969a). The paradoxical phase of sleep (PS; see details on behavioral and electrophysiological tonic and phasic aspects in a review by Jouvet, 1967a), during which a striking increase in the threshold of reticular arousal is observed (Hubel, 1960; Benoit and Bloch, 1960), seems to be not simply a deepened stage of the slow wave sleep (SWS), but a qualitatively distinct state, as suggested by the duality of structures and mechanisms underlying these two states. Actually, serotonincrgic neurons concentrated in the raphe system appear to be involved in the mechanism of SWS, while different circumscribed zones of the pons (lateral part of the RF and noradrenergic neuronal population located in the locus coeruleus ) seem responsible for phasic and tonic phenomena of PS (Jouvet, 196% ) . That sleep is also due to an active, and not only to a purely passive process, is further indicated by the analysis of unitary and mass evoked responses during stimulation of inhibitory structures or following transections interrupting the ascending inhibitory systems. Likewise, the studies of fluctuations in thalamic and cortical evoked responses during various stages of natural sleep and wakefulness in chronically implanted animals support the hypothesis of an active process. This review will consider the influences of the diffuse, unspecific structures (especially involved in wakefulness and sleep states) on the responsiveness of various neothalamocortical systems, and their interactions with the effects induced by stimulating or interrupting the specific pathways.

ASCENDING REGULATING CEREBRAL SYSTEMS

91

II. Testing Responses

A. SENSORYTHALAMOCOHTICAL SI’SI-I.;~

1. Potentials Eookecl hy Ceutlal Stimuli At first, modifications of 1,chavior and spontaneous EEG rhythms were used as signs of ascending activation. Later, experiments independently carried out by Rremer and Stoupel (1958) and Dumont and Dell ( 1958) pres(wtcd direct proof (see Magoun: 1959) of RF-induced potentiation of thalamocortical responsiveness, defined by an input-output relationship. These studics showed the presence of a dramatic enhancement of responses evoked in cortical sensory areas by shocks delivered to appropriate thalamic nuclei or afferent pathways chiring arousal induced by high-frequency R F stimulation. The potential cvoked in the visual cortex by a shock applied to thc afhwnt pathway was used mainly a s test response. This choice \YRS suggested by the classic and very systematic pattern of the surfaee-recorded response in the visual cortex, consisting of five compntmts: three fast positive deflections. preceding a slow. biphasic ( l)ositi\ie-negative) sequence ( Chang and Kaada, 1950). The origin of these deflections (numbered 1-5; see Fig. 1) was debated hotly. The nature of the fast, spikelike components 2 and 3 w a s rc~garcledby Bremer (1958) as uncertain. Combined analysis of the surfacr3-recorded mass response and unitary discharges suggcstccl thc prcsynaptic nature of deflection 2 and revealed that unit discharges related to component 3 can bc either pre- or postsynaptic (\Vitli.n and Ajmone-Marsan, 1960). There was agreement, on the other hand, concerning the origin of the other componcnts. Thc initial (no. 1) deflection (“radiation” potential) is presynaptic to the cortex and represents the mass discharge of thalamocortical imirons in the optic radiation; it can be recorded from the white matter after an extensive ablation of thc visual cortex. The slow components 4 and 5 are undoubtedl!. intracortical events and lwtray successive depolarizatioiis of neurons (in the deep and superficial layers) and apical deiidritic arborization. As pointed out by Lsndnu and Clare (1956) and by Bremer and Stoupel (1957), fast positive dcflcctions ( 1-3) preceding the biphasic slow sequcnce of surfaccl-rccmled shock-evoked responses

92

M. STERIADE

FIG. 1. Reticular influences on lateral geniculate ( LC ) and visual cortex ( VC ) centrally evoked responses. Etice'phale isole' cat. Testing shocks applied to the optic tract ( A ) , optic radiation 1-2 nun above the thalainic relay ( B ) and deep layers of the visual area ( C ) . Effects of high-frequency (300/sec) mesencephalic reticular stimulation depicted in the right column. Time: 2 msec (and 0.2 msec in B). Vertical bar: 0.3 mV. In this and subsequent figures, negativity upward. Note: marked fluctuations of postsynaptic ( 3-5) cortical components with constant presynaptic ( 1 ) deflection ( B ); selective enhancement of the postsynaptic ( r , and r2) LG responses during reticular arousal ( A ) and increase of the cortical response independent of geniculate events ( B and C ) (from Steriacle, 1969a).

ASCENDING REGULA1'1KG CEREBRAL SYSTEMS

93

may be seen also at the level of the auditory, somesthetic, and associational cortical areas. A similar, sometimes identical, pattern of shock-evoked responses in the visual cortex is shown for the somesthetic area in Figs. 8-10. The origin of the various components are ascribed to the same events as for those of visual responses. This is also valid for cortical responses evoked by strong stimuli applied to the surface or by direct stimulation of deep cortical layers. In this case, spikelike positive deflections recorded at the surface of the somesthetic cortex are reminiscent of similar potentials described a t the level of the visual area by G. H. Bishop and Clare (1953). Hence, no special mention will be made of responses in auditory and somesthetic areas. Many investigators have used as testing potentials the primary cortical responses evoked by a shock to the appropriate thalamic nuclei ( lateral geniculate body, LG; medial geniculate body, MG; ventrobasal complex, VPL and VPM ) , the amplitude of deflection 1 being used as an index of the thalamic output. However, in such cases some divergent results have been ascribed to a varying mixture of presynaptic and/or postsynaptic thalamic stimulation. Therefore, more recently, the responsiveness of the thalamic relays was dissociated from cortical responsiveness by applying stimuli before the thalamic nuclei (e.g., to the optic nerve, ON; chiasma; optic tract, OT; or to the leniniscal pathways) and after these relays, to the optic, auditory, or somesthetic radiations. Thus, amplitude of cortically recorded deflection 1 from prethalamic stimulation and of components 4-5 from radiation stimulation could be used to measure the synaptic transmission at the thalamic and cortical relays, respectively. In addition, more direct information concerning the control of the thalamic responsiveness can be derived from simultaneous recording of the specific thalamic nuclei when prethalamic stimulation is used. Tlw orthodromic LG response consists of an initial component reprwcnting the action potentials of presynaptic axons (tract: t ) , whilc subsequent components represent postsynaptic events (relayed components: rl and r , ) . This pattern (P. 0. Bishop, 1964) is depicted in Fig. 1A. As in the case of cortical responses, t and r components may be extrapolated for analysis of shock-evoked response in other specific thalamic relay nuclei (see Fig. 8 A ) . The small size and fragility of neurons in the visual cortex account for scarcity of data coiiwrning correlations between various componrnts of the surfacc.-recoldr.d iiiass response and simultnne-

94

h.i. STERIADE

ously recorded iiitracellular responses. However, some indications may be obtained from data of Watanabe et al. (1966) comparable to those found at the level of motor cortex, reporting relations between surface-positive and surface-negative waves with excitatory ( EPSPs) and inhibitory ( IPSPs) postsynaptic potentials (see later).

2. Pe riph e ra 11y Elicited Responses The primary responses of somesthetic and auditory cortices to natural (peripheral ) stimuli are well-known short-latency, initially positive potentials ( see review by Albe-Fessard, 1957). Flash-evoked responses are still a matter of controversy, especially concerning the origin of waves subsequent to the early biphasic deflection. In deeply anesthetized preparations, the afterdischarge is simplified, usually consisting of a single cortical wave; a “second complex” was found in good time relation with responses recorded from the optic disc, and was thought to be of retinal origin (Bignall and Rutledge, 1964). The analysis of flash-evoked responses in unanesthetized, acute preparations or in behaving animals, used for studies on reticular stimulation effects or slcepwakefulness variations in responsiveness, becomes very difficult. In these experimental conditions, “postprimary” ( b ) components (Fig. 4 ) were ascribed to an intrinsic cortical mechanism, in contrast to the initial ( a ) wave, having a retinal determination (for details, see review by Steriade, 1968).

B. MOTOR CORTEXRESPONSESA N D PYRAMIDAL TRACTEVOKED DISCHARGES Mass responses evoked in the cat’s motor cortex by shocks to the specific ( ventrolateral, VL ) thalamic nucleus, the cerebellothalamic pathway, or the cerebellar hemispheres, are most prominent at the level of the precruciate gyms, especially in its lateral part (Combs and Saxon, 1959; Sakata et ul., 1966; Schlag and Villablanca, 1967; Steriade, 1969b). In contrast with cortical sensory responses whose general configurations are identical whether due to thalamic or prethalainic stimulation, a VL-induced motor cortex response consists of a short-latency (1.2-1.5 msec to onset), biphasic ( positive-negative ) slow sequence, sometimes preceded by a small amplitude, spikelikc deflection of presynaptic origin ( see Fig. 9R,1 and Fig. 12R), while thc responsc evoked in the same

ASCENDING REGULATING CEREBRAL SYSTEMS

95

cortical area by a shock to the cerebellothalamic pathway at the level of the red nucleus ( K N ) is composed of two well-defined surface-positive slow waves; the second deflection appears at 3.5 msec and is followed by a negative component. Simultaneous recording of the VL response to rubral stimulation may reveal a potential consisting of a presynaptic deflection representing the action potential of the cerebellothalamic tract ( t ) and a negative postsynaptic (relayed, T ) coinponcmt, the onset of the latter preceding the cortical first positive wave (Fig. 12A). For the VL-induced response a second positive wave may occur at a latency of 8-10 msec a t stimulus intensities of 2-3 times threshold, and is associated with different intracellular evcnts than the first positive wave ( vide: infra ) , Motor cortex responses to VL shocks can be differentiated from other centrally evoked potentials in the same region (by VPL or unspecific thalamic stimulation ) by their cortical distribution and pattern of responses to single or low-rate stimuli. The VL-evoked response with an early, inconstant, b r i d deflection is distributed over the anterior sigmoid gyrus, whereas the VPL-evoked response begins with two or three well-defined, fast positive components (see difference in Fig. 9 ) , and is more evident over the postcruciate region. Short-latency, initially positive responses to single shocks differentiate VL-evoked potcntials from responses evoked by stiniulating unspecific thalamic nuclei. Besides, augmenting responses evoked by VL shocks at 7-12lsec with an initial positivity and characterized by the appearance of a second surface-positive component followed by a negative wave (Fig. 9R,2) may be dissociated from long-latency, surface negative recruiting responses of the motor cortex ( Brookhart and Zanchetti, 1956; Purpura and Housepian, 1961; Spencer and Brookhart, 1961). Development of a second positive wave during augmenting ( lO/sec), analogous to those elicited by VL stimulation, can be induced in the sigmoid gyri by stimulating the white matter, cvcn following removal of the entire thalamus (except LG, MG, and posterior part of the veiitrobasal complex) and most of the hcad of the caudate nucleus (Schlag and Villablanca, 1967). Thus, the thalamus does not seem to be necessary for the production of cortical incremental responses, and the second positive wave might be regarded in such preparations ;is a sign of cortical elaboration. However, upon repetition of the V L stimuli at 100 msec interval, unit activity recorded in the same

96

M. STERIADE

thalamic nucleus was changed into an “augmenting” pattern even in decorticate preparations ( Schlag and Villablanca, 1968). It seems, therefore, that both specific thalamic nuclei and cortical receiving areas are able to generate an incremental activity of the “augmenting” type. Intracellular recordings showed a succession of EPSPs and a subsequent IPSP evoked by a single shock to the VL nucleus. The first EPSP has a latency of 14 msec, leads frequently to a discharge, and corresponds to the primary surface-positive wave. With increasing intensity of stimulation, an IPSP may appear with a latency of 5-10 msec, shortening the second EPSP and occurring in association with a second surface-positive deflection or even during the primary surface negativity (Purpura et al., 1964; Creutzfeldt et al., 1966; Grossman et al., 1967). This hyperpolarizing potential might be due to recurrent collateral inhibition (Nacimiento et al., 1964). Augmenting responses are accompanied by attenuation of IPSPs and progressive enhancement of late EPSPs (Purpura et al., 1964). It was suggested by Creutzfeldt et al. (1966) that mass evoked responses consist of different intracortical excitatory patterns : besides the synchronous afferent volley, the primary EPSP and the corticofugal spike discharge compose the primary surface positivity; the primary negativity was regarded as a mixture of transients from the secondary EPSP to the IPSP; later, the IPSP remains the determinant factor for the descending limb of the surface-negative wave and for the subsequent positivity. The pyramidal tract (PT) discharges evoked by a shock to the VL nucleus consist of a direct ( D ) , unrelayed component, at 0.50.7 msec, likely elicited by stimulus spread to the internal capsule or by activation of collateral axons to the VL (Clare et al., 1964), and of relayed (monosynaptic, R1, at 1.5-1.8 msec; and shortlatency polysynaptic, R2, at 3-3.5 msec) components, as described by Amassian and Weiner (1966). Depending on the site of the VL stimulating electrode, the D wave is well defined (Fig. 12C) or two relayed components may appear without an antecedent D discharge. The PT response to direct stimulation of the motor cortex ( Fig. l2D) consists of a direct ( D ) wave (with a latency of 0.5-0.8 msec, at the level of the medulla), ascribed to discharge of the Betz cells and a series of later, indirect deflections ( I waves), at 1.5-3 msec, resulting from impulses delayed by intercalated synapses of interneurons (Patton and Amassian, 1954; Gorman, 1966).

ASCENDISG REGCTL.4TING CEREBRAL SYSTEMS

97

Ill. Unspecific a n d Specific Influences on Thalamocortical Complexes

A. VISUALRELAYS 1. Diguse Regulutwn a. Shock-Euoked Responses. Reticulocortical activation is associated with an increase in excitability of the whole thalaniocortical visual complex, as shown by an increase in both pre- and postsynaptic cortical components evoked by stimuli applied to the LG or the optic chiasma. Although the surface-negative deflection 5 is specially enhanced during RF-induced arousal, thus suggesting a prevalent reticular influence on the superficial dendritic n e t ~ o r k , ~ the increase of the slow-positivc (axosomatic) wave 4 can be obtained even following elimimation of the negative wave by topical application of pentobarbital. These were the conclusions drawn by Bremer and Stoupel (1959) and Dumont and Dell (1960) from their pioneer experiments. Reticular facilitation of synaptic transmission at the level of the LG, evidenced by elective incrcasc of OT-evoked postsynaptic ( r , and r L ) geniculate components during RF stimulation (Fig. 1A) , was fully confirmed in a subsequent analysis of alterations induced by R F stimulation on OT-elicitcd unitary discharges in the geniculocortical radiation (Suzuki and Taira, 1961), and in studies concerning the effects of natural arousal or alertness in the behaving cat (Walsh and Cordeau, 1965; Dagnino et al., 1965a) and monkey (Doty et al., 1964). Facilitation of geniculate responses was also obtained by conditioning stimulation of the posterior hypothalamus, even after extensij e destruction of the upper RF (Chi and Flynn, 1968). Reticular influences on the visual thalamic relay may be mediated by fibers arising from the RF or intralaminar thalamic nuclei (Nauta and Whitlock, 1954; Scheibel and Scheibel, 1958). The presence of axo-axonic synapses of nonvisual source on optic fibers entering the LG was shown by Szentagothai (1962). It may be the morphological basis of the RF-induced depolarization of the terminals of optic fibers within the LG, i.e. a presynaptic inhibitory mechanism, evidenced in the cat (Angel et al., 1965) An alternative or joint explanation, taking into account the contribution of IPSPs to the surface-negative \vH\’~ 5 , might suggest that an increase in these IPSPs would be reflected in an enhancement of the wave 5.

98

M. STERIADE

and monkey ( Pecci-Saavedra et al., 1966). Enhancement of optic radiation response may result from dominance of a postsynaptic excitatory influence from RF. Similar response changes (increase in the postsynaptic LG response, and increase in the OT antidromic response to LG stimulation suggesting a presynaptic inhibition of OT terminals ) were observed simultaneously with spontaneous pontogeniculocortical spikes that occur during rapid eye movements of the PS (Bizzi, 1965; Iwama and Sasakura, 1965). Facilitation induced by stimulation of the rostra1 R F can be exerted at the cortical level independently of the LG events, as shown by an increase in deflections 3 to 5 evoked by a shock to the optic radiation above the thalamic relay (Cavaggioni and Goldstein, 1965; see also Fig. 1 B ) and by enhancement of responses evoked by shocks to the cortical surface, deep cortical layers, or white matter just below the visual cortex (Bremer and Stoupel, 1959; our own observations, Fig. 1C). Potentiation of cortical postsynaptic components occurs only in the first 4-10 sec of R F stimulation and is replaced beyond this interval by diminution of testing responses below the control values (Courtois and Cordeau, 1969; Fig. 2).4 The short-term facilitatory effects of R F stimulation may explain why enhancement of cortical responses to an optic radiation shock occurs during transition from SWS to arousal in chronically implanted animals (Cordeau et d.,1965; Walsh and Cordesu, 1965), in agreement with Bremcr's and Dell's data obtained in the ence'plzale isole' preparation. On the other hand, the samc data account for decrease of visual cortical postsynaptic responses during the steady state of natural waking (Cordeau et al., 1965; Walsh and Cordeau, 1965; see also for the somesthetic cortex: Allison and Goff, 1968), in agreement with previous investigations (Long, 1959; Evarts et a!., 1960), and with other studies showing the decrease of cortical responses to optic radiation, transcallosal or direct cortical stimulation during natural arousal (Okuma and Fujimori, 1963; Palestini et al., 1964; Dagnino et nl., 1965a; Baldissera et al., 1966). The decrease of the ' The reason for striking differences from one preparation to another (concerning the time course of the early 0s. subsequent effects) could not be entirely explained. In some of our experiments, long-lasting enhancement ( u p to 2 0 4 0 sec) was observed in the visual cortex, without detection of possible causes inducing this exception to the rule, which essentially could be confimied.

99

ASCENDING REGULATIKG CEREBRAL SYSTEMS

cortical postsynaptic biphasic sequcmce ( components 4 and 5 ) during the tonic state of waking is not to be explained by an occlusive, but presumably by an activc inhibitory phenomenon, since reduction of cortical response to an optic radiation shock during wakefulness compared with SWS is not associated with increase, but, contrariwise, with reduction of thr global frequencies of spon-.\/--

%

-

u - 7 -*,

1

CONTROL P€RIOD

180

-

P

I6Ol 140-

I

u & 2

:2

-‘*r

120

-~7 ) -

--b.----

POlENlIALI EVOKED BY SllMULAllON OF OPTIC RADIATION AROUSAL B Y SUSIAINED RETICULAR FORMATION I l l M U L A l l O N

;

100

g u+J

2

:

TIME IN SECONDS

80-

5 0

40

20

~

-

0 -

FIG. 2. Changes in cortical responsiveness during transition from sleep to wakefulness following reticular stimulation in a drowsy encbphale isold cat. The curve follows the amplitude of wave 4 of the visual cortex response evoked by a shock applied to the optic radiation. The line drawn as abscissa at the 100%level represents the average amplitntle of 25 potentials recorded during the control period of observation ( 50 sec ). Samples of individual potentials are shown in the top right-hand corner, the first row showing amplitudes during the control period; R F stimulation is applied at arrow. Note the phasic enhancement of cortical potentials (during die first 10 sec of RF stimulation), followed by a decrease in cortical responsiveness corresponding to the maintained state of wakefulness (from Courtois and Cordeau, 1969).

taneous unitary discharges (Evarts, 1960; see also Hubel, 1959; Evarts, 1962). A more complex situation was found at the level of the motor area where alterations of spontaneous unitary discharges during waking and sleep are in close relation with the neuronal size ( Evarts, 1965; vide infm ) , Ascending inhibitory mcdianisms balance facilitatory ones. Chemical or high-rate electrical stimulation of synchronizing struc-

100

M. STEFUADE

tures lying in the lower brainstem (Courville et al., 1962) and in the ventral part of the pons (Demetrescu and Demetrescu, 196213) was found to induce diminution of shock-evoked intrinsic cortical waves, while potentiation by release from inhibition could be achieved following inactivation ( Cordeau, 1962) or transections in front of these zones (Demetrescu et al., 1965). Depression of visual thalamocortical responses was also obtained during high-rate stimulation of the caudate nucleus in cerveau i.soZd or RF lesioned animals (Demetrescu and Demetrescu, 1962a) to avoid a downward activation of the activating mesencephalic RF (Koizumi et al., 1958). A “disinhibited preparation,” with brain stem ( midpontine or rostropontine ) transection and large bilateral destruction of the caudate nucleus, exhibiting striking increase of geniculostriate responses and susceptibility to protracted epileptic afterdischarges, was described by Demetrescu et al. (1965). Inhibition of the OT-evoked response in the visual cortex was also induced by stimulating the hippocampus (Redding, 1967). This effect might be mediated to the cingulum through the circuit of Papez (1937) and then by fibers running to the visual area, as described in the monkey by Yakovlev and Locke ( 1961). Whereas interruptions of various inhibitory ascending systems result in additive enhancement of neocortical responses, lowering of high responsiveness in associational or primary cortices can be induced by adding the destruction of a facilitatory (posterior hypothalamus) structure after that of an inhibitory one ( anterior hypothalamus) ( Kreindler and Steriade, 1966). Recovery of responsiveness can be achieved by lesion of an inhibitory structure ( caudate nucleus ) after elimination of tonic facilitatory influences arising from the posterior hypothalamus ( Steriade, 196710). Thus, incessant corticipetal unspecific activating impulses and tonic ascending unspecific inhibition are mutually counteracting. This hypothesis was further supported by analysis of the recovery cycle of cortical responses to LG or optic radiation stimulation in different phases of sleep and wakefulness. Enhancement of both kinds ( facilitatory and inhibitory) of diffuse influences during arousal was suggested by increase of the first, and decrease of the shortly delayed (7-15 msec), potential from a pair of geniculocortical responses ( Demetrescu et al., 19S6). These data, resulting from analysis of cortical responses to LG stimulation, might be partially ascribed ( as concerning the enhancement of facilitatory

ASCENDING REGULAHNG CEREBRAL SYSTEMS

101

influences during waking) to an increased responsiveness of the stimulated thalamic relay, considering the increase in amplitude of the first (presynaptic) deflection. On the other hand, reduction of the second response from a pair of cortical potentials during waking (compared with SWS) was not regarded as classic refractoriness, since great subnormality of the second response occurs during waking even at long ( 100 msec or even 200 msec) delays, whereas during sleep the test response shows complete recovery at 100 msec (Evarts et al., 1960). Besides, refractoriness has been believed to be due to active inhibition, considering the appearance of “supramaximal” shortly delayed responses after damage of inhibitory structures (Demetrescu et nl., 1965). The striking difference between wakefulness and PS responsiveness (although single responses, or first from a pair, were found by all investigators to be similarly enhanced ) consists in enhancement, during PS, of the shortly delayed response following a first maximal response (Demetrescu et al., 1966). It was proposed that a progressive reduction of inhibitory influences occurs as sleep deepens, simultaneous with a sudden decrease of facilitatory influences during the onset of sleep with EEG spindles (accounting for the decreased responsiveness in this phase) and an increased amount of facilitation anew toward the PS. This tentative representation of competitive influences throughout the sleep-wakefulness continuum concerned the responsiveness at cortical level. Since the testing shock was applied in these experiments to the thalamic relay, some of above alterations might be due to fluctuations in excitability of the thalamic synapses at the stimulated point. Thus, diminution of the first (“radiation”) spike during onset of sleep could be regarded as an inhibition exerted by diffusely projecting thalamic nuclei at the LG level, because similar changes were observed during light sleep with EEG spindles at the VPL (Favale et al., 1965) and VL (Steriade et d., 196913) (see Fig. 12), and well-defined IPSPs were induced in these specific nuclei by low-rate stimulation of midline thalamus or during spontaneous 6-12/sec spindles (Purpura et al., 1965, 1966; Maekawa and Purpura, 1967; see Section 111, C and D ) . Besides, by using wave 1 from chiasma stimulation and wave 4 from radiation stimulation, Walsh and Cordeau (1965) have reported that thalamic excitability declines whereas cortical excitability rises when changing from arousal to a relaxed state. The enhancement of shock-evoked responses during PS in compari-

102

hl. STERIADE

son with SWS and even with arousal (Benoit, 1964; Demetrescu et al., 1966) may be attributed to striking increase of the LG responsiveness during PS, while intrinsic cortical responsiveness remains at approximately SWS lcvel (Palestini et al., 1964; Dagnino et al., 1965a) or even declines slightly (Walsh and Cordeau, 1965). Further study (especially investigations at the unitary level) is required to complete the picture of responsiveness at different (thalamic and cortical) relays and to ascertain their participation in various processes of sleep-wakefulness cycle. Anyhow, up-to-date studies have suggested that diffuse regulation of visual thalamic and cortical responsiveness results from ascending active inhibition paralleling facilitatory influences, and these studies have attempted to assess the respective contribution of different mechanisms in various stages of awareness. b. Photically Evoked Responses. The reduction of flash-evoked responses in the visual cortex during some processes (“attention” or “distraction”) due to reticular activation was ascribed by Hern6ndez-Pe6n and his colleagues (1956a, 1957) to a subcortical blockade, considering the simultaneous decrease of the LG response. Subsequent studies failed to confirm this observation. Thalamic blockade was not regarded as determinant for the diminution of cortical responses (Bremer and Stoupel, 1959; Bremer et al., 1960). Moreover, an increase of the LG photically elicited responses was observed during reticular arousal, simultaneous with depression of cortical potentials evoked by single or low-rate flashes (Steriade and Demetrescu, 1960). Naquet et al. (1960) and Fernhndez-Guardiola et al. ( 1964) claimed that pupillary dilatation accounts for the increase of the LG potentials during arousal, since no alterations of the LG photic responses were found in their experiments after administration of atropine. Nevertheless, the RFinduced facilitation of synaptic transmission at the LG level was repeatedly confirmed with microelectrode recordings in atropinized animals, both during phasic (Taira and Okuda, 1962; Ogawa, 1963) and tonic (Maffei et al., 1965; Maffei and Rizzolatti, 1965) states of wakefulness. It explains the decrease in latency of the early flash-evoked deflection in the visual cortex during arousal in the cat ( Steinberg, 1965) and its increase during sleep in man (Ebe and Mikami, 1962). The RF-induced “occlusion” of the early biphasic deflection evoked in the visual cortex by a single flash was regarded as a “masking phenomenon,” depending on the nature of the testing

ASCENDISC REGULA1‘1N(; CEREBRAL SYSTEMS

103

stimulus, i t . , photically elicited impulses are not synchronous enough (as are those induced by a shock to the central pathway) to overcome the relative refractoriness of cortical neurons during highrate R F stimulation (Bremer, 1961; see also the discussion in Gauthier et ul., 1956). This hypothesis is supported by experiments of Steriadc and Denietrescu ( 1967a) showing that short (15-30

20 msec FIG. 3. Effects of reticular stiiniilation on cortical responses to trains of pulses applied to the lateral genicrilate body. Encdphale isoE cat. Note: appearance of late ( off) response at 200/sec; suppression, during reticular arousal, of late response to 2OO/sr~c pulse-train, contrmting with enhancement of first geniculocortical response and of 20 insec delayed geniculocortical responses (at 50/sec). See also text ( fro111 Steriade and Demetrescu, 1967a).

msec) trains of electrical pulscs at 200-1000/sec. applied to the OT or LG simulate the naturally desynchronized photically elicited impulses, inducing a “multiple” cortical response which consists of an early potential evoked by the first pulse in the train, identical with the shock-evoked response, and ;I lute biphasic wave at the end of the train, strikingly rc~semblirig the peripherally elicited response. Enhancement of the. early potential and suppression of the late response during reticular stitnulation ( Fig. 3) stress that differences in pattern and brhavior of these components depend on

104

M. STERIADE

FIG. 4. Reticular and steady light potentiation of flash-evoked cortical afterdischarge. Four different experiments ( enckplzule koU cats). A, B, and D: recordings from the surface of the visual cortex. C: simultaneous recordings from the visual cortex (VC) and lateral genicuhte body ( LG). Contralateral testing flashes (10 iiisec duration) in darkness and during steady light applied to the same retina as stimulated by the flash, Left column: test responses, Middle column: effects of conditioning short trains of fast pulses (400/sec in A, 350/sec in B-D) to the inesencephalic R F (see arrow for location in a typical experinient); in D, the pulse-train was applied to the R F after full development of the early biphasic sequence. Right column: effects of the pulse-train alone (lack of response in A; varialde and inconsistent late response

ASCEXDING REGULATING CEREBRAL SYSTEMS

105

synchronous zjerms temporally tlispersecl ( flashlike) impulses inducing each of them. The decrease of a single flash-evoked cortical response during reticular arousal, changing into facilitation of rhythmic responses to photic stimuli above 5/scc (Steriade and Demetrescu, 1960), was also ascribed to the greater synchrony of afferent volleys resulting from repetitive stimulation ( Rremer, 1961). Unspecific facilitation of cortical unit discharges evoked by fast photic stimuli (rise in critical fusion freqwncy ) was induced by stimulating the midline thalamic nuclei ( Creutzfeldt and Grusser, 1959). The diffuse thalamic ascending facilitation of cortical responsiveness does not seem to be dependent on downward reticular activation (cf. Schlag et nl., 1961), since it persists in high cerveau 6 0 l L preparations ( Steriade and Demetrescu, 1960). Reticular potentiation of cortical responses to flashes above a certain rate was confirmed in subsequent experiments on animals ( Bremer, 1961; Kaneko et n?., 1961; Narikashvili, 1963). Enhancement of rhythmic photic responses above 51 sec was also observed during auditory stimulation or arithmetic computation in inan ( Steriade et d., 1961). Reticular facilitation may occur also for intrinsic cortical responses to single flashes, providing that synchronous stimuli are used. Apparent contradictions between the results of Akimoto et al. (1961) and those of Creutzfeldt ct crl. (1961), i.e. enhancement and decrease of photically evoked responses during reticular arousal, were explained by differences in parameters of photic stimuli ( Creutzfeldt, 1961) . With well-synchronized (short duration and very bright) photic stimuli, an elective potentiation of intracortically elaborated postprimary components or fast afterdischarge evoked by a singlc flash was induced by conditioning train of pulses applied to the mesencephalic RF, simultaneous with decrease of the early biphasic deflection (Fig. 4) (Steriade, 1967a; Steriade et al., 1968). Reduction of the initial wave is not due to a depression of the corticipetal volley, since the LG response is not __

in B ) . Time: 20 msec. Vertical Imn: 0.3 mV. Note: different patterns of potentiation induced by steady light and RF conditioning stimulation on the cortical b postprimary complex; additice effect of light and reticular stimulation (A, B ) ; lack of correlation between RF-induced enhancement of cortical fast afterdischarge and LG oscillatory w'ivei ( C ) ; reticular facilitation of the 2, complex independent of the early wave ( D ) (froni Steriade et al., 1968).

106

M. STERIADE

depressed (Fig. 4C), but very likely represents a cortical “occlusion.” On the other hand, potentiation of subsequent components has no causal relation with LG events (Fig. 4C) and is independent of reduction of the first one, as shown by increase of postprimary waves by applying the trains of RF pulses after full development of the early deflection (Fig. 4D). It may be supposed that differences in synchronization between impulses generating the initial and subsequent components account for opposite behavior under R F stimulation. Facilitation of unit discharges evoked by single flashes in visual cortex neurons could also be obtained during EEG arousal induced by protractecl reticular stimulation or even by a single shock applied to the RF 50-100 msec before the testing photic stimulus (experiments of Steriade and Ionescu; see Fig. 81 in Steriade’s monograph, 1969a ) . Reticular potentiation of flashevoked postprimary components in cat can be related to previous similar findings, namely: enhancement of a “secondary” photic response in the rabbit visual cortex during RF stimulation or following administration of amphetamine (Fuster and Docter, 1962), increase in unit discharges evoked by flashes at 80-100 msec latency during natural waking in cat ( Evarts, 1963), augmentation of fast sensory afterdischarge during alertness in monkey (Hughes, 1964), and decrease in latency of late responses during wakefulness in man ( Gastaut et aZ., 1963). Some effects induced by stimulation of the superior colliculus and potentiating selectively the cortical “secondary” waves evoked by flashes (Brown and Marco, 1966-1967) might be due to activation of the RF. This is suggested by the reticular-like facilitation of the shock-evoked response obtained by collicular stimulation (Bremer, 1966; Brown and Marco, 1967) and the fact that a cortical negative shift was associated with these effects (Brown and Marco, 1966-1967), very similar to that induced by stimulation of the RF or diffuse thalamic system (Bonnet, 1957; Brookhart et ul., 1958). 2. Specific Ascending Control Background retinal illumination potentiates the cortical responses evoked by shocks to the central pathway (effect of Chang, 1952) and intracortically generated components evoked by photic stimuli (Steriade and Demetrescu, 1966; Steriade and Ionescu, 1967). This process is mediated by the specific pathway, and it is

ASCENDING REGULATING CEREBRAL SYSTEMS

107

not due to unspecific activation through optic collateralization toward reticular structures; since the phenomenon was discovered under deep barbiturate anesthesia ( Chang, 1952) which blocks the RF, the process also persists in the high cerveau is016 preparation (Steriade and Demetrescu, 1966) and it was not seen for shock-evoked responses outside thc visual cortex ( Arduini and Goldstein, 1961; Steriade, 1964; Cavaggioni and Goldstein, 1965). The photic enhancement of geniculocortical responses was attributed by Chang (1952) to an increase in the LG excitability, since potentiation of the optic radiation response to an LG shock was still obtained following cortical ablation. However, the cortical level of photic potentiation i n d r p n d e n t of geniculate events was suggested by the prevalent increasc of the postsynaptic components of cortical response evoked by a shock to the chiasma (Dumont and Dell, 1960), and is supported by the experiments of Schoolman and Evarts (1959) and of Cavaggioni and Goldstein (1965) showing a selective enhancement of intrinsic cortical responses to an optic radiation shock ( t o avoid a concomitant effect at the thalamic relay), and by the potentiation of striate and parastriate responses to transcallosal ( Ajmone-Marsan and Morillo, 1963) or direct stimulation of the visual cortex during steady light ( Steriade, 1964, 1968).5 The results concerning LG responsiveness are rather difficult to reconcile, because most authors have observed an increase in amplitude of the first (“radiation”) spike of the shock-evoked response in the visual cortex under steady light, i.e., an enhancement of geniculate excitability, while Cavaggioni and Goldstein (1965) reported an inhibition at the LG (decrease of the first spike), simultaneous with facilitation at the cortical level. It was thought to represent a means of preventing cortical saturation. These puzzling results, together with the enhancement, during steady retinal illumination, of cortical postprimary complex in spite of a global decrease of the LC, response to a flash of light (Steriade and Demetrescu, 1966; Steriade and Ionescu, 1967; Ionescu, 1969), indicate that the potentiation induced by continuous illumination at cortical level is so powerful that it can be seen even with a reduction of the LG output. Besides the facilitatory summation of the presynaptic impulses Phasic (transient) enhancemerit can be induced by a conditioning flash, while tonic potentiation was obtained in the original study of Chang by continuous illumination.

108

M. STERIADE

of photic origin and the shock applied directly to the central neurons ( Chang, 1952), potentiation induced by steady illumination was also ascribed to a release from inhibition exerted at central ( geniculate and/ or cortical) levels by spontaneously discharging retinal elements in darkness (Arduini and Hirao, 1960). Both mechanisms (facilitation and release from inhibition) may coexist, considering the two retinofugal populations of neurons, altered in opposite direction by illumination (Granit, 1955; Kuffler et al., 1957; Arduini and Cavaggioni, 1960). The prevalence of one of these two mechanisms may depend on experimental conditions; thus, release from the inhibitory dark-discharge seems untenable in deep barbiturate anesthesia when retinal spontaneous activity is strikingly reduced (Kuffler et al., 1957). On the other hand, release from inhibition is supported by enhancement of shock-evoked responses in the visual cortex observed during reversible retinal inactivation ( Arduini and Hirao, 1960), following interruption of the visual pathway (Posternack et al., 1959) or after bilateral enucleation of the eyes (Sakakura and Doty, 1969). The effects of the OT transection on LG-evoked single or paired cortical responses paradoxically resembled those exerted by continuous retinal illumination (usually the enhancement was even larger than that induced by steady light; see Fig. 6B) and were in contrast with the unspecific RF-elicited influences ( Steriade and Demetrescu, 196%; Fig. 6 ) . These data suggest that, besides the unspecific (reticular) control, the responsiveness of the LG and visual cortex depends on tonic facilitatory and inhibitory influences ascending from the specific pathway. The enhancement of the flash-evoked cortical postprimary waves by application of background illumination, simultaneous to the decrease of the flash-evoked OT and LG responses, could be explained (besides the alternative suggestion of a very powerful facilitation at cortical synapses, vide supra) by the reduction of retinal phasic output resulting from this experimental condition. Actually, striking potentiation of flash-evoked unit discharges in the visual cortex neurons (extracellular recordings) could be obtained in both conditions of (1) background illumination, and ( 2 ) decreasing the intensity of the testing flash in darkness from maximal to middle (optimal) or lower intensity range (Fig. 5 ) . It suggested that “inhibition exerted by strong flashes in darkness (direct geniculocortical inhibitory projections or intracortical re-

ASCENDING REGULATIZG CEHERR.IL SYSTEMS

109

current inhibitory pathway) in;iy be reduced during steady light” (Steriade, 1968, p. 202). The above data and suggestion are in complete agreement with recent findings of Creutzfeldt et al. (1969) showing (by means of intracellular recordings) that the secondary excitation evoked by flashes in primary visual area was stronger with background illumination than without and also “suggesting a stronger primary inhibition” ( p. 133) when flashes were applied without background illumination.

FIG.5. EEects of decreasing the flash intensity and of background illuniination on nerironal discharges evoked by flashes in the visual cortex. EncLphak isold cat. Time: 20 msec. 1, 2, and 3 are flashes (arrows) of different intensities (0.5, 0.25 and 0.1% joules, respectively). Note maximal discharge at middle intensity in dark, and striking enhancement of neuronal discharges during steady light at maximal antl middle intensities (from Steriade, 1968).

The effects exerted by stimulation of associative thalamic nuclei on visual cortex responses (Morillo, 1961; Battersby and Oestherreich, 1963; Brown and Marco, 1967) were believed to be due to specific activity, directly projected from pulvinar and lateral posterior nuclei to the suprasylvian gyrus and then to the visual cortex, rather than through projections by diffuse multisynaptic pathways.

110

M. STERIADE

FIG.6. Specific potentiation and its interaction with unspecific effects on the excitability cycle of visual thalamocortical responses. TWOexperiments ( A and B ) on enchphale isoU cats. In A, a pulse at 50 msec was added to a pair of pulses separated by 10 msec, resulting in a group of three shocks delivered to the LG, in order to obtain simultaneous information concerning both short (10 msec) and 50 msec-delayed visual cortical responses. Note: increase of short-delayed responses during steady light (right column) and of 50 msec-delayed responses during R F high-rate stimulation, in spite of the

ASCENDING R E G U L A l I S C CEREBRAL SYSTEMS

111

3. Interactions between Specific. mid Unspecific Influences

The effects of both reticular and light potentiation may result in very similar enhancement of the shock-evoked cortically generated waves (Schoolman and Evarts, 1959; Cavaggioni and Goldstein, 1965) and may explain that both effects can be additive, such as seen by unit discharges (Fuster, 1961) and by intrinsic cortical components evoked by electrical shocks (Dumont and Dell, 1960) or photic stimuli (Steriade, 1967a) (see also Fig. 4A,B). However, different or even opposite effects of unspecific and specific stimulations are shown by: ( i ) dissimilar pattern of effects induced by steady light and RF conditioning stimuli on cortical responses to flashes; although both kinds of influences cause a reduction of the early deflection and an enhancement of the afterdischarge, light induces mainly a latency decrease of the postprimary ( b ) complex, while RF stimulation especially amplifies it (Steriade et al., 1968) (Fig. 4A,R); ( i i ) opposite alterations induced by specific us. unspecific stimulation on the second cortical response evoked by paired LG stimulation; at short delays (7-15 msec ) , steady light or OT high-frequency stimulation enhance the second response, while RF stimulation supresses this specific potentiation achieved by enlargement of infraliminally excited fringe of neurons; at longer delays, specific facilitation is no longer seen on the second response and, moreover, steady light impairs the enhancement induced at these delays by reticular stimulation (Fig. 6A). On a background of high LG and cortical responsiveness following OT transection, R F stimulation is still effective in decreasing the second, short-delayed response ( Fig. 6B), thus suggesting that unspecific ascending inhibition is not hindered by specific phenomena occurring in the visual pathway (Steriade and Demetrescu, 196713). equal enhancement of the first response from the pair; when specific and unspecific stimulations were simultaneously given, steady light impaired ( especially the intracortically generated negative wave ) the potentiation induced by R F stimulation. See also text. Note in B: different types of effects induced by RF and steady light on the second, short-delayed response; optic tract transection enhanced ipsilateral LG-induced responses in the visual cortex in a similar direction as steady light; R F btiinulation was still effective to depress the short-delayed response on the hackground of disinhibition induced by interruption of the specific pathway. ( Modified from Steriade and Demetrescu, 1967b.)

112

hf. STERIADE

FIG. 7. Reticular influences on auditory cortical responsiveness. Three different experiments (A, B, and C ) on eticdpphale isole‘ cats. Shocks applied every 2 sec to the white matter, just below the primary auditory cortex (anterosuperior ectosylvian area). In A and C, siinultaneous recordings of visual cortex responses evoked by shocks to the white matter. A: a series of evoked responses; note much greater “spontaneous” Huctwations in the visual than in the auditory cortex. B: test responses (left column) and effects exerted by a conditioning pulse-train (300/sec; 300 msec in duration) applied to the mesencephalic RF (right column); testing shock followed a t 20 msec the end

ASCENDING REGULATING CEREBRAL SYSTEhIS

113

R. AUDITORYRELAYS

Facilitation of thalamic transmission and inhibition of fast intracortical shock-evoked responses during wakefulness were reported in chronically implanted cats by Dagnino et al. (1965b). Similar findings, i.e., inhibition of primary cortical responses evoked by h o c k s to the auditory radiation, simultaneous with facilitation of synaptic transmission in thc, thalamic relay, were recently reported in the ence‘phale isole‘ monkey (Symmes and Anderson, 1967). However, clear-cut signs of a short-term facilitatory effect exerted by reticular arousal on auditory cortical responsiveness may be detected if the testing shock follows at convenient delays the RF conditioning stimulation. Thus, reticular potentiation of responses in anterosuperior ectosylvian gyrus evoked by shocks to the auditory radiation (just below the deep cortical layers) is expressed at delays between 50 and 300-400 msec and only for the first response in a series (in contradistinction with much greater duration of the early, facilitatory phase in the visual cortex6), being thereafter followed by obvious depression (Fig. 7B,C) ( Steriade and Cordeau, unpublished data). Our findings suggest that RF-induced decrease of cortical auditory responses reported by Symmes and Anderson (1967) was likely obtained since the testing shock was delayed by an interval exceeding 900 msec (from the beginning of the RF stimulation) and since evoked responses were averaged, thus preventing the possibility to detect the early phase of reticular potentiation. The RF-induced potentiation of

’Visual cortex responses are also much more influenced by “spontaneous” arousal than are those simultaneously recorded in the auditory or other cortical areas (Fig. 7 A ) . It could explain why reticular potentiation is most easily obtained at the level of the visiial area, a fact which was emphasized by Bremer and Stoupel (1959). of the pnlse-train; note the enhancement of the first response in the series (bottom) and progressive depression (helow the control values, see response a t the top) of the cortical postsynaptic component (arrow). C: test responses in visual and auditory cortices (left column) and the effects exerted by a conditioning pulse-train ( 300/sec; 700 msec in duration) applied to the RF; the testing shock was applied 100 msec after the ortset of the conditioning RF stimulation; three responses are depicted from the bottom to the top: lst, 4th, and 7th of the series; note (like in B ) the enhancement of the first and progressive depression of the postsynaptic component, marked by arrow. Time (horizontal bar) : 5 insec ( Steriade and Cordeau, unpublished observations ).

114

M. STERIADE

cortical unit discharges evoked in the auditory area was also observed by Akimoto et al. ( 1961). Previous divergent results showing the increase (Bremer and Stoupel, 1959) or the decrease (Desmedt and La Grutta, 1957) of the MG-evoked primary cortical response during RF stimulation may be explained (besides prevalent pre- or postsynaptic thalaniic testing stimulation, implying different alterations at the stimulated point ) by recording during the early or the subsequent phase of the RF effects. The modifications undergone in different behavioral conditions by click-evoked potentials at central levels are to be studied with stabilized acoustic input, taking into account the orientation of the pinna and the activation of the middle ear muscles during reticular, sensory, or spontaneous arousal reaction and motor activity (Galambos and Rupert, 1959; Hugelin et al., 1960; Starr, 1964). Invariability of the acoustic input can be secured by administration of curarizing agents in acute preparations (Hugelin et al., 1960) or by using earphones to deliver the clicks and by cutting the intraaural muscles in behaving animals (Worden et al., 1964; Baust et aZ., 1964 ) . Under these experimental conditions, the microphonic potential recorded at the round window and click-evoked responses of the dorsal cochlear nucleus are not reduced during R F stimulation and throughout the sleep-wakefulness cycle (Hugelin et al., 1960; Baust et al., 1964). In addition, click-evoked responses recorded from superior olive and inferior colliculus were found to be highly stable during the sleep-waking cycle ( Wickelgren, 1968). Data of other authors showing the attenuation of cochlear responses by “attention” or RF stimulation (Hernhndez-Pe6n et al., 1956b) and during rapid eye movements of the PS (Jouvet, 1962) may be ascribed to middle ear muscles contraction. Similarly, the reduction in amplitude of the MG responses to clicks during RF stimulation was no longer obtained after cutting the tendons of intraaural muscles or following administration of curarizing drugs and may be, thus, regarded as a peripheral phenomenon (Hugelin et al., 1960). Some signs of facilitation were found at the MG relay in cats under flaxedil or with bilateral tenotomy of the middle ear muscles during RF stimulation (Steriade and Demetrescu, 1962) or natural waking, possibly resulting from a considerable increase in background activity (Berlucchi et al., 1967). Differences between large and constant enhancement of shock-evoked potentials, and slight, inconsistent increase of peripherally elicited MG responses

ASCENDING REGULATING CEREBRAL SYSTEMS

115

following RF stimulation in ence‘phale isole‘ monkeys were assigned to differences in synchronization of the testing afferent volley (Symmes and Anderson, 1967), such as the original explanation of Bremer (1961). All the data, with one exception (Gellhorn et al., 1954), showed that cortical responses to single clicks are reduced under the waking action of ascending activating systems (Bremer and Bonnet, 1950; Desmedt and La Grutta, 1957). This diminution during arousal, in comparison with SWS, was confirmed more recently in chronically implanted cats, independently of middle ear muscle contraction (see Fig. 5, threshold stimulation, in Chin et al., 1965; and Fig. 4 in Berlucchi et al., 1967). On the contrary, cortical responses to fast rhythmic clicks are increased during RF stimulation (Steriade and Demetrescu, 1962). The following of very fast sensory stimuli during R F stimulation is probably due to the decrease in the refractory period of cortical evoked responses, as shown by alterations of the recovery cycle during arousal (Schwartz and Shagass, 1962; Steriade and Demetrescu, 1962). Cortical facilitation was believed to depend on the increased MG responsiveness (Symmes and Anderson, 1967). However, if only the altered input to the cortex is considered, it is difficult to cxplain why the auditory cortex does not react as a whole during transition from sleep to wakefulness. Recordings from several points in the ectosylvian gyms revealed a complex cortical reorganization during reticular arousal, with increased responses in some points simultaneous with depression in neighboring areas; it was designed as a “commutation phenomenon” ( Steriade and Demetrescu, 1962, 1964) and, subsequently, it was found also in the cat (Steriade and Demetrescu, 1964) and monkey (Hughes, 1964) visual cortex. During PS, the reduction of click-evoked responses in the auditory cortex of animals with intact intra-aural muscles (Winters, 1964; Chin et al., 1965) was found to affect, in tenotomized animals, only the late negative component,‘ while the initial negative deflection was tonically increased (Berlucchi et al., 1967). I t was not dependent on the increase of thalamic transmission at this stage of sleep ( Dagnino et al., 1965b ), since the increased responsiveness of the auditory area I was seen to occur in absence of changes at the geniculate level (Berlucchi et a)., 1967). ‘This fact might be correlated with complete suppression of auditory potentials recorded from the brainstern R F during PS (Huttenlocher, 1960).

116

M. STERIADE

As far as the modulations exerted by the cerebellar auditory area on click-evoked responses in the ectosylvian gyrus are concerned, the failure of the MG to respond to cerebellar stimulation (Steriade and Stoupel, 1960; Infantellina et al., 1966) and disappearance of responses evoked by cerebellar stimulation in secondary auditory cortex following electrolytic lesion of the mesencephalic R F showed that, at least for this auditory area, cerebellofugal influences are exerted through unspecific ascending systems of the reticular core. A tentative hypothesis concerning the cerebellar projections to the auditory area I and suggesting an ascending pathway including the inferior colliculus ( Steriade and Stoupel, 1960) was supported by experimental evidences of Teramoto and Snider ( 19f36). C. SOME~THETIC RELAYS Since some investigators attempted to study the thalamic and cortical responsiveness during arousal and sleep by means of cutaneous testing stimuli, first of all it is necessary to give a brief account on the events at the first synapse. Responses evoked in the medial lemniscus by cutaneous stimulation are unchanged or even increased on light arousal from SWS, but could be depressed during strong arousal accompanied by clear-cut behavioral reactions (Favale et d.,1965). These results are in partial agreement with the data of Hernhdez-Pe6n et al. (1965) showing the RF-induced depression of the evoked potentials at the first synapses (gracilis and cuneatus nuclei),s and in full harmony with studies on other sensory systems (see Section I V ) emphasizing that an increased responsiveness may be replaced by a depressed state beyond a certain degree of alertness. During PS, a depression of the orthodromic lemniscal response was observed to occur only with large bursts of rapid eye movements (Carli et al., 1967a). This phasic depression is due to both presynaptic and postsynaptic events in the cuneate nucleus (Carli et al., 1967b), induced by vestibular nuclei through the roundabout way of the sensorimotor cortex ( Carli et al., 1 9 6 7 ~ )Similar . inhibitory phenomena were reported during PS at the first synapses of the trigeminal pathway, but they 'Reticular depression of the activity evoked in the cuneate nucleus was ascribed to presynaptic inhibition ( Cesa-Bianchi and Sotgiu, 1969), which is in agreement with anatomical data showing axoaxonic contacts in the ventral half of the nucleus (Walberg, 1965), where presynaptic depolarization was found.

ASCENDING RECXJLAIINC; CEREBRAL SYSTEMS

117

were believed to arise in thcx midbrain tegnientum (HernhdezPe6n et al., 1965). The afferent transmission at the second station (VPL) is facilitated on arousal from light sleep and becomes maximal during PS.

FIG. 8. Reticular influences on centrally evoked responses in the VPL nucleus and somesthetic cortex. Enckphale isole' preparations. Testing shock applied to the medial lemniscns ( A ) and underneath the primary somesthetic cortex ( B ) . Effects of high-rate (250/se c ) stimulation of the mesencephalic R F depicted in the right column. In R , a conditioning pulse-train ( 1 sec in duration, delivered each 2 sec) was applied to the R F just before the testing shock. Time: 2 msec (horizontal bar in B ) . Vertical bar in A: 0.3 mV. Note in A: selective reticular enhancement of the postsynaptic ( r ) VPL response; B: RF-induced enhancement of c~rticalpostsynaptic activity (arrow).

This picture is supported by analysis of the VPL postsynaptic response evoked by a shock to the medial lemniscus in various stages of natural sleep and wakefulness (Allison, 1965; Favale et al., 1965) or during reticular arousal (Fig. SA), and may be also inferred from the bchavior of the prosynaptic (no. 1 ) component

118

M. STERIADE

evoked in the somesthetic cortex by stimulating the VPL nucleus (see Fig. 10). Similar results were obtained in man, the component 1 of somatosensory responses (“radiation” potential) being usually smallest during SWS (Goff et al., 1966). The decrease of the VPL responsiveness during SWS may be partially explained in the light of data showing prominent IPSPs and EPSP-IPSP sequences in the VPL relay cells during spontaneously developing 6-12/ sec spindle waves in sensorimotor cortex (Maekawa and Purpura, 1967). The cortical responses evoked in the somesthetic area by peripheral stimulation was decreased, but VPL-elicited potentials were not changed during reticular or sensory arousal (Gauthier et al., 1956). Reticular enhancement of cortical responses evoked by shocks to the VPL (Bremer and Stoupel, 1959) might be attributed to facilitation at the thalamic relay, considering that responses evoked in the somesthetic area by shocks to the radiation or the callosal fibers were found to be depressed during natural waking (Favale et al., 1963, 1965). However, the responsiveness of the somesthetic cortex is considerably enhanced on natural arousal from sleep (early phase of waking), in comparison with the steady state of waking (Allison and Goff, 1968). This time-course of arousing effects might explain some contradictory results and might account for potentiation of cortical responses to stimulation of somesthetic radiation only with conditioning pulses to the RF at short delays (Fig. 8b) (Steriade and Cordeau, unpublished data) and for depression in a subsequent phase (steady state of waking?). Discrepancy in results of Allison (1965) and Favale et al. ( 1965) reporting, respectively, increase and decrease of cortical responsiveness from SWS to PS were ascribed in the subsequent study of Allison and Goff (1968) mainly to the difference in location of the stimulating electrode: the presynaptic (no. l ) and postsynaptic (no. 3-5) cortical waves evoked by sub-S, stimulation are smaller, while the first (afferent volley) and the slow surfacenegative component (no. 5 ) evoked by a shock near the VPL tend to be larger during PS compared with SWS. These findings suggested a causal relationship between alterations of the wave 1 and those of the subsequent, postsynaptic components, resulting either from changes in excitability of VPL neurons (increased responsiveness during PS) or from an increased depolarization of thalamocortical terminals and a consequent increase in presynaptic cortical inhibition during PS.

ASCENDING REGULATING CEREBRAL SYSTEMS

119

The increase during S W S arid PS of peripherally evoked primary responses in the cortical somesthetic area ( Albe-Fessard et al., 1964; Allison et al., 1966) may reflect the competitive alterations induced in these stages of sleep at the first synapse, VPL nucleus, and cortical relay. Unspec6c and associative responses to cutaneous stimulation are depressed or abolished during arousal and PS compared with light sleep. It is a common observation at the level of the centromedian nucleus ( Albe-Fessard et al., 1961), bulbar RF, VL nucleus, caudate, pallidum, putamen, suprasylvian and anterior marginal gyri (Guilbaud, 1968), and in the cortical midline region, late potentials resembling in this region the K-complex in man (Goff et al., 1965) . The depression of extra-lemniscal responses during wakefulness and PS was attributed to an occlusive phenomenon due to striking increase of unitary discharge in the RF (Huttenlocher, 1961) or to an active corticofugal control exerted by cortical areas and PT on nonprimary responses (Massion and Meulders, 1960; Ascher, 1965; Denney and Thompson, 1967), considering that depression of such responses during waking and PS may be prevented by sensorimotor decortication ( Guilbaud, 1968). D. MOTORSYSTEM 1. Difuse Regulation

If cortical responses to stimulation of the appropriate thalamic nuclei are tested, the responsiveness of the motor thalamocortical complex is apparently in contrast to that of all sensory neothalamocortical complexes. Indeed, an obvious dissociation has been revealed between diminution of the VL-evoked motor cortex response during reticular arousal ( Steriade, 1969b) and well-known enhancement of sensory thalamocortical potentials (Fig. 9). This differentiation between motor and sensory thalamocortical responses appeared also during natural arousal and sleep in the chronically implanted cat (Steriade et al., 1969b) : VPL-induced responses of the somesthetic area were increased during arousal and PS, while VL-induced motor cortex responses were decreased on arousal from drowsiness or light sleep, and were again reduced during PS (Fig. 10). In acute experiments, the depression of motor thalamocortical potentials was obtained by high-rate stimulation of dorsal and ventral RF points, from the mesencephalon to the medulla, in front

120

M. STERIADE

FIG. 9. Reticular influences on somestlietic and motor thalamocmtical responses. Encbpphale isole' cat. Somesthetic cortex responses evoked by single shocks to the VPL nucleus ( A ) and motor cortex responses to single (B,1) and lO/sec (B,2) VL stimuli. Time: 2 msec. Vertical bar: 0.5 mV. Note that enhancement of the VPL-somesthetic cortical response during high-rate (250/sec) R F stimulation contrasts with diminution of the VL motor cortex response and striking reduction of the second positive wave developing during augmenting responses.

ASCENDING REGULATING CEREBRAL SYSTEMS

121

FIG.10. Alterations of the VL-evoked motor cortex responses and VPLevoked somesthetic cortex responses during natural arousal ( 1 ), drowsiness with EEG spindles ( 2 ) , slow wave sleep ( 3 ) and paradoxical sleep (4)in chronically implanted cat. Time ( oscillographic records ) : 2 msec. Vertical bar: 0.3 mV. Note: increase of the presynaptic (component no. 1) cortical response to VPL stimulation dnring arousal compared with light sleep, and maximal enhancement during PS; opposite alterations between somesthetic and motor thalamocortical potentials in all the stages of sleep and wakefulness. See also text (from Steriade et al., 196%).

of the inferior olive. By exploring a more caudal area, at midolivary level, potentiation of the VL-motor cortex response was obtained by high-rate stimulation of a critical area located 2-2.5 mm below the dorsal zone (central gray) exerting a depressive influence, similar to that induced from the upper R F points (Fig. 11) . Simultaneous EEG recordings disclosed that a synchronizinglike reaction is obtained from this critical medullar area, while the

122

M. STI!XIADE

FIG.11. Opposite influences, arising from different areas of the lower brainstem, on motor cortex responsiveness. Cat under flaxedil. Left column: VLelicited responses at the surface and in the depth of the precruciate area. Right column: depressive and enhancing effects of high-rate (250/sec) stimulation at the top (1, decrease), middle (2, decrease), and bottom (3, increase) arrows. Time: 2 msec. Vertical bars: 0.25 mV (surface) and 0.2 mV (depth). (Modifled from Steriade, 1969b).

ASCENDING REGULAIING CEREBRAL SYSTEMS

123

classic desynchronization ( arousal ) reaction is associated with depression of the VL-motor cortex response by stimulating the rostra1 RF (Steriade, 1969b). Together with data resulting from studies on sensory (especially visual ) systems ( see Section III,A), these findings show the nonhomogeneous functional organization of the RF, consisting of critical areas in the lawer brainstem with ascending influences opposite in sign to those exerted by the remaining structures belonging to the so-called “activating” system. To check whether the depression of cortical response was not dependent on an influence exerted primarily at the stimulated thalamic nucleus, intrinsic cortical components, PT relayed waves, and VL responses evoked by shocks to the cerebellothalamic pathway had to be simultaneously recorded during RF-induced arousal or natural sleep and wakefulness. a. V L Responsiveness. Reticular stimulation not only suppresses the inhibitory effects of midline thalamic low-rate (7/sec) pulses on VL mass (Cohen et al., 1962) and intracellular (Purpura et al., 1965) responses evoked by shocks to the brachium conjunctivum, but also may increase VL focal potentials over control values ( Frigyesi and Purpura, 1964) . The RF-induced arousal electively facilitates the postsynaptic VL response to red nuclear stimulation, without significant alteration of the presynaptic spike (see Fig. 12A) (Steriade, 196913). Investigation of the VL responses to rubral stimuli in chronically implanted cats confirmed data resulting from acute ( ence‘phale isolk ) preparations, showing an elective increase of the postsynaptic wave on arousal from light sleep and a maximal enhancement during PS (Fig. 13) (Steriade et al., 196913). This picture is very similar to that obtained at the level of other (sensory) thalamic relays. In particular, there was a striking reduction, and sometimes a complete disappearance, of the postsynaptic VL response during drowsiness with EEG spindles (see Fig. 13,2). The decrease of the VL field potential during light sleep might be closely connected with great temporal dispersion of unitary discharges evoked in this nucleus by prethalamic stimulation in this stage of sleep, while neuronal discharges are evoked during wakefulness with short and stable latencies (Filion et al., 1969). Such findings may be correlated with IPSPs evoked in the VL nucleus by low-rate midline thalamic stimulation and the blockade of these inhibitory potentials

124

M. STERIADE

FIG.12. Alterations of centrally evoked responses in the ventrolateral (VL ) nucleus, motor cortex, and pyramidal tract ( a t the medullar level) during reticular arousal. Five different experiments ( A l , 2 and B: encbpphale isold preparations; C and D: cats under flaxedil). Testing shocks applied to the cerebellothalamic pathway a t the level of the red nucleus ( A ) , VL nucleus ( B and C), and motor cortex ( D ) . Effects of high-rate R F stimulation depicted in the right column. Time: 2 msec. Vertical bar: 0.3 mV. Note: reduction of the cortically elaborated response (second positive-negative wave, 2 ) to red nuclear stimulation during RF arousal, simultaneous with enhancement of the

ASCENDING REGULATING CEREBRAL SYSTEMS

125

during high-frequency RF stimulation (Fig. 14) (Purpura et al., 1966). Analysis of spontaneous unit activity also revealed that VL relay cells ( monosynaptically activated by shocks to the cerebellothalamic pathway and antidromically invaded by motor cortex stimulation) and VL neurons transsynaptically activated by stimulation of the precruciate gyrus ( cortico-VL pathway) are facilitated during wakefulness. Their spontaneous discharges are increased and regularized on arousal, and progressively reduced during SWS : bursts of spikes with periods of silence appear during 10-12Isec spindles and striking depression can be observed during occurrence of highamplitude slow ( P 6 l s e c ) waves (Steriade and Apostol, 1969). b. Motor Cortex Responsiueness. When simultaneously tested during R F stimulation, cortical and VL responses evoked by a shock to the red nucleus show different alterations, increase of the postsynaptic VL response being in sharp contrast with reduction of the intracortically generated ( second positive-negative ) component (Fig. 12A). Thus, RF-induced depression at the motor cortex level is so powerful that it appears even when the VL output is increased ( Steriade, 1969b). Decrease in cortical responsiveness appears not only with a protracted reticular stimulation, but also with a conditioning train of pulses to the RF, immediately preceding the testing shock (Steriade et a/., 1969b). Simultaneous recordings of potentials evoked in all (visual, auditory, somesthetic, and motor ) areas by stimulation of appropriate radiation showed that the motor cortex responsiveness was in sharp contrast with that of all sensory cortices, being tonically depressed from the very beginning of the RF conditioning pulse-train and showing no sign of early facilitation (Steriade and Cordeau, unpublished data). Some indications concerning the cortical level of the reticular inhibitory influences resulted also from reduction of the initial surface-positive slow wave evoked by VL stimulation (Fig. 9BJ; Fig. 12B), which was correlated to the primary evoked EPSP and the corticofugal spike discharge (Creutzfeldt et uZ., 1966). The reticular inhibitory VL postsynaptic ( r ) response ( A ); reticular depression of VL-evoked cortical biphasic sequence without alteration of the early (presynaptic, see arrow) spike ( B ) ; RF-induced suppression of relayed (R1and Rz) or indirect ( I ) polysynaptic component of the pyramidal tract response to VL ( C ) or motor cortex ( D ) stimulation, without depression of unrelayed or direct waves. ( A l , B, and D from Steriade, 19691); A2 and C from Steriade et al., 1969b).

126

M. STEFUADE

FIG. 13. Modifications undergone by rubrally evoked responses in the ventrolateral ( V L ) nucleus and motor cortex during wakefulness ( l), drowsiness with EEG spindles ( 2 ) , slow wave sleep ( 3 ) , and paradoxical sleep ( 4 ) . Time: 2 msec. Vertical bar: 0.3 mV. Note: striking reduction of the VL postsynaptic ( r ) component during EEG spindles ( 2 ) and maximal enhancement during paradoxical sleep ( 4 ) , without concomitant alterations of the presynaptic ( t ) spike; increase of cortically generated wave ( 2 ) during slow wave sleep, in spite of the VL output reduction, evidenced by decrease of the VL relayed component and early ( 1 ) cortical deflection. See also text ( froni Steriade et al., 1969b).

ASCENDING REGULATING CEREBRAL SYSTEMS

127

influence on cortical motor area was further demonstrated by analysis of diflerent components of the PT responses, regarded by Zanchetti and Brookhart (1955) as an index of excitability of the PT neurons. Relayed ( especially polysynaptic ) components of the PT response to VL stimulation and indirect ( I ) delayed waves evoked by direct cortical stimuli were diminished during

100msec

FIG. 14. Intrathalamic synaptic events associated with suppressing effect of high-frequency ( 50/sec) stimulation of brain stem RF on thalamocortical recruiting responses. A: unidentified VL neuron exhibits characteristic EPSPIPSP sequences during evoked recruiting responses to 7/sec midline thalamic stimulation. B: R F stimulation blocks recruiting responses a t the cortex and markedly attenuates synchronizing IPSPs. C and D: from another VL neuron; C, 7/sec midline thalamic stimulation alone, and D, marked attenuation of IPSPs and increase in cell discharges during simultaneous low-frequency thalamic stimulation and high-frequency reticular stimulation. Time bar in B: 100 insec. (From Purpura et al., 1966.)

reticular or natural arousal, without alteration of the unrelayed or direct ( D ) waves ( Fig. lZD,D), thus suggesting that interneurons in the motor cortex are more susceptible to depression during arousal than pyramidal neurons ( Steriade, 1969b). However, one might wonder whether selective depression of the PT polysynaptic delayed discharges is not partially due to “occlusion,” taking into account that, in opposition with inactivity of the largest PT cells during waking, spontaneous discliarges of the smallest PT neurons are more active during waking (Evarts, 1965).

128

M. STERIADE

Fluctuations in responsiveness of the motor cortex were subsequently studied by analysis of single neuronal evoked discharges during natural waking and sleep in the unrestrained, chronically implanted Mucucu mulatta (Steriade and Lamarre, 1969). From this preliminary communication it can be shortly said that dis-

W

sws

A t

B

5 msec

FIG. 15. Changes in responsiveness of the precentral motor cortex during sleep and wakefulness, in the unrestrained, chronically implanted monkey ( Macaca muluttu). Extracellular unit recordings. Testing shocks (arrows) applied to the ipsilateral VL nucleus ( A ) and to the homotopic point of the contralateral precentral gyrus ( B ). Ten superimposed sweeps. Note great decrease ( B ) or even disappearance ( A ) of the evoked neuronal discharges during wakefulness ( W), as compared with synchronized sleep (SWS) (Steriade and Lamarre, in preparation).

charges evoked at 1.5-4 msec latency in the precentral motor cortex by shocks applied to the VL nucleus were reduced or even abolished on arousal from SWS (Fig. 15A). Transsynaptically activated neurons by callosal stimulation of homotopic points in the contralateral precentral gyrus (some of them could be identified

ASCESDISC REGITLATIhC; CEKEBRAL SYSTEMS

129

as PT neurons) were also strikingly diminished during waking (Fig. 15B). In contrast, discharges of neurons activated by stimulation of the somesthetic pathway were obviously facilitated on waking. A study of spontaneous unit activity during sleep-wakefulncss cycle in the same preparation ( unrestrained, chronically im-

A

:,I

\..

FIG. 16. Patterns of spontaneous unit activity in the precentral motor cortex during sleep and wakefulness, in the unrestrained, chronically implanted monkey (Macaca rnidatta). Extracellular unit recordings. In A2 and B,, the firing rate of the unit was measured by an electronic counter and displayed over successive intervals of 10 seconds and 1 second, respectively. A: two units activated by VL (1) or transcallosal ( 2 ) stimulation. Note clear-cut increase of spontaneous firing during EEG synchronization associated with behavioral drowsiness or sleep. B: two units ( 1 and 2 ) modulated by gentle passive movements of the contralateral extremities. Note: increase of spontaneous firing during waking ( W ) and when the monkey's eyes open ( Lamarre and Steriade, in preparation).

planted monkey) showed similarly that neurons which could be modulated by VL or transcallosal stimulation decreased their spontaneous firing on arousal, while those which could be activated by passive movements or shocks to the lemniscal system were spontaneously more active during wakefulness (Fig. 16) ( Lamarre and Steriade, in preparation).

130

M. STERIADE

Our findings reporting inhibitory influences of RF-induced or natural arousal on motor cortex responsiveness are in disagreement with facilitation of VL-elicited cortical and PT discharges during RF stimulation reported by Akimoto and Saito (1966). but are supported by data of Klee (1966) showing a sustained increase of membrane potential and the decrease of shock-evoked responses in the motor cortex during R F arousal. Reticular-like depression of motor cortex PT responsiveness was induced by conditioning trains of fast pulses to unspecific thalamic nuclei (Buser et al., 1965) and powerful, long-lasting IPSPs have been recorded in the motor cortex upon stimulation of the diffuse thalamic system (Pollen, 1964; Purpura et al., 1964). Inhibitory effects exerted by caudate stimulation on motor cortex neurons (Spehlmann et al., 1960; Klee and Lux 1962; see also Hull et al., 1967) might be due to costimulation of unspecific thalamocortical fibers in the internal capsule, since capsular stimulation abolishes the VL-induced motor cortex and PT responses, while stimulation restricted to the caudate nucleus does not affect these evoked potentials (Marco and Brown, 1966). By comparative analysis of cortical responses to VL and red nuclear stimulation during natural arousal and various stages of sleep (Steriade et al., 1969b), clear-cut differences have been disclosed between the behavior of potentials evoked in the motor cortex by thalamic (Fig. 10) and prethalamic (Fig. 13) stimuli. Diminution of rubrally evoked response during EEG spindles and behavioral drowsiness, contrasting with increase of the VL-evoked potential, is explained by striking inhibition of synaptic transmission at the VL relay (as shown by disappearance of simultaneously recorded postsynaptic VL component ) and by predominantly postsynaptic testing stimulation of the thalamic nucleus. These differences at cortical level depending on the location of testing stimulation might be partially ascribed to the fact that a significantly large population of cortical units are evoked orthodromically by VL and are not evoked by rubral stimulation (Blum et al., 1968). With partial recovery of the VL output during SWS (evidenced by slight increase of the post-synaptic VL response as compared with drowsiness, see Fig. 13, 2 and 3 ) , motor cortex responses to VL and rubral stimuli are similarly enhanced in this stage of sleep, thus further suggesting cortical disinhibition. It must be again emphasized that disinhibition succeeds in eiihancing

ASCENDING REGULATING CEREBRAL SYSTEMS

131

cortical responses in spite of the reduction of the VL output during SWS as compared with wakefulness. We have difFiculty in reconciling our data showing increase of cortical responses to VL stimuli during both stages of synchronized sleep and enhancement of intracortically generated responses to rubral stimuli during SWS, with the results of Frommer and Livingston (1963) who reported a decrease in amplitude of cortical responses evoked by shocks to the region of the interpositus nucleus during synchronized sleep. the more as no obvious differences were observed in their study between drowsiness with EEG spindles and SWS, as is depicted in our Fig. 13. The enhancement of rubrally evoked cortical response during PS over values obtained in SWS was ascribed to an increased synaptic transmission through the intercalated VL relay at this deepest stage of sleep, since cortical response to postsynaptic VL stimulation and pyramidal tract response ( I waves) to direct stimulation of the motor cortex are depressed during PS (see Fig. 10 in Steriade et al., 196913). Hodes and Suzuki (1965) observed that cortically induced movements of the vibrissae and pinna have a lower threshold during PS than during SWS, concluded that “neocortical cells which initiate cortico-spinal movements have a lower stimulation threshold during desynchronized sleep, and ascribed appearance of jerks during PS “to the exaggerated responsiveness of the cortical centers” ( p. 245). These conclusions (opposite to our data, above mentioned) conflict with the pioneer observations of Tarchanoff (1894) and with more recent findings of Marchiafava and Pompeiano ( 1964) which showed convincingly that muscular jerks occur during desynchronized sleep even after destruction of the sensorimotor cortex and pyramidal tract, and that “motor response induced by repetitive stimulation of the pyramidal tract is altogether abolished during desynchronized sleep” ( p. 502). 2. Specific (Cerebellar) Znfluences The facilitatory specific influences exerted by the cerebellum on motor cortical area (see references in Dow and Moruzzi, 1958) is opposite in sign to unspecific (reticular) ascending inhibition. A conditioning stimulation of the cerebellothalamic pathway is to be preferred in order to avoid interference between excitatory influences of the nucleus interpositus on the contralateral red

132

M. S'IXRIADE

FIG. 17. Double, inhibitory vs. facilitatory, control of motor cortex responsiveness, exerted by unspecific vs. specific ascending influences. Two experiments (cats under flaxedil). Time: 2 msec. Vertical bars: 0.25 mV (Mot.), 0.05 mV (Pyr. in A1,2) and 0.1 niV (Pyr. in A3,4). A: pyramidal tract relayed responses to VL stimulation ( 1 ) and their depression ( 2 ) during highrate RF stimulation at the top arrow of the caudal electrode; contrariwise, motor cortex and pyramidal tract discharges were enhanced in the same experiment by conditioning short trains of pulses (300/sec, subliminal for evoking a cortical response) to the magnocellular (bottom arrow of the caudal elec-

ASCENDING REGULATINC; CEREBRAL SYSTEMS

133

nucleus and inhibitory corticonuclear mechanisms ( Tsukahara et nl., 1964; Ito et nl., 1964; see also review by Massion, 1967). In this condition, cortical and PT relayed discharges evoked by VL stimulation are obviously enhanced by conditioning trains of fast pulses applied to the magno- or parvocellular part of the red nucleus, in fair opposition with the decrease of motor cortex responsiveness by stimulating in the same experiment the neighboring mesencephalic KF ( Steriade, 1969b; Fig. 17A). The interruption of tegmentothalamic connections, leaving intact the cerebellothalamic facilitatory pathway, releases the motor cortex from tonic reticulocortical inhibition and induces a striking enhancement of cortical responsiveness ( Fig. 17B ) . These data suggest a double, inhibitory us. facilitatory, tonic control of motor cortex responsiveness, exerted by unspecific us’. specific ascending influences. IV. Final Remarks

As shown in the preceding section, neothalamocortical complexes are controlled by ascending influences arising in unspecific structures and specific afferent pathways. Since the most powerful effects exerted on the responsiveness of specific thalamic nuclei and cortical areas were obtained by stimulating the ascending diffuse systems and during the sleep-wakefulness cycle, these final remarks will concern themselves with the general events occurring in these experimental situations and some speculative psychophysiological correlations. Facilitation of synaptic events during arousal is a common finding for all (sensory and motor) specific thalamic relays, thus providing an improved transmission of afferent messages toward cortical receiving areas. Visual cortex exhibits an increased responsiveness during the transition from sleep to wakefulness (Bremer and Stoupel, 1959; Dumont and Dell, 1960), as shown by enhancement of cortically generated responses independent of geniculate events (see again Fig. 1B,C). Enhancement of fast rhythmic photic responses by reticular (Steriade and Demetrescu, 1960) trode) or parvocellular ( rostra1 electrode) part of the red nucleus (compare control in 3 with rubral effects in 4 ) . R: obvious enhancement of VL-evoked motor cortex and pyramidal tract discharges following a transection of reticulothalamic connections, sparing the specific cere1)ellothalamic pathway, Time: 2 msec. (Modified from Steriade, 1969b.)

134

M. STERIADE

or unspecific thalaniic ( Creutzfeldt and Grusser, 1959) stimulation, together with facilitation of visual recovery cycle during arousal ( Lindsley, 1958), are physiological correlates of some integrative perceptual processes, evidenced by improvement of tachistoscopic performances in monkey (Fuster, 1958; Fuster and Uyeda, 1962) and visuomotor time reaction in man (Lansing et al., 1956). Dishabituation is also a sign of reticulocortical activation at the level of the visual (Mancia et al., 1959; Cavaggioni et al., 1959) and auditory (Steriade and Demetrescu, 1962) systems. In man, increase of responses to rhythmic photic stimuli during arousal was sometimes associated with increase in subjective experiences ( Steriade et al., 1961). It must be emphasized that facilitation during the transient phase of arousal affects intrinsic cortical responses to photic stimulation, i.e., postprimary components or fast sensory afterdischarge (see Steriade et al., 1968, and Fig. 4 ) . Such waves may carry crucial information regarding the responsiveness of a given sensory system (Hughes, 1964) and are regarded as reflecting activities related to information processing or perhaps early storage (Shagass et al., 1965). Some diverging results reporting the reduction of photic driving during arousal reaction in cat (Morrel et al., 1957) or during nonvisual stimulation in man (Carels, 1962) may be explained by differences in the level of arousal preceding the conditioning stimulation. When the already “activated EEG is associated with high responsiveness, an additional activation may induce the reduction of responses to photic (Jane et al., 1962; Floru et al., 1964) or acoustic (Chin et aZ., 1965) stimuli. It was proposed that an optimum level of activation is required for most efficient performances, which, when exceeded, interferes with efficiency (Malmo, 1959). Depression of sensory evoked responses beyond a certain level of activation and deterioration of motor performances with slightly increased voltages of R F stimulation recall the states of “obtusion” due to strong reticulocortical activation, sometimes leading to an arrest-reaction (Hunter and Jasper, 1949). Cortical responsiveness was found to be increased during the phase of transition from sleep to wakefulness in all sensory areas, but mostly at the level of the visual and somesthetic areas. Bremer and Stoupel ( 1959) had already emphasized that RF-induced potentiation is much greater in the visual, than in the other sensory, areas. These might be provisionally related to the outstanding role

ASCENDING REGULATING CEREBRAL SYSTEMS

135

of a visual messages in the orienting reaction of the high mammals which are used in these experiments. Simultaneous depression of the events recorded in the motor cortex might be regarded as a sign of “expectancy” and preparation of activity following integration of sensory messages. Since all thalamic nuclei (including VL) are facilitated during arousal, it may be alternatively inferred that brain processes underlying performances of the orienting reaction require a high level of responsiveness up to and including the thalamic relays. It is without saying that these speculations are rather tenuous without further simultaneous examination of behavior and electrophysiological activities. As far as the diminution of evoked responses during the tonic state of waking is concerned, it was associated in the visual cortex with significant reduction of spontaneous discharges, thus possibly resulting in relative enhancement of discriminative processes b y increase of signal-to-noise ratio (Evarts, 1960). If a generalized reaction may be conceived as resulting at first from diffuse facilitation of thalamic relays during arousal, then discriminative awareness of the outside world, which follows the short-term hiatus induced by any novel stimulus, implies ascending inhibitory influences to restrict excitability processes and to concentrate perceptual integration and motor activity for biological necessities of a given moment. Indeed, waking may be associated ( besides increase of facilitatory influences) with enhancement of active inhibition in the visual (Evarts, 1960; Cordeau et al., 1965; Dagnino et al., 1965a; Demetrescu et al., 1965), auditory (Symmes and Anderson, 1967), somesthetic (Favale et al., 1965; Allison and Goff, 1968), and motor ( Steriade, 1969b; Steriade et al., 1969b) areas, highly increasing the cortical efficacy and preventing discharge of crude, uncontrolled energy ( Demetrescu et al., 1966). These conclusions drawn from studies on cortical responsiveness are in agreement with investigations on the spontaneous activity of single neurons. Taken together, these data support the prior view of Jasper (1958) stating that a complex reorganization of spatial and temporal patterns of neuronal discharges is associated with wakefulness, implying that ascending inhibition parallels excitation during the state of waking. On the other hand, a progressive disinhibition is seen during sleep. It occurs at the cortical l e ~ e l ,an assumption which is supported by reduction during sleep of inhibitory events induced by a shock to the LG radiation on spontancoas discharges of the visual

136

M. STERIADE

cortex neurons (Evarts, 1960). Release from inhibition during SWS was also suggested by recovery of the delayed potential from a pair of cortical mass responses evoked by stimulation of the visual (Evarts et d., 1960; Demetrescu et al., 1966) or motor (Steriade et al., 1969b) systems. Maximal increase of responsiveness during PS was explained, besides disinhibition, by transient increase of facilitatory influences, although it is rather difficult to ascertain whether striking enhancement of evoked responses was not due to postsynaptic LG facilitation ( see Section II1,A). Actually, an increased responsiveness was found to occur in all thalamic relays at the deepest stage of sleep, possibly related to the elaboration of dreams. This change, which is in opposite direction as compared with the decreased thalamic responsiveness during SWS, suggested that partial sensory deaff erentation may contribute to the onset of sleep, but it is not a necessary condition for the maintenance of sleep (Walsh and Cordeau, 1965), in good agreement with the active theory of sleep. Finally, dissimilar and even opposite alterations of thalamic and cortical responsiveness during SWS and PS add further arguments to the dual theory of sleep. ACKNOWLEDGMENT

The preparation of the inanuscript and some of the experiments reported in this review were supported by a research grant, MA-3689, from the Medical Research Council of Canada, which is gratefully acknowledged. I wish to thank my colleagues with whom a part of the personal experiments quoted above was done (in alphabetical order): V. Apostol, M. Demetrescu, D. Ionescu, G. Iosif and Y. Lamarre. I also wish to express my thanks to Professors F. Bremer (Brussels) and J. P. Cordeau (Mont&al) for helpful criticism on the manuscript. REFERENCES Ajmone-Marsan, C., and Morillo, A. (1963). Arch. Itul. B i d . 101, 1. Akert, K., Koella, W. P., and Hess, R., Jr. (1952). Am. J. Physiol. 168, 260. Akimoto, H., and Saito, Y. (1966). In “Correlative Neurosciences” (T. Tokizane and J. P. Schadk, eds.), Part A, pp. 32.3351. Elsevier, Amsterdam. Akimoto, H., Saito, Y., and Nakamura, Y. (1961). In “The Visual System: Neurophysiology and Psychophysics” ( R. Jung and H. Kornhuber, eds. ), pp. 363-372. Springer, Berlin. Albe-Fessard, D. (1957). J. Physiol. ( P u r k ) 49, 521. Albe-Fessard, D., Mallart, A., and Alkonard, P. ( 1961). Compt. Rend. 252, 1060. Albe-Fessard, U., Massion, J., Hall, H., and Rosenblith, W. ( 1964). Contpt. Rend. 258, 1060. Allison, T. ( 1965). Electroencephulog. CZin. Nezrrophysiol. 18, 131.

ASCENDING REGULATING CEREBRAL SYSTEMS

137

Allison, T., and Goff, G. D. (1968). Arch. Ital. Biol. 106, 41. Allison, T., Goff, W. R., and Sterman, M. B. (1966). Electroencephalog. Clin. Neurophysiol. 21, 461. Amassian, V. E., and Weiner, H. (1966). In “The Thalamus” (D. P. Purpura and M. D. Yahr, eds.), pp. 255-282. Colntnbia Univ. Press, New York. Angel, A., Magni, F., and Strata, P. (1965). Arch. ltal. Biol. 103, 668. Arduini, A., and Cavaggioni, A. (1960). Boll. SOC. Ztal. Biol. Sper. 37, 1393 4rduini, A., and Goldstein, M. H., Jr. (1961). Arch. Ital. Biol. 99, 397. Arduini, A., and Hirao, T. (1960). Aich. Ifal. Biol. 98, 182. Ascher, P. ( 1965). Thesis, Paris. Baldissera, F., Cesa-Bianchi, G., and hlancia, M. ( 1966). Arch. Ztul. Bic~l. 104, 247. Batini, C., Moruzzi, G., Palestini, M., Rossi, G. F., and Zanchetti, A. (1959) Arch. ltul. Biol. 97, 1. Battersby, W. S., and Oesterreich, R. E. (1963). Electroencephalog. Clin. Neurophysiol. 15, 849. Baust, W., Berlucchi, G., and Moruzzi, G. ( 1964). Arch. Ital. Biol. 102, 657. Benoit, 0. (1964). Compt. Rend 258, 1076. Benoit, O., and Bloch, V. ( 1960). J. Physiol. (Paris) 52, 17. Berlucchi, G., Maffei, L., Mornzzi, C . , and Strata, P. (1964). Arch. Ital. Bio!. 102, 372. Berlucchi, G., Munson, J. B., and Rizzolati, G. (1967). Arch. Ital. Biol. 105, 118. Bignall, K. E., and Rutledge, L. T. (1964). J. Neurophysiol. 27, 1048. Bishop, G. H., and Clare, M. H. (1953). J . Neurophysiol. 16, 1. Bishop, P. 0. (1964). Intern. Rec. N c w o ~ ~ o 6, ~ . 191. Bizzi, E. (1965). Physiologist 8, 11.3. Blum, B., Halpern, L. M., and Ward, A. A., Jr. ( 1968). Exptl. Neurol. 20, 156. Bonnet, V. (1957). Arch. Intern. Physiol. Biochim. 65, 506. Bonvallet, M., and Allen, M. B., Jr. ( 1963). Electroencephalog. Clin. Neurophyswl. 15, 969. Bonvallet, M., and Bloch, V. (1961). Science 133, 1133. Bremer, F. (1935). Compt. Rend. Soc. Biol. 118, 1235. Bremer, F. (1958). Physiol. Rev. 38, 357. Bremer, F. (1961). In “The Nature of Sleep” ( C . E. W. Wolstenholme and M. O’Connor, eds. ), pp. 30-50. Churchill, London. Bremer, F. (1966). Compt. Rend. 262, 2484. Bremer, F., and Bonnet, V. ( 1950). Electroencephalog. Clin. Neurophysiol. 2, 389. Bremer, F., and Stoupel, N. (1957). Arch. Ital. Btol. 95, 3. Bremer, F., and Stoupel, N. (1958). Acta Neurol. Psychiat. Belg. 58, 401. Bremer, F., and Stoupel, N. (1959). Arch. Intern. Physiol. Biochim. 67, 240. Bremer, F., Stoupel, N., and Van Reeth, P. C. ( 1950). Arch. ltul. Biol. 98, 229. Brookhart, J. M., and Zanchetti, A. ( 1956). Electroencephalog. Clin. Neiirophysiol. 8, 427. Brookhart, J. M., Arduini, A,, Mancia, M., and Moruzzi, G. (1958). J. Neurophysiol. 21, 499. Brown, T. S., and Marco, L. A. ( 1966-1967). Bruin Res. 3, 302.

138

M. STEWADE

Brown, T. S., and Marco, C. A, ( 1967). Electroencephalog. Clin. Neurophysiol. 22, 150. Buchwald, N. A., Wyers, E. J., Okuma, T., and Hewer, C . (1961). Electroencephabg. Clin. Neurophysiol., 13, 509. Buser, P., Encabo, H., and Lamarche, M. (1965). Arch. Ital. Biol. 103, 448. Carels, G. ( 1962). Electroencephalog. Clin. Neurophysiol. 14, 714. Cali, G., Diete-Spiff, K., and Ponipeiano, 0. (1967a). Arch. Ital. Biol. 105, 31. Carli, G., Diete-Spiff, K., and Pompeiano, 0. ( 19678). Arch. Ital. Biol. 105, 52. Carli, G., Diete-Spiff, K., and Pompeiano, 0. ( 1 9 6 7 ~ )Arch. . Ital. B i d . 105, 83. Cavaggioni, A., and Goldstein, M. H., Jr. (1965). Arch. Ital. Biol. 103, 397. Cavaggioni, A., Giannelli, G., and Santibanez-H., G. (1959). Arch Ital. Biol. 97, 266. Cesa-Bianchi, M. G., and Sotgiu, M. L. (1969). Brain Res. 13, 129. Chang, H. T. (1952). J. Neurophysiol. 15, 5. Chang, H. T., and Kaada, B. (1950). J. Neurophysiol. 13, 305. Chin, J. H., Killam, E. K., and Killam, K. F. (1965). Electroencephalog. Clin. Neurophysiol. 18, 567. Chi, C. C., and Flynn, J. P. (1968). Electroencephalog. Clin. Neurophysiol. 24, 343. Clare, M. H., Landau, W. M., and Bishop, G. H. (1964). Exptl. Neurol. 9, 262. Cohen, B., Housepian, E. M., and Purpura, D. P. (1962). Exptl. Neurol. 6, 492. Combs, C. M., and Saxon, S. V. (1959). Exptl. Neiirol. 1, 583. Cordeau, J. P. (1962). Rev. Can. Biol. 21, 113. Cordeau, J. P., and Mancia, M. ( 1959). Electroencephalog. Clin. Neurophysiol. 11, 551. Cordeau, J. P., Moreau, A., Beaulnes, A , , and Laurin, C. (1963). Arch. Ital. Biol. 1101, 30. Cordeau, J. P., Walsh, J., and Mahut, H. (1965). In “Aspects anatomo-fonctionnels de la physiologie du sommeil” ( M . Jouvet, ed.), pp. 473-503. C.N.R.S., Paris. Courtois, A., and Cordeau, J. P. ( 1969). Brain Res. 14, 199. Courville, J., Walsh, J., and Cordeau, J. P. (1962). Science 138, 973. Creutzfeldt, 0. D. ( 1961 ). I n “The Visual System: Neurophysiology and Psychophysics” ( R . Jung and H. Kornhuber, eds.), pp. 372-373. Springer, Berlin. Creutzfeldt, O., and Grusser, 0. J. (1959). Proc. 1st Intern. Congr. Neurol. Sci., Brussels, 1957 pp. 349-355. Perganion Press, Oxford. Creutzfeldt, O., Rosina, A., Ito, M., and Probst, W. (1969). J. Neurophysiol. 32, 127. Creutzfeldt, 0. D., Spehlmann, R., and Lehman, D. (1961). In “The Visual System: Neurophysiology and Psychophysics” (R. Jung and H. Kornhuber, eds.), pp. 351-363. Springer, Berlin. Creutzfeldt, 0. D., Watanabe, S., and Lux, H. 1). (1966). Electroencephalog. Clin. Neurophysiol. 20, 1. Dagnino, N., Favale, E., Loeb, C., and Manfredi, M. (1965a). J. Neurophysiol. 28, 443.

ASCENDING REGULATING CEREBRAL SYSTEMS

139

Dagnino, N., Favale, E., Loeh, C., a i d hlanfredi, M. (1965b). Erperientin 21, 459. Dell, P., and Padel, Y. ( 1965). Elect,.oeiice/illcllog. C h . Neurophysiol. 18, 725. Dell, P., Bonvallet, M., and Hugelin, A. (1961). I n “The Nature of Sleep” (G. E. W. Wolstenholme and M. O’Connor, eds.), pp. 86-107. Churchill, London. Dement, W. ( 1958). Electroencephalog. Clin. Neurophysiol. 10, 291. Demetrescu, M., and Demetrescu, M. ( 196% ). Electroencephalog. Clin. Neurophysiol. 14, 37. Demetrescu, M., and Denietrescu, hl. ( 19621,). Electroencephalog. Clin. Neurophysiol. 14, 602. Demetrescu, M., Demetrescu, M., and Iosif, G. ( 1965). Electroencephulog. Clin. Neurophysiol. 18, 1. Demetrescu, M., Demetrescu, M., and Iosif, C.. ( 1966). Electroencephalog. Clin. Neurophysiol. 20, 450. Denney, D., and Thompson, R. F. ( 1967 ). Electroencephalog. Clin. Neurophysiol. 23, 248. Desmedt, J. E., and La Grutta, G. (1957). J. Physiol. ( L o n d o n ) 136, 20. Doty, R. W., Kiniura, D. S., and Mogenson, C . J. ( 1964). Erptl. Neurol. 10, 19. Dow, R. S., and Moruzzi, G. (1958). “The Physiology and Pathology of the Cerebellum.” Univ. of Minnesota Press, Minneapolis, Minnesota. Dumont, S., and Dell, P. ( 1958). J. Physiol. ( P a r i s ) 50, 261. Dumont, S., and Dell, P. ( 1960). EZectroencephalog. Clin. Neurophysiol. 12, 769. Ebe, M., and Mikami, T. ( 1962). Tohokrr J. Exptl. Med. 77, 383. Evarts, E. V. (1960). Federation Proc. 19, 828. Evarts, E. V. (1962). J. Neurophysiol. 25, 812. Evarts, E . V. (1963). J. Neurophysiol. 26, 229. Evarts, E. V. (1965). In “Sleep Mechanisms” ( K . Akert et al., eds.), pp. 81-89. Elsevier, Amsterdam. Evarts, E. V., Fleming, T. C., and Hrittenlocher, P. R. (1960). Am. J. Physiol. 199, 373. Favale, E., Loeb, C., and Manfredi, M. (1963). Boll. Soc. Ztal. Biol. Sper. 39, 1513. Favale, E., Loeb, C., Manfredi, M., and Sacco, G. (1965). Electroencephulog. Clin. Neurophysiol. 18, 354. Fernandez-Guardiola, A,, Harmony, T., and Roldan, E. (1964). Electroencephalog. C h . Neurophysiol. 16, 259. Filion, M., Lamarre, Y., and Cordeau, J . P. (1969). 1. Physiol. (Paris), in press. Floru, R., Sterescu-Volanschi, M., and Bittman, E. (1964). Rev. Roumuine Physiol. 1, 63. Frigyesi, T. D., and Purpura, D. P. (1964). Erptl. Neurol. 10, 305. Frommer, G. P., and Livingston, R. B. (1963). Science 139, 502. Fuster, J. M. (1958). Science 127, 150. Fuster, J. M. (1961). Science 133, 2011. Fuster, J. M., and Docter, R. F. (1962). J. Neurophysiol. 25, 324. Fuster, J. M., and Uyeda, A. A. (1962). Exptl. Neural. 6, 384. Galambos, R., and Rupert, A. (1959). J. Acoust. Soc. Am. 31, 349.

140

M. STJ3RIADE

Gastaut, H., Beck, F., Faidherbe, J., Franck, G., Fressy, J., Rkmond, A,, Smith, C., and Werre, P. (1963). In “Brain Mechanisms” ( G . Moruzzi et al., eds. ), pp. 374-392. Elsevier, Amsterdam. Gauthier, C., Parma, M.,and Zanchetti, A. ( 1956). Electroencephalog. Clin. Neurophysiol. 8, 237. Gellhorn, E., Koella, W. P., and Ballin, H. M. (1954). J. Neurophysiol. 17, 14. Goff, W. R., Sterman, M. B., and Allison, T. (1965). Physiologist 8, 174. Goff, W. R., Allison, T., Shapiro, A., and Rosner, B. S. (1966). Electroencephalog. Clin. Neurophysiol. 21, 1. Gorman, A. L. F. (1966). J. Neurophysiol. 29, 547. Granit, R. (1955). “Receptors and Sensory Perception.” Yale Univ. Press, New Haven, Connecticut. Grossman, R. G., Clark, K., and Whiteside, L. (1967). Brain Res. 5, 273. Guilbaud, G. ( 1968). Thesis, Paris. Heath, C. A., and Hodes, R. (1952). Tram. Am. Neurol. Assoc. 77, 204. Hemlindez-PBon, R., Scherrer, J., and Velasco, M . (1956a). Acta Neurol. Latinoam. 2, 8. Hernhdez-Pkon, R., Scherrer, H., and Jouvet, M. (1956b). Science 123, 331. Hemlindez-PAon, R., Guzman-Flores, C., Alcaraz, M., and Fernandez-Gnardiola, A. (1957). Acta Neurol. Latinoam. 3, 1. Hernlindez-PBon, R., Charez-Ibarra, G., Morgane, P. J., and Tinio-Iaria, C. ( 1963 ) . Exptl. Neurol. 8, 93. Hernlindez-Pkon, R., O’Flaherty, J. J., and Mazzuchelli-O’Flaherty, A. L. (1965). Exptl. Neurol. 13, 40. Hess, W. R. ( 1949). “Das Zwischenhirn: Syndrom, Lokalisationen, Funktionen.” Benno Schwabe, Basel. Hodes, R., and Suzuki, J. I. ( 1965). Electroencephalog. Clin. Neurophysiol. 18, 239. Hubel, D. H. (1959). J. Physiol. (London) 147, 226. Hubel, D. H. (1960). Arch. ltal. Biol. 98, 171. Hugelin, A., Dumont, S., and Paillas, N. (1960). Electroencephalog. Clin. Neurophysiol. 12, 797. Hughes, J. R. ( 1964). Intern. Reu. Neurobiol. 7 , 99. Hull, C. D., Buchwald, N. A., and Vieth, J. (1967). Brain Res. 6, 12. Hunter, H., and Jasper, H. ( 1949). Electroencephalog. Clin. Neurophysiol. 1, 405. Huttenlocher, P. R. ( 1960). Electroencephalog. Clin. Neurophysiol. 12, 819. Huttenlocher, P. R. (1961). J. Neurophysiol. 24, 451. Infantellina, F., Sanseverino, E. R., and Sperti, L. (1966). Arch. Sci. Biol. (Bologna) 50, 1. Ionescu, D. A. (1969). Exptl. Brain Res. 7, 275. Ito, M., Yoshida, M., and Ohata, K. (1964). Experientia 20, 575. Iwama, K., and Sasakura, H. (1965). Proc. Japan Acad. 41, 499. Jane, J. A,, Smirnov, G. D., and Jasper, H. H. (1962). Electroencephalog. Clin. Neurophysiol. 14, 344. Jasper, H. H. (1958). In “Reticular Formation of the Brain” ( H . H. Jasper et al., eds.), pp. 319-331. Little, Brown, Boston, Massachuset~. Jouvet, M. (1962). Arch. ltal. Biol. 100, 125.

ASCENDING REGULATING CERE:BRAL SYSTEMS

141

Jouvet, M. (1967a). Physiol. Rev. 47, 117. Jouvet, M. (1967b). Res. Publ., Assoc. Res. Nervous Mental Disease 45, 86. Jonvet, M., and Michel, F. (1960). ConLpf. Rend. SOC. B i d . 154, 636. Kaada, B. R., Thomas, F., Alinars, E., and Wester, K. (1967). Elfctroencephalog. Clin. Neuroph~siol.22, 220. Kaneko, Z., Hidaka, Y., Hishikawo, Y ., Shiinixu, A., and Miyazaki, K. ( 1961 ). Folia Psychiat. Neurol. Japoii. 15, 206. Klee, M. R. ( 1966). In “The Thalanius” ( D . P. Purpura and M. D. Yahr, eds.), pp. 287-317. Colunibia Univ. Press, New York. Klee, M. R., and Lux, H. L). ( 1962). Arch. Psychiat. Newenkrankh. 203, 667. Kleitman, N. (1929). Physiol. Reo. 9, 624. Koizumi, K., Ushiyama, J., and Brooks, C. h?. (1958). Am. J. Physiol. 195, 1. Kreindler, A., and Steriade, M. (1964). Arch. Ital. Biol. 102, 576. Kreindler, A., and Steriade, M. (1966). Bruin Res. 1, 279. Kuffler, S. W., Fitzhugh, R., and Barlow, €1. B. (1957). J. Gen. Physiol. 40, 683. Landau, W. A?., and Clare, hl. H . (1956). E/ectroencephalog. Clin. Neurophysiol. 8, 457. Lansing, R. W., Schwartz, E., and Lindsley, D. B. (1956). Am. Psychologist 11, 433. Lindsley, 1). B. ( 1958). In “Reticular Formation of the Brain” ( H . H. Jasper et al., eds.), pp. 513-534. Little Brown, Boston, Massachusetts. Lindsley, U. B., Schreiner, L., Knowles, W. B., and Magoun, H. W. (1950). Electroencephalog. Clin. Neurophysiol. 2, 483. Long, R. G. (1959). J. Neurophysiol. 22, 412. Maekawa, K., and Purpura, D. 1’. (1967). Brain Res. 4, 308. Maffei, L., and Rizzolatti, G. (1985). Arch. ltal. Biol. 103, 609. Xlaffei, L., Moruzzi, G., and Rizzolatti, G. (1965). Arch. Ital. B i d . 103, 596. Magnes, J., Moruzzi, G., and Pompeiano, 0. (1961). Arch. Ztal. Biol. 99, 33. Magni, F., Moruzzi, G., Rossi, G. F., and Zanchetti, A. (1959). Arch. Ztal. Biol. 97, 33. Manzoni, T., Sapienza, S.,and Urbano, A. (1968). Arch. Ital. Biol. 106, 61. Marchiafava, P. L., and Poinpeiano, 0. (1964). AT&. Ztal. Biol. 102, 500. hlagoun, H. W. (1959). Proc. 21st Intern. Congr. Physwl., Buenos Aires, 1959 Symp., p. 46. Mdmo, B. (1959). Psychol. Reo. 66, 376. Mancia, M., Meulders, M., and Santibanez-H., G. (1959). Arch. Ztal. Biol. 97, 378. Marco, L. A,, and Brown, T. S. (1966). Exptl. Neurol. 14, 8. Massion, J. (1967). Physiol. Reu. 47, 383. Massion, J., and Meulders, M. (1960). J . Phqsiol. (Paris) 52, 972. Morillo, A. ( 1961). Ekctroencephulog. C h i . Neurophysiol. 13, 9. Morrell, F., Naquet, R., and Gastaut, H. (1957). J. Neurophysiol. 20, 574. Moruzzi, G . (1958). Arch. Ital. B i d . 96, 17. Moruzzi, G. ( 1960). Electroencephalog. C h i . Neurophysiol. Suppl. 13, 231. Moruzzi, G. (1964). Harvey Lectures 58, 23-3. Moruzzi, G., and Magoun, H. W. ( 1949). Ebctroencephalog. Clin. Neurophysiol. 1, 455.

142

M. STERIADE

Nacimiento, A. C., Lux, H. D., and Creutzfeldt, 0. D. (1964). Arch. Ges. Physiol. 281, 152. Naquet, R., R6gis, H., Fischer-Williams, M., and Fernandez-Guardiola, A. (1960). Brain 83, 52. Narikashvili, S. P. (1963). In “Brain Mechanisms” (G. Moruzzi et ul., eds.), pp. 155-183. Elsevier, Amsterdam. Nauta, W. J. H. (1946). J . Neurophysiol. 9, 285. Nauta, W. J. H., and Whitlock, D. G. (1954). In “Brain Mechanisms and Consciousness” ( E . D. Adrian et al., eds.), pp. 81-104. Blackwell, Oxford. Ogawa, T. (1963). Science 139, 343. Okuma, T., and Fujiniori, M. (1963). Foliu Psychiat. Neurol. Jupon. 17, 25. Palestini, M., Pisano, M., Rosadini, G., and Rossi, G. F. (1964). Exptl. Neurol. 9, 17. Papez, J. (1937). A.M.A. Arch. Neurol. Psychiut. 38, 725. Patton, H. D., and Aniassian, V. E. (1954). J. Neurophysiol. 17, 345. Pecci-Saavedra, J., Wilson, P. D., and Doty, R. W. (1966). Nature 210, 740. Pollen, D. A. ( 1964). Electroencephalog. Clin. Neurophysiol. 17, 398. Pompeiano, O., and Swett, J. E. (1962). Arch. Ital. Biol. 100, 311. Posternack, J. M., Fleming, T. C., and Evarts, E. V. (1959). Science 129, 39. Purpura, D. P., and Housepian, E. M. ( 1961). Electroencephulog. Clin. Neurophysiol. 13, 365. Purpura, D. P., Shofer, R. J., and Musgrave, F. S. (1964). J. Neurophysiol. 27, 133. Purpura, D. P., Scarff, T., and McMurtry, J. G. (1965). J . Neurophysiol. 28, 487. Purpura, D. P., McMurtry, J. G., and Maekawa, K. (1966). Brain Res. 1, 63. Ranson, S. W. (1939). A.M.A. Arch. Neurol. Psychiut. 41, 1. Ranson, S. W., and Magoun, H. W. (1939). Ergeb. Physiol. 41, 56. Redding, F. K. (1967). Electroencephulog. Clin. Neurophysiol. 22, 74. Roytbak, A. I. (1960). Electroencephulog. Clin. Neurophysiol. Suppl. 13, 91. Rosina, A., and Mancia, M. ( 1966). Ebctroencephulog. Clin. Neurophysiol. 21, 157. Rossi, G. F., Favale, E., Hara, T., Giussani, A., and Sacco, G. (1961). Arch. Ital. Biol. 99, 270. Sakakura, H., and Doty, R. W. (1969). Ebctroencephalog. Clin. Neurophysiol. 27, 687. Sakata, H., Ishijama, T., and Toyoda, Y. (1966). Japan. J . Physiol. 16, 42. Scheibel, M. E., and Scheibel, A. B. (1958). In “Reticular Formation of the Brain” (H. H. Jasper et al., eds.), pp. 31-55. Little, Brown, Boston, Massachusetts. Schlag, J., and Villablanca, J. (1967). Bruin Res. 6, 119. Schlag, J., and Villablanca, J. (1968). Brain Res. 8, 255. Schlag, J., Chaillet, F., and Herzet, J. P. (1961). Science 134, 1691. Schoolman, A., and Evarts, E. V. (1959). J . Neurophysiol. 22, 112. Schwartz, M., and Shagass, C. ( 1962). Electroencephulog. Clin. Neurophysiol. 14, 11. Shagass, C., Schwartz, M., Brown, J. C. N., and Straumanis, J. (1965). Rept. U.S. P.H.S. Grant MH02635.

ASCENLXNG REGULATING CEREBRAL SYSTEMS

143

Spehlmann, R., Creutzfeldt, 0. D., and Jung, R. (1960). Arch. Psychiat. Neruenkrunkh. 201, 332. Spencer, W. A., and Brookhart, J. M. (1961). J. Neurophysiol. 24, 26. Starr, A. (1964). Exptl. Neurol. 10, 191. Steinberg, R. H. ( 1965). Electroencephulog. Clin. Neurophysiol. 18, 378. Steriade, M. ( 1964). Ekctroencephulog. Clin. Neurophysiol. 17, 600. Steriade, M. ( 1967a). In “Actualit& neurophysiologiques” (A. M. Monnier, ed. ), Vol. 7, pp. 109-139. Masson, Paris. Steriade, M. ( 1967b). Ekctroencephulog. Clin. Neurophysiol. 22, 577. Steriade, M. (1968). Bruin Res. 9, 169. Steriade, M. (1969a). “Physiologie des voies et des centres visuels.” Masson, Paris. Steriade, M. ( 196913). Electroencephalog. Clin. Neurophysiol. 26, 25. Steriade, M., and Apostol, V. (1969). Unpublished data. Steriade, M., and Demetrescu, M. (1960). J. Newophysiol. 23, 602. Steriade, M., and Demetrescu, M. ( 1962). Electroencephalog. Clin. Neurophysiol. 14, 21. Steriade, M., and Demetrescu, M. (1964). Federation Proc. 23, Trans]. Suppl., T101. Steriade, M., and Demetrescu, M. ( 1966). Electroencephulog. Clin. Neurophysiol. 20, 576. Steriade, M., and Demetrescu, M. (1967a). Exptl. Neurol. 19, 265. Steriade, M., and Demetrescu, M. (1967b). Ekctroencephalog. Clin. Neurophysiol. 23, 429. Steriade, M., and Ionescu, D. (1967). Exptl. Bruin Res. 4, 2%. Steriade, M., and Lamarre, Y. (1969). J. Physiol. (Paris), in press. Steriade, M., and Stoupel, N. ( 1960). Electroencephalog. Clin. Neurophysiol. 1% 119. Steriade, M., Stoica, I., and Stoica, E. (1961). Rev. Neurol. loS, 187. Steriade, M., Belekhova, M., and Apostol, V. (1968). Bruin Res. 11, 276. Steriade, M., Constantinescu, E., and Apostol, V. (1969a). Bruin Res. 13, 177. Steriade, M., Iosif, G., and Apostol, V. (196913). J. Neurophysiol. 32, 251. Sterman, M. B., and Clemente, C. D. (1962a). Exptl. Neurol. 6, 91. Sterman, M. B., and Clemente, C. D. (1962b). Exptl. Neurol. 6, 103. Suzuki, H., and Taira, N. (1961). Japan. J. Physiol. 11, 641. Symmes, D., and Anderson, K. V. (1967). Exptl. Neurol. 18, 161. Szentagothai, J. (1962). In “Information Processing in the Nervous System” (R. W. G r a r d and J. W. DuyfF, eds.), pp. 119-135. Excerpta Med. Found., Amsterdam. Taira, N., and Okuda, J. (1962). Tohoku 1. Exptl. Med. 78, 76. Tarchanoff, J. (1894). Arch. Ital. Biol. 21, 318. Teramoto, S., and Snider, R. S. (1966). Exptl. Neurol. 16, 191. Tsukahara, N., Toyama, K., and Kosaka, K. (1964). Expedentiu 20, 632. Velasco, M., and Lindsley, D. B. (1965). Science 149, 1375. Walberg, F. (1965). Exptl. Neurol. 13, 218. Walsh, J. T., and Cordeau, J. P. (1965). Exptl. Neurol. 11, 80. Watanabe, S., Konishi, M., and Creukfeldt, 0. D. (1966). Exptl. Brain Res. 1, 272.

144

M. STERIADE

Wickelgren, W. 0. (1968). J . Neurophysiol. 31, 757. Wid&, L.,and Ajmone-Marsan, C. (1960). Arch. Ital. Biol. 94 248. Winters, W. D. (1964). Electroencephalog. Clin. Neurophysiol. 17, 234. Worden, F. G., Marsh, J. T., Abraham, F. D., and Whittlesey, J. R. B. (1964). Electroencephalog. Clin. Neurophysiol. 17, 524. Yakovlev, P. I., and Locke, S. (1961). Arch. Neurol. 5, 364. Zanchetti, A., and Brookhart, J. M. (1955).J. Neurophysiol. 18, 288.

THEORIES OF BIOLOGICAL ETIOLOGY OF AFFECTIVE DISORDERS By John M. Davis Laboratory of Clinical Science, and Clinical Research Branch, National Institute of Mental Health, Eetherda, Maryland

I. Introduction . . . . . . . . . . 11. The Biogenic Amine Hypothesis . . . . . . 111. Synthesis and Metabolism of Biogenic Amines . . . IV. Antidepressants . . . . . . . . . A. Monoamine Oxidase Inhibitors . . . . . . B. Tricyclic Antidepressants . . . . . . . V. Reserpine-Induced Depression . . . . . . . VI. Electroconvulsive Shock Therapy . . . . . . VII. Lithium . . . . . . . . . . . VIII. Amphetamine . . . . . . . . IX. Central Receptors and Depression . . . . . . X. Catecholamine Metabolism in Depressed and Manic Patients XI. Indole Metabolism in Depressed Patients . . . . XII. Biogenic Amine Levels in Human Brain . . . . . XIII. Experimental Drugs and the Biogenic Amine Hypothesis . XIV. Electrolytes . . . . . . . . . XV. Steroids in Depression . . . . . . . . XVI. Discussion . . . . . . . . . . . References . . . . . . . . .

.

. .

.

. .

.

. . . . .

. .

. .

. . .

. . .

. .

.

145 147 148 149 149 150 152 154 154 156 156 157 159 159 160 161 163 165 167

I. Introduction

The existence of effective physical and pharmacological treatments of mania and depression provides an important avenue for investigation into the etiologies of these disorders. It is germane to consider the modes of discovery of the antimanic and antidepressant drugs, as well as their etiological implications. The body of data on electrolyte and steroid alterations in the affective disorders will then be discussed, focusing on findings of etiological, rather than descriptive, signscance. In this review, we will first discuss evidence relating to the catecholamine hypothesis, and then briefly review the work on steroids and electrolyte metabolism. In the late 194Os, the Australian investigator Cade (1949) noted that lithium produced sedation in experimental animals. Conse145

146

JOHN M. DAVIS

quently, he tried lithium as a treatment for mania and found it had a dramatic therapeutic effect. It is probable that the effect of lithium in experimental animals represented toxicity, since the therapeutic dose of lithium produces few behavioral changes in the normal animal. Thus the clinical trial was carried out for the wrong reasons. Like many so-called chance discoveries, this one was not entirely serendipitous, since Cade had been interested in investigating the possibility that there might be toxic substances in the urine of manic patients which might be of etiological importance, and had been injecting urates, such as lithium urate, into animals. It was no accident that he applied his new sedative to manic patients. Lithium was, therefore, the first therapeutically active psychotropic drug introduced to western medicine, yet it unfortunately remained almost unknown for some 10 to 15 years (Gershon and Yuwiler, 1960; Gershon, 1968; Baastrup and Schou, 1967). Eight years after the discovery of lithium, two different types of drugs were found to reduce depression. Imipramine was tested in schizophrenic patients in the hope that it might prove to be a superior antipsychotic drug. It was found to be ineffective as an antipsychotic drug, but it did have antidepressant properties (Kuhn, 1958). Iproniazid ( a monoamine oxidase inhibitor) was first used in the treatment of tuberculosis, where it was noted clinically to have mood-elevating properties. This, in turn, led to a trial in depressed patients (Crane, 1956, 1957; Loomer et al., 1957). Reserpine, at this time, was used in high doses to treat hypertension, and it was observed that depression occurred in as many as 15%of hypertensive patients treated with this drug (Bunney and Davis, 1965). The existence of drugs that either caused or relieved manic and depressive disorders raised the question of whether a common denominator could be found in the action of these drugs. Such a common denominator might be a clue to the etiology of these syndromes and, hence, might be important in uncovering events in the pathogenesis of manic and depressive illnesses. The discovery of the therapeutic effects of the monoamine oxidase (MAO) inhibitors in the treatment of depression, and the discovery of the antihypertensive and antipsychotic effects of reserpine, stimulated work by basic scientists in elucidating the pharmacological action of these drugs. It was found that the most likely mechanism of action of the M A 0 inhibitors was the inhibition of the enzyme monoamine oxidase, an enzyme which breaks

147

BIOLOGY OF AFFECTIVE DISORDERS

down biogenic amines, such as norepinephrine and serotonin. Reserpine, on the other hand, interferes with the storage of these amines in intracellular storage vesicles. Hence, monoamine oxidase inhibitors raise the brain levels of biogenic amines in many species ( by inhibiting their degradation) while reserpine depletes the brain of these amines. The fact that these drugs either produce or relieve depression, and alter amines in opposite ways, suggested to many investigators that perhaps both mania and depression may be caused by alterations in biogenic amine metabolism (Jacobsen, 1964). The possibility that an alteration in amines may be involved in depression was mentioned in several animal studies involving amines (Rosenblatt et d.,1960; Everett and Toman, 1959). In addition, Pare and Sandler (1959), in line with these observations, began to treat depressed patients with 3,4dihydroxyphenylalanine ( dopa), a precursor of norepinephrine, and with 5-hydroxytryptophan (SHTP), a precursor of serotonin. II. The Biogenic Amine Hypothesis

Table I shows the biogenic amine hypothesis of depression in outline form. Reserpine lowers brain levels of serotonin and catecholamines, and often causes depression. The M A 0 inhibitors increase brain levels of amines by inhibiting their degradation, while the tricyclic antidepressants may make more norepinephrine or TABLE I EFFECTS OF DRUGSON AMINESAND DEPRESSION ~~

D w

Action

Result

Tricyclic antidepres sants

Block uptake of norepinephrine (NE) and 5-hydroxytryptamine (5-HT) Elevate brain levels of NE and 5-HT Increases turnover of NE Increases net uptake of NE and 5-HT, and reduces nerve stimulated release of NE and 5-HT Tryptamine autagonist Depletes brain of N E and 5-HT 8-Adrenergic blocker

Relieve depression

MA0 inhibitors ECT Lithium

Methysergide Reserpine Propranolol

Relieve depression Relieves depression Relieves mania

Relieves mania Causes depression Causes depression

148

JOHN M. DAVIS

serotonin available at the receptor site through inhibition of reuptake. Both of these latter classes of drugs are beneficial in depression ( Davis, 1965; Davis et al., 1968~).Electroconvulsive therapy ( ECT ) increases norepinephrine synthesis and utilization, and also relieves depression. Lithium may act by increasing the net reuptake or accumulation of norepinephrine or serotonin, thereby suppressing mania (Freedman, 1966; Schildkraut et al., 1968). During the last decade there have been rapid and impressive advances in our understanding of the biological functions of amines. The volume of information available is so great that it is impossible even to begin to summarize it here. Thus, the discussion will be limited to a brief introduction to the biology of these compounds. The evidence most relevant to each aspect of the biogenic amine hypothesis will be discussed; that is, evidence supplied by studies of reserpine, MA0 inhibitors, tricyclic antidepressants, ECT, lithium, and amphetamines. I l l . Synthesis and Metabolism of Biogenic Amines

Norepinephrine (NE) is a catecholamine located in the terminals of postganglionic sympathetic fibers where it functions as a neurotransmitter; it is also located in specifk structures in the brain, where it probably serves the same function. The synthesis of norepinephrine involves the hydroxylation of the amino acid tyrosine to 3,4-dihydroxyphenylalanine ( dopa), which is decarboxylated to 3,4-dihydroxyphenylethylamine ( dopamine ) , which, in turn, is hydroxylated (aliphatic chain) to norepinephrine in synaptic vesicles. The first step in the synthetic pathway, the tyrosine to dopa step, is the rate-limiting reaction in the sequence. Norepinephrine is stored in intracellular synaptic vesicles and is present in body fluids and cell sap. It is continually being synthesized, broken down, and released (Glowinski and Baldessarini, 1966). The breakdown of norepinephrine involves two pathways: (1) It may be initially deaminated by monoamine oxidase and then O-methylated (by means of the enzyme catechol-O-methyl transferase, COMT) to form 3-methoxy-4-hydroxymandelic acid (VMA ) or 3-methoxy-4-hydroxyphenylglycol( MHPG). (Armstrong et al., 1957; Axelrod et al., 1959; Maas and Landis, 1967; Schanberg et al., 1967a, 1968). ( 2 ) Circulating and released norepinephrine may be initially methylated by COMT to normetanephrine and then later be deaminated (Kopin and Gordon, 1962, 1963; Kopin, 1964).

BIOLOGY OF AFFECTIVE DISORDERS

149

Both norepinephrine and serotonin are conserved by a process in which nerves take up released norepinephrine (or 5-HT) so that it may be re-used. Thus, resynthesis and rebinding can keep up with constant nerve stimulation. The action of catecholamines is terminated by rebinding and by physical removal and destruction outside the nerve cell (Dengler et at., 1961). The functions of norepinephrine in the brain may be related to the drive or motivational systems (Stein, 1967; Poschel and Ninteman, 1963). Serotonin (5-hydroxytryptamine, 5-HT) is localized in areas of the brain which are concerned with emotions and sleep. It may act as either a neurotransmitter or as a neuromodulator. It is synthesized from tryptophan by hydroxylation to 5-hydroxytryptophan and decarboxylation to 5-hydroxytryptamine (serotonin, 5-HT) . Serotonin is deaminated by monoamine oxidase to 5-hydroxyindoleacetic acid (5-HIAA). Since depression and mania are disease states characterized by alterations in affect and drive and associated with marked sleep disturbances, and since amines are localized in and appear to function in areas related to drive, emotion, and sleep, additional supportive evidence is provided for the hypothesis that a relationship does exist between amines and these two disease states. IV. Antidepressants

A. MONOAMINE OXIDASE INHIBITORS The monoamine oxidase (MAO) inhibitors probably act by the irreversible inhibition of the enzyme monoamine oxidase, which intracellularly deaminates NE and 5-HT. This may lead to the accumulation of higher levels of these amines in the brains of many species, in the cell sap, and possibly in synaptic vesicles. The elevated levels may result in more transmitter substance being available at the receptor site. The evidence that these drugs act by inhibition of monoamine oxidase rather. than by other mechanisms is based primarily on the fact that there are many MA0 inhibitors, differing in chemical properties and structure; yet all of these drugs share the common properties of inhibiting monoamine oxidase and relieving depression. Evidence that therapeutic doses of MA0 inhibitors do indeed alter MA0 and increase brain amines (5-HT, NE, dopamine) in humans is provided by autopsy studies (Ganrot et d.,1962). Of interest is the observation that the

150

JOHN M. DAVIS

elevation of 5-HT occurs several weeks after the onset of treatment with MA0 inhibitors, the time at which the antidepressant effect occurs (Maclean et al., 1965). Clinical improvement in depressed patients has been correlated with the degree of monoamine oxidase inhibition achieved during drug administration (Shaffer et al., 1964; Dunlop et al., 1965; Feldstein et al., 1965). Unfortunately, assays of peripheral monoamine oxidase inhibition in man may not necessarily measure the degree of enzyme inhibition in the brain (Levine, 1966). Clinically effective doses of the monoamine oxidase inhibitors decrease urinary excretion of 3-methoxy-4-hydroxymandelic acid (VMA), a major deaminated metabolite of norepinephrine, and of 5-hydroxyindoleacetic acid (5-HIAA) (Sjoerdsma et al., 1958; v. Studnitz, 1959; Schildkraut et aZ., 1964; Schopbach et al., 1964). Hendley and Snyder (1968) measured the effects of MA0 inhibitors on MA0 activity and on net uptake into brain tissue, and noted a better correlation between clinical antidepressant activity and net uptake, than between clinical activity and MA0 activity. These findings raise the interesting possibility that there may be a unitary explanation for the mode of action of the different antidepressants. The rank order for effectiveness of MA0 inhibition in this system is phenelzine, tranylcypromine, isocarboxazid, pargyline, nialamide, and iproniazid. Tranylcypromine is a highly effective inhibitor of metaraminol net uptake, while phenelzine and pargyline are moderately effective inhibitors of uptake. Isocarboxazid, nialamide, and iproniazid are ineffective. If iproniazid is an effective antidepressant, then at least one antidepressant exerts its action by a mechanism different from uptake inhibition. There is evidence that iproniazid is an effective antidepressant ( Davis, 1965; Davis et al., 1968c), although the clinical evidence is not definitive. Tranylcypromine is an effective antidepressant, and it is tempting to speculate that its uptake-inhibiting properties may play a role in its mechanism of action.

B. TRICYCLIC ANTIDEPRESSANTS The potentiation by imipramine and similar drugs of the action of NE in a variety of systems has been explained (Sigg, 1959) by the observation that imipramine blocks the uptake of NE into the nerve ending, thereby leaving newly released NE at the receptor

BIOLOGY OF AFFECTIVE DISORDERS

151

site for a longer time (Dengler et aZ., 1961; Hertting et al., 1961).A similar mechanism exists for the neuronal uptake of serotonin. In work using a preparation of broken off nerve endings, the author and Dr. Colburn (Davis et al., 1968a) have demonstrated that nerve endings from rat brain concentrate NE and 5-HT against a 50 (or more) to 1 concentration gradient, a process which is essentially the re-uptake mechanism. Similar work in brain slices has also demonstrated that brain tissues can take up serotonin in a manner similar to NE (Blackburn et al., 1967; Ross and Renyi, 1967). Using the analogy of the transport-carrier system to enzymatic kinetics, we have studied the net uptake of NE, 5-HT, and metaraminol into brain nerve ending particles, using different substrates (Colburn et al., 1968). The tricyclic antidepressants inhibit uptake in a competitive manner in this system (Davis et al., 1968b). Evidence that the inhibition of uptake of NE seen in synaptosomes, in peripheral systems, and in brain slices also occurs in the living brain comes from studies based on the ventricular or intracisternal injection of labeled amines (Glowinski and Axelrod, 1965, 1966). When administered prior to the injection of labeled NE, imipramine increases brain levels of labeled normetanephrine, the metabolite which may reflect the level of activity of norepinephrine at adrenergic receptors (Schanberg et d.,196%; Schildkraut et al., 1967). In the interpretation of studies based on the administration of labeled compounds, care must be taken to determine that the amine is localized in those structures where it occurs and functions endogenously. If a substance is localized in tissue where it does not normally occur, such as nonneuronal tissue, studies using it as a tracer may not reflect physiological events. It should be noted that tricyclic drugs do increase the urinary excretion of O-methylated amines and decrease the excretion of VMA (Schildkraut et al., 1964, 1966a; Schildkraut, 1965). Surprisingly, platelets, structurally and functionally, have many similarities to nerve endings, and may in fact provide a more accessible model of the nerve ending. In platelets drawn from patients under treatment with imipramine, Murphy et al. (1969) showed that the tricyclic drug decreases uptake of 5-HT into platelets in comparison with platelets drawn from controlled drug-free patients (Davis et aZ., 196813). Thus, there is evidence from a study on depressed

152

JOHN M. DAVIS

patients that the tricyclic drug dose given does function in the manner expected from animal studies. V. Reserpine-Induced Depression

A significant percentage of hypertensive patients treated with high doses of reserpine develops depressions similar to those of endogenous depressive reactions ( Achor et al., 1955; Muller et al., 1955; Lemieux et al., 1956; Harris, 1957; Ayd, 1958; Bemstein and Kaufman, 1960). We have recently reviewed the literature and documented 187 cases of reserpine-induced depression ( Bunney and Davis, 1965). In some studies, the incidence of these depressions was as high as 15%.It was shown that hypertensive patients treated with reserpine showed statistically si&cantly more depressive reactions than did similar patients who were treated with other antihypertensive drugs. For example, in a study by Lemieux et al. ( 1956), reserpine-induced depressions occurred in 30/195 cases (IS%),whereas no cases of depression were seen in patients treated with other antihypertensive drugs. In addition, it was found that the incidence of reserpine-induced depression was dose-related, with most cases occurring in patients receiving 0.75 mglday or more of reserpine. Generally, the onset of depression occurred between 1 and 7 months after the initiation of reserpine therapy, although it may occur as early as 1 week or as late as 14 months. The depression usually subsides when the patient is taken off reserpine and recurs when reserpine therapy is resumed. Tetrabenazine, a drug similar to reserpine in reducing brain norepinephrine and serotonin, has also been reported to be associated with depression ( 0. Lingjaerde, 1963) . Reserpine depletes the brain of serotonin and norepinephrine. A variety of evidence suggests that reserpine disrupts the storage process by which the intracellular vesicles store norepinephrine. Norepinephrine is released into the cell sap where it is destroyed by mitochondria1 monoamine oxidase; hence, most of the transmitter is inactivated intracellularly, without exerting any physiological effect (Kopin and Gordon, 1962, 1963). In animals, reserpine produces a syndrome of sedation. This has been suggested to be analogous to depression in man (Sulser et al., 1964, 1966, 1967). Pretreatment with tricyclic drugs will convert this sedation to an overactivity, catecholamines being necessary for the reversal. In addition, the behavioral effect is related to rate rather than to

BIOLOGY OF AFFEcL?vE DISORDERS

153

amount of norepinephrine release (Sulser et al., 1964, 1966, 1967). The sedative effect of reserpine is not correlated with total brain amine levels, but rather with the relative inability of the brain to accumulate amines (Haggendal and Lindqvist, 1964). When the hydroxylation of tryptophan is inhibited by p-chlorophenylalanine, brain serotonin levels are reduced to 10%of normal without marked change in NE content, a condition which does not generally lead to sedation (Koe and Weissman, 1966). But if reserpine is added, and the brain further depleted of serotonin, the animals do become sedated. On the other hand, if the animals are specifically depleted of NE by the use of a synthesis inhibitor, a-methyl-p-tyrosine, sedation is observed (Weissman et al., 1966). It must be stressed that any single interpretation of evidence in such a complicated field as catecholamine and serotonin metabolism must be viewed with caution; consequently, several interpretations of reserpine sedation have been offered (Brodie et al., 1966). Reserpine sedation can be reversed in animals by dopa (Carlsson et at., 1957; Carlsson, 1964). It has been shown by Degkwitz et al. (1960) that patients treated with reserpine become tired and lethargic, and this feeling of tiredness or lethargy can be reversed for several hours by treatment with L-dopa. The evidence suggests that NE deficit is involved in reserpine sedation; however, it cannot be assumed that reserpine sedation is a valid model of depression. Because reserpine releases both NE and S H T , it does not provide evidence which could relate the etiology of depression to a deficit of any single amine. Although catecholamines are involved in the reserpine-reversal test, it may be dopamine, rather than NE, which produces the behavioral effects (Creveling et al., 1968; Everett and Wiegand, 1962). In contrast, the synthesis inhibitors, a-methyl-p-tyrosine and p-chlorophenylalanine, are specific in their action, in the sense that the former inhibits the synthesis of norepinephrine and the latter, serotonin. They have been used in man in the treatment of pheachromocytoma and carcinoid syndrome, respectively. Although behavioral changes have been noted following the use of both drugs, neither drug has consistently produced a syndrome identical to clinical depression (Engelman et al., 1967, 1968; Cremata and Koe, 1966; Sjoerdsma et aZ., 1965). But this cannot be taken as definitive evidence that amines are not involved in depression, since there may not have been complete inhibition of synthesis at the doses used in human

154

JOHN M. DAVIS

subjects. Careful psychiatric and psychological study which would characterize quantitatively and qualitatively the mental effects of these drugs would be helpful. The experimental drug Ro 4-6861 lowers serotonin and 5-hydroxyindoleacetic acid concentrations in the brain without altering norepinephrine or dopamine concentrations. If depression is associated with low serotonin levels, the administration of this drug should result in a worsening of the depressive condition. However, in a preliminary study done by van Praag et al. (1968),it was suggested that this drug may be effective in treating depression, a finding which would be contrary to the serotonin-deficit hypothesis of depression while still consistent with the catecholamine hypothesis. Only preliminary information is available at present, and we must eagerly await further data. The drug a-methyldopa has been used to treat hypertension in man. This drug depletes the brain of norepinephrine through a mechanism which is different from that of reserpine. The metabolites of a-methyldopa, a-methyldopamine, and a-methylnoradrenaline are false transmitters and displace norepinephrine from nerve endings. a-Methyldopa has been noted to cause depression in a few patients (Bunney and Davis, 1965). VI. Electroconvulsive Shock Therapy

Radioactively labeled norepinephrine was intracisternally injected prior to electric shock by Schildkraut, Schanberg, Breese, and Kopin (1!367), who observed that levels of labeled NE in the brain are decreased and normetanephrine levels increased in shocked animals relative to control values. This would suggest that the rate of neuronal discharge of NE in the brain may be increased by electric shock. Furthermore, Kety et al. (1967), using techniques based on the measurement of specific activity of labeled amines (injected intracisternally) 24 hours after a series of electroshock treatments, found a more rapid fall of specific activity in the postECT animals as compared to controls (no ECT). This suggests that following ECT there is an increase in the synthesis and utilization of NE. ECT is the most effective therapy for severe depression (Davis, 1965; Davis et al., 196&). VII. Lithium

Schildkraut, Schanberg, and Kopin (1966b) found that lithium administered to animals prior to injection of intracisternal NE

BIOLOGY OF AFFECL?VE DISORDERS

155

decreased the levels of labeled normetanephrine and increased the levels of tritiated deaminated catecholamine metabolites in the brain (Schanberg et al., 196%). It is noteworthy that the pharmacological studies of the effects of lithium treatment on amines were undertaken as a sole result of the clinical effects of lithium on mania. Colburn, Robinson, and the author (Davis et al., 1968a) showed that lithium did not alter M A 0 levels in brain nerve endings, or in brain mitochondria, nerve membranes, or subcellular fragments. One hypothesis on the mechanism of action of lithium is that it alters the transport of norepinephrine and/or serotonin through the nerve membrane, either by increasing the uptake of amines or by decreasing their release so that there is a greater net movement of biogenic amines from outside to inside the nerve. In a functional sense, this would be a change in the opposite direction from that produced by imipramine, although, of course, the exact mechanism of lithium action may be quite different from that of imipramine. The author and Dr. Colburn isolated broken-off nerve endings from brains of animals pretreated with lithium and compared these to nerve endings from brains of control animals (Colburn et al., 1967). Initially it was found that lithium pretreatment increased the net uptake of NE into brain. In a subsequent experiment, it was demonstrated for nerve endings in a sodium-potassium-magnesium buffered system that a small increase in net uptake of NE, metaraminol, and serotonin occurs after lithium pretreatment. The increased net uptake of NE seen in lithium-pretreated tissue is present in nerve ending particles which have also been treated with reserpine. One can hypothesize that, even though reserpine, by interfering with storage, reduced the total synaptosomal accumulation of NE, the increased accumulation in the lithium-pretreated synaptosomes is present at all concentrations of reserpine. Corrodi et al. (1967) observed in lithium-pretreated animals an enhancement of the decrease in norepinephrine after the inhibition of tyrosine hydroxylase. Stern et al. (1967) found an increase in norepinephrine turnover following lithium pretreatment using the a-methyltyrosine technique. Katz, Chase, and Kopin (1968) incubated brain slices from both lithium-pretreated and untreated animals with labeled norepinephrine or serotonin, and subjected these slices, as well as unincubated slices, to a mild electrical stimulation. Lithium inhibited the stimulation-induced release of amines. If NE net uptake is increased, either by an increase in

156

JOHN M. DAVIS

spontaneous release or by a decrease in stimulation-induced release, more norepinephrine may be exposed to intracellular destruction; hence there could be an increase in turnover. Matussek and Linsmayer (1968) studied the interaction between lithium pretreatment and the hyperactivity produced in rats by the reserpine reversal test. Lithium pretreatment prevents the hyperactivity induced by a combination of desipramine and Ro 4-1254 ( a reserpine-like drug ) and prolongs amphetamine-induced stereotyped behavior. Although more information is needed for an understanding of the role of lithium in amine transport and metabolism, the interaction shown here provides another link in the implication of amines in the etiology and treatment of affective disorders. VIII. Amphetamine

Acute administration of amphetamine causes elation in man and hyperactivity in animals, an effect which, in animal studies, has been shown to depend on the presence of catecholamines in the brain ( Weissman et al., 1966). Possibly because tachyphylaxis develops, amphetamine is not an effective antidepressant in man (General Practitioner Research Group, 1964;Davis et aZ., 1968~).In large doses, it depletes the brain of NE, and prolonged use of this drug is sometimes followed by depression and fatigue (McLean and McCartney, 1961; Moore and Lariviere, 1963; C. B. Smith, 1965). J. Fawcett and his collaborators (1968)have noted that a behavioral response of elation to a 3-day trial of amphetamine predicts which depressed patients will respond to tricyclic drugs. The action of amphetamine on a pharmacological level is quite complex, and the drug may have several mechanisms of action. It has been shown to decrease the rebinding of catecholamines and 5-HT, to release catecholamines, and possibly to have a direct action on the receptor site, as well as to act as an inhibitor of monoamine oxidase in certain situations. IX. Central Receptors and Depression

W. G. Dewhurst and Marley (1964,1!365a,b,c) have performed a series of experiments to identify brain receptors for amines using classic techniques such as specific antagonism and cross tachyphylaxis, and using the young chick, whose blood-brain barrier allows passage of hydroxylated amines during the &st few weeks

BIOLOGY OF A F F E m DISORDERS

157

of life (Allen and Marley, 1967; Key and Marley, 1962; Marley, 1!368a,b,c; Marley and Morse, 1967; Marley and Stephenson, 1968; Vane, 1961; Vane et al., 1961). They classify cerebral amines into two main functional groups: the excitant (type A ) amines and the depressant (type C ) amines (W. G. Dewhurst, 196813). Type C amines have the general formula R-CH( OH) CH,NH( CH, ), where R is a water-soluble group, such as a catechol group. Type A amines appear to act on a specific receptor similar to the peripheral tryptamine receptor, whereas type C amines may act on a different receptor, possibly similar to the peripheral alpha receptor but not necessarily identical to it. Methysergide is a specific competitive antagonist of type A amines. Type C receptors are not influenced by beta-adrenergic blockers, antihistamines, or anticholinergic agents. W. G. Dewhurst ( 1 W a ) and Hagkovec and SouEek (1968) independently treated manic patients with methysergide and observed a dramatic response within 24 to 48 hours. In addition to suggesting that methysergide may be an important addition to the therapeutic armamenterium, this finding has considerable theoretical significance in that it provides another body of data implicating amines in affective disorders. As the central pharmacology of methysergide is better understood, it may provide clues as to the nature and quality of the defect involved in mania. Weil-Malherbe ( 1968) has critically discussed the theories of Dewhurst, noting that the effect of the amines may be unphysiological since they may reach sites where they do not normally occur, may release other amines, or may act as false transmitters. Further research is needed for a full understanding of central receptors; however, it is of great interest that the theoretical notion in part led to the suggestion of a new drug for the treatment of mania. X. Catecholamine Metabolism in Depressed and Manic Patients

Neither 5-HT, NE, nor the norepinephrine metabolite normetanephrine, passes in appreciable amounts through the bloodbrain barrier, and hence studies of unchanged urinary 5-HT, NE, or normetanephrine can tell little of what actually happens in the brain. Because it is possible that any change in brain NE metabolism in depression may be related to body changes, it is relevant to study NE excretion in depressed or manic patients; however, suita-

158

JOHN M. DAVIS

ble caution is indicated in the interpretation of results from this type of study. In longitudinal studies of patients with manic-depressive disorders, urinary excretion of norepinephrine and epinephrine was observed to be greater during the manic phase than during the depressive phase ( Strom-Olsen and Weil-Malherbe, 1958; Shinfuku et al., 1961). In a longitudinal study of manic patients receiving lithium, it was found that during the manic phase, larger amounts of catecholamines and their metabolites were excreted, compared with periods of relatively normal behavior. These changes seem to relate to changes in the degree of mania rather than to the lithium treatment (Davis et aZ., 1968a). In a cross-sectional study of manic patients by Bergsman ( 1959), norepinephrine and epinephrine excretions were elevated above control values, but resting catecholamine excretion was not found to be significantly altered in depressed patients, although these patients did have a lower catecholamine response to insulin stress than did control subjects. Sloane et al. (1966) found on admission in comparison to discharge a decrease in NE and epinephrine excretion in depressed patients, but a group of 5 manic patients had significantly higher levels of dopamine excretation, both at admission and at discharge, than did patients in other diagnostic categories. The author, in collaboration with Bunney, Weil-Malherbe, and Smith (Bunney et al., 1967), found that psychotic depressives excreted larger amounts of catecholamines and their metabolites, particularly norepinephrine and VMA, than did neurotic depressives. Normetanephrine and metanephrine were reported to be elevated in association with agitated behavior in a study of depressed and other patients (Nelson et al., 1966). Furthermore, the excretion of norepinephrine and normetanephrine was markedly increased with the occurrence of psychotic delusions in depressed patients (Schildkraut et al., 1965). Sachar et al. ( 1963) reported increased catecholamine excretion during acute psychotic turmoil in schizophrenic patients. Hence, turmoil, psychosis, or muscular activity, as well as affective state, may be associated with alterations in the excretion of norepinephrine and metabolites. Rosenblatt and Chanley (1965) infused radioactively labeled norepinephrine into a group of depressed patients and observed an increase in the ratio of amines to deaminated metabolites in the patients classified as manic-depressive, in contrast to that observed in patients with

BIOLOGY OF AFFECTIVE DISORDERS

159

reactive depressions or other disorders. The interpretation of urinary studies is somewhat problematical at this time (Karki, 1956). Since it has been postulated that depression is associated with low brain NE levels, and since it has been suggested that significant quantities of brain catecholamines are metabolized to 3-methoxy4-hydroxyphenylglycol ( MHPG ) , the amount of MHPG excreted in the urine may reflect metabolism of NE in brain (Maas and Landis, 1967). In a pilot study, it was found that the urinary levels of MHPG were statistically significantly lower in a group of seriously depressed patients than in controls (Maas et al., 1968). The significance of these results is dependent upon the demonstration that MHPG levels in urine do indeed reflect NE metabolism in brain. XI. lndole Metabolism in Depressed Patients

Rodnight (1961) found that the urinary excretion of tryptamine was significantly decreased in patients suffering from depression in comparison with both normals and schizophrenic controls. A. Coppen et al. (1965b) found a decreased excretion of urinary tryptamine during depression in comparison with that observed after recovery. Furthermore, after a large tryptophan load (50 mglkg), they observed the excretion of lower amounts of tryptamine during depression in comparison with recovery, a finding consistent with the hypothesis that urinary tryptamine is produced by decarboxylation of tryptophan in the kidney. A. Coppen and his collaborators (1965a) injected S-hydr~xytryptophan-l~C intravenously and measured the rate of expiration of radioactive carbon dioxide. In initial studies, it seemed that the rate of decarboxylation was slowed in depression, but subsequent studies have shown no consistent trend (A. Coppen, 1967). XII. Biogenic Arnine levels in Human Brain

The most direct test of the biogenic amine hypothesis as related to levels would be the measurement of levels in human brain. Shaw et al. (1967) found brains from suicides to have lower 5-HT levels than did brains from controls. The author, in collaboration with J. N. Davis, Bourne, and Colburn, found suicide brains to have lower levels of 5-HIAA (the breakdown product of 5-HT) than did controls, in brains supplied by Bunney, Coppen, and Shaw (Bourne et d.,1968). No differences in 5-HT and NE were found in this

160

JOHN M. DAVIS

study; of course, these results are subject to the considerable limitation of autopsy material. In addition, two research groups have independently found depressed patients to have lower 5 H M levels than controls (Dencker et al., 1966b; Ashcroft et al., lSeS), but no marked difference in NE levels were observed (Dencker et al., 1966a). Because brain MA0 may increase with age, agematched controls may be necessary for a definitive study (Davis et aL, 1968a). XIII. Experimental Drugs and the Biogenic Amine Hypothesis

The critical link in the chain of evidence supporting the amine hypothesis is the consistency with which drugs that alter amines in the brain also affect depression. Critical evidence which would tend to disprove the hypothesis would arise from the discovery of drugs that had the same actions on brain biogenic amines as tricyclic drugs, yet failed to be effective clinical antidepressants. EXP 561; was predicted on the basis of preclinical screening to be an effective antidepressant by the use of such tests as the potentiation of peripheral exogenous catecholamines and antagonism of reserpine-like syndrome. The drug was not found to be an effective antidepressant in a trial in 10 depressed patients (Gershon et al., 1968) Similarly, BL-KR 1402potentiated peripheral exogenous catecholamines and prevented reserpine-induced ptosis, but failed to be effective clinically in 10 depressed patients (Hekimian et al., 1968). Likewise, gamfexine showed a similar preclinical screening profile and failed to benefit depressed patients (Gershon et al., 1967). Similar results are said to have been observed with cyprolidol (Hekimian et al., 1968).It should be remembered that it is di5cult to establish clinical efficacy for antidepressants; hence, it is possible that a study using a small sample of patients could yield a false . example, the drugs might not negative (Davis et al., 1 9 6 8 ~ )For have been used in the proper dose. Furthermore, the early drug evaluations of gamfexine and EXP 561 have not been uniformly negative, since a few patients, but not necessarily a significant number, have been helped by these drugs (Lehmann et al., 1967; Schiele et al., 1966). Furthermore, even though these drugs do have the animal preclinical screening profiles of antidepressants, they

.

4-phenyl-bicycle( 2,2,2 )octan-l-amine hydrochloride monohydrate. a

N-acetonyl-N,N-dimet ylbenzylammonium chloride.

BIOLOGY OF AFFECTIVE DISORDERS

161

might not necessarily alter amine function in the brain of man in ways identical to those of imipramine. A second way to use drugs to test the biogenic amine hypothesis is by the administration of precursors. The author has given up to 400 mg iv of L-dopa to depressed patients without any therapeutic success, a result consistent with the experiences of a number of authors (Klerman et al., 1963; Pare and Sandler, 1959; Schildkraut et al., 1963; Turner and Mcrlis, 1964). 5-Hydroxytryptophan, the precursor of 5-HT, also fails to lessen depression consistently, although both dopa and 5-HTP have been reported to help a few patients (Ingvarsson, 1965; Matussek et al., 1966; Kline and Sachs, 1963; Kline et aZ., 1964; Pare and Sandler, 1959). Since depression is a disorder that often improves spontaneously, it is necessary to show a drug superior to a control substance before definitive conclusions can be drawn. Furthermore, it may be difficult to reach therapeutic levels in the brain since dopa does not pass through the blood-brain barrier easily, being in part metabolized in the blood vessels (Bertler et al., 1966). Tryptophan, when given with MA0 inhibitors, can shorten the response latency, and when given alone, can lead to improvement in depression, a finding consistent with the serotonin-deficit hypothesis of depression (A. Coppen et al., 1963, 1967; Pare, 1963). XIV. Electrolytes

Water and electrolyte balance in depressed patients has been subject to investigation for many years, with varying degrees of control of relevant variables (Altschule, 1953; Ueno et al., 1961; Cade, 1962; Klein and Nunn, 1945; Klein, 1950; Strom-Olson and Weil-Malherbe, 1958; Crammer, 1959; A. Coppen and Shaw, 1963; Gibbons, 1963). Russell (1960) performed a controlled balance study of depressed patients treated with ECT and found no statistically significant change in water, sodium, or potassium balance with recovery, although nonsignificant trends were in favor of a slight loss of sodium (5.5 mEq/day) and retention of water (20 mEq/ day). Gibbons (1960) found that the 24-hour exchangeable sodium decreased in depressed patients with recovery. A similar trend was reported by A. Coppen and Shaw (1963), but the differences in this study were statistically significant only in the female patients and after corrections were made for body weight. The “total ex-

162

JOHN M. DAVIS

changeable sodium” did not change with recovery (A. Coppen et al., 1962). A. Coppen and Shaw (1963) have derived a method of calculating values for extracellular and residual sodium (intracellular plus exchangeable bone sodium ) . The measured volume of distribution of isotopic bromine ( 82Br)is multiplied by a correction factor (0.9) to estimate extracellular space. This number is then multiplied by the measured plasma sodium level to give an estimated value for extracellular sodium. Subtraction of this value from measured exchangeable sodium gives residual sodium. Thus the amount of residual sodium is a small number representing the difference between two large numbers ( total exchangeable and extracellular sodium). Therefore, any error or extraneous factor affecting either of the two large values could profoundly influence the value for residual sodium (e.g., the entrance of bromine into cells or into the GI tract, or hormonal or dietary changes of unlabeled body sodium in any compartment). By the use of this technique, A. Coppen et al. (1966) found that depressed patients had elevated residual sodium in comparison with the values measured in recovery. The residual sodium values were even higher in manic patients. There was little change in potassium between sickness and recovery, although the possibility exists that depressed patients, both in relapse and in remission, may have a constantly altered potassium or magnesium metabolism (Gibbons, 1960; A. Coppen and Shaw, 1963; Cade, 1964). A number of investigators have reported that depressed patients showed decreased extracellular fluid or total body water during depression, in comparison with measurements made after recovery ( Altschule and Tillotson, 1949; Dawson et al., 1956; Brown et al., 1963; A. Coppen and Shaw, 1963; Hullin et al., 1967a). A. Coppen et al. (1966) found that total body water and its partition between intracellular and extracellular water did not change significantly in mania, but Hullin et al. (196%) observed total body water to be greater in the manic phase of cyclic manic-depressive disease. The increased sodium retention in depression may be secondary to the elevated cortisol observed in depression (Gibbons, 1963; A. Coppen and Shaw, 1963; Durell and Schildkraut, 1966; Baer et al., 1969). Further work is needed to define more reliably how consistent and free of measurement artifacts these alterations in

BIOLOGY OF AFFECTIVE DISORDERS

163

electrolytes and water are, to identify more precisely what and where the abnormalities are, and to determine whether these abnormalities are primary to disorders of affect or secondary to changes in glucocorticoids and mineralocorticoids, or in adrenergic functioning ( Shaw, 1966). XV. Steroids in Depression

Clinical studies indicate that a significant proportion of patients with Cushing’s syndrome or under long-term treatment with ACTH or cortisone show mental changes such as euphoria, depression, suicidal tendencies, or overt psychosis ( Borman and Schmallenberg, 1951; Stefanini et al., 1950; Spillane, 1951; Clark d al., 1952, 1953; Lidz et al., 1952; Rome and Braceland, 1952; Trethowan and Cobb, 1952; Glaser, 1953; Goolker and Schein, 1953; Rees, 1953; Fleminger, 1955; Quarton et al., 1955; Cobb, 1960; Pearson and Eliel, 1950; Galdston et al., 1951). Addison’s disease is associated with symptoms such as depression, apathy, inability to concentrate, insomnia, drowsiness, irritability, apprehension, and sleep disturbances (Engel and Margolin, 1941; Cleghorn, 1951). The association of mental changes with abnormalities of adrenal function raises the question of the role of glucocorticoids in affective disorders. In 1949, Reiss et al. reported an increase in P-hydroxy-17-ketosteroids in the depressive phase. Rizzo et al. (1954) studied a cyclothymic patient who, during a period of manic behavior, had a low glucocorticoid excretion which returned to normal upon recovery. Furthermore, during the manic period, ACTH challenge raised glucocorticoid values to normal, as did two psychological situations which were quite threatening to the patient’s self-esteem. Bryson and Martin (1954) studied a cyclic manic-depressive patient who showed increased urinary 17-ketosteroids during the depressed phase, with decreased levels during the manic phase of illness. In the first of the larger studies, subsequent to these initial case studies, high plasma 17-hydroxycorticosteroid levels were found in patients with depression (Board et al., 1956, 1957), particularly in those patients who showed inability to cry, intense stress associated with their depressive affect, or extensive personality disintegration. Increased plasma steroid levels ( 11-hydroxycorticosteroids,17hydroxycorticosteroids, and cortisol ) have been generally observed

164

JOHN M. DAVIS

in depression, returning to normal after recovery (Gibbons and McHugh, 1962; P. S. Lingjaerde, 1964; Anderson and Dawson, 1965; Bridges and Jones, 1966; Doig et al., 1966; McClure, 1966a; Brooksbank and Coppen, 1967; Hullin et al., 1967; Sachar, 196%; Crane and Wolfman, 1960; Ferguson et al., 1964; Pryce, 1964; Stenback et al., 1966; Cameron and Dawson, 1967; Curtis et a?., 1960; Fullerton et al., 1968; Michael and Gibbons, 1963). By using an isotope dilution technique, Gibbons (1964, 1966) found that the mean cortisol secretion rate was elevated during depression and fell to normal after recovery. Jakobson et al. (1986) noted an increased response of 11-deoxycorticoids to the metyrapone test in depression, which suggests increased ACTH secretion. The diurnal curve of steroid excretion seems to persist in many depressed patients, and shows either a proportionate increase or a slight tendency toward a greater relative increase in the early morning hours (Bridges and Jones, 1966; Doig et al., 1966; McClure, 1966a,b; Knapp et al., 1967). Occasional flattening or reversal of the diurnal curve has been observed, though, in a few depressed patients (Sakai, 1960; Brooksbank and Coppen, 1967; J. A. Fawcett and Bunney, 1967; Green, 1967; Conroy et al., 1968; Butler and Besser, 1968). Dexamethasone, particularly when given in high doses, is able to suppress corticosteroid excretion in many depressed patients; however, cases have been reported of dexamethasone, particularly in low doses, failing to suppress corticosteroid excretion (Gibbons and Fahy, 1966; J. A. Fawcett and Bunney, 1967). Recently Carroll et al. (1968) and Butler and Besser (1968) reported a group of depressed patients who were resistant to dexamethasone suppression of 11-hydroxycorticosteroids. These findings are consistent with an alteration in the CNS regulation of steroid production in depression. Longitudinal studies provide a more precise understanding of the correlation between depression and steroids. In longitudinal studies, some but not all depressed patients show a good correlation between day-to-day variations in urinary excretion of 17hydroxycorticosteroids and psychological ratings of depression and anxiety (Bunney et al., 1 9 6 5 ~ ). Sachar et al. (1963; Sachar, 1967a,b) found that elevations in urinary 17-hydroxycorticosteroid excretion occurred in association with acute psychotic turmoil and disruption of ego defenses in acute schizophrenic patients and in depressed patients. Bunney et al. (196%) isolated episodes of depressive crisis associated with a

BIOLOGY OF AFFECTIVE DISORDERS

165

very high excretion of 17-hydroxycorticosteroids; these episodes were characterized by stressful events which challenged the patient’s defenses and frequently resulted in marked ego disintegration. Sachar et al. (1967) observed that elevated levels of 17hydroxycorticosteroids occurred during those periods of the therapy when the patients experienced confrontation of loss. Elevated excretion of 17-hydroxycorticosteroids has been reported by Bunney and Fawcett (1965) to occur in patients prior to suicide. There seems to be no consistent pattern of steroid secretion in manic patients, with some studies finding low steroid excretion during the manic phase (Reiss et al., 1949; Bryson and Martin, 1954; Rizzo et al., 1954; Bunney and Hartmann, 1965; Bunney et al., 1965a), some finding normal levels (Brooksbank and Coppen, 1967; P. S. Lingjserde, 1964), and others finding elevated levels ( Schwartz et al., 1966; Bliss et al., 1956). The present evidence suggests a correlation among urinary steroid excretion, mental distress, and ego disintegration. Although their cortisol excretion has been found to be high generally, it has been suggested that depressed patients suffer from a relative adrenal insufficiency. Kurland (1964a,b) found that 17ketogenic steroids correlated with clinical ratings of depression, and Jakobson et al. (1966) found an elevation of compound S in depressed patients. Kurland (1965) reported some success in treating depressive affect with prednisone. Furthermore, a patient with Addison’s disease who developed an agitated psychotic depression and was unresponsive to twelve electroconvulsive shock treatments responded dramatically to ECT after treatment with hydrocortisol (Cumming and Kort, 1956). McClure and Cleghorn (1969) note in an exploratory study of seventeen patients that pretreatment for a week with dexamethasone shortens the lag period before the therapeutic response to MA0 inhibitors or tricyclic drugs. The steroid changes in depression have generally been thought to be secondary to mental distress and ego disintegration; however, the work of McClure and Cleghorn ( 1969), Kurland (1965),and Cumming and Kort (1956) raise the hypothesis of an etiological role for the steroids. XVI. Discussion

There are a number of lines of circumstantial evidence implicating amines in depression, but when it is reviewed critically, the

166

JOHN' M. DAVIS

evidence is indirect and indefinite. The antidepressants are clinically neither specific nor highly effective. They do benefit patients to a degree greater than placebo, but fail to help, or help only partially, a great many patients. This may be due to the possibility that depression is not a single diagnostic entity. Mania and depression are not necessarily bipolar. The cause of mania may be depression associated with an alteration of 5-HT and not NE levels, or vice versa. Such a disturbance of amines could cause depression or, on the other hand, it could be secondary to depression. Either amine may play a role in neurotransmission in inhibitory or excitatory fibers. Evidence related to the MA0 inhibitors, tricyclic drugs, reserpine, and lithium fails to implicate clearly either NE or 5-HT. The reduced levels of 5-HT or 5-HIAA in brain and/or CSF, and the therapeutic efficacy of tryptophan and methysergide, implicate serotonin. The observation that propranolol can cause depression implicates NE. It may be inappropriate to designate a preferred amine since most of the evidence is not definitive; for example, 5-HTP may release NE. The amine hypothesis, in its most simple form, is that depression is associated with low levels of biogenic amines in the brain, while mania is associated with high levels. Thus, for example, the catecholamine hypothesis has been stated to equate low levels of brain norepinephrine with depression, and high levels with mania. In this sense, the norepinephrine and the serotonin hypotheses could be said to be competing hypotheses. It is certainly heuristically useful to have explicit hypotheses about depression; however, alternate views exist, e.g., that an alteration in amine function or metabolism is involved, which may or may not be related to either total brain levels or levels at the synapses. Conceivably there could be a defect in one or more places in the chain of events which make up the functioning of biogenic amines. The defect could be in synthesis of amines, storage, release, neuronal re-uptake, axonal transport, or the receptor site. It could involve norepinephrine, serotonin, dopamine, or possibly other amines or neurotransmitter substances. It need not necessarily be a disturbance of any one amine, but could be a disturbance of another or of a family of amines, or of the balance between different amines. Most of the psychotropic drugs alter dopamine, which may play a role in depression. Also, most of the psychotropic medications have anticholinergic properties, which suggests that cholinergic systems are important in de-

BIOLOGY OF AFFECIII% DISORDERS

167

pression. The balance between neurotransmitters may be important. The drugs may act on many transmitters and produce their antidepressant effect through action on several rather than one. They may alter the balance of release, re-uptake, and storage, rather than simply cause depletion. Evidence for the functional amine hypothesis, rather than the depletion hypothesis, comes from the observation that reserpine, in certain situations, can lessen depression in spite of its depleting effect. Specifically, it has been shown that reserpine treatment, when added to a previous course of tricyclic drugs, can be of benefit in depression (Dick and Roch, 1967; HaZkovec and RyZBnek, 1967; Poldinger, 1963). Reserpine may release amines which tend to remain in the synaptic cleft because of the inhibition of re-uptake. It is difficult to integrate the electrolyte findings at this time. Colburn and I have found that physiological levels of Na+ and K’ are necessary for optimal amine uptake, but the changes in electrolytes necessary to alter uptake are beyond physiological limits. The steroid changes are of interest because, in animals, high cortisol levels can deplete the brain of 5-HT, which in turn could cause depression. In that case, the steroid change, which could be triggered by psychological events, could in turn alter central 5-HT synthesis and cause depression. It has been found that triiodothyronine also shortens the lag period before therapeutic response to tricyclic drugs, so that thyroid hormone, as well as steroids, is involved in some depressions (Prange et al., 1967). There are many possible interrelationships between the various abnormalities alleged to be related to manic-depressive disease. However, the actual relationship, if any, remains a question for continued chemical research. There are many promising leads at this time, but the depressive disorders remain diseases in search of an etiology. REFERENCES Achor, R. W. P., Hanson, N. O., and GifFord, R. W., Jr. (1955). J . Am. Med. Assoc. 159, 841. Allen, D. J., and Marley, E. (1967). Brit. J . Pharmacol. 31, 290. Altschule, M. D. (1953). “Bodily Physiology in Mental and Emotional Disorders.” Grune & Stratton, New York. Altschule, M. D., and Tillotson, K. J. (1948).Am. J. Psychiat. 105, 829. Anderson, W.McC., and Dawson, J. (1965). J. Psychosomat. Res. 9, 237. Armstrong, M. D., McMillan, A., and Shaw, K. N. F. (1957). Biochim. Biophys. Acta 25, 422.

168

JOHN M. DAVIS

Ashcroft, G. W., Crawford, T. B. B., Eccleston, D., Shaman, D. F., MacDougall, E. J., Stanton, J. B., and Binns, J, K. (1966). Lancet 11, 1049. Axelrod, J., and Inscoe, J. K. (1963). J. Phamcol. Exptl. Therap. 141, 161. Axelrod, J., Kopin, I. J., and Mann, J. D. (1959). Biochim. Biophys. Acta 36, 576. Ayd, F. J., Jr. (1958). N. Y. State 1. Med. 5 4 354. Baastrup, P. C., and Schou, M. (1987). Arch. Gen. Psychiut. 16, 162. Baer, L., Durell, J., Bunney, W. E., Levy, B. S., and Cardon, P. V. (1969). J. Psychiut. Res. 6, 289. Berpman, A. (1959). A& Psychid. Neurol. Scand. 34, Suppl. 133, 1. Bemstein, S., and Kaufman, M. R. (1960). J. Mt. Sinai Hosp., N . Y. 27, 525. Bertler, A., Falck, B., Owman, C., and Rosengrenn, E. (1966). Phumcol. Rev. 18, 369. Blackbum, K. J., French, P. C., and Merrills, R. J. (1987). Life Sci. 6, 1653. Bliss, E. L., Migeon, C. J,, Branch, C. H. H., and Samuels, L. T. (1956). Psychosomut. Med. IS, 56. Board, F., Persky, H., and Hamburg, D. A. (1956). Psychosomut. Med. 18, 324. Board, F., Wadeson, R., and Persky, H. (1957). A.M.A. Arch. Neurol. Psychiut. 78, 612. Borman, M. C., and Schmdenberg, H. C. (1951). J. Am. Med. Assoc. 146, 337. Bourne, H. R., Bunney, W. E., Jr., Colburn, R. W., Davis, J. M.,Davis, J. N., Shaw, D. M., and Coppen, A. J. (1968). Lancet I& 805. Bridges, P. K., and Jones, M. T. (1968). Brit. J . Psychiat. IlQ, 1257. Brodie, B. B., Comer, M. S., Costa, E., and Dlabac, A. (1966). J. Pkumcol. Exptl. Therap. 152, 340. Brooksbank, B. W. L., and Coppen, A. (1967). Brit. J. Psych&. 113, 395. Brown, D. G., Hullin, R. P., and Roberts, J. M. (1963). Brit. J. Psychiat. 109,

395. Bryson, R. W., and Martin, D. F. (1954). Lancet 11, 365. Bueno, J. R., and Himwich, H. E. (1967). Psychosomatics 8, 82. Bunney, W. E., Jr., and Davis, J. M. (1985). Arch. Gen. Psychiat. 13, 483. Bunney, W. E., Jr., and Fawcett, J. A. (19%). Arch. Gen. Psychiat. 13, 232. Bunney, W. E., Jr., and Hartmann, E. L. (1965). Arch. Gen. Psychiat. 12, 611. Bunney, W. E., Jr., Hartmann, E. L., and Mason, J. W. (1965a). Arch. Gen. Psychiat. 12, 619. Bunney, W. E., Jr., Mason, J. W., Roatch, J. F., and Hamburg, D. A. (196513). Am. 1. Psychiat. 122, 72. Bunney, W. E., Jr., Mason, J. W., and Hamburg, D. A. (1965c). Psychosomut. Med. 27, 299. Bunney, W. E., Jr., Davis, J. M., Weil-Malherbe, H., and Smith, E. R. B. (1967). Arch. Gen. Psychiut. 16, 448. Butler, P. W. P., and Besser, G. M. (1988). Lumet I, 1234. Cade, J. F. J. ( 1949). Med. I. Australia 2, 349. Cade, J. F. J. (1962). Med. J. A U S ~ T U2,~911. ~U Cade, J. F. J. (1984). Med. J. AuStlaZia 1, 195. Cameron, I. A., and Dawson, J. (1967). J. Psychosomut. Res. 11, 157.

BIOLOGY OF AFFECTIVE DISORDERS

169

Carlsson, A. (1964).Progr. Brain Res. 8, 9. Carlsson, A., Lindqvist, M., and Magnusson, T. (1957). Nature 180, 1200. Carroll, B. J., Martin, F. I. R., and Davies, B. (1968). Brit. Med. J. 111, 285. Clark, L. D., Bauer, W., and Cobbs, S. (1952). New Engl. J. Med. 246, 205. Clark, L. D., Quarton, G. C., Cobb, S . C., and Bauer, W. (1953). New Engl. J. Med. 249, 178. Cleghorn, R. A. (1951). Can. Med. Assoc. 1. 65, 449. Cobb, S. (1960). J. Neroous Mental Disease 130, 97. Colburn, R. W., Goodwin, F. K., Bunney, W. E., Jr., and Davis, J. M. (1967). Nature 215, 1395. Colbum, R. W., Goodwin, F. K., Mulphy, D. L., Bunney, W. E., Jr., and Davis, J, M. (1968). Biochem. Phurmucol. 17, 957. Conroy, R. T. W. L., Hughes, B. D., and Mills, J. N. (1968). Brit. Med. I. 111, 405. Coppen, A. (1965). Brit. 1. Psychiat. 111, 1133. Coppen, A. (1967). Brit. J. Psychiat. 113, 1237. Coppen, A., and Shaw, D. M. (1963).Brit. Med. 1. II, 1439. Coppen, A., and Shaw, D. M. (19e7). Lancet 11, 805. Coppen, A., Shaw, D. M., and Mangoni, A. (1962). Brit. Med. 1. 11, 295. Coppen, A., Shaw, D. M.,and Farrell, J. P. (1963). Lancet I, 79. Coppen, A., Shaw, D. M., and Malleson, A. (1965a). Brit. J. Psychiat. 111, 105. Coppen, A., Shaw, D. M., Malleson, A., Eccleston, E., and Gundy, G. (1965b). Brit. J. Psychiat. 111, 993. Coppen, A., Malleson, A., and Shaw, D. M. ( 1 9 6 5 ~ )Lancet . I, 682. Coppen, A., Shaw, D. M.,Malleson, A,, and Costain, R. (1966). Brit. Med. J. I, 71. Coppen, A., Shaw, D. M., Henberg, B., and Maggs, R. (1967). Lancet 11, 1178. Coppen, A. J. (1960). J. Neurol., Neurosurg., Psychiat. [N.S.] 23, 156. Corrodi, H., Fuxe, K., Hokfelt, T., and Schou, M. (1967). Psychophumacologia 11, 345. Crammer, J. L. (1959). Lancet I, 1122. Crane, G. E. (1956). 1. Nervous Mental Diseuse 124, 322. Crane, G. E. (1957). Psychid. Res. Rept. 8, 142. Crane, G. E., and Wolfman, M. (1960). J. Nervous Mental Diseuse 130, 134. Cremata, V. Y., Jr., and Koe, B. K. (1966). Clin. P h u m ~ c o l .Therap. 7, 768. Creveling, C. R., Daly, J., Tokuyama, T., and Witkop, B. (1968). Biochem. P h a m c o l . 17, 65. Cumming, J., and Kort, K. (1956). Can. Med. Assoc. J . 74, 291. Curtis, G. C., Cleghorn, R. A., and Sourkes, T. L. (1960). J. Psychosomut. Res. 4, 176. Davis, J. M. (1965). Arch. Gen. Psychiat. 13, 552. Davis, J. M., Colburn, R. W., and Robinson, D. (1968a). Unpublished data. Davis, J. M., Colburn, R. W., Murphy, D. L., and Bunney, W. E., Jr. (1968b). Sci. Proc. 124th Ann. Meeting Am. Psychiat. Assoc., Boston p. 228. Am. Psychiat. Assoc., Washington, D. C. Davis, J. M., Herman, G. L., and Schildkraut, J. J. (1968~).In "Psycho-

170

JOHN M. DAVIS

pharmacology: A Review of Progress” (D. H. Efron, ed.), p. 719. U. S. Govt. Printing Office, Washington, D. C. Dawson, J., Hullin, R. P., and Crocket, B. M. (1956).1. Mental Sci. 102, 168. Degkwitz, R., Frowein, R., Kulenkampff, C., and Mob, U. (1960). Klin. Wochschr. 38, 12D. Dencker, S. J., Haggendal, J., and Malm, U. (1966a).Lancet I I ,754. Dencker, S. J., Malm, U., Roos, B.-E., and Werdinius, B. (1966b). J. Neurochem. 13, 1545. Dengler, H. G., Spiegel, H. E., and Titus, E. 0. (1961).Nature 191, 816. Dewhurst, K. E., El Kabir, D. J., Exley, D., Harris, G. W., and Mandelbrote, B. M. (19s8).Lancet I& 1160. Dewhurst, W. G. (1965).J. Psychosomat. Res. 9, 115. Dewhurst, W. G. (1968a). Lancet I& 514. Dewhurst, W. G. (1968b).Nature 219, 506. Dewhurst, W. G. ( 1 9 6 8 ~ )In . “Studies in Psychiatry” (M.Shepherd and D. L. Davis, eds.), Chapter 14, p. 289. Oxford Univ. Press, London and New York. Dewhurst, W. G., and Marley, E. (1964). Ciba Found. Symp. Animal Behaviour Drug Action p. 175. Dewhurst, W. G., and Marley, E. (1965a). Brit. J. P h a m c o l . 25, 671. Dewhurst, W. G., and Marley, E. (196%). Brit. J. Phumcol. 25, 682. Dewhurst, W. G.,and Marley, E. (196%). Brit. J. P h u m o l . 25, 705. Dick, P., and Roch, P. (1967). In “First International Symposium on Antidepressant Drugs” (S. Garattini and M. N. G. Dukes, eds.), Excerpta Med. Found. Intern. Congr. Ser., No. 122,p. 311. Excerpta Med. Found., Amsterdam. Doig, R. J., Mummery, R. V., Wills, M. R., and Elkes, A. ( 1966). Brit. J. Psychiat. 112, 1263. Dunlop, E., DeFelice, E. A., Bergen, J. R., and Resnick, 0. (1965).Psychosomdics 6, 1. Durell, J., and Schildkraut, J. J. ( 1966). In “American Handbook of Psychiatry” (S. Arieti, ed.), Vol. 3, p. 423. Basic Books, New York. Engel, G. L., and Margolin, S. G. ( 1941). A.M.A. Arch. Neurol. Psychiat.

45, 881. Engelman, K., Lovenberg, W., and Sjoerdsma, A. (1967).Neru Engl. J. Med. 277, 1103. Engelman, K., Honvitz, D., JBquier, E., and Sjoerdsma, A. (1968). J. Clin. Inuest. 47, 577. Everett, G. M., and Toman, J. E. P. (1959).B i d . Psychiat., Proc. Sci. Sessions SOC.Biol. Psychiat., Sun Francisco, 1958, Vol. 1, p. 75. Grune & Stratton, New York. Everett, G. M., and Wiegand, R. G. (1962). Proc. 1st. Intern. Phurmucol. Meeting, Stockholm, 1961 Vol. 8, p. 85. Pergamon Press, Oxford. Fawcett, J., Maas, J. W., Siomopoulos, V., and Dekirmenjian, H. (1968). Psychosomat. Med. 30, 550. Fawcett, J. A., and Bunney, W. E., Jr. (1967).Arch. Gen. Psych&. 16, 517. Feldstein, A., Hoagland, H., Oktem, M. R., and Freeman, H. (1965).Intern. 1. Neuropsychiat. 1, 384.

BIOLOGY OF AFFEcIlvE DISORDERS

171

Ferguson, H. C., Bartram, A. C. G., Fowlie, H. C., Cathro, D. M., Birchall, K., and Mitchell, F. L. (1964). Acta Endocrit~l.47, 58. Fink, M., Klein, D. F., and Kramer, J. C. (1965). Psychopharmologia 7, 27. Fleminger, J. J. (1955). 1. Mental Sci. 101, 123. Freedman, D. X . (1966). In “Psychiatric Drugs” (P. Solomon, ed.), p. 32. Grune & Stratton, New York. Fullerton, D. T., Wenzel, F. J., Lohrenz, F. N., and Fahs, H. (1968). Arch. Gen. Psychiat. 19, 674. Galdston, M., Weisenfeld, S., Benjamin, B., and Rosenbluth, M. B. (1951). Am. 1. Med. 10, 166. Ganrot, P. O., Rosengren, E., and Gottfries, C. G. (1962). Experientia 18, 260. General Practitioner Research Group. ( 1964). Practitioner 192, 151. Gershon, S. (1968). Am. J. Psychiat. 124, 1452. Gershon, S., and Yuwiler, A. (1960). J. Neuropsychiat. 1, 229. Gershon, S., Hekimian, L. J., and Floyd, A. (1967). Current Therup. Res. 9, 349. Gershon, S., Hekimian, L. J., and Floyd, A. (1968). ArzneimitteGForsch. 18, 243. Gibbons, J. L. (1960). Clin. Sci. 19, 133. Gibbons, J. L. (1963). Postgrad. Med. J. (London) 39, 19. Gibbons, J. L. ( 1964). Arch. Gen. Psychiat. 10, 572. Gibbons, J. L. (1966). 1. Psychosomat. Res. 10, 283. Gibbons, J. L., and Fahy, T. J. ( 1966). Neuroendocrinology 1, 358. Gibbons, J. L., and McHugh, P. R. (1962). J. Psychiat. Res. 1, 162. Glaser, G. H. (1953). Psychosomat. Med. 15, 280. Glowinski, J., and Axelrod, J. (1965). 1. P h a mcol. Exptl. Therup. 149, 43. Glowinski, J., and Axelrod, J. (1966).PhurmacoZ. Rev. 18, 775. Glowinski, J., and Baldessarini, R. J. (1966). Phurmacol. Rev. 18, 1201. Goolker, P., and Schein, J. (1953).Psychosomat. Med. 15, 589. Green, R. (1967). Intern. J. Neuropsychiat. 3, 133. Haggendal, J., and Lindqvist, M. (1964). Acta Physiol. S c a d . 60, 351. Hams, T. H. (1957). Am. J. Psychiat. 113, 950. Haslovec, L., and RySBnek, K. ( 1967). Psychophamcologia 11, 18. Ha&ovec, L., and Soukk, K. (1988). Nature 219, 507. Hekimian, L. J., Gershon, S., and Floyd, A. (1968). Cusrent Therap. Res. 10, 282. Hendley, E. D., and Snyder, S. H. (1968). Nature 220, 1330. Hertting, G., Axelrod, J., and Whitby, L. G. (1961). J. Phumc ol. Erptl. Therap. 134, 148. Hollister, L. E., Overall, J. E., Shelton, J., Pennington, V., Kimbell, I., and Johnson, M. (1967). Arch. Gen. Psychiat. 17,486. H u b , R. P., Bailey, A. D., McDonald, R., Dransfield, G. A., and Milne, H. B. ( 1967a). Brit. 1. Psychiat. 113, 573. Hullin, R. P., Bailey, A. D., McDonald, R., Dransfield, G. A., and Milne, H. B. ( 196713). Brit. J. Psychiat. 113, 584. Hullin, R. P., Bailey, A. D., McDonald, R., Dransfield, G. A,, and Milne, H. B. (1967~).Brit. J. Psychiat. 113, 593.

172

JOHN M. DAVIS

Ingvarsson, C. G. (1965). AszneimittebForsch. 15, 849. Jacobsen, E. ( 1964). In “Depression” (E. B. Davis, ed.), p. 208. Cambridge Univ. Press, London and New York. Jakobson, T., Stenback, A., Strandstrijm, L., and Rim6n, R. (1988). J. Psychsomat. Res. 0, 363. Karki, N. T. (1956). Actu Physiol. Scud. 39, Suppl. 132, 1. Katz, R. I., Chase, T. N., and Kopin, I. J. (1963). Sdence 162, 486. Kety, S. S., Javoy, F., Thierry, A.-M., Julou, L., and Glowinski, J. (1967). Proc. Nutl. Acud. Sci. U. S . 58, 1249. Key, B. J., and Marley, E. (1962). Electroencephulog. Clin. Neurophysiol. 14,w. Klein, R. (1950). 1. Mental Sci. 96, 293. Klein, R., and Nunn, R. F. (1945). J. Mental Sci. 91, 79. Klerman, G. L., Schildkraut, J. J., Hasenbush, L L., Greenblatt, M., and Friend, D. G. (1963). J. Psychiut. Res. 1, 289. Kline, N. S. (1958). J. Clin. Erptl. Psychoputhol. G Qzrort. Rev. Psychiat. Neurol. 19, Suppl. 1, 72. Kline, N . S., and Sacks, W. (1963). Am. J. Psychiut. 120, 274. Kline, N. S., Sacks, W., and Simpson, G. M. (1984). Am. J . Psychiut. 121, 379. Knapp, M. S., Keane, P. M., and Wright, J. G. (1967). Brit. Med. J. I& 27. Koe, B. K., and Weissman, A. (1966). Federation Proc. 25, 452. Kopin, I. J. (1964). Phumcol. Reo. 16, 179. Kopin, I. J., and Gordon, E. K. (1962). J. Phurmcol. Erptl. Therap. 138, 351. Kopin, I. J., and Gordon, E. K. (1963). 1. Pharmucol. Erptl. Therap. 140, 207. Kuhn, R. (1958). Am. J. Psychiut. 115, 459. Kurland, H. D. (1964a). Proc. SOC. Exptl. Biol. Med. 115, 723. Kurland, H. D. (1904b). Arch. Gen. Psychiut. 10, 554. Kurland, H. D. (1965). Am. J. Psychiut. 122, 457. Lehman, H. E., Ban, T. A., and Debow, S. L. (1967). Current Therap. Rar. 9, 306. Lemieux, G., Davignon, A., and Genest, J. (1956). Can. Med. Assoc. J . 74, 522. Levine, R. J. (1966). Biochem. Phumcol. 15, 1645. Lidz, T., Carter, J. D., Lewis, B. I., and Surratt, C. (1952). Psychosomat. Med. 14, 363. Lingjaerde, 0. (1963). Actu Psychiat. Scud. 39, Suppl. 170, 1. Lingjaerde, P. S. (1964). Brit. J. Psychiat. 110, 423. Loomer, H. P., Saunders, J. C., and Kline, N. S. (1957). Psychfut. Res. Rept. 8, 129. Maas, J. W., and Landis, D. H. (1967). Fedemtion Proc. 26,463. Maas, J. W., Fawcett, J., and Dekimenjian, H. (1968). Arch. Gen. Psychiut. 19, 129. McLean, J. R., and McCartney, M. (1961). Proc. SOC. Exptl. Biol. Med. 107, 77. Maclean, R., Nicholson, W. J., Pare, C. M. B., and Stacey, R. S. (1965). Lancet 11, 205. McClure, D. J. (1966a). J . Psychosomat. Res. 10, 189. McClure, D. J. (1966b). J. Psychosomat. Res. 10, 197.

BIOLOGY OF AFFECTIVE DISORDERS

173

McClure, D. J., and Cleghorn, R. A. (1989). In ‘The Future of the Brain Sciences” (S. Bogoch, ed. ), p. 525. Plenum Press, New York. Marley, E. (1968a). I n “Studies in Psychiatj’ (M. Shepherd and D. L. Davies, eds.), Chapter 13, p. 253. Oxford Univ. Press, London and New York. Marley, E. ( 196813). In “The Scientific Basis of Medicine Annual Reviews,” (Brit. Postgrad, Med. Fed., ed.), p. 359. Oxford Univ. Press (Athlone), London and New York. Marley, E. (1968~).Lancet II, 357. Marley, E., and Morse, W. H. (1967). Brit. J. Phumcol. 31, 367. Marley, E., and Stephenson, J. D. (1968).J . Physiol. (London) 196, IMP. Matussek, N., and Linsmayer, M. (1968). Life Sci. 7 (Part I), 371. Matussek, N., Pohlmeier, H., and Riither, E. (1966). Klin. Wochschr. 44, 727. Maynert, E. W., and Levi, R. (1964). J. Pharmacol. Exptl. Therap. 143, 90. Michael, R. P., and Gibbons, J. L. (1963).Intern. Reu. Neurobiol. 5, 243. Moore, K. E., and Lariviere, E. W. (1963).Biochem. Phurmucol. 12, 1283. Muller, J. C., Pryor, W. W., Gibbons, J. E., and Orgain, E. S. (1955). J. Am. Med. Assoc. 159, 836. Murphy, D. L., Colburn, R. W., Davis, J. M., and Bunney, W. E., Jr. (1969). Life Sci. 8 (Part I), 1187. Nelson, G. N., Masuda, M., and Holmes, T. H. (1966). Psychosomat. Med. 28, 216. Pare, C. M. B. (1963). Lancet IT, 527. Pare, C. M. B. (19%). I n “Scientific Basis of Drug Therapy in Psychiatry” (J. Marks and C. M. B. Pare, eds. ), p. 103. Pergamon Press, Oxford. Pare, C. M. B., and Sandler, M. (1959). J. Neurol., Neurosurg., Psychiat. [N.S.] 22, 247. Pare, C. M. B., Rees, L., and Sainsbury, M. J. (1962). Lancet I& 1340. Pearson, 0. H., and Eliel, L. P. (1950). J. Am. Med. Assoc. 144, 1349. Persson, T., and Roos, B.-E. (1967). Lancet Ill, 987. Persson, T., and Waldeck, B. (1968). Acta Phurmacol. Tmicol. 26, 383. Poldinger, W. (1963).Psychopharmucologia 4, 308. Poschel, B. P. H., and Ninteman, F. W. (1963). Life Sci. 2, 782. Prange, A. J., Jr., McCurdy, R. L., and Cochrane, C. M. (1967). J. Psychiat. Res. 5, 1. Prange, A. J., Jr., Wilson, I. C., Rabon, A. M., and Lipton, M . A. (1968). Sci. Proc. 124th Ann. Meeting Am. Psych&. Assoc., Boston p. 325. Am. Psychiat. Assoc. Washington,D. C. Pryce, I. G. (1964). Brit. J. Psychiat. 114 90. Quarton, G. C., Clark, L. D., Cobb, S., and Bauer, W. (1955). Medicine 34, 13. Rees, L. (1953). J. Mental Sci. 99, 497. Reis, D. J., and Gunne, L.-M. (1965). Science 14, 450. Reiss, M., Hemphill, R. E., Gordon, J. J., and Cook, E. R. (1949). Biochem. J. 45, 574. Rizzo, N. D., Fox, H. M., Laidlaw, J. C., and Thorn, G. W. (1954). Ann. l n t e d Med. [N.S.] 41, 798. Rodnight, R. (1961). Intern. Rev. Neurobiol. 3, 251.

174

JOHN M. DAVIS

Rome, H. P., and Braceland, F. J. (1952). J. Am. Med. Assoc. 148, 27. Rosenblatt, S., and Chanley, J. D. (1%). Arch. Gen. Psychiat. 13, 495. Rosenblatt, S., Chanley, J. D., Sobotka, H., and Kaufman, M. R. (1960). J. Neurochem. 5, 172. Ross, S. B., and Renyi, A. L. (1967). Life Sd. 6, 1407. Russell, G. F. M. (1960). Clin. Sci. 19, 327. Sachar, E. J. (1967a). Arch. Gen. Psychiut. 17, 544. Sachar, E. J. (1967b.) Arch. Gen. Psychiut. 17, 554. Sachar, E. J., Mason, J. W., Kolmer, H. S., and Artiss, K. L. (1963). Psychosomut. Med. 25, 510. Sachar, E. J., Mackenzie, J. M., Binstock, W. A., and Mack, J. E. (1967). Arch. Gen. Psychiat. 16, 461. Sakai, M. (1960). Yokohama Med. Bull. 11, 352. Schanberg, S. M., Schildkraut, J. J., and Kopin, I. J. (1967a). Federation Proc. 26, 464. Schanberg, S. M., Schildkraut, J. J., and Kopin, I. J. (1967b). Biochem. PhamMcol. 1% 393. Schanberg, S. M., Schildkraut, J. J., Breese, G. R., and Kopin, I. J. (19688). Biochem. P h a m w 2 . 17, 247. Schiele, B. C., Tredici, L. M., and Zimmermann, R. L. (1966). Current Therap. Res. 8, 565. Schildkraut, J. J. (1965). Am. J . Psychiat. 122, 509. Schildkraut, J. J., Klerman, G. L., Friend, D. G., and Greenblatt, M. (1963). Ann. N . Y. Acad. Sci. 1107, 1005. Schildkraut, J. J., Klerman, G. L., Hammond, R., and Friend, D. G. (1964). 1. Psychiat. Aes. 2, 257. Schildkraut, J. J., Gordon, E. K., and Durell, J. (1965). 1. Psychiat. Res. 3, 213. Schildkraut, J. J., Green, R., Gordon, E. K., and Durell, J. (196%). Am. J . Psychiat. 123, 690. Schildkraut, J. J., Schanberg, S. M., and Kopin, I. J. ( 1 W b ) . Life S c i . 5, 1479. Schildkraut, J. J., Schanberg, S. M., Breese, G. R., and Kopin, I. J. (1967). Am. J . Psychiat. 124, 600. Schildkraut, J. J., Davis, J. M., and Klerman, G. L. (1968). In “Psychopharmacology: A Review of Progress” (D. H. Efron, ed.). p. 625. U. S. Govt. Printing Office, Washington, D. C. Schopbach, R. R., Kelly, A. R., and Lukaszewski, J. S. (1964). Progr. Brain Res. 8,207. Schwartz, M., Mandell, A. J., Green, R., and Ferman, R. (1966). Brit. J . Psychiat. 112, 149. Shaffer, J. W., Freinek, W. R., McCusker, J. K., and DeFelice, E. A. (1964). J. New Drugs 4, 288. Shaw, D. M. (1966). Brit. Med. J. II, 262. Shaw, D. M., and Coppen, A. (1966). Brit. J. Psychiat. 112, 269. Shaw, D. M., Camps, F. E., and Eccleston, E. G. (1967). Brit. J. Psychfat. 113, 1407. Shinfuku, N., Omura, M., and Kayano, M. (1961). Y m g o Acta Med. 5, 109.

BIOLOGY OF AFFECTIVE DISORDERS

175

Sigg, E. B. (1959). Can. Psychiat. Assoc. J. 4, S75. Sjoerdsma, A., Gillespie, L., Jr., and Udenfriend, S. (1958). Lancet 11, 159. Sjoerdsma, A., Engelman, K., Spector, S., and Udenfriend, S. (1965). Lancet U, 1092. Sloane, R. B., Hughes, W., and Haust, H. L. (1966). Can. Psychiat. ASSOC. J. 11, 6. Smith, C. B. (19ss). J. Phurmacol. Exptl. Therap. 147, 96. Spillane, J. D. ( 1951). Bruin 74, 72. Stefanini, M., Roy, C. A., Zannos, L., and Dameshek, W. (1950). J . Am. Med. Assoc. 144, 1372. Stein, L. (1967). I n “First International Symposium on Antidepressant Drugs” (S. Garattini and M. N. G. Dukes, eds.), Excerpta Med. Found. Intern. Congr. Ser., No. 122, p. 130. Excerpta Med. Found., Amsterdam. Stenback, A,, Jakobson, T., and Rimbn, R. (1966). J. Psychosomat. Res. 9, 355. Stem, D. M., Fieve, R., Neff, N., and Costa, E. (1967). Pharmacologist 9, 210. Strom-Olsen, R., and Wed-Malherbe, H. (1958). J. Mental Sci. 104, 696. Sulser, F., Bickel, M. H., and Brodie, B. B. (1964). J. Phurmacol. Ezptl. Therap. 144, 321. Sulser, F., Owens, M. L., and Dingell, J. V. (1966). Life Sci. 5, 2005. Sulser, F., Owens, M. L., and Dingell, J. V. (1967). Pharmacologist 9, 213. Trethowan, W. H., and Cobb, S. (1952). A.M.A. Arch. Neurol. Psychiat. 67, 283. Turner, W. J., and Merlis, S. (1964). Diseases Nervous System 25, 538. Ueno, Y., Aoki, N., Yabuki, T., and Kuraishi, F. (1961). Folia Psychiat. Neural. Japon. 15, 304. Vane, J. R. ( 1961). Cibu Found. Symp. Adrenergic Mechanisms p. 356. Vane, J. R., Collier, H. 0. J., Come, S. J., Marley, E., and Bradley, P. B. ( 1961). Nature 191, 1068. van Praag, H. M., Korf, J., van Woudenberg, F., and Kits, T. P. (1988). Psychophamnacologia 13, 145. v. Studnitz, W. (1959). Scand. J. Clin. 6 Lab. Invest. 11, 224. Waal, H. J. (1967). Brit. Med. J . 11, 50. Weil-Malherbe, H. (1968). Lancet I& 219. Weil-Malherbe, H., Axelrod, J., and Tomchick, R. (1959). Science 129, 1226. Weissman, A., Koe, B. K., and Tenen, S. S. (1966). J. Phumcol. Exptl. Therap. 151, 339.

This Page Intentionally Left Blank

CEREBRAL PROTElN SY NTHESlS INHIBlTORS BLOCK LONG-TERM MEMORY' By Samuel H. Barondes2 Departments of Psychiatry and Molecular Biology, Albert Einstein College of Medicine, Bronx, New York

I. Introduction . . . . . . 11. Additional Background . . . . 111. Inhibitors of Protein and RNA Synthesis . . . . . IV. Behavioral Assays V. Consolidation and Redundancy . . VI. Inhibition during Training . . . VII. Inhibition Shortly after Training . . VIII. Inhibition Long after Training . . IX. Inhibition of RNA Synthesis . . . X. Short-Term and Long-Term Memory . XI. Conclusion . . . . . . . References . . . . . . .

.

.

.

.

. .

. .

. .

. .

.

.

.

.

. . .

. . . .

. . . .

. . . .

. .

. .

. .

. .

. .

. .

. .

. .

. . . . . . . .

, , . .

1 1 1 1 1 1 1

77 78 80 83 85 86 91 192 1 9 5 1 9 8 201 203

1. Introduction

Memory storage is one of the principal functions of higher nervous systems. Yet, until recently it has eluded intensive biological investigation. The reason for this is quite apparent. The storage of memory in the nervous system is believed to be mediated by an alteration in the functional relationship among large numbers of neurons. Lack of knowledge about general processes of cellular regulation and the control of intercellular relationships would preclude serious study of mechanisms for the formation of new functional interneuronal relationships. Recent advances in our understanding of biological regulatory processes have, however, made possible the initiation of attempts to understand the mechanisms by which memory is stored in the brain. An important finding of this research is that general biological regulatory mechanisms are frequently mediated by the synthesis of new protein. In the course of such studies a number of inhibitors of protein or ribonucleic acid synthesis were discovered. Indeed, studies with these Supported by U.S.P.H.S. Career Development Award #K-3-MH-18232. 'Present address: Department of Psychiatry, The Medical School, University of California at S a n Diego, La Jolla, California.

177

178

SAMUEL H. BARONDES

inhibitors helped establish the critical role of protein synthesis in cellular regulation. Because it could be readily imagined that new functional interneuronal relationships could also be established by protein synthesis (Barondes, 1965), these inhibitors have been employed for studies of the possible role of cerebral protein synthesis in memory storage. There is now considerable evidence that inhibitors of cerebral protein synthesis interfere with memory storage. Since all such experiments involve the introduction of a complex reagent with multiple effects and the measurement of alterations in a complex process, the interpretation of the results are diflicult, particularly for the casual reader in this field. The major point of this review will be to focus on the important variables which influence the types of results which have been found. II. Additional Background

It is important to recognize that even if cerebral protein synthesis can be implicated in the mechanism of long-term memory storage, this is only the first step in determining the biochemical basis of this process. The synthesis of new brain protein could mediate memory in a variety of ways. One alternative which has been suggested is that there are specsc molecules which, in some manner, code for specific bits of behavioral information. Proponents of this position draw analogies with deoxyribonucleic acid (DNA), where memory of genetic information is encoded in specific base sequences, and with antibody' formation, where specific protein molecules are synthesized in response to a specific antigen. However it does not seem necessary to postulate the existence of specific memory molecules if one accepts that interneuronal relationships at synapses are the important units of coding in the central nervous system and that the molecular mechanisms in learning could operate by influencing these relationships. Because of the large number of neurons and the finding that each may make contact with thousands of other neurons, the potentiality for detailed storage of information by facilitation or inhibition of a series of already existent synaptic relationships is certainly present. Synthesis of brain protein could lead to the formation of new functional interneuronal relationships in a number of ways. Either the presynaptic neuron or the postsynaptic neuron could be involved. Although there might be a major outgrowth of nerve end-

PROTEIN SYNTHESIS INHIBITORS AND MEMORY

179

ings or dendrites which generate the formation of new synaptic connections, the same functional consequence would be produced more parsimoniously by facilitating preexisting synaptic connections through changes in enzymatic composition on the presynaptic or postsynaptic side. Increased synthesis, by the presynaptic cell, of an enzyme that synthesizes neurotransmitter might make the presynaptic terminal more effective in activating the postsynaptic neuron. Increased synthesis of an enzyme in the postsynaptic cell which destroys an inhibitory neurotransmitter could also facilitate synaptic connections. Likewise increased synthesis of receptor protein in the postsynaptic neuron or even changes in some other limiting component of this neuron (Bullock, 1!368) could make a functional synapse develop. Studies with inhibitors of cerebral protein synthesis thus far provide no information which would permit a choice among these and other alternatives. However, such work makes tenable this whole class of possibilities and, by defining the temporal limits during which such synthetic activity must occur, will be of some value in attempts to identify the proteins whose synthesis may be involved in synaptic facilitation for memory storage. Although the search for spec& proteins whose synthesis might lead to the facilitation of new synaptic connections has been compared to looking for a needle in a haystack, attempts have been made to determine whether demonstrable changes in overall brain ribonucleic acid (RNA) or protein synthesis occur during a learning experience. Some of these studies have failed to demonstrate changes in these processes during learning. However, there is considerable evidence from the work of Glassman and his colleagues (Zemp et d.,1966, 1967; Adair et d.,1968a,b) that there is markedly increased incorporation of labeled uridine into all classes of cerebral RNA during a period of training, and from the work of H y d h and Egyhazi (1962, 1963, 1W) that there are marked changes in the base composition of neuronal and glial RNA following a period of training. Adair et al. (1968b) present evidence that the observed changes are correlated with learning but not with simple activation of the animal. Hydkn and Lange (1968) have also shown evidence for substantial increases in protein synthesis in CA3 neurons of the hippocampus of the rat during a period of training. These studies are consistent with the participation of cerebral RNA and protein synthesis in the memory storage

180

SAMXEL H. BARONDES

process. What is noteworthy about these changes is that they are very large. If the observed increases in macromolecular synthesis are indeed related to a specific memory storage process, the results suggest that there is increased synthesis in a large number of neurons and also increased synthesis of a large number of different RNA and protein molecules. However, it is not clear, at present, how much of this increased synthesis might be related specifically to memory storage. 111. Inhibitors of Protein and RNA Synthesis

The synthesis of each protein is directed by specific messenger RNA. Increased synthesis of a particular protein may occur either as a consequence of increased synthesis of the appropriate messenger RNA or increased translation (“read out”) of already synthesized messenger RNA. The former process may be blocked by the administration of actinomycin-D, an inhibitor of DNA-dependent RNA synthesis. However, this drug does not block the synthesis of proteins whose messenger RNA is already present in the cell but is not, for some reason, being translated. In bacterial systems the synthesis of specific protein is usually regulated by the synthesis of specific messenger RNA. This may not always be true in mammalian systems. Therefore, the fact that a process is not inhibited by actinomycin-D does not necessarily mean that protein synthesis is not involved. All it means is that if protein synthesis is involved the synthesis of new RNA is not required to direct synthesis of this new protein. Another group of drugs, including puromycin and the glutarimide derivatives, act directly on the protein synthesizing system and would therefore inhibit the synthesis of protein directed either by newly synthesized RNA or by RNA already present in the cell. These protein synthesis inhibitors differ in their mechanism of action. Puromycin resembles amino-acyl transfer RNA which is the form in which amino acids are incorporated into protein. Puromycin is itself incorporated into growing polypeptide chains with the formation of peptidyl-puromycin (Nathans, 1964), an abnormal polypeptide, which is then released from the ribosomes which are the site of protein synthesis. This abnormal product might adversely affect cellular metabolism. The glutarimide derivatives which include cycloheximide and acetoxycycloheximide also interfere with protein synthesis but by a different mechanism of action. These

PROTEIN SYNTHESIS INHIBITORS AND MEMORY

181

drugs stop ongoing protein synthesis but upon their removal normal proteins are completed, and no abnormal polypeptides are released into the cell. Indeed, these drugs inhibit incorporation of puromycin into polypeptide chains and, when administered with puromycin, prevent the formation of peptidyl-puromycin (Siege1 and Sisler, 1963). This difference in the mechanism of action of these drugs may be of considerable importance in interpreting the differences which they exert on brain function and on memory. In addition to differences in their mechanism of action, these two classes of inhibitors of protein synthesis differ in other ways as well. There is evidence that puromycin directly inhibits 3’,5’ cyclic adenosine monophosphate (AMP) phosphodiesterase (Appleman and Kemp, 1966) which degrades cyclic AMP. Since this compound may play an important role in brain function, puromycin may interfere with brain function by this action which is unrelated to its effect on cerebral protein synthesis. Puromycin also has a direct effect on glycogen synthetase ( Sovik, 1967). In contrast with puromycin, the glutarimide derivatives have no known direct effects on these enzymes. However, the possibility that they have unknown effects unrelated to inhibition of protein synthesis cannot be excluded. Both puromycin (Kanfer and Richards, 1967) and acetoxycycloheximide ( Barondes and Dutton, 1969) interfere with ganglioside synthesis in the brain. Although the mechanism for this action is not known it may be due to interference with the synthesis of protein acceptors for gangliosides. Puromycin and the glutarimide derivatives have strikingly different effects on the brain, and these differences may explain differences in their behavioral effects. Puromycin produces abnormal cerebral electrical activity (Cohen et al., 1966) and decreases the threshold to overt seizures observed after administration of pentylenetetrazol (Cohen and Barondes, 1967). These abnormalities were initially observed in mice and have subsequently been found in goldfish ( Agranoff, 1969). In mice the decreased seizure threshold produced by puromycin is antagonized by cycloheximide (Cohen and Barondes, 1967). This may explain certain protective actions of cycloheximide on specific behavioral effects of puromycin (which will be considered later). A similar protective action has not been observed in goldfish ( Agranoff, 1969). Puromycin aminonucleoside (which does not inhibit cerebral protein synthesis) decreases seizure threshold in goldfish, and hydrolyzed puromycin (which has no

182

SAMUEL H. BARONDES

effect on protein synthesis) has some effect on seizure threshold in mice but less than that of intact puromycin (Cohen and Barondes, 1W7).The protective action of cycloheximide in mice implies that peptidyl-puromycin may be a convulsion-producing agent since cycloheximide can interfere with production of this product (Siege1 and Sisler, 1963). Puromycin has also been shown to produce mitochondrial abnormalities in mice (Gambetti et al., 1968a,b). The mitochondrial swelling produced by this drug can be antagonized by acetoxycycloheximide, thus implying that peptidyl-puromycin is also responsible for this toxic action (Gambetti et al., 1968a,b). There is evidence that the delayed action of puromycin on memory may be due to the persistence of peptidyl-puromycin in the brain for long periods of time (L. B. Flewer and Flexner, 1968b). Because of these multiple toxic effects which puromycin has on the brain, we have concluded that puromycin is of only limited value for studying the hypothesis that cerebral protein synthesis is required for memory. It has, however, proved to be a useful agent for studying phases of memory storage. It is important to keep in mind that the different results with puromycin and cycloheximide may be related to these special effects of puromycin. In addition to toxic side effects which administration of any drug might have, administration of inhibitors of protein synthesis may have “side effects” related to their primary mode of action. When inhibition of cerebral protein synthesis is established in a memory experiment to attempt a blockade of the synthesis of cerebral protein required for long-term memory storage, other ongoing cerebral protein synthesis is blocked also. This ongoing synthesis is presumably directed toward replacement of proteins which are being metabolized and also toward other cellular regulatory processes which may be proceeding independently at the time of the experiment. Interference with these processes would ultimately impair brain function. However, the half-life of most brain proteins tends to be in the range of days (Lajtha, 1964). Therefore, inhibition of cerebral protein synthesis for a number of hours might have no significant effect on ongoing function since there should be only slight depletion of “constitutive” brain protein whose replacement has been blocked. Nonetheless it is never possible to rule out completely the possibility that interference with ongoing brain protein synthesis has a nonspecific effect on the memory storage process.

PROTEIN SYNTHESIS INHIBITORS AND MEMORY

183

Thus, even with an apparently nontoxic short acting drug like cycloheximide some degree of caution is required in interpreting results. With drugs with known toxicity like puromycin and with very long acting toxic agents like actinomycin-D interpretation of results is considerably more difficult. IV. Behavioral Assays

Although attempts are being made to investigate correlates of behavior in simpler isolated nervous system preparations (Brown and Noble, 1967; Kandel and Spencer, 1968), studies on memory thus far have been concerned primarily with the behavior of whole organisms. Therefore, measurements of the formation and storage of memory are indirect. Although we may be concerned with the biochemical mechanism which mediates the development of new functional synaptic relationships, the assay is a behavioral one. This poses the great problem that the behavioral measurements used may not be linearly correlated with the underlying processes which we hope to study. Is an animal who receives two training trials in a specific task receiving twice as much training as an animal that has received only one trial, or is he receiving much more than twice as much or much less? Likewise, does an animal who exhibits 80%retention on an arbitrarily defined retention scale remember twice as well as an animal who exhibits 40% retention, or does his performance reflect only slightly superior storage or really vastly superior storage? Until we clearly understand the mechanism of memory storage, inferences made from observations of performance must be regarded as highly qualitative. They demonstrate only tendencies but do not reflect the underlying memory storage process in a truly quantitative sense. In choosing the behavioral situations for studies of this kind great emphasis is placed on the degree of training required to produce measurable memory. To study the effects of inhibition of cerebral protein synthesis on memory it is preferable to establish inhibition for only as short a time as possible to obviate the possibility of secondary effects of inhibition of cerebral protein synthesis. Therefore, it is desirable to employ tasks which can be learned relatively quickly-that is, in a small number of training trials during one training session. Particularly useful for such studies is discrimination training in a T maze. Mice are trained to escape shock by choosing one of the limbs (left or right) of a

184

SAMUEL H. BARONDES

T maze to a specified criterion of consecutive correct responses. If one wishes to train the animal to perform a correct response consistently at some subsequent time the training is usually continued until the animal makes 9 out of 10 consecutive correct responses. However, for many tasks this constitutes substantial “overtraining.” We have found that position discrimination training to escape shock to a criterion of only 3 out of 4 consecutive correct responses is sufEicient to produce a high level of retention for at least 6 weeks thereafter (Barondes and Cohen, 1967b). In this task, training to a higher criterion of consecutive correct responses gives the animal the opportunity to rehearse the correct response for a large number of additional trials. This has the effect of rendering the training less susceptible to disruptive agents. In mice the task can be made more difficult by requiring that the animal choose the correct limb of a T maze not on the basis of a position cue but on the basis of the presence of light. Mice learn to choose the lighted limb (randomly alternated) of a maze to any given criterion in more trials than they do when a position cue is provided (Cohen and Barondes, 1968a). The more difficult task is apparently more sensitive to the effect of inhibitors of protein synthesis (Cohen and Barondes, 1968a). Other investigators have chosen to employ an avoidance rather than an escape criterion. In the work of Flewer et al. (1963) mice are trained to a given criterion to avoid shock by choosing the correct limb of a maze. They are scored as having made an incorrect response if they do not run to the correct side of the maze before the shock is applied. This requires many more trials than does escape training, and a criterion of 9 out of 10 consecutive correct responses constitutes considerable overtraining. Agranoff et al. (1965) who work with goldfish have also used an avoidance criterion by training them to avoid shock by crossing to the other side of their tank upon presentation of a light cue. In the situation they usually employ, the fish are trained for only a relatively small number of trials so that at the end of the training session they are not consistently performing the correct response. In addition to these active avoidance tasks, so called passive avoidance tasks have been used in these studies (Barondes and Jarvik, 1964; Geller et at., 1969). In these tasks the animal is trained, often in one trial, to avoid shock by inhibiting a response. An appetitive task in which water-deprived mice are trained in a position discrimination for water reinforcement has also been used (Cohen and Barondes, 1968b).

PROTEIN SYNTHESIS INHIBITORS AND MEMORY

185

Differences in the training procedure must be kept in mind in evaluating the results of these studies. Since the degree of inhibition of cerebral protein synthesis in all these experiments is not total, it would be expected that the greater the degree of learning the more likely the effects of the inhibitor on the memory storage process could be obscured by the overtraining. V. Consolidation and Redundancy

Two observations from earlier studies of memory using other disruptive techniques are particularly relevant to the interpretation of studies of the effect of inhibitors of cerebral protein synthesis. The first is the phenomenon of consolidation. Its existence is inferred from studies in which electroconvulsive shock, administered at increasing times after a learning experience, has progressively less effect on subsequent memory. These and related studies have been reviewed recently (Weissman, 1967; McGaugh, 1966). Administration of electroconvulsive shock immediately after training, particularly in a one-trial training situation, produces marked impairment of memory when the animal is tested at subsequent times. Electroconvulsive shock within minutes after training generally produces effects less marked than when given immediately after training. If the treatment is made hours after training there is generally no effect on subsequent retention. Studies of this type suggest that the memory process is not completed with completion of training. Rather a process is believed to continue beyond training, which renders memory increasingly resistant to the effects of electroconvulsive shock. Some have suggested that this provides evidence for the existence of a “labile” phase of memory which is susceptible to electroconvulsive shock and a “stable” phase of memory which is resistant, and inferences have been drawn about the possible existence of two independent processes of memory storage. Although these inferences may well be correct, and they are indeed supported by studies with inhibitors of cerebral protein synthesis which will be discussed below, the existence of two different processes is not proved by experiments of this type. All that is shown is that the memory storage process changes with time after training, possibly by strengthening of the storage due to the continued activity of a single process initiated by training. In research on the effect of inhibitors of cerebral protein synthesis on memory, it is important to keep these observations in mind, since it may well be critical that cerebral protein synthesis be inhibited

186

SAMUEL H. BARONDES

for a considerable period of time after a training experience in order to interfere with what appears to be a relatively prolonged memory storage process. A second general phenomenon was discovered in studies of the effects of ablation of portions of the brain after training on subsequent retention. In classic studies on the effects of cerebral ablation Lashley (reprinted, 1960) found that impairment in retention of maze learning was correlated with the amount of cerebral tissue that was removed rather than with the site of removal. This gave rise to the notion that memories may be stored in many different regions of the brain-in other words, there is “redundant” storage of memories throughout the brain. One view of this finding is that repeated training leads to a variety of perceptions of the training situation and that these different experiences are stored in separate but overlapping neuronal circuits. Alternatively, large portions of the brain may have to behave in concert to store a specific piece of information so that progressive removal of pieces of cerebral cortex would progressively impair the highly organized general cerebral activity. A consideration of these alternatives has been presented by John (1967). Although we do not clearly understand the phenomenon of redundancy, it has important implications for studies of the effects of inhibitors of cerebral protein synthesis on memory. If there is a redundant storage in many circuits in different parts of the brain it is critical that generalized inhibition of cerebral protein synthesis be established to interfere with memory. If repeated training calls into play additional circuits or generates still further cerebral protein synthesis in relevant circuits, the need for extensive inhibition of cerebral protein synthesis would be required. Because of the phenomena of consolidation and redundancy it would seem necessary to establish inhibition of cerebral protein synthesis to a marked degree and for a considerable duration to insure that all the participating cells are extensively inhibited and also that inhibition is maintained long enough to insure that the critical process is interfered with. These inferences are supported by studies with inhibitors of cerebral protein synthesis. VI. Inhibition during Training

A particularly useful approach to the study of the possible role of cerebral protein synthesis in memory is to administer an inhibi-

PRO=

SYNTHESIS INHIBITORS AND MEMORY

187

tor prior to training so that marked inhibition is established during training. In this way, protein synthesis dependent memory storage processes which occur during training could be blocked. This approach has the potential disadvantage that the drugs might interfere with the training process per se. Fortunately, cycloheximide and acetoxycycloheximide have no detectable effect on training in mice in the situations studied, as described below. In these studies the inhibitor has been administered either intracerebrally or subcutaneously. The former mode of administration has the advantage that a smaller total amount of the drug must be administered to achieve marked inhibition of cerebral protein synthesis. The latter has the advantage that onset of action is much more rapid since it is not dependent on diffusion through brain substance from intracerebral injection sites. Whereas intracerebral injections of acetoxycycloheximide do not produce maximal inhibition of cerebral protein synthesis for at least a few hours after injection (Barondes and Cohen, 1967b), subcutaneous administration produces maximal inhibition within about 10 minutes (Barondes and Cohen, 1968a). With the latter method the

-

Acetoxycycloheximide

---a

0.15M NaCl

"3 0 b

- l , l2 I4 I 6I I8 I 1I I0 I 1 I 2I u 14 16 18 20

0

TRIAL

FIG.1. Percentage of correct responses made on each trial during training in a light-dark discrimination. Mice were injected intracerebrally with acetoxycycloheximide or saline 5 hours before training to escape shock by choosing the lighted limb of a T maze to a criterion of 5 out of 6 consecutive correct responses. The numbers in parentheses indicate the cumulative percentage of mice which had learned to criterion by the indicated trial. (For details see Cohen and Barondes, 1968a.)

188

SAMUEL H. BARONDES

animal need not have been under a condition of inhibition of protein synthesis for a prolonged period of time before training. The results with both modes of administration are identical. Mice whose cerebral protein synthesis was inhibited 90-95% by preinjection with cycloheximide or acetoxycycloheximide learn normally. They require precisely the same number of trials to reach any criterion in a T maze where the correct side is either marked by position (Barondes and Cohen, 1967b, 1968a) or by the presence of light ( Cohen and Barondes, 1968a). Furthermore, the progressive increase in performance of correct responses of a group from trial to trial is identical in acetoxycycloheximide and saline-treated groups (Fig. 1). These findings suggest that the learning process is not interfered with by marked inhibition of cerebral protein synthesis. This conclusion is further supported by the h d i n g that retention, when measured 3 hours after training in several maze tasks, is completely normal (Fig. 2) (Barondes and Cohen, 1967b, 1968a; Cohen and Barondes, 1968a,b). Therefore, not only is learning normal but memory for at least 3 hours after training (here designated Short-term” memory) is not affected by marked inhibition of cerebral protein synthesis. However, when such mice are tested 6 or more hours after training (here designated “long-term” memory) retention is found to be markedly impaired (Fig. 2 ) (Barondes and Cohen, 1967b, 1968a; Cohen and

c

.

8 40I

I

I

3

I

I

I

6 DAYS

z+ FIG. 2. Effect of acetoxycycloheximide on cerebral protein synthesis and memory. Mice were injected subcutaneously with 240 pg of acetoxycycloheximide 30 minutes before training in a light-dark discrimination as described in Fig. 1. Different groups were tested for retention at each of the indicated times. (For details, see Barondes and Cohen, 1968a.)

PROTEIN SYNTHESIS INHIBITOFS AND MEMORY

189

Barondes, 1968a,b). This impairment is apparently permanent since groups tested 6 weeks after training also show marked impairment of retention by comparison with controls which show very good retention ( Barondes and Cohen, 196%). Therefore, inhibition of cerebral protein synthesis during training interferes with the formation of long-term memory. Demonstration of these effects is dependent on the establishment of a fairly extensive degree of inhibition of cerebral protein synthesis and also on the degree of training administered. Mice trained to a criterion of 3 out of 4 consecutive correct responses in a position discrimination have markedly impaired long-term memory if more than about 90%of cerebral protein synthesis was inhibited during training. However, if less than about 8% of cerebral protein synthesis was inhibited no impairment of memory was observed (Barondes and Cohen, 1967b). When mice received further training to a criterion of 9 out of 10 consecutive correct responses in a position discrimination, memory was normal even though 90-95% inhibition of cerebral protein synthesis was achieved during training (Barondes and Cohen, 1967a,b). Because total inhibition of cerebral protein synthesis was not achieved in this experiment, it is presumed that with marked overtraining the residual capacity for cerebral protein synthesis was su5cient to lead to long-term memory storage. When mice were trained to choose the lighted limb of a maze to a criterion of 9 out of 10 consecutive correct responses marked impairment of long-term memory was still observed (Cohen and Barondes, 1968a). This task is more dficult than the position discrimination, requiring more trials to reach any given criterion. Apparently a more difficult task is more sensitive to interference by inhibition of cerebral protein synthesis even though training is conducted to an identical criterion. When training to a criterion of 15 out of 16 consecutive correct responses was given in a light-dark discrimination there was much less impairment of memory (Cohen and Barondes, 1968b). An amnesic effect of puromycin has also been observed in studies of this kind, but the results have been more di5cult to interpret because of the other effects of puromycin on the brain as indicated above. When mice were injected intracerebrally with puromycin prior to training they showed normal acquisition in an escape task in which a position discrimination was taught to a

190

SAMUEL

H. BARONDES

criterion of 9 out of 10 consecutive correct responses (Barondes and Cohen, 1966). In the latter task retention was normal 15 minutes after training but then deteriorated so that by 3 hours after training and thereafter it was found to be markedly impaired (Barondes and Cohen, 1966). These results differ from those observed with the glutarimide derivatives in two regards. First, the impairment of memory was detectable within 45 minutes of training with puromycin but was not detectable until about 6 hours after training with cycloheximide. Indeed, addition of cycloheximide to the solution of puromycin which was injected antagonized the amnesic effect of puromycin when measured 3 hours after training (Barondes and Cohen, 1967a). The puromycin effect on memory 3 hours after training was also antagonized by pretreatment with the anticonvulsant diphenylhydantoin ( Cohen and Barondes, 1967). These findings suggest that the puromycin effect is mediated in part by abnormalities in cerebral electrical activity which can be antagonized by diphenylhydantoin and that peptidyl-puromycin, whose synthesis can be antagonized by cycloheximide, may be at least in part responsible for the abnormality observed. Other effects on memory attributed to peptidyl-puromycin will be discussed subsequently. Puromycin also differed from the glutarimide derivatives in that it was effective in impairing memory even when training in a position discrimination was continued to a criterion of 9 out of 10 consecutive correct responses (Barondes and Cohen, 1966, 1967a). Apparently the amnesic effect of the disturbances in cerebral electrical activity when coupled with the amnesic effect of inhibition of cerebral protein synthesis can act together to antagonize memory formation despite overtraining. Agranoff et a2. (1966) found that administration of puromycin, prior to training, impaired memory in goldfish tested 4 days later. Shashoua (1968) has also reported amnesic effects of administration of pwomycin prior to training in goldfish. Interpretation of these results are difficult because abnormalities in electrical activity were also observed in this class of organisms ( Agranoff, 1969). Although the puromycin experiments are difticult to interpret, the results with the glutarimide derivatives clearly suggest that cerebral protein synthesis is not required for learning or shortterm memory, but is required for long-term memory. The time when this critical protein synthesis might occur will be considered in the next series of experiments.

PRO"

SYNTHESIS INHIBITORS AND MEMORY

191

VII. Inhibition Shortly after Training

The disruptability of memory by administration of electroconvulsive shock for some period after learning has been interpreted to mean that the form in which memory was stored was changing during this period. If the cerebral protein synthesis, which is apparently required for long-term memory storage, was occurring during this period, one would expect that the establishment of inhibition of cerebral protein synthesis during this period might impair long-term memory. In experiments in which inhibitors are administered prior to training, marked inhibition persists long after training (Fig. 2). Therefore, a protein synthesis dependent process which occurred after training would be impaired by administration of the inhibitor prior to training. Studies of effects of establishing inhibition of cerebral protein synthesis in mice by intracerebral injection of acetoxycycloheximide after training were difficult to interpret since considerable time is required for diffusion of the drug and generalized inhibition to occur. The fact that intracerebral injections given immediately after training had no effect (Barondes and Cohen, 196%) is not surprising since maximal inhibition was not established until hours later. However, since subcutaneous injections have a far more rapid onset of action, the effects of administration by this route were studied (Barondes and Cohen, 1968a). It was found that subcutaneous administration of acetoxycycloheximide immediately after training or 5 minutes after training of a light-dark discrimination in a maze had a slight but significant effect on long-term memory storage, whereas administration 30 minutes after training was without effect (Fig. 3 ) . Using a one-trial passive avoidance situation, which has been shown to be susceptible to electroconvulsive shock for as much as several hours after training (Kopp et al., 1966) subcutaneous injection of cycloheximide immediately after training produced a marked amnesic effect (Geller et al., 1969). An amnesic effect was also observed if injections were given 30 minutes but not 2 hours after training (Geller et al., 1969). From these studies it appears that the protein synthesis apparently required for long-term memory storage may go on for at least minutes after training in mice, although much may have already occurred during the 6-8 minutes of training in the maze task ( Fig. 3 ) . Intracranial injections of puromycin or acetoxycyclo-

192

SAMUEL H. BARONDES

100

5 ao L

# , l

x

-

60

ul

z(3 40

2 20

0-l

s

0 0

0

TIME INJECTED (minutes)

FIG.3. Effect on memory of subcutaneous administration of acetoxycycloheximide at times before or after training. All groups were tested for retention 7 days later. (For details, see Barondes and Cohen, 1968a.)

heximide in goldfish immediately after training produced marked impairment of retention (Agranoff et al., 1965, 1966). The drugs, like electroconvulsive shock, were progressively less effective if given at various intervals within an hour after training and totally ineffective if given longer than that after training (Agranoff et al., 1965, 1966). Repeated injections of puromycin after each of a series of daily training sessions had significant effect on subsequent performance in goldfish (Potts and Bitteman, 1967). The period after training during which memory in the goldfish can be affected by protein synthesis inhibition appears to be somewhat longer than that found in the experiments on maze learning in the mouse but may be quite comparable to that found in the one-trial passive avoidance situation in the mouse. The impression that is gained from these experiments is that cerebral protein synthesis for longterm memory storage may occur over a fairly prolonged period of time after the training experience. It is presumed that during this period memory is stored by a short-term process which is not dependent upon cerebral protein synthesis, and which may persist for 3-6 hours in mice and possibly for longer periods of time in goldfish as will be discussed later. VIII. Inhibition long after Training

In the studies just reviewed it was found that administration of acetoxycycloheximide long after training had no effect on subse-

PROTEIN SYNTHESIS INHIBITORS AND MEMORY

193

quent memory storage. This is indeed an important control for studies of this kind, since it indicates that protein-synthesizing capacity of the brain must be disrupted at about the time of training in order to interfere with long-term memory storage, and that inhibition at a later time, which would have the same nonspecific effects on brain function, has no effect on the storage of older experiences. Unfortunately, the results with puromycin have been far more complex. In a series of experiments the Flexners and their colleagues (J. B. Flexner et al., 1963; L. B. Flexner et al., 1964, 1965) found that in mice bitemporal injections of puromycin up to 3 days after training in an avoidance task produced impaired performance on testing days or weeks later. Injections in other brain regions 1 day after training were ineffective, and if the bitemporal injections were postponed until 1 week after training they too had no effect. This suggested that some abnormality in the temporal regions of the brain could interfere with some aspect of memory storage for up to several days after training but not 1 week after training. Subsequent experiments suggested that puromycin was not acting by inhibiting the synthesis of protein required for long-term memory storage. First, acetoxycycloheximide, in doses which had at least as great an effect on protein synthesis, had no effect if given 1 day after training (L. B. Flexner and Flexner, 1966; L. B. Flexner et al., 1966). Second, intracerebral injections of saline given days after puromycin injections reversed the effect of puromycin injections made 1 day after training (J. B. Flexner and Flexner, 1967; L. B. Flexner and Flexner, 1968a). This effect of saline injections has since been confirmed (Rosenbaum et al., 1968). These results suggest that the effect of puromycin given 1 day after training is on “retrival” of memory rather than on its “deposition.” The Flexners presently believe that the effect of puromycin given 1 day after training is due to formation of peptidyl-puromycin which interferes with the ?etrival’’ process. This is suggested by their finding that acetoxycycloheximide injected together with puromycin 1 day after training antagonizes this effect of puromycin (L. B. Flexner and Flexner, 1966). It is believed that this is due to inhibition of formation of peptidyl-puromycin by acetoxycycloheximide. Although the mechanism of the effect of saline injections given after puromycin is not known, the Flexners (L. B. Flexner and Flexner, 19681,) have presented evidence that peptidyl-puro-

194

SAMUEL H. BARONDES

mycin may persist in the brain for very long periods after puromycin injection and imply that the binding of this product to some component in the brain or the metabolism of peptidyl-puromycin may be influenced by subsequent injections of saline. It should be noted that if puromycin is given prior to training subsequent intracerebral injections of saline do not antagonize its amnesic effect (L. B. Flexner and Flexner, 1968a). This is presumably due to the fact that puromycin in this situation may be impairing memory irreversibly by preventing the synthesis of cerebral protein and also by producing disturbances in cerebral electrical activity during and following the learning situation ( Cohen et al., 19S6). Furthermore, intracerebral injections of saline do not restore memory in mice trained while cerebral protein synthesis was inhibited by acetoxycycloheximide ( Rosenbaum et al., 1968). Injections of saline could not, of course, compensate for the lack of cerebral protein synthesis during and shortly after training which is presumably the basis for the amnesic action of acetoxycycloheximide. The effect of puromycin injections long after training on subsequent performance apparently is not a general phenomenon since in goldfish only injections of puromycin shortly after training have amnesic effects (Agranoff et al., 1965). Although the experiments in which puromycin is given days after training shed little light on the role of protein synthesis in memory, I believe they are extremely important in another regard. They were the first experiments that suggested that in mice there is a change in some aspect of memory storage for several days after training since there was sensitivity to bitemporal injections of puromycin only within the first few days after training but not at 1 week. Subsequent experiments with diisopropylfiuorophosphate (Deutsch et at., 1966) and with actinomycin-D (Squire and Barondes, 1970, described below) provide further support for some change in memory storage for days after training. The mechanism of this change is not known. Since multiple puromycin injections that affect the whole brain and not just the temporal regions have effects if given 1week after training it was proposed (Flexner et al., 1963) that the locus of storage spreads from the temporal region to other parts of the brain during the week after training. An alternative is that temporal lobe processing is required for days after training but not longer, so that at 1 week far more general disruption of brain function must be produced with puromycin to impair performance.

PROTEIN SYNTHESIS INHIBITORS AND MEMORY

195

The experiments with delayed puromycin injections were initially conceived of as possibly testing the participation of cerebral protein synthesis in the “maintenance” of memory. Unfortunately, almost nothing is currently known about this aspect of memory. When memory is stored, is it stored forever? If so, are the molecular changes which mediate memory permanent, or is there turnover? It may well be that if some specific group of cerebral proteins must be synthesized for long-term memory storage these proteins have a finite survival time and must be regenerated by some mechanism. Inasmuch as intracerebral injection of acetoxycycloheximide may inhibit cerebral protein synthesis markedly for many hours, it would appear that such proteins survive at least for many hours since injection of the drug at any time beyond the period immediately around training has no apparent amnesic effect. Because it has not been possible to markedly inhibit cerebral protein synthesis for prolonged periods (days) without producing death, it has not been possible to determine if there is a “maintenance” phase of memory storage with regeneration of appropriate proteins with half-lives which could be as short as several days. IX. Inhibition of RNA Synthesis

Protein synthesis may be regulated by increased synthesis of messenger RNA or by increased “translation” of existent “stable” messenger RNA. If one accepts that cerebral protein synthesis is required for long-term memory it is important to attempt to determine the role of RNA synthesis in this process. Unfortunately, inhibitors of RNA synthesis that are presently available are not as well suited for behavioral studies as the glutarimide inhibitors of protein synthesis. Two classes of inhibitors of RNA synthesis have been used-analogs of constituents of RNA such as 8-azaguanine which are incorporated into RNA and render it abnormal, and inhibitors which block RNA transcription such as actinomycin-D. The first study of effects of an inhibitor of macromolecular synthesis on memory was done by Dingman and Sporn ( 1962) who administered 8-azaguanine into the subarachnoid space of rats before or 30 minutes after training in a water maze to a criterion of 15 consecutive correct responses. They found that the drug-treated animals injected prior to training had impaired acquisition. If the drug was given 30 minutes after training it had no effect. Because the drug impaired acquisition it was not

196

SAMUEL H. BARONDES

possible to decide whether RNA synthesis was required for acquisition or whether some other action of the drug (such as interference with other guanosine-dependent reactions ) was interfering with performance of the task. Since actinomycin-D had no detectable effect on acquisition in a number of situations (Cohen and Barondes, 1966), it seems unlikely that the 8-azaguanine was preventing acquisition by preventing the normal synthesis of RNA required for the acquisition process. Therefore, 8-azaguanine probably has other effects which make its usefulness limited for studies of this kind. Interpretation of other effects of this drug which have been related to memory (Chamberlain et al., 1963) are also limited by this daculty. Actinomycin-D also has serious drawbacks. Injection of the relatively large doses of the drug required to produce marked inhibition of cerebral RNA synthesis produces illness and death within hours to days after the drug is given (Barondes and Jarvik, 1964; Cohen and Barondes, 1966). The drug must be injected intracerebrally since large subcutaneous doses have little effect (Barondes and Jarvik, 1964). With the large doses required to inhibit about 95% of cerebral RNA synthesis, substantial evidence of toxicity is already apparent within 10 hours of administration of the drug. With smaller doses, onset of evidence of illness is delayed somewhat. Neurological abnormalities are apparent within 1 day of injection, and marked neuronal abnormalities have been observed (Appel, 1965; Koenig and Lu, 1967). Because of rapid onset of toxic effects with large doses, testing could not be conducted reliably more than a few hours after the training procedure, since it was necessary to pretreat the animals about 5 hours prior to training with the drug to allow for diffusion from intracerebral sites. As with inhibitors of cerebral protein synthesis, administration of actinomycin-D intracerebrally prior to training, in amounts which inhibited approximately 95% of cerebral RNA synthesis during training, had no effect on acquisition (Cohen and Barondes, 1966). Two to 4 hours after training, when the animals were still testable, they were found to have intact memory when training had been conducted in a Y maze or a double T maze to a criterion of 9 out of 10 consecutive correct responses (Cohen and Barondes, 1966). This does not exclude a possible role for RNA synthesis in long-term memory since, in this experiment, there was probably considerable "overtraining" and also since, studies with the glutari-

PROTEW SYNTHESIS INHIBITORS AND MEMORY

197

mide derivatives suggest that the short-term memory process persists for more than 3 hours. Actinomycin-treated mice, which received relatively brief training in passive avoidance situations, showed normal retention when tested up to 3 hours after training (Barondes and Jarvik, 1964), but here again studies at longer times after training could not be done. In a subsequent study (Barondes and Cohen, 1967b), doses of actinomycin-D which inhibited approximately 70-75% of cerebral RNA synthesis were administered prior to training in a position discrimination to a criterion of 3 out of 4 consecutive correct responses. With the lower dose of the drug it was possible to test for retention 24 hours later, since the animals still showed no incapacitating signs of illness at this time. Retention here was normal. However, it is not possible to conclude that cerebral RNA synthesis is not required for long-term memory since the level of inhibition produced may well have been inadequate to impair retention in this behavioral situation. In a study of memory 24 hours after passive avoidance training, Goldsmith (1967) found substantial retention in actinomycin-D treated rats, but it is not clear what degree of inhibition was achieved during training. Meerson et al. (1966) have reported that intracerebral injections of actinomycin-D prior to training produce impaired performance 1 day later. Since neuronal abnormalities and systemic toxicity would be expected to be well developed by this time (Appel, 1965; Koenig and Lu, 1967), and since even small doses have been shown to produce abnormalities in cerebral electrical activity (Nakajima, 1969), it is not possible to draw conclusions from these experiments. To evaluate the possible role of toxic effects it is necessary to compare the effect of administration of actinomycin-D before or immediately after training with its effects if given somewhat longer after training. The toxic effects would be expected to be identical at the time of testing one or more days later, and if the effects on memory are only observed with injections closely related to the time of training, this might be taken as evidence that RNA synthesis is required for long-term memory storage. Agranoff et al. (1967) have found that intracranial injection of actinomycin-D in goldfish immediately after training markedly impairs memory when tested 3 days later, but injections 1 hour after training were ineffective as was the case with intracranial administration of acetoxycycloheximide or puromycin. In contrast, they found that administration of cytosine arabinoside, an

198

SAMUEL H. BARONDES

inhibitor of DNA synthesis, was without effect (Casola et d., 1968). Recent studies using minimal training in mice (Squire and Barondes, 1970) indicate that delayed injections of actinomycin-D may produce effects similar to puromycin. Mice injected in the temporal regions of the brain with 1pg or 30 pg actinomycin-D 3 hours before or 1 day after training in a position discrimination to a criterion of two correct responses had no retention when tested 1 day after injection. Injections in the frontal regions 1 day after training or injections in the temporal regions 7 days after training had no effect. Because of the marked neuronal damage produced, these experiments shed little light on the possible role of cerebral RNA synthesis in memory. They are of interest primarily because they provide further support for the finding that there is a change in some aspect of memory between 1 and 7 days after training as discussed in Section VIII. On the basis of available evidence the firmest conclusion that can be drawn is that cerebral RNA synthesis does not appear to be required for learning or short-term memory. Experiments have been done which are consistent with a requirement for cerebral RNA synthesis during or around the time of training for long-term memory storage, but the role of toxic effects in producing impaired performance cannot be properly evaluated. X. Short-Term and long-Term Memory

The studies with the glutarimide antibiotics provide evidence for the existence of a short-term memory process which appears to be independent of cerebral protein synthesis and a long-term process which appears to be dependent on cerebral protein synthesis. This is inferred from the finding that mice trained while 95%of their cerebral protein synthesis is inhibited remember normally for at least 3 hours after training but have markedly impaired memory 6 hours after training and thereafter, The long-term process appears to be dependent on cerebral protein synthesis during training or within minutes thereafter or both; but its absence, produced by inhibition of cerebral protein synthesis during training, is not detectable until the short-term process has decayed. Implicit in this formulation is the concept that the short-term and long-term processes are based on different mechanisms. Some have questioned this and prefer to view memory as a unitary process which might continue to be “strengthened” for some time

PROTEIN SYNTHESIS INHIBITORS AND MEMORY

199

after training. However, it does not seem useful to call a process unitary if it has one component which is independent of cerebral protein synthesis and another which is dependent on cerebral protein synthesis. It might be true that both short-term and longterm memory are mediated by cerebral protein synthesis. In this view, short-term memory could be stored by the limited proteinsynthesizing capacity that escapes inhibition during training since total inhibition is not produced by these drugs. If this were true, then this protein would have to be substantially degraded within 6 hours after training since short-term memory has decayed by this time. One would not expect such rapid turnover for protein used for memory storage. Furthermore, since establishment of prolonged inhibition 30 minutes after training in a maze has no effect on memory measured 6 hours later (Barondes and Cohen, 1968a), the protein whose synthesis is presumed to mediate memory storage does not appear to turn over particularly rapidly. One would have to conclude from this that either the protein which mediates shortterm memory is more labile than that for long-term memory, or that such a large amount is made that substantial decay during a 6-hour period does not produce detectable amnesia. Although the data do not exclude a unitary process based completely on progressive cerebral protein synthesis during training and for some period thereafter, they are suggestive of different short-term and longterm processes. Recent experiments have attempted to investigate a possible conversion of short-term to long-term memory in mice. This was made possible by subcutaneous administration of a short acting dose of cycloheximide before training. With this dose more than 90% of cerebral protein synthesis is inhibited during training but less than 20% is inhibited 3 hours later, a time when short-term memory is still detectable. Even though the information acquired from training is still available at a time when cerebral proteinsynthesizing capacity has largely recovered, long-term memory is not spontaneously formed (Barondes and Cohen, 196813). This suggests that in addition to a relatively intact cerebral proteinsynthesizing capacity, and the retention of information in a shortterm form, an additional factor may be necessary for the generation of long-term memory. One major difference between mice in this situation and mice during the time of training is that the latter are in a state of

200

SAMUEL H. BARONDES

“arousal.” For this reason we studied the effect of various “arousal producing manipulations” on the possible conversion of short-term to long-term memory. It was found that cycloheximide-treated mice, whose cerebral protein-synthesizing capacity has largely returned to normal 3 hours after training, were capable of forming longterm memory if injected at this time with amphetamine or corticosteroids or if given foot shock (Barondes and Cohen, 1968b). All these factors presumably produce arousal, so that the combination of arousal (which is typical of the learning experience) plus persistence of the information and recovery of cerebral proteinsynthesizing capacity could lead to what we believe may be a “conversion” of short-term to long-term memory. If protein synthesis inhibition was reinstated prior to injection of amphetamine 3 hours after training, long-term memory was not observed (Fig. 4). Therefore this conversion apparently depends on intact cerebral protein-synthesizing capacity as does usual long-term memory 70 60

!.-

. Amphetamine : 3hr

50

7. Y)

s 40 rrn

p

.-

30

0 >

m

8

20

~ - - ~ ~ + o x y c y c l o h e x i m i d:e2Vzhr Amphetamine : 3hr

10

0

I

I

0.25

0.50

I

I

I.o

Amphetamine ( m g / k g )

FIG.4. Effect of amphetamine on memory and its antagonism by acetoxycycloheximide. All mice were injected with 0.12 gm/kg of cycloheximide 30 minutes before training. To reinstate marked inhibition of cerebral protein synthesis one group received acetoxycycloheximide 2% hours after training. Both groups received amphetamine 3 hours after training and retention was measured 7 days later. (For details, see Barondes and Cohen, 1968b.)

PROTEIN SYNTHESIS INHIBITORS AND MEMORY

201

storage. Likewise, if the amphetamine was given 6 hours after training, a time when the short-term memory process was no longer detectable, long-term memory was not formed ( Barondes and Cohen, 1968b). The conclusion from these studies is that long-term memory storage requires information, an intact cerebral proteinsynthesizing capacity, and the presence of an appropriate state of arousal. The arousal may instruct some regulatory process in the brain to “print” the information in a long-term form, as has been suggested by Livingston ( 1967). Results on the effects of puromycin in goldfish also suggest the existence of different mechanisms for short-term and long-term memory. Injection of puromycin 1minute after training produces a detectable impairment 6 hours later but marked impairment is not observed until several days later (Davis and Agranoff, 1966). If one accepts that in this situation puromycin is impairing memory by interfering with the synthesis of protein required for memory, these experiments imply that the protein synthesis-independent phase (short-term memory) may persist, at some level, for days after training. It would appear that the protein synthesis-dependent phase may not begin in the goldfish, until the animal is removed from the training apparatus. If goldfish are retained in the training apparatus for 1 hour after training long-term memory remains susceptible to injections of puromycin (Davis and Agranoff, 1966; Davis, 1968), whereas it is resistant to puromycin if the goldfish were removed from the training apparatus for an identical period. These experiments have been taken as evidence that there must be a “trigger” which directs the onset of the protein synthesis-dependent long-term memory storage process. These results, like those in mice, suggest that there are different mechanisms for short-term and long-term memory storage. XI. Conclusion

There are two major conclusions which can be drawn from this body of work-that the brain can function well enough for learning to occur despite marked inhibition of either its RNA or proteinsynthesizing capacity; and that long-term memory storage is apparently dependent on cerebral protein synthesis during training or within minutes thereafter. The evidence for the participation of cerebral protein synthesis in long-term memory storage rests pri-

202

SAMUEL H. BARONDES

marily on results with cycloheximide and acetoxycycloheximide. I will conclude by considering objections to the interpretations we have made from studies with these drugs. These objections are of two kinds: (1) The glutarimide derivatives not only inhibit cerebral protein synthesis but also have some other action on the brain which is responsible for their effect on memory; and (2) inhibition of ongoing “constitutive” protein synthesis impairs replacement of the brain proteins undergoing degradation, and this produces abnormal cerebral functioning which impairs memory storage. Although these objections are compelling and can never be absolutely ruled out, the available experiments provide many arguments against them. The objection that the amnesic effect of a glutarimide derivative is related to some “side effect” is argued against by the relative effectiveness of cycloheximide and acetoxycycloheximide. These drugs are closely related structurally but differ very markedly in their potency as inhibitors of cerebral protein synthesis. It is necessary to give about 10 to 20 times as much cycloheximide as acetoxycycloheximide either intracerebrally or subcutaneously to achieve identical inhibition of cerebral protein synthesis. The amnesic effects of the drugs are correlated with the degree of inhibition of cerebral protein synthesis produced rather than with the absolute amount of drug given (Table I). Therefore, the hypothetical side effect would have to have the same dose relationship as the protein TABLE I EFFECTOF ANALOGSOF CYCLOHEXIMIDE ON MEMORYO

% Inhibition D w None Cycloheximide Cycloheximide Acetoxycycloheximide Isocycloheximide

6-Dihydrokocycloheximide

Dose (rg)

of protein synthesis

% Savings

-

-

73

100 15 15 100 75

89 56 91

0 4

33 70 28 71 69

0 Mice (15 in each group) were injected in each side of the brain with 15 microliters of saline containing the indicated drugs. The total dose given is shown. Four hours later they were trained in a light-dark discrimination as described in Fig. 1. Retention was determined 7 days later. Cerebral protein synthesis was estimated in groups of 3 mice (Barondes and Cohen, 1968a) in the interval between 4 and 435 hours after injection. (Unpublished data of Cohen and Barondea.)

PROTEIN SYNTHESIS INHIBITORS AND MEMORY

203

synthesis inhibiting action of these two drugs. Furthermore, other derivatives of the glutarimide series which have little or no effect on cerebral protein synthesis have no effect on memory when administered in similar doses (Table I ) . From these experiments an unrelated side effect remains possible but does not appear to be particularly likely. The second objection requires that there are critical brain proteins required for normal function which are present in limiting concentrations and turn over very rapidly. Thus one can imagine a brain protein with a half-life of 1 hour whose level must be rigidly maintained for normal functioning of some general brain process not specifically related to long-term memory storage. When acetoxycycloheximide is administered subcutaneously 5 minutes prior to training, replacement of this brain protein would be immediately blocked so that, during training or within minutes thereafter, brain function would already be somewhat abnormal and impairment of long-term memory would result. Administration of the inhibitor some time after training would also interfere with the replacement of this protein, but now the memory storage process would have been set in motion, so that nonspecific abnormalities in brain function might have no effect. Clearly our experiments are consistent with a requirement for synthesis of cerebral protein either specifically for memory storage or for replacement of a limiting and rapidly turning over constitutive brain protein which nonspecifically influences the memory storage process. However, the existence of this hypothetical rapidly metabolized protein whose precise concentration is required for normal general brain function is not suggested by the apparently normal learning found even after hours of profound inhibition of cerebral protein synthesis produced in some experiments (Barondes and Cohen, 1967b, 1M8a). In contrast, the possibility that protein is qzwcificdly synthesized to store each long-term memory implies the existence of a highly efficient mechanism used by cells in a variety of situations for regulatory purposes. It is for this reason that we interpret our results, and those of others, as indicating that cerebral protein synthesis during training and within minutes thereafter is specifically required for a long-term memory storage process. REFERENCES Adair, L. B., Wilson, J. E., Zemp, J. W., and Glassman, E. (1968a). Proc. Natl. Acad. Sci. U. S. 61, 808.

'

204

SAMUEL H. BARONDES

Adair, L. B., Wilson, J, E., and Glassman, E. (1968b). Proc. Natl. Acad. Sci. U. S . 61, 817. Agranoff, B. W. (1969). In ”Protein Metabolism of the Nervous System” (A. Lajtha, ed.). Plenum Press, New York (in press). Agranoff, B. W., Davis, R. E., and Brink, J. J. (1965). Proc. Natl. Acad. Sci. U. S. 54, 788. Agranoff, B. W., Davis, R. E., and Brink, J. J. (1966). Brain Res. 1, 303. Agranoff, B. W., Davis, R. E., Casola, L., and Lim, R. (1967). Science 158, 1600. Appel, S . H. (1965). Nature 207, 1163. Appleman, M. M., and Kemp, R. G. (1966). Biochem. Biophys. Res. Commun. 24,564.

Barondes, S. H. ( 1965). Nature ZOS, 18. Barondes, S. H., and Cohen, H. D. (1966). Science 151, 594. Barondes, S. H., and Cohen, H. D. (1967a). Bruin Res. 4 44. Barondes, S. H., and Cohen, H. D. (1967b). Proc. Natl. Acad. Sci. U . S . 58, 157. Barondes, S. H., and Cohen, H. D. (1968a). Science 160, 556-557. Barondes, S. H., and Cohen, H. D. (196Sb). Proc. Nutl. Acad. Sci. U . S . 61, 9m29. Barondes, S . H., and Dutton, G. (1969). J. Neurobiol. 1, 99. Barondes, S. H., and Jarvik, M. E. (1964). J. Neurochem. 11, 187. Brown, B. M., and Noble, E. P. (1967). Bruin Res. 6, 363. Bullock, T. H. (1968).Proc. Natl. Acad. Sci. U.S . 60, 1055. Casola, L., Lim, R., Davis, R. E., and Agranoff, B. W. (1968). Proc. Nutl. Acad. Sci. U . S . 60, 1389. Chamberlain, T. J., Rothschild, G. H., and Gerard, R. W. (1963). Proc. Natl. Acad. Sci. U. S . 49, 918. Cohen, H. D., and Barondes, S. H. (1966).J . Neurochem. 13, 207. Cohen, H. D., and Barondes, S. H. (1967). Sdence 157, 333-334. Cohen, H. D., and Barondes, S. H. (1968a). Nature 218, 271-273. Cohen, H. D., and Barondes, S . H. (1968b). Comm. Behuu. Biol. Al, 337. Cohen, H. D., Ervin, F., and Barondes, S . H. (1966). Science 154, 1557. Davis, R. E. (1968). J. Comp. Physiol. €5, 72. Davis, R. E., and Agranoff, B. W. (1966). Proc. Natl. Acad. Sci. U . S.55,555. Deutsch, J. A., Hamburg, M. D., and Dahl, H. (1966). Science 151, 221. Dingman, W., and Sporn, M. B. (1962). J. Psychiat. Res. 1, 1. Flexner, J. B., Flexner, L. B., and Stellar, E. (1963). S d a c e 141, 57. Flexner, J. B., and Flexner, L. B. (1967). Proc. Natl. Acad. Sci. U . S . 57, 1651. Flexner, L. B., and Flexner, J. B. (1966). Proc. Nut2. Acad. Sci. U . S. 55, 369. Flexner, L. B., and Flexner, J. B. (1968a). Science 159, 330. Flexner, L. B., and Flexner, J. B. (1968b). Proc. Nutl. Acad. Sci. U . S . 60, 923. Flexner, L. B., Flexner, J. B., Roberts, R. B., and de la Haba, G. (1964). Proc. Natl. Acad. Sci. U.S . 52, 1165. Flexner, L. B., Flexner, J. B., and Stellar, E. (1965). Exptl. Neurol. 13, 284. Flexner, L. B., Flexner, J. B., and Roberts, R. B. ( 1966). Proc. Nutl. A d . Sd. U. S. 56, 730. Flexner, L. B., Flexner, J. B., and Roberts, R. B. (1967). Science 155, 1377.

PROTETN SYNTHESIS INHIBITORS AND MEMORY

205

Gambetti, P., Gonatas, N. K., and Flexner, L. B. (1968a). J . CeU Biol. 36, 379. Gambetti, P., Gonatas, N. K., and Flexner, L. B. (1968b). Science 161, 900. Geller, A., Robustelli, F., Barondes, S. H., Cohen, H. D., and Jarvik, M. E. (1969). Psychophurmcologia 14, 371. Goldsmith, L. J. (1967). J. Comp. Physiol. Psychol. 63, 126. Hydhn, H., and Egyhazi, E. (1962). Proc. Natl. Acad. Sci. U. S . 48, 1366. Hyd&, H., and Egyhazi, E. (1963). Proc. Natl. Acad. Sci. U. S . 49, 618. HydAn, H., and Egyhazi, E. (1964). Proc. Natl. Acad. Sci. U . S . 52, 1030. HydQ1, H., and Lange, P. W. (1968). Science 159, 1370. John, E. R. (1967). “Mechanisms of Memory.” Academic Press, New York. Kandel, E. R., and Spencer, W. A. (1968). Physiol. Rev. 48, 65. Kanfer, J., and Richards, R. L. (1967). J . Neurochem. 14, 513. Koenig, H., and Lu, C. (1967). Trans. Am. Neurol. Assoc. 92, 250. Kopp, R. Z., Bohdanecky, Z., and Jarvik, M. E. (1966). Science 153, 1547. Lajtha, A. (1964). Intern. Rev. Neurobiol. 6, 1. Lashley, K. ( 1960). I n “The Neuropsychology of Lashley” (F. A. Beach et al., eds.), p. 478. McGraw-Hill, New York. Livingston, R. B. (1967). In “The Neurosciences-A Study Program” (G. Quarton et al., eds.), p. 499. Rockefeller Univ. Press, New York. McGaugh, J. L. (1966). Science 153, 1351. Meerson, F. Z., Kruglikov, R. I., and Goryacheva, I. A. (1968). Dokl. Akad. Nauk. SSSR 170, 741. Nakajima, S. (1969). J. Comp. Physiol. Psychol. 67, 457. Nathans, D. (1964). Proc. Natl. Acad. Sci. U. S . 51, 585. Potts, A., and Bitterman, M. E. (1967). Science 158, 1594. Rosenbaum, M., Cohen, H. D., and Barondes, S. H. (1968). Comm. Behau. Biol. A2, 47. Shashoua, V. E. (1988). Nature 217, 238. Siegel, M. R., and Sisler, H. D. (1963). Nature 200, 675. Sovik, 0. (1967). Biochim. Biophys. Acta 141, 190. Squire, L., and Barondes, S. H. (1970). Nature (in press). Weissman, A. (1967). Intern. Reu. Neurobiol. 10, 167. Zemp, J. W., Wilson, J. E., Schlesinger, K., Boggan, W. O., and Glassman, E. (1966.). Proc. Natl. Acad. Sci. U. S . 55, 1423. Zemp, J. W., Wilson, J. E., and Glassman, E. (1967). PTOC.Natl. Acad. Sci. U . s. ss, 1120.

This Page Intentionally Left Blank

THE MECHANISM OF ACTION OF HALLUCINOGENIC DRUGS ON A POSSIBLE SEROTQNIN RECEPTOR IN THE BRAIN By J. R. Smythies, F. Benington, and R.

D. Morin

Department of Psychiatry, University of Edinburgh, and the Neurosciences Research Program, Massachusetts Institute of Technology, Boston, Massachusetts; Department of Psychiatry, University of Alabama, Birmingham, Alabama

I. Introduction: Characteristics of Hallucinogens . . . . 11. Specification of a Serotonin Receptor Site . . . . . 111. Specification of the Hallucinogens as Central Serotonin Antagonists A. Mescaline Analogs . . . . . . . . . B. Tryptamine Derivatives . . . . . . . . C. Lysergic Acid Diethylamide . . . . . . . D. As-Tetrahydrocannabinol ( THC ) . . . . . . E. Harmine . . . . . . . . . . . IV. Mitomycin . . . . . . . . . . . . V. Type BO Hallucinogens . . . . . . . . . VI. Generalization of the Hypothesis . . . . . . . VII. A Possible Role of RNA . . . . . . . . . VIII. Testing of the Hypothesis . . . . . . . . . IX. Summary . . . . . . . . . . . . References . . . . . . . . . . . . Note Added in Proof . . . . . . . . . .

207 21 1 215 215 219 220 221 223 223 223 224 225 230 231 232 233

1. Introduction: Characteristics of Hallucinogens

Lysergic acid diethylamide (LSD) is one of the most potent biologically active substances known, inducing its psychoactive effect at a dose of less than 1 pg/kg. Gaddum (1953) and Woolley and Shaw (1954) suggested that its potent antiserotonin properties (as determined on peripheral systems) might also be responsible for its central actions. The subsequent demonstration that 2-bromLSD is just as active an anti-serotonin agent as LSD in peripheral systems, and yet is much weaker as a psychoactive compound, cast doubt on this theory. However, more recent evidence suggests that, after all, LSD does exert its central effects via the serotonin mechanisms in the brain. Freedman (1963) has shown that LSD raises the level of 5-hydroxytryptamine (5-HT) in the brain and decreases the level of its main metabolite, S-hydroxy207

208

J. R. SMYTHIES, F. BENINGTON, AND R. D. MORIN

indoleacetic acid (5-HIAA). Diaz et al. (1968) have shown that LSD decreases brain 5-HT turnover rate. Aghajanian et al. (1968) have demonstrated that intravenous LSD, in very small dosage, markedly inhibits the firing of the 5-HT containing cells of the raphe nuclei. Neither of these results was obtained with 2-bromLSD or other nonhallucinogenic relatives of LSD, and the latter result was obtained with mescaline and dimethyltryptamine (DMT) and in higher dosage (Fig. 1).LSD inhibits the effect of 5-HT in the isolated superior colliculus (Kawai and Yamamoto, 1968), and it inhibits the potential produced by 5-HT, and not by acetylcholine ( ACh) in the cortex (Roberts and Straughan, 1967). LSD acts as an 5-HT agonist in the lateral geniculate system (Curtis and Davis, 1961) . Pretreatment of animals with parachlorphenyalanine ( PCPA) , the speci6c 5-HT depleter, markedly potentiates the behavioral disruption produced by LSD (Freedman, 1969); pretreatment of animals and man with reserpine does the same, whereas prolonged administration of a monoamine oxidase inhibitor does the opposite. Marchbanks et al. (1964) have shown that LSD inhibits the binding of 5-HT by synaptosomes. Various studies have

.. (C)

FIG. 1. The formulas of some typical hallucinogens: ( a ) mescnline, ( b ) dimethyltryptamine, (c) d-LSD, and (d) tetrahydroharmine.

ACTION OF HALLUCINOGENIC DRUGS

209

indicated that at low doses LSD can excite rather than inhibit 5-HT mechanisms. Welsh and his co-workers (Wright et al., 1962; Welsh, 1968) have tested a number of hallucinogens on the clam heart, where both 5-HT and dopamine are known to be transmitters (the former controlling the force of the beat, and the latter its rate). All hallucinogens proved to be 5-HT agonists in proportion to their effectiveness as hallucinogens, and none of them affected the response to catecholamines in any way. But what was remarkable in their findings was that LSD produced a maximum biological response at a concentration of M (if given long enough ( 4 hours) to work). This strongly suggests that there are only a limited number of trigger zones in the cell that the LSD molecule has to find. This great biological activity stands in marked contrast to the chemically inert nature of the LSD molecule. Its most remarkable property is its intense T cloud energy contributed by the resonance of the conjugated structure of its four rings. Snyder and Merrill (1966) have demonstrated that LSD possesses a high HOMO (highest occupied molecular orbital) energy, and so the molecule is very lipophilic. The diethylamide side chain is very unreactive. The only points at which LSD can form hydrogen bonds are the indole nitrogen (and even this is not necessary since 1-methyl-LSD is an active hallucinogen) and the amide carbonyl oxygen which can act as an acceptor. However, its structure is stereospecifically defined, and the other three stereoisomers are quite inactive. Any change in the diethylamide side chain, removal of the methyl group or the double bond of the D ring leads to great loss or abolition of activity. The only known position where substitution is tolerated is at the 1position (e.g., methyl, acetyl, and larger groups). Substitution at 2 by bromine or oxygen reduces or abolishes hallucinogenic activity. All this suggests that LSD must fit some receptor site so #closely that it can maintain itself there tenaciously, mainly by weak interactions. The mere inactivation, in the clam heart, of a few receptor sites for serotonin out of the large number of receptor sites on the surface of the muscle cell at the neuromuscular junction could hardly have much noticeable result. The trigger zone may be at a strategically more important site. Siegal and Salinas (19SS) have recently reported that 5-HT inhibits RNA polymerase both in vivo brain slice and in vitro without interfering with its synthesis. They

210

J. R. SMYTHIES, F. BENINGTON, AND R. D. MORIN

also reported that 5-HT interacts strongly with nucleic acids. Furthermore several workers have suggested recently that the biogenic amines not only may act as ordinary “transmitters” whose only function is to depolarize or hyperpolarize membranes but that in addition they may cross into the postsynaptic cell there to act on some biochemical mechanism (Weiss, 1960; Costa, 1959; also see Salmoiraghi et al., 1965). Kety (1969) has suggested that the biogenic amines may mediate control of protein synthesis; the latter is known to be concerned with the consolidation of the memory trace in the brain. Recent data (Barondes and Cohen, 1968; Flexner and Flexner, 1968; H y d h and Lange, 1968) indicate that memory depends on the initiation of protein synthesis at the time of the actual learning situation, or at the most within a very few minutes of it. Thus there may be some messenger from effective (reinforced) synaptic activity to the DNA/RNA system. The only candidates so far suggested for this role have been electrical fields to a repressor protein capable of changing its configuration in an electrical field (Hydkn, 1968) and potassium ions acting on soluble RNA (John, 1967). However, Schmitt (1967) has written as follows: “The action of transmitters may not be limited to topochemical reactions on the external membrane of the neuron, as pictured in conventional receptor theory; they in fact may exert their effect intracellularly by repression or activation of gene expression.” Thus it is possible that 5-HT and norepinephrine (or their metabolites) could act as signals to nucleic acids. This would place the synthesis of new protein in the neuron directly under the control of specific transmitters. This new protein could alter the reactivity of the synapses concerned or even of the whole neuron, and this mechanism could provide an important mechanism of learning. There is considerable evidence that amine transmitters are concerned in emotional reactions in the brain and in reinforcement mechanisms so important for learning. Herskowitz (1967) reports that work on the neurons responding to dark-light schedules in the Californian sea have indicated that DNA-dependent RNA synthesis is linked with the long-term electrical activity of neurons. Neuhoff ( 1968) has reported that LSD (300pg a day for 4 days) in rabbits caused a change in RNA base ratios in the hippocampal increase in cytosine, 4% decrease in guanine). There neurons (11% was also an over-all increase in RNA and of tyrosine-containing protein in these neurons.

ACTION OF HALLUCINOGENIC DRUGS

211

II. Specification of a Serotonin Receptor Site

It seemed worthwhile to investigate how serotonin could interact with the DNA/RNA system and how LSD and the other hallucinogens could interfere with this mechanism. A possible clue here is provided by the formula of the antibiotic mitomycin (Fig. 2 ) . This is an analog of serotonin that prevents the separation of DNA strands on replication by forming a covalent link between two base pairs. The probable mechanism is as follows. The mitomycin molecule links two base pairs by disrupting the hydrogen bond between guanine and cytosine (probably involving the 0-6 position of guanine) and itself forming covalent bonds to the bases by means of its exceedingly active aziridine ring with one base and the active locus in the long side chain to the other. This prevents the separation of the DNA strands and leads to cell death. The number of cross-links produced is always very lownot more than one per 106 or lo7 daltons (Waring, 1968). This suggests that it cannot attack any guanine-cytosine link but only special ones. Thus the site could normally be occupied by serotonin in a specific relationship to the DNA molecule that has some functional significance. Occupancy of the site by mitomycin, which can bond like 5-HT, introduces its highly reactive groups just in the right relation to the guanine-cytosine link so that its attack on this link is expedited. There are only three possible ways in which the mitomycin molecule can attack this link and thus, by derivation, there are three possible locations for the serotonin receptor site in relation to the DNA molecule.

(1) The mitomycin molecule can intercalate between the base pairs. (2) It could attack the link from the side: This would locate the 5-HT receptor site in some protein molecule (possibly the protein moiety of RNA polymerase ) wrapped closely around the OH I

FIG.2. The formula of activated mitomycin. R, can be -OH, -OCHI, or -NH2. The sites of cleavage are marked (Iyer and Szybalski, 1964).

212

J. R. SMYTHIES, F. BENINGTON, AND R. D. MORIN

DNA molecule. However, this approach is rather unrewarding as nothing is known about the protein at this site. ( 3 ) A third possible site is at the very beginning of the DNA molecule. The first phosphate group may be anchored to a short additional segment of the ribose-phosphate chain, or alternatively to amino acid. However, for various reasons this proved unsatisfactory so we turned our attention to (1). Important factors involved in the attachment of 5-HT to its receptor site are its lipophilicity, the energy of its r-cloud, steric factors, and its capacity to form hydrogen bonds (or similar electrostatic interactions). Serotonin possesses three sites for such bonding-the hydroxyl group, the indole nitrogen, and the primary amine group on the side chain. This suggests that at the 5-HT receptor there are three atoms so located and oriented that they can form H-bonds (or electrostatic interactions) with these three sites on 5-HT (Fig. 3). If the side chain is placed in the position shown in Fig. 3 and an external bonding chosen: and if the “down” bonding of the freely rotatable ring hydroxyl (in the plane of the ring) is likewise selected, the three bonding atoms in the site will be arranged in a triangle with sides approximately 9.0 X 8.0 X 7.0 A in length ( as measured from the midpoint of the bond) ; we can call these 01, 02,and 03.The hydroxyl hydrogen bonds to 01, the amine group to 02,and the indole NH to O3 (Fig. 3). The objective now is to detail the specifications and mechanisms of action of the 5-HT receptor site such that blocking by the motley collection of molecules constituting the hallucinogenic drugs may be explained. After much work with molecular models we found that such a specification is possible based on the type of site described above. We will describe first the mechanism of a possible

1

H

8 FIG.3. Postulated orientation of 5-HTat its binding site. This configuration of 5-HT gives the closest “stacking” effect required for intercalation into a nucleic acid.

ACITON OF HALLUCINOGENIC DRUGS

213

FIG.4. Intercalation site in DNA with 0,and 0,shown in their hydrated state, viewed from the side.

reaction with DNA and then the mechanism of a similar reaction with helical double-stranded RNA. Possibly both these mechanisms may represent different biological functions of transmitters or their metabolites. In the DNA molecule there are two atoms located in a position where they could hydrogen bond to molecules intercalated between the base pairs-namely the ring oxygens of deoxyribose on each side, one in the upper left-hand corner of the gap between the base pairs and one in the lower right-hand corner (Fig. 4) (Fuller and Waring, 1964). Because these are located in the hydrophilic portion of the DNA molecule it may be that these atoms would normally be hydrated. If water molecules are hydrogen bonded here with straight bond angles, two hydrogen atoms can be located facing each other approximately 8.0 A apart (measured from the middle of the H-bond) (Fig. 4). If only one molecule of water is bonded this gives a distance of approximately 9.0A between the available H-bonds. The two H-bonds attached directly to the deoxyribose oxygens (without any intermediate water) would be approximately 10.5A apart (Fig. 5). If a molecule of serotonin is intercalated at this site, there are four possible ways of doing it but only one way in which there is a maximal stacking effect (i.e., maximum contact between the Tclouds of serotonin and the base pairs above and below) and hydrogen bonding by its side chain N directly to 0,and by its ring hydroxyl hydrogen to 0;via water (Fig. 6 ) . This locates O,, to which the indolic N is postulated to bond, in some protein wrapped around the DNA molecule. The function of 5-HT at this site would appear to be to inhibit RNA polymerase by binding the two strands of DNA together and so preventing replication. The water mole-

214

J. R. SMYTHIES, F. BENINGTON, AND R. D. MORIN

FIG.5. Plan of the site viewed from above showing the range of horizontal rotation of the water molecules bonded to 0,and 02.The positions for bonding ot serotonin (1) and DOM ( 2 ) are shown by the water H bond, on 0 2 . On 0 1 serotonin bonds directly to 0, as shown and DOM to the water H bond in the position shown. 2,5-Dimethoxyamphetamineand mitomycin bond like DOM.

cules give a fair range of flexibility to the specification of the molecules that can bond at the site-for instance, both the 2,4 and the 2,5 methoxylated amphetamines can bind with good bond angles, but not the 2,3 nor the 2,6. It is now possible to specify a central serotonin agonist. This is any compound that fulfills the steric requirements, that can be transported to the site, and that binds to 01,O,, and 0,. It is also possible to recognize two main types of antagonist:

A. Compounds binding to O3 by an indolic N and failing to bond properly to 0,and/or 0,. B. Compounds bonding to 0,and 0, sufficiently strongly to resist displacement by 5-HT, but not to 0,. In this category we can recognize three subdivisions depending on the degree of hydration of 0, and 0, required for fit: BO-The antagonist connects directly to 0, and 0,; B1-the antagonist connects directly to one oxygen and via a water molecule to the other; B2-the antagonist connects

I

FIG. 6. Mode of bonding of serotonin at this site; black triangle = N.

ACTION OF HALLUCINOGENIC DRUGS

215

to both 0, and 0, via water molecules. Using this classification we can now examine the hallucinogens. 111. Specification of the Hallucinogens as Central Serotonin Antagonists

A. MESCALINEANALOGS Owing to the invaluable work of Shulgin et al. (1969) we now have available a quantity of structure-activity relationship data on the ring methoxylated phenylisopropylamines ( amphetamines ) in the human. The phenylethylamines themselves are too susceptible to attack by amine oxidase to give reliable data. Protection of the amine group from amine oxidase, however, by a-methylation allows us to obtain a more reliable index of the activity of these compounds on the brain. Data from the rat (Smythies et al., 196713) are also useful once it is realized that the metabolism of these compounds is different in man and rat. The latter uses mainly 4-hydroxylation and the former conjugation. Thus a compound not substituted in the 4 position may be active in man and yet inactive in the rat, in which the compound is rapidly inactivated by 4-hydroxylation. Conversely, 4-methoxy compounds will tend to be more active in rat than man. An examination of the fit of these compounds into the site yielded the following results: Monomthoxy compounds. The 2-substituted compound could not meet the requirements, and its N-methyl derivative is used therapeutically ( Orthoxine ) . The 3-substituted compound has not yet been tested. Dimethoxy compounds. Neither the 2,3 nor the 2,6 compounds could meet the criteria and should be inactive. The 2,5 compound can form good H-bonds (with the methoxy groups oriented facing in different directions) as can the 2,4 and the 3,s compounds (in these the methoxy groups point toward each other). The 3,4 compound should act as the 4 (with possibly some steric hindrance). Table I compares the predictions that can be made from the hypothesis and the reported activity of these compounds. It should be noted that to count as good binding the ring system of the molecule should be roughly where the ring system of 5-HT would be while occupying the site, i.e., sharing a maximum stacking effect. Trimethoxy compounds. The effect of adding extra methoxy groups has two conflicting results. On the one hand, adding methoxy

216

J. €3. SMYTHIES, F. BENINGTON, AND R. D. MORIN

TABLE I COMPARISON BETWEEN PREDICTED ACTIVITYAND REPORTEDACTIVITY'OF DIMETHOXY COMPOUNDS _

Methoxy substitution

Predicted activityb

2 3 4 2,3 294 2,5 2,6 315

0 0

314

+0 ++ + !o, t?

+

_

_

_

Reported activityc ? ? 6 ? 5

8 1 ? NH, > OH (in order of potency of resulting compound). This suggests this group is not being used for hydrogen bonding (if it were, the potency should be reversed). V. Type BO Hallucinogens

It will be apparent that no type BO hallucinogens have yet been described, that is, compounds capable of bonding to both 0, and 0,directly without a linking water molecule. The “widest” putative hallucinogen is harmaline, but this is not quite wide enough and its C ring N is in the wrong orientation to bond to 0,. However, two compounds that clearly could bond directly to 0, and 0, are ethidium and proflavine (Figs. 14 and 15), which are known to act as antibiotics by intercalation between base pairs. The fact that the

224

"-YNHz J. R. SMYTHIES, F. BENINGTON, AND R. D. MORIN

b,

\ / (a)

HZN

H

NH,

(b)

FIG.14. Formulas of ( a ) ethidium and (b) proflavine.

NH, groups in each compound are separated by the same distance, together with the fact that on intercalation both compounds can in fact bond in the model, does suggest that their intercalation is stabilized by H-bonding to O1and 0, as was suggested by Fuller and Waring (1964).But there is no evidence that these compounds are hallucinogenic. However, one agent capable of producing a psychosis that fills the specification to be a type BO hallucinogen is quinacrine. This is sterically quite similar to d-LSD with respect to the relation of the x-cloud and the diethylamide side chain. VI. Generalization of the Hypothesis

Two separable aspects of this hypothesis should be borne in mind-( 1) the identification of the receptor site in terms of three atoms (0,,O,, and 0,) whose only specifications are their relative location, their capacity to form hydrogen bonds, and their bonding angles, and ( 2 ) the further identification of these atoms. The particular identification suggested above may be supplemented by others. The similarity between O1and O2and the deoxyribose oxy-

FIG.15. Fit of proflavine into site.

ACTION OF HALLUCINOGENIC DRUGS

225

gens of DNA (as described) may be no more than an unremarkable coincidence. The real 0,, 02, and 0, may belong to a similar site in membrane protein with some helical structure concerned with opening or closing ionic channels, or with some helical RNA. There may even be serotonin receptor sites in more than one such locus. The matter can only be settled by further experiment in such fields as electron microscope autoradiography and nucleic acid chemistry. The reactions generally termed “hydrogen bonding” in this presentation may also include electrostatic interactions with the same steric requirements-for example, a protonated amino group could react with an 0 bearing a negative charge. However for our purpose the precise nature of the bonding is less important than the stereochemistry. VII. A Possible Role of RNA

Investigations using a model of helical RNA kindly loaned by Dr. W. Fuller indicate that the main stereochemical relationships described above between 5-HT and its central antagonists and DNA hold as well for RNA. The extra hydroxyl group is so located that it can also form an H-bond to the primary amino group of 5-HT. The intercalation site in helical RNA ( A configuration) has a different shape from that of DNA in B configuration. The former is a rhomboid and the latter a rectangle (viewed from the side). This entails that 0, and 0, are about 1.5A closer together in RNA than in DNA. Thus 5-HT could bond in the intercalation site in helical RNA without a water link and moreover it can form no less than four hydrogen bonds (NH to 0,: NH to 0 of ribose hydroxyl: OH to 0,: H of ribose hydroxyl to 0 of 5-HT OH), and maximum rr-cloud overlap. Likewise type B, hallucinogens such as 4-methoxy amphetamine can bind in helical RNA with no water link, and type B, hallucinogens need only one water link. The hydroxyl groups in the groove of the RNA helix would appear to offer less strong attachment to the ethylamide side chain of LSD and the complex ring system of THC. Morgan and Austin (1968) have shown that RNA is present in synaptosomes in two forms. One is mitochondria1 RNA and the other is located either in membrane or in the cytoplasm and has properties akin to ribosomal RNA. However, it has not yet been established on which side of the synaptic cleft this latter type of

226

J. R. SMYTHIES, F. BENINGTON, AND R. D. MORIN

RNA is located. Ribosomes have not been located in the axon terminal, but they could exist incorporated in the membrane in such a manner that they would not be seen by electron microscopy. On the other hand, the RNA may be located in the postsynaptic region in the form of the subsynaptic ribosomes described by Bodian (1966) in the spinal cord and that may be associated with the spine apparatus. In either event, the function of 5-HT and NE could be to inhibit or initiate the synthesis of membrane protein (or proteins) at the level of RNA that would alter the electrical characteristics of the synapse so that learning would result: ". . . the existence of subsynaptic ergastoplasm in the adult indicates a continued role of the adult neuron in adaptive adjustments of synapses, such as formation of new contacts" (Bodian, 1966). Furthermore, if RNA is present and actually incorporated in the subsynaptic membrane, part of it may be available to access from the outer surface of the cell so that 5-HT and NE could react with it without actually crossing the membrane and thus affect ionic channels or charge transfer properties of the membrane. Evidence that LSD binds to nucleic acids by intercalation has been presented by Wagner (1969), Yielding and Sterglanz (1968), and Smythies and Antun (1969). Further reports that RNA is present in membrane have been given by Kaspar and Kashing (1969), King and Fitschen (1968), Takagi and Ogato ( 1968), and Tulegenova et al. ( 19f38).A further stereochemical analysis has been made (Smythies, 1969a,b) of the relationship between drugs known to act on neurons and RNA. This has led to the general hypothesis that the channels conducting ions through neuronal and muscle membranes, as well as those carrying amines into synaptic vesicles are composed of helical RNA (or ribonucleoprotein ) . A segment of double-stranded helical RNA orientated vertically through the membrane would form a potential channel capable of transporting Na+ ions into (or K+ ions out of) the neuron. When the base pairs are hydrogen bonded they would form an internal shutter closing the channel. If the hydrogen bonds were disrupted, the two RNA strands would separate thus opening the channel. Hoffman and Lad& (1964) point out that nucleic acids are normally insulators, but if they lie in an electric field orientated parallel to their long axis, and a powerful electron acceptor or donor intercalates at one end, the charge transfer resulting (the addition or removal of an electron) would supply sufEcient energy

ACTION OF HALLUCINOGENIC DRUGS

227

(12.4 eV) to disrupt the base pair hydrogen bonds. The charge transfer sets up a migration of r-electrons and the terminal base pairs become polarized and repel each other. This could set up a cooperative disruption of the rest of the hydrogen bonds joining the base pairs and so the channel would open. A study of the molecular configurations of acetylcholine, its agonists and antagonists, as well as NE, histamine, and glutamate, and their agonists and antagonists indicates that they all have the required stereochemical structure to bond to helical RNA. All the molecules in these groups have the common property of possessing charged atoms M A apart with the correct bond angles to bind in the way described. Thus the general hypothesis suggests that the receptor sites on the neuronal membrane for all transmitters may be made up of a portion of the helical RNA that itself makes up the main part of the channels for conducting ions through membranes. The attachment of the transmitter leads to a charge transfer process that disrupts the hydrogen bonds joining the base pairs and the channels opens. The factors that specify which particular transmitter is active at the site could include the following: (1) the particular base pairs involved, ( 2 ) the amount of RNA left unmasked by the protein moiety of ribonucleoprotein or other molecules, (3) steric hindrance imposed on binding by the same moieties, and ( 4 ) the sense in which the helix is wound. As an example under ( 3 ) , d-LSD as well as its three inactive stereoisomers all bind equally well to native DNA. The factor making only the d-LSD active may depend on steric factors imposed by some of the molecules, e.g., prostaglandin (see below) bound to the RNA. A role for the prostaglandins. The prostaglandins are thought to play an auxillary role in transmitter function without being transmitters themselves. Molecular models indicate that prostaglandins with fully staggered side chains bear a close stereochemical relationship to helical RNA. Two prostaglandin molecules can bond to a four base-pair segment of RNA. One limb of each prostaglandin molecule binds along one ribose-phosphate chain; the other binds along the outer base-pair (of the four) to the other side. Each molecule can form no less than five hydrogen bonds (in the case of 19-hydroxy-prostaglandins ) to elements in RNA ( COOH to ribose OH; 9-0 to ribose OH; ll-OH to guanine NH;S O H to cytosine 0; 19-OH to ribose hydroxyl). The two molecules form a highly lipophilic rhomboid-shaped ring around the primary receptor site (the

228

J. R. SMYTHIES, F. BENINGTON, A N D R. D. MOFUN

bonding sites on the two middle base pairs of the four). Without the prostaglandin molecules, many molecules of the transmitter (such as ACh) would be wasted by bonding to the many phosphate and ribose anionic sites in the normally highly hydrophilic surrounds of the receptor site. With the two prostaglandin molecules in place, these are blanketed off and the transmitter is limited to the effective bonding site. On this hypothesis prostaglandins should only be relevant to minor-groove transmitters (such as ACh and glutamate). They should not be necessary for adrenergic sites in the major groove. The upper prostaglandin molecule can only bind to a CG base pair and the lower one to a GC pair. It is also possible to demonstrate that reserpine can form a very close attachment to helical RNA (but not DNA). The lipophilic tetrahydroharmine moiety could intercalate between base pairs and bond like 5-HT to which it is sterically equivalent. The complex ring system terminating in the trimethoxybenzene moiety now fits neatly down the minor groove of the helix with two additional hydrogen bonds to the hydroxyl groups of ribose to give a total of four hydrogen bonds, multiple weak interactions and the strong r-cloud interaction for binding. If the RNA was forming a channel as described, reserpine binding in this location and manner would clearly block it mechanically. Likewise, yohimbine could also bind to helical RNA but with an interesting difference. Yohimbine is a truncated stereoisomer of reserpine such that reserpine could only bind to helical RNA wound left-handed and yohimbine could only bind to helical RNA wound right-handed. Thus it is possible that the RNA in channels carrying ions are wound right-handed (and thus are blockable by yohimbine) and those carrying amines into synaptic vesicles are wound left-handed (and thus blockable by reserpine). It is further possible to demonstrate that ouabain, which blocks Na+ channels, could physically block an RNA based channel by forming five hydrogen bonds to three adjacent base-pairs. Furthermore veratridine, which potentiates the transport of Na+ across membrane, can form no less than 12 hydrogen bonds to unpaired RNA, and thus could prevent the RNA from reconstituting the helical form and would thus hold the channel open. The stereochemical details of this mechanism have been detailed elsewhere (Smythies, 1969a,b). Likewise tetrodotoxin has a similar array of oxygen and hydroxyl groups in the precise steric codguration

ACTION OF HALLUCINOGENIC DRUGS

229

required to bind to single stranded RNA, only the radius of the circle around which they are aligned is much smaller than is the case for veratridine. This suggests that tetrodotoxin may react with an acutely flexed portion of single stranded RNA or would block a channel running down the middle of a piece of helical RNA with disrupted base pairs. Molecular models indicate that the following putative transmitters can bind only to particular groupings of base-pairs, which therefore can specify the receptor site of the prostaglandin-RNA complex.

(1) Muscarine. The active stereoisomer can bind only to a GC: GC combination in the narrow groove (ring 0 to guanine NH; N+ to cytosine 0) with an additional bond from its hydroxyl group to the adjacent prostaglandin carboxy OH. ( 2 ) Nicotinic site. If both nitrogen atoms of nicotine are protonated this will bind only to a CG: GC combination (to both cytosine 0 s ) . ( 3 ) Acetylcholine. In the muscarinic site this binds like nicotine (noncarbonyl 0 to guanine NH; N' to cytosine 0; carbonyl 0 to prostaglandin carboxy OH). In the nicotinic CG:GC site, it binds differently to nicotine, i.e., by its carbonyl 0 to the guanine NH and N+ to cytosine 0. ( 4 ) Glutamate. This can bind to a GC:CG combination ( 0 to guanine NH in each case). (5) Glycine can bond across any single base pair (0 to NH; NH, to 0). ( 6 ) Adrenergic site. 2-Substituted phenothiazines like chlorpromazine and perphenazine can intercalate between base pairs only from the major groove, since the intercalation site, seen from above, is trapezoidal in shape. This locates the long phenothiazine side chain (with its charged groups 3 A apart in the fully staggered form) running down the floor of the major groove binding to successive NH or 0 groups which are themselves located 3 A apart. Since phenothiazines are mainly antiadrenergic in their action, this suggests that the adrenergic receptor site may also be in the major groove. A good site is provided by a UA:UA base pair combination, The ring hydroxyls bond to the two uridine 0 s and this locates the p-hydroxyl and amino groups correctly to bond to the two double-bonded oxygens of two successive phosphate groups.

230

J. R. SMYTHIES, F. BENINGTON, AND R. D. MORIN

VIII. Testing of the Hypothesis

The advantage of this hypothesis is that it postulates alternative but precise mechanisms of the mode of action of serotonin and the interactions of this amine by the hallucinogens. This specificity enables us to make quite specific predictions that can easily be tested by experiment.

1. Hallucinogens may protect RNA polymerase against inhibition by 5-HT. 2. Both should also bind to DNA (or helical RNA). 3. N,N-dimethyl derivatives of type B1 antagonists should be prepared and tested for activity (e.g., 4 and 3,4 substituted amphetamines). They should be less active. Deaminated derivatives should be inactive. 4. l-Methyl and l-benzyl derivatives of type A antagonists should be inactive. 5. Other substituents besides the methoxy group capable of hydrogen bonding could be tried in the amphetamine series such as NH, or S-CH,. p-Amino-N-methylamphetamine has been reported to be hallucinogenic (Usdin and Efron, 1967). It has already been shown that the methyl group, incapable of H-bonding, is ineffective by itself. 6. THC should inhibit central 5-HT mechanisms as does LSD. Freedman (1969) has already shown that it increases brain 5-HT levels. 7. LSD (and other hallucinogens) as well as 5-HT should offer protection against attack by mitomycin on DNA. 8. The hypothesis that mitomycin and serotonin act at the same site or sites can be tested by a variety of methods. 9. Ribonuclease may depolarize neurons. These predictions can be tested by a variety of techniques. New putative hallucinogens can be tested on humans, or in suitable animal tests (Smythies et el., 1967b). If the rat is used for testing derivatives of amphetamine with an unsubstituted 4 position, these derivatives should be protected from hydroxylation by methylation, otherwise misleading results will be obtained. These compounds can also be tested as central 5-HT antagonists since the hypothesis equates central 5-HT antagonism and hallucinogenic activity. The interaction among 5-HT, hallucinogens and nucleic acids, and

ACTION OF HALLUCINOGENIC DRUGS

231

enzymes associated with nucleic acids, as well as with various antibiotics, should be further explored. IX. Summary

1. It is proposed that serotonin acts at a site in the brain in which nucleic acid is involved. The action may involve RNA in the membrane or DNA or RNA inside the cell. Helical RNA provides a better stereochemical “fit” than does DNA. 2. Serotonin may intercalate between base pairs and further attach by hydrogen bonding to the two sugar ring oxygens (one via a hydroxyl ion) that appear at the site from each side, as well as to a third atom located in some other molecule (perhaps protein) in close apposition to the nucleic acid. In RNA no water link is necessary. 3. Using this model it is possible to define the specifications of a central 5-HT antagonist and it is possible to show that this can account for the known structure-activity relationship data for the hallucinogenic drugs (central 5-HT antagonists). 4. There are a number of interesting chemical relationships between 5-HT and some antibiotics known to act on nucleic acide.g., mitomycin, vinblastine, violacein, chloroquine, and quinacrine. 5. The function of RNA in membrane may be concerned with local protein synthesis, charge transfer, or ionic channels. The function of serotonin at its ribonucleoprotein receptor site may be to initiate or modulate these processes. It is also possible that 5-HT (or its metabolites) has an additional intracellular site of action on RNA or even DNA in the axon terminal or postsynaptic site. Other transmitters such as NE may have similar sites of action. 6. The hypothesis is highly specific and is capable of undergoing experimental tests, a number of which are suggested. 7. The specification of possible receptor sites for other transmitters on RNA is given and an auxillary role for the prostaglandins has been suggested. ACKNOWLEDGMENTS We are grateful to Professor Francis 0. Schmitt for the Neurosciences Research Program Work Session that initiated this hypothesis; a report on this Work Session is contained in the N R P Bulletin, Vol. 12, and to Helmut Neumann for advice concerning the exact form of THC. We thank the following for the benefit of much helpful discussion and advice: R. B. Barlow, G . S. Boyd, Melvin Calvin, George and Ruth Clayton, E. Costa, Harden McConnell,

232

J . R. SMYTHIES, F. BENINGTON, AND R. D. MORIN

Theodore Melnechuk, Marshall Nirenberg, M. J. Waring, and J. H. Welsh. Mr. Frank Ferguson of the Ealing Corporation, Cambridge, Mass. kindly lent the molecular models on which many of these determinations were made. &FERF,NCES

Aghajanian, G. K., Foote, W. E., and Sheard, M. H. (1968). Science 161,706. Barondes, S . H., and Cohen, H. D. (1968). Sdence 160, 556. Bodian, D. (1966). Science 151, 1093. Costa, E. (1959). Intern. Rev. Neurobiol. 2, 175. Curtis, D. R., and Davis, R. (1961). Nature 192, 1083. Diaz, P. M., Ngai, S. H., and Costa, E. (1968). Advan. Pharmacol. 6,75. Flexner, L. B., and Flexner, J. B. (1968). Science 159, 330. Freedman, D. X. (1963). Am. J. Psychht. 119, 843. Freedman, D. X. (1969). NRP Bull. (in press). Fuller, W., and Waring, M. J. (1964). Ber. Bunsenges. Physik. Chem. 68, 805. Gaddum, J. H. (1953). J. Physiol. (London) 121, 15. Herskowitz, I. H. (1967). “Basic Principles of Genetics,” p. 249. Nelson, London. Hoffer, A., and Osmond, H. (1967). “The Hallucinogens.” Academic Press, New York. Hoffman, T. A., and Ladik, J. (1964). Cancer Res. 21, 474. Hyd&, H. ( 1968). Personal communication. Hydh, H., and Lange, P. W. (1968). Science 159, 1370. Iyer, V. N., and Szybalski, W. (1964). Science 145, 55. John, E. R. (1967). “Mechanisms of Memory.” Academic Press, New York. Kasper, C. B., and Kashing, D. M. (1969). Federation Proc. 28, 404. Kawai, N., and Yamamoto, C. (1968). Bruin Res. 7, 325. Kety, S. S. (1969). In “Beyond Reductionism” (A. Koestler and J. R. Smythies, e d s . ) , p. 33 Macmillan, New York. King, H. W. S., and Fitshscen, W. (1968). Biochim. Biophys. Actu 155, 32. Marchbanks, R. M., Rosenblatt, F., and O’Brien, R. D. (1964). Science 144, 1135. Morgan, I. G., and Austin, L. (1968). J. Neurochem. 15, 41. Neuhoff, V. (1968). Umchau Wiss. Tech. 67,1968. Poschel, B. P. H., and Ninteman, F. W. (1963). LifeS d . 10, 782. Roberts, M., and Straughan, D. W. (1967). J. Neurophyswl. 193, 269. Salmoiraghi, G. C., Costa, E., and Bloom, F. E. (1965). Ann. Rev. Phurmacol. 5, 213. Schmitt, F. 0. (1967). In “The Human Mind” (J. D. Roslansky, ed.), p. 111. North-Holland Publ., Amsterdam. Shulgin, A. T., Sargent, T., and Naranjo, C. (1967). In “Ethnopharmacologic Search for Psychoactive Drugs” (D. H. Efron et ul., eds.), p. 202. U.S. Public Health Serv. Shulgin, A. T., Sargent, T., and Naranjo, C. (1969). Nature 221, 537. Siegal, F. L., and Salinas, A. (1968). Fedmution Proc. 27, 464. Smythies, J. R. (1969a). Comm. Behao. Biol. (in press). Smythies, J. R. ( 1969b). Neurosci. Res. Program Bull. 12 (in press).

ACTION OF HALLUCINOGENIC DRUGS

233

Smythies, J. R., and Antun, F. (1969). Nature 223, 1061. Smythies, J. R., Bradley, R. J., and Johnston, V. S. (1967a). Life Sci. 6, 1887. Smythies, J. R., Bradley, R. J., Johnston, V. S., Benington, F., Morin, R. D., and Clark, L. C., Jr. (1967b). Nature 216, 128. Snyder, S. H., and Merrill, C. R. (1966). In “Amines and Schizophrenia” (H. E. Himwich, S. S. Kety, and J. R. Smythies, eds.), p. 229. Pergamon Press, Oxford. Takagi, M., and Ogata, K. (1968). Biochem. Biophys. Res. Commun. 33, 55. Tulegenova, L. S., Rodionova, N. P., and Shapov, V. S. (1968). Biochim. Biophys. Acta 166, 265. Usdin, E., and Efron, D. H. (1967). “Psychotropic Drugs and Related Compounds.” U.S. Public Health Serv. Vojtechovsky, M. ( 1967). Personal communication. Wagner, T. E. (1969). Nature 222, 117. Waring, M. J. (1968). Nature 219, 1320. Weiss, P. (1960). Arch. Neurol. 2, 595. Welsh, J. H. (1968). Personal communication. Woolley, D. W., and Shaw, E. (1954). Science 119, 587. Wright, A. M., Moorhead, M., and Welsh, J. H. (1962). Brit. 1. PhannacoZ. 18, 440. Yielding, K. L., and Sterglanz, H. (1968). Proc. SOC. Exptl. B i d . Med. 128, 1096. Note added in proof: In helical RNA the 4-substituted amphetamines can bind like 5-HT to 0, and 02.However, there is no obvious fit for 2,5 compounds. Hence, the methoxylated amphetamines remain a puzzle, particularly as we have now shown that N,N-dimethyl-DOM is quite inactive and must therefore use its N group to bind. We have also shown that the 2,5 compound has a higher native fluorescence than the 2,3 or 2,4. Therefore, these groupings may act partly sterically and partly by modulating the quantum chemical properties of the molecule.

This Page Intentionally Left Blank

SIMPLE PEPTIDES IN BRAIN By lsamu Sano Department of Neurology, The Institute of Higher Nervous Activity, Osaka University Medical School, Fukushimo-ku, Osaka, Japan

I. y-Glutamyl Peptides . . . . . . . . . . A. Glutathione . . . . . . . . . . . B. Other y-Glutamyl Peptides . . . . . . . . 11. y-Aminobutpyl and p-Alanyl Peptides . . . . . . A. The Occurrence and Distribution of y-hninobutyryl and pAlanyl Peptides . . . . . . . . . . B. Biosynthesis and Degradation of y-Aminobutyryl and p-Alanyl . . . . . . . . . . . . Peptides C. Physiological Aspects of y-Aminobutyryl and p-Alanyl Peptides . . . . . . . . . . D. Clinical Aspects 111. N-Acetyl Amino Acids and Peptides . . . . . . . A. N-Acetylaspartic Acid (NAA) . . . . . . . B. N-Acetylglutamic Acid (NAG) . . . . . . . . . . . C. N-Acetyl-a-aspartylglutamic Acid (NAAG) IV. Proteases and Peptides . . . . . . . . . References . . . . . . . . . . . .

237 237 237 245 246

249 251 252 252 252 256 256 257 259

Although polypeptides such as substance P and related compounds, neurohypophyseal hormones, and the releasing factors of the anterior hypothalamus have been the subjects of extensive studies in the last decade, relatively few studies have been made of the occurrence of smaller oligopeptides in the brain and of their possible function. In 1961 there were only three characterized brain oligopeptides, namely glutathione, homocarnosine, and carnosine. Tallan et al. (1954) showed that acid hydrolysis of protein-free acid extracts of cat brain resulted in an increase of various amino acids (Table I ) , which was partially accounted for by glutamine, glutathione, N-acetylaspartic acid and homocarnosine, but there was an unaccounted-for fraction which suggested the presence of unknown peptides or their derivatives. Automatic amino acid analysis of protein-free extracts of mammalian brain also demonstrated the presence of unknown ninhydrin-positive compounds which disappeared after acid hydrolysis. In spite of such provocative findings and of the possibilities of their physiological importance, there 235

236

ISAMU SANO

TABLE I AMINOACIDSIN ACIDEXTRACT OF CAT BRAIN BEFORE AND AFTER HYDROLYSIS")

Constituent

HydroEx- lyzed tract extract

Aspartic acid Aspargine Glutamic acid Glutamine a-NHradipic acid Glycine Sarcosine Alanine @-Manine 8-NHpisobutyric acid a-NHm-butyric acid 7-NHz-butyric acid Valine Leucine Isoleucine Serine Threonine Methionine Cystine

29.7 144 1.4 128 186 50 0.2 10.1 24.1 2.7 8.4 8.4 0.2 0.6 0.1 1.3 0.2 23.4 24.2 2.1 1.7 1.9 1.8 1.3 1.2 7.9 7.6 L 2.6 1.5 0.2 1 .o

a

Constituent Taurine Felenine Glutathione Proline Phenylalanine Tyrosine Tryptophan His tidine 1-Methylhitidine 3-Methylhistidine Ornithine Lysine Arginine Creatinine Citrulline Urea E thanolamine Phosphoethanolamine Glycerophosphoethanolamine

Extract 24.0 0.6 27.1 1.6 1.2 1.2 0.7 0.9 0.3 0.3

0.6 2.0 1.4 (57) 0.4 25 20.7 41.9 2.9

Hydrolyzed extract 22.9

L

1.0 0.8 0.5 0.5 0.4 0.3 3.0 1.6 86.5

-

20 26.1 11.7

-

Tallan et al., 1954. The values represent mg/100 gm wet weight of tissue.

have been surprisingly few attempts to isolate and identify these peptides, in marked contrast to the thousands of reports on the isolation, identification, function, and synthesis of the higher polypeptides in the brain. The nomenclature of the di- and tripeptides which will be reviewed here is ambiguous. Their classification as dior tripeptides as a group may be justfied, since naturally occurring peptides can be separated grossly into three classes: di- or tripeptides, biologically active peptides having 7-15 amino acid residues which include most peptide hormones and antibiotics, and the larger polypeptides, which have very diverse biological functions. The purpose of this review is to describe the occurrence, distribution, and metabolism of oligopeptides which have been found in the mammalian brain up to the present time. The physiological and

SIMPLE PEPTIDES IN BRAIN

237

pharmacological actions of some of them and of compounds related to them also are reviewed, although unfortunately there is almost nothing known about this. It is hoped that this review may stimulate some interest in this respect. I. y-Glutarnyl Peptides

A. GLUTATHIONE The presence of glutathione in the brain has been known for some time, and there are many reports on its alterations under various conditions. Only a brief account can be given here. Changes of glutathione in the developing brain of the cat were reported recently by Berl and Purpura (1963). They showed that a maximal increase occurred during the period of maximal development of nervous networks in the neocortex. Developmental changes in different brain areas also were reported by the same authors (Berl and Purpura, 1%6). Glutathione levels increased almost simultaneously in cerebellum, brain stem, and hippocampus, while the mesodiencephalon had a peak at a later stage. Similar developmental changes were observed in rat brain (Oja and Piha, 1966). The administration of psychotropic drugs such as chloropromazine, imipramine, and reserpine had no signscant effects on glutathione levels in brain (Tallan, 1962). Recently it was reported that there is a decrease of glutathione levels in brain caused by mescaline LSD, ethanol, and chloroform (Cernoch and Preininger, 1966; Varma et al., 1968). Although the importance of glutathione as a cofactor of many enzyme systems is well known, there are very few reports on its role in brain function. Hotta and Seventko (1968) proposed that the hexose monophosphate shunt pathway plays a role in maintaining glutathione in the reduced form and hence protects the sulfhydryl groups for the maintenance of cellular integrity.

B. OTHER7-GLUTAMYL PEPTIDES Paper and ion-exchange chromatography and paper electrophoresis of a fraction containing acidic ampholites of brain tissue extracts have revealed the presence of small amounts of many ninhydrin-positive compounds in addition to glutamic acid, aspartic acid, and glutathione. Among various unknown ninhydrin-positive compounds found in extracts of bovine brain, many y-glutamyl

238

ISAMU SANO

peptides have been isolated and identified in this laboratory (Kakimoto et al., 1964, 1965; Kanazawa et at., 1965a,b; Sano et al., 1966). These studies will be briefly summarized here. In some typical experiments, 75 fresh bovine brains weighing 28.5 kg were extracted with trichloroacetic acid, and ampholites and organic cations were absorbed on Amberlite IR-120 (hydrogen form), followed by elution with ammoniacal ethanol. The acidic ampholites were separated from the dried residue of the eluate by absorption on Dowex-1 (acetate form), followed by elution with dilute acetic acid. This was chromatographed on an Amberlite IR-120 column with pyridine-acetic acid buffers. Fractional elution of the ninhydrin-positive compounds was monitored by paper electrophoresis of aliquots of the fraction to obtain elution patterns of 13 fractions, as illustrated in Fig. 1. Each fraction was further purified by ion-exchange chromatographic techniques using Dowex-50 or Dowex-1 and pyridine-acetic acid buffers, to yield 17 fractions containing single ninhydrin-positive substances as shown in Fig. 1. The substances 1-4 were obtained as crystalline materials and were identified by elemental analysis, acid hydrolysis, N-terminal determination, infrared spectra, paper chromatographic, and electrophoretic properties in comparison with synthetic materials. To obtain authentic compounds for comparison, they had to be synthesized. Synthesis of 7-glutamyl peptides was done by coupling N-carbobenzoxyglutamylhydrazide with sodium salts of the required amino acid by azide reaction in cold water solution followed by catalytic hydrogenation. Substances 5-8 were identified by acid hydrolysis, N-terminal determination, and comparison of paper chromatographic and electrophoretic characteristics with synthetic

Star

FIG.1. Elution pattern of ninhydrin-positive substances.

SIMPLE PEPTIDES IN BRAIN

239

materials. For substances 9-13, only amino acid composition and N-terminal amino acids were determined. ‘The results are summarized in Table 11. The identification of the first four peptides, y-glutaminylglutamic acid, y-glutamylglutamine ( Kakimoto et al., 1964), y-glutamylglycine ( Kanazawa et al., 1965a), and y-glutamyl-L-P-aminoisobutyric acid (Kakimoto et al., 1965) were unequivocal, and the glutamic acid residues of these peptides were of the L-configuration, as determined by measuring their optical rotation. The structures of y-glutamylserine, y-glutamylalanine, y-glutamylvaline, and S-methylglutathione were confirmed by their synthesis ( Kanazawa et aZ., 1965b). Of considerable interest is the fact that these eight compounds, which were the major unknown acidic ninhydrin-positive compounds, are all y-glutamyl peptides. The approximate concentrations of these peptides in the bovine brain are given in Table 111, along with their physical properties (Sano et a,?.,1966). The concentrations of these compounds are less than one hundredth of that of glutamic acid. Substance 10 contains glutamyl, a-aminobutyryl, and glycyl residues, with a glutamyl residue at the N-terminal end. It may be ophthalmic acid ( y-glutamyl-a-aminobutyrylglycine) , which occurs in the eye lens (Waley, 1956). Substance 14 was obtained as a crystalline material but its identification remains evasive. Substance 15 contains a N-terminal aspartyl residue and a C-terminal glutamic acid, but was identical neither with synthetic a-aspartylglutamic acid nor /3-aspartylglutamic acid. Substance 16 behaved like cysteinylglutamic acid but its identity is still uncertain. Substance 17 was obtained as a crystalline material and its structure was identified as S-cysteinylglutathione. Because of the possibility of oxidative disulfide formation during isolation procedures, Sano et at. (1966) considered it an artifact, although the endogenous occurrence of such a disulfide is quite possible. The occurrence of the above mentioned y-glutamyl peptides has not been hitherto described in mammalian tissues, although some of them are known to occur elsewhere. yGlutamylalanine (Virtanen and Berg, 1954; Virtanen and Mitikkala, 1960) and y-glutamylvaline (C. J. Morris et al., 1964) occur in plants and in human urine ( Buchanan et d., 1962), and y-glutamyl-D-p-aminoisobutyric acid (C. J. Morris et al., 1961) occurs in iris bulb. The presence of various y-glutamyl peptides in the brain is of special interest because

S

~

OF

TABLE I1 PROPERTIES Y AND IDENTIFICATION OF PURIFIED ACIDICCOMPOUNDS IN

Amino acids

No. 1

released by acid hydrolysis Glu

N-terminal amino acid residue Glu

Identical with

OH

Glu-OH

Glu-OH

I

2

Glu

Glu

NHa

I

I

H-Glu-OH

H-Glu-OH

NHz

NHz

Glu-OH

Glu-OH

I

BOVINEBRAIN"

Compound prepared for comparison

OH

I

THE

OH

I

H-Glu-Glu-OH

I

I

H-Glu-OH

3

Glu

Glu

GlY 4

Glu

Glu

PUB

OH

Gly-OH

Gly-OH

H-Glu-OH

H-Glu-OH

H-Glu-Gl y-OH

L-BAIB

D-BAIB

I

BAIB

I

H-Glu-OH

I

I

H-Glu-OH

I

I

H-Glu-OH

I

OH H-GIu-D-PAIB 5

Glu Ser

Glu

Ser

I

H-Glu-OH

Ser

I

H-Glu-OH

OH

I

H-Glu-Ser-OH

s:

Glu

6

Glu

I

Ala

7

Glu

Glu

Val Glu

8

Glu

H-Glu-OH

H-Glu-OH

Val-OH

Val-OH

H-Glu-OH

H-Glu-OH

-

I

MeCys-Gly-OH

-

MeCys-Gly-OH

B

H-Glu-OH I

i? M

Glu

9

I'

H-Glu-OH I

MeCys GlY

-

AleOH

AlaOH

Glu

-

-

OH

I

H-Glu-Gly-OH NH,

GlY

2

I

H-Glu-Gly-OH

GluaA3A

10

Glu

Probably ophthalmic acid

!

E -

GlY ~~

11

Glu

Glu

Probably polyglutamic acid

12

Glu

Glu

Probably polyglutamic acid

-

13

Glu

Glu

Probably polyglutamic acid

-

a

Abbreviations: PAIB

=

j3-arninoisobutyric acid, aABA = a-aminoisobutyric acid, MeCys = S-methylcysteine.

c.3

e

TABLE I11 CONCENTRATIONS OF 7-GLUTAMYLPEPTIDES IN THE BOVINEBRAINA N D THEIRPHYSICAL PROPERTIESO ~

~

~

~

_

Rr value in solvent systems

_

_ ~

_

_

Peptides

Concentration (dgm wet tissue)

Melting point

aD

A

B

C

Mobility in paper electrophoresis

Y-Glu-Glu Y-Glu-GluNHz Y-Glu-Gl y Y-Glu-BAIB y-Glu-Ala 7-Glu-Ser 7-Glu-Val S-met$hylglutathione

7-10 7-10 0.7-1 0.3-0.4 0.5-1 0.2-0.4 0.1-0.2 0.1-0.2

191.5-192 193-194 192-192.5 196-200 185- 187 175-180 207 194-195

+6.8b +9.8b

0.02 0.09 0.10 0.16 0.14 0.12 0.20 0.20

0.29 0.12 0.24 0.41 0.38 0.19 0.58 0.33

0.23 0.17 0.18 0.30 0.38 0.13 0.70 0.25

10.5 8.3 9.5 1.6 8.5 10.8 7.1 6.9

-

+24.6"

-

+8.@ -226.1'

a Solvent systems of paper chromatography: A = pyridine, acetone, 3 N NHlOH (10:6:5), B = isopropyl alcohol, formic acid, water (8: 1:l), C = n-butyric acid, acetic acid, water (4: 1:1).Paper electrophoresis, 100 V/cm, pyridine, acetic acid, water 1:10: 189 (pH 3.6). Mobility is expressed as cm/30 minutes toward anode. Rt values of glutamic acid was 8, 34, and 23 in solvent systems A, B, and C, respectively, and mobility 3.8 cm. b atSD (c = 1 in 1N HCl). CasD (c = 1 in HzO).

0

SIMPLE PEPTIDES IN BRAIN

243

of the abundance of glutamic acid, glutamine, and y-aminobutyric acid (GABA) in mammalian brain, and their structural resemblance to y-aminobutyryl peptides which are found only in brain tissue. This is in contrast to the abundance of P-aspartyl peptides in human urine as was demonstrated by Buchanan et aZ. (1962) and Kakimot0 and Armstrong (1961).The formation of these peptides from higher polypeptides during the isolation procedures is not probable because of the mild conditions of the isolation procedures, and because there is no evidence for the presence of y-glutamyl linkages in proteins. The formation of these y-glutamyl peptides in the brain is probably catalyzed by 7-glutamyl transferase which is known to catalyze the transfer of the y-glutamyl residue of glutathione to free amino acid (Hanes et al., 1952). Another possibility is that such peptides are formed by direct linkage of glutamic acid with other amino acids in the presence of y-glutamylcysteine synthetase, since cysteine is replaced by other amino acids as a substrate of the enzyme (Strumeyer and Bloch, 1960; Mandeles and Bloch, 1955; ClifFe and Waley, 1958). The y-glutamyl peptides also may be partially degraded by a reversible reaction of the above enzyme. An enzyme system which catalyzes the breakdown of 7-glutamyl amino acids into pyrrolidone carboxylic acid and amino acid has been described by Connell and Hanes (19!X) and by C l 8 e and Waley (1961), but it was not characterized as to its substrate specificity and other properties. Kakimoto et nl. (1967) purified an enzyme catalyzing the breakdown of 7-glutamylglutamine into pyrrolidone carboxylic acid and glutamine, and which has a fairly high substrate specificity. It is distributed widely in mammalian tissues, including the brain. It catalyzes the breakdown of y-glutamylglutamine, y-glutamylalanine, and y-glutamylserine, but not other di- or tripeptides including a-glutamyl peptides. 7-Glutamylglycine was the best substrate for the former lactamase, while the latter does not split the peptides. 7-Glutamyl peptide lactamase seems to be a group of isoenzymes. As will be described below, y-aminobutyryl peptides are formed by the coupling of y-aminobutyryl-AMP with an amino acid. It would be interesting, however, to know whether y-glutamyl peptides are the precursors of y-aminobutyryl amino acids, since the removal of the a-carboxyl group of glutamyl residue should result in the formation of an y-aminobutyryl moiety. The possibility that homocarnosine is formed from

244

ISAMU SANO

glutamylhistidine in the presence of tissue extracts of various mammalian organs has been examined in this laboratory, with negative results (Taniguchi, 1967). A similar unsuccessful attempt was made by Taniguchi et at. (1967) to explain the formation of carnosine by decarboxylation of p-aspartylhistidine, a normal constituent of human urine. The metabolism of 7-glutamyl peptides is summarized in Fig. 2. The turnover rates of these metabolic routes are not known. The activity of y-glutamylglutamine lactamase is high in the kidney and liver and about one tenth of this in the brain, but the activity in the brain is still so high that any y-glutamylglutamine that appeared in the brain would be degraded in a second. The intracellular distribution of y-glutamyl peptides and their catabolic enzymes have not been studied. It would be interesting to see if the formation and breakdown of the y-glutamyl peptides are coupled with the metabolic turnover of glutathione. One difficulty encountered in the investigation of these peptides is that there are no simple methods for their determination in tissues. The physiological functions of the y-glutamyl peptides are far from elucidation. Intraperitoneal injection had no behavioral effects in mice (Kakimoto et a l , 1966), but their application to retinal preparations showed a change in neuronal activity caused by photic stimulation similar to the effect of L-glutamic acid, namely an increase in the frequency and intensity of postsynaptic discharges. Nothing is known about their biochemical and physiological importance in the brain. Glutamic acid Cysteine y-

Glutamylcysteine

Glycine Cysteinylgly cine

Glutathione

y-Glutamyl amino acids

Amino acids

Pyrrolidone carboxylic acid

FIG.2. Metabolism of y-glutamyl peptides.

SIMPLE PEPTIDFS IN BRAIN

245

II. y-Aminobutyryl and P-Alanyl Peptides

Since the identification and discovery of the unique distribution of GABA in the brain (Roberts and Frankel, 1950a; Awapara et al., 1950; Udenfriend, 1950), the biochemistry and physiology of this amino acid have been studied by many investigators. GABA in the central nervous system (CNS) is at present known to occupy a bypassed section (GABA-shunt) at the point of connection of the tricarboxylic acid cycle with amino acid metabolism through 2-0x0glutarate and glutamic acid. From physiological considerations, Florey’s factor I was found to consist mainly of GABA (Bazemore et al., 1956; Brockman and Burson, 1957), and it was shown they have identical neurophysiological effects ( McLennan, 1957). From many physiological and biochemical studies, it is generally agreed that GABA is probably an inhibitory neurotransmitter at the crustacean neuromuscular junction (Otsuka et al., 1966) and also in some synapses of the vertebrate CNS, especially of the cerebellar inhibitory neurons (Otsuka et al., 1968). Derivatives of GABA include y-guanidinobutyric acid ( Irreverre et al., 1957) which has been isolated from calf brain and which is formed by transamidation with arginine ( Pisano et al., 1957; Pisano and Udenfriend, 1958); y-aminobutyrylcholine ( Kiriaki et al., 1958) which has also been identified in dog brain; homopantothenic acid (Bizerte et al., 1955) which may be a constituent of human urine, although the experimental evidence is not conclusive; 7-butyrobetaine ( Engeland and Kutscher, 1910) and y-amino-p-hydroxybutyric acid (Seo, 1957) which were reported to be formed from y-GABA, but the evidence for their occurrence in the CNS needs confirmation. The first evidence of the natural occurrence of 8-alanine in vertebrate tissues was the discovery of carnosine (Gulewitsch and Amiradzibi, 1900; Sif€erd and Du Vigneaud, 1935) and anserine (/3-alanyl-l-methylhistidine) ( Ackermann et a,?.,1929; Behrens and Du Vigneaud, 1937) in muscle. The free form of p-alanine was also found in various tissues (Roberts et al., 1950; Synge, 1951; Hulme and Arthington, 1950). Simple derivatives of p-alanine such as pantothenic acid (Maas and Novelli, 1953; Maas, 1956) and coenzyme-A (Baddiley, 1955) have been found in addition to the compounds listed above. P-Alanine is known as a product of decarboxylation of aspartic acid by certain bacteria (Virtanen and

246

ISAMU SANO

Laine, 1937; Virtanen et al., 1938). In animals, Fink et a2. (1951, 1952, 1956) demonstrated that rat liver slices catalyzed the formation of p-alanine from uracil and p-aminoisobutyric acid from thymine. The administration of uracil caused a significant increase in the amount of @-alanine excreted in urine while that of aspartic acid resulted in no change of p-alanine excretion (Takao, 1967).

A. THEOCCURRENCE AND DISTRTBUTI~N OF ~-AMINOBUTYR~ AND P-ALANYLPEPTIDES In addition to the GABA derivatives mentioned above, the occurrence of free peptides of this amino acid in the brain has been postulated from studies of hydrolysis of brain acid extracts, and Pisano et al. ( 1961 ) purified and identified y-aminobutyrylhistidine (homocarnosine), which was reported to be distributed in higher concentrations in human brain white matter than in the gray matter TABLE IV DISTRIBUTION OF HOMOCARNOSINE IN HUMANBRAIN^ Homocarnosine (pmoles/100 gm) ~

Frontal Gray White Parietal Gray White Temporal Gray White Occipital Gray White Corpus callosum Caudate nucleus Putamen Thalamus Hypothalamus Pons Cerebellum Gray White Medulla oblongata 0

Kanasawa and Sano, 1967.

26.3 1 8 . 2 37.8 k 10.1 37.8 f 5 . 6 35.7 f 12.8 24.6 f 9 . 8 26.6 1 1 0 . 0 34.7 6 . 5 33.4 f 8 . 6 15.3 k 5 . 1 24.1 f 7 . 9 44.0 f 5 . 6 5 6 . 3 k 15.5 60.6 f 5 . 6 30.2 k 9 . 7 51.8 k 4 . 7 52.4 k 15.4 22.8

247

SIMPLE PEI'TIDES IN BRAIN

(Abraham et al., 1962). Later, our group (Kanazawa et at., 1965c) also isolated this peptide from bovine brain, and its structure was confirmed by its optical rotation and the identity of its infrared spectrum with that of the authentic compound. Studies of its regional distribution, using electrophoretic analyses, of the peptide in human brain (Kanazawa and Sano, 1967) revealed no significant difFerence in concentration between white and gray matter. A relatively high concentration was found in thalamus, hypothalamus, and cerebellum in comparison with that of white matter (Table IV). The pattern of distribution was similar to that of GABA in brain (Berl and Waelsch, 1958). The occurrence and levels of homocarnosine in brains of various vertebrates ( Kanazawa and Sano, 1967) were studied phylogenetically. The peptide appears at the amphibian stage in the evolutionary tree, with the exception of cat and dog which have very little amounts of the compound in the CNS (Table V). It is interesting to compare the evolutionary appearance of the peptide with that of GABA, which is known to occur in plants. GABA is distributed in high concentrations in the CNS from early stages of evolution, as well as in the nervous systems of insects and crustaceans (Nakajima and Kakimoto, 1968). This gap may depend on the evolution of carnosine synthetase, although its occurrence in the brain has not been reported, but it is still not explained why dog and cat have large amounts of carnosine in muscle but no homocarnosine in the CNS. Homoanserine ( 7-aminobutyryl-1-methylhistidine ) had also TABLE V HOMOCARNOSINE IN BRAINOF APIIMALS~ Homocarnosine (pmoles/100 gm) Crustacean Fish (mackerel) Chicken Rat Guinea pig Rabbit Cat Dog Kanazawa and Sano, 1967.

0.5 0.5 3.2 0.7 5.6 f 1.5 5.9 1.8 11.4 f 1 . 9 0.5 0.5

+

*

248

ISAMU SANO

been suggested to exist in brain (Pisano et al., 1!361), and its isolation from bovine brain was recently reported by Nakajima et al. ( 1967). y-Aminobutyryl-a-lysine ( Nakajima et al., 1969) also was found in vertebrate brain and its distribution is similar to that of GABA and homocarnosine. Another y-aminobutyryl peptide was found in bovine brain by our group. It is a dipeptide of GABA and an unknown basic amino acid. This unknown amino acid was recently isolated in our laboratory, and analysis gave the formula of CloH,3N,03, but the structure is still unknown. The amino acid is abundant in the brain, less in liver and kidney, apd none in other organs. P-Alanylhistidine ( carnosine) has long been known to be present in skeletal muscle. Many carnosine derivatives such as anserine ( P-alanyl-l-methylhistidine ) ( Ackermann et al., 1929; Behrens and Du Vigneaud, 1937), balenine ( p-alanyl-3-methylhistidine ) ( Pocchiari et al., 1962), and ophidine ( P-alanyl-2-methylhistidine) (Kendo, 1942; Ono and Hirohata, 1956) have been identified in muscle. The structure of the last compound was recently corrected to be the same as that of balenine by nuclear magnetic resonance and N-methyl analysis ( W O E et al., 1968). Another p-alanyl dipeptide, p-alanyl-a-lysine, was recently added to the above list of muscle peptides (Matsuoka et al., 1969). Its formation in muscle occurs in muscle of chicks fed diets containing 2%lysine (Kalyanker and Meister, 1965). The vertebrate CNS also contains carnosine (Abraham et al., 1962; Tsunoo et al., 1966), anserine (Tsunoo et al., 1966), and palanyl-a-lysine ( Matsuoka et al., 1969) but in lower concentration than their 7-aminobutyryl analogs. Another P-alanyl dipeptide, containing the above-mentioned unidentified basic amino acid at the carboxyl terminal, also was found in bovine brain in our laboratory. Carnosine (Perry and Jones, 1961; Hamilton and Dickinson, 1964; Dickinson and Hamilton, 1966) and anserine (Dickinson and Hamilton, 1966) have been reported in human cerebrospinal fluid (CSF) in very low concentration. Abraham et al. (1962:) found very small amounts (0.1 pmole per 100 ml) of homocarnosine in human CSF. However, Dickinson and Hamilton (1966) found carnosine but not homocarnosine in human CSF. Recently, Perry et al. (1968a) have reported that the concentration of the peptide in CSF is higher in children than in adults.

249

SIMPLE PEPTJDES IN BRAIN

B. BIOSYNTHESIS AND DEGRADATION OF Y - A M I N O B ~ ~ Y L AND P - h N Y L PEPTJDES Carnosine synthetase (L-histidine: palanine ligase (AMP); E.C.6.3.2.11) has a broad substrate specificity, and catalyzes the formation of various y-aminobutyryl and P-alanyl dipeptides (Table VI) (Kalyanker and Meister, 1959; Stenesh and Winnick, 1960; Winnick and Winnick, 1959a,b).

-

+ ATP + Ena Ma++Enz-j3-alanine-AMP + PPi Em-8-alanine-AMP + thistidine + L-carnosine + Enz + AMP 8-Alanine

Another enzyme, carnosine-N-methyltransferase ( S-adenosylmethionine: carnosine N-methyltransferase; E.C.2.1.1.12) (Winnick and Winnick, 1959a,b; McManus, 1962) mediates the synthesis of anserine from carnosine. The activity of carnosine synthetase is TABLE VI ENZYMATIC SYNTHESIS OF CARNOSINE AND RELATED AND 7-AMINOBUTYRYL PEPTIDES" P-ALANYL

Amino acids and amines

Product (mpmole)

8-Alanyl peptides PHistidine D-Hktidine 1-Methyl-Dthistidine 2-Methyl-1rhistidine 3-Methyl-thistidine Histamine IrLysine n-Lysine 5-Hydroxylysine L-Ornithine a,e-Diaminopimelic acid L-qyDiaminobutyric acid m,B-Diaminopropionic acid tArginine D- Arginine Agmatine tCitrulline tcanavarine Glycine

26.5 0 15.8 21.8 14.4 0 15.2 0 12.4 13.7 6.12 4.59 2.29 13.6 0 0 0 11.8 0

a

Kalyankar and Meister, 1959.

Amino acids and amines

Product (mpmole)

Peptides of histidine 8-Alanine DL-fl-Aminobutyric acid 8-Aminoisobutyric acid 7-Aminobutyric acid e-Aminovaleic acid e-Aminoc,aproicacid Ira-Aminobutyric acid ~ a y-Diaminobu , tyric acid DIrcr,fl-Diaminopropionic acid 7-Aminobutyryl peptides tLysine tOrnithine >Histidine n-Histidine l-Methyl-DL-hktidine 3-Methyl-~-hktidine tArginine

60.3 35.0 22.2 41.3 15.0 17.9 0 20.6 7.85 16.4 8.80 24.3 0 11.3 6.60 4.34

250

ISAMU SANO

markedly higher in pectoral muscle of young chicks than in adult muscle, and this is also the case with carnosine N-methyltransferase. Another possible synthetic pathway of these peptides is decarboxylation of P-aspartylhistidine, which was found as a normal constituent of human urine (Kakimoto and Armstrong, 1961). However, isotopic studies have contested this pathway ( Aonuma et al., 1968). Whereas the formation of these peptides as well as the synthetic enzymes is established in skeletal muscle, the occurrence of the synthetic enzyme has not been reported in the CNS. The activity of carnosine synthetase with GABA as substrate is about 50%of that with p-alanine. The concentrations of the peptides in various tissues therefore are functions of the concentration of the component amino acids and of the synthetic and degrading enzymes. Indeed, neither homocarnosine nor GABA occur in muscle. Abraham et al. (196l), however, demonstrated the formation of homocarnosine in muscle of rat injected with GABA. Homoanserine may also be formed from GABA and l-methylhistidine in brain by carnosine synthetase. Another synthetic pathway, mediated by carnosine N-methyltransferase, was demonstrated in guinea pig and rat brain by McManus (1962). The concentration of y-aminobutyryl-a-lysine is much less than that of homocarnisine. Since the activity of carnosine synthetase with lysine is 60% that with histidine, and the concentrations of lysine and histidine in brain are comparable, the marked difference in the amounts of the two peptides in brain is difficult to understand. Many factors such as storage, degradation, and intracellular distribution mechanisms must be considered. Degradation of the histidine dipeptides such as carnosine and homocarnosine is effected by carnosinase ( aminoacyl-L-histidine hydrolase: E.C.3.4.3.3) (Hanson and Smith, 1949) which acts on the dipeptides containing imidazole amino acids in the carboxyl terminal and also on the amides of these dipeptides. Carnosinase activity has been found in various tissues of vertebrates, including stomach, spleen, uterus, kidney, liver, heart, diaphragm, bladder, muscle, lung, and brain of rat (Wood, 1957). It was isolated and purified from hog kidney and characterized to be a metalloprotein in which zinc may be bound (Hanson and Smith, 1949; Rosenberg, 1960). Zinc and magnesium activate the enzyme activity in uitro. Anserine and homoanserine are hydrolyzed by the same enzyme in

251

SIMPLE PEPTIDES IN BRAIN

mammals. However, anserinase (aminoacyl-l-methyl-L-histidine hydrolase) has been demonstrated in fish (Jones, 1955, 1956). The lysine dipeptides such as y-aminobutyrl-a-lysine and palanyl-a-lysine are hydrolyzed by a carboxypeptidase ( Kumon, 1969), which mediates hydrolysis of peptides having ornithine, lysine, and arginine in the carboxyl terminal; the amides of these peptides, however, are not substrates for the enzyme. The substrate specificity is different from that of carboxypeptidase B. The activity of the enzyme has been found in various tissues including brain of rat, and in higher activity in muscle and kidney. The difference of the action of the degradation enzymes on histidine and iysine dipeptida may be one reason for the significant differences in the concentrations of these peptides in brain. C. PHYSIOLOGICAL ASPECTS

OF

y

-

A

AND

~P-ALANYL ~

PEPTIDES

The physiological significance of these dipeptides is not clear yet. As Udenfriend et al. (1961) pointed out, the principal interest in carnosine and anserine and their homologues lies in the occurrence of these dipeptides only in excitable tissues; the former in muscle and the latter in the central nervous system. They may have a common role in excitable tissues. The effect of carnosine and anserine on muscle preparations has been considered as being due to their buffering action as ampholites (Davey, 1957), which stabilize the pH of anaerobically contracting muscle (Davey, 1960), or of their ability to chelate traces of metals (Jencks and Hyatt, 1959). Recently, these peptides, as well as imidazole, have been known to stimulate acetylcholine synthesis of insect muscle in vitro (Severin et al., 1966) and to inhibit peroxide oxidation in mitochondria or homogenates of rat cardiac muscle ( Neifakh, 1966). Bocharnikova and Petushkova (1967a) also reported that imidazole, imidazole amino acids, carnosine, and anserine inhibited rat muscle myosin adenosinetriphosphatase ( ATPase) activity in vitro in the presence of a high concentration of ATP and Ca++.These imidazole derivatives protected the ATPase activity of rat muscle myosin in vitro during thermal denaturation or during preincubation with p-chloromercuribenzoate or dinitrophenol (DNP), in contrast with Ca++-facilitated heat denaturation ( Bocharnikova and Petushkova, 196713). In spite of

~

~

252

ISAMU SANO

these experiments, their physiological role in muscle and brain remains to be elucidated.

D. CLINICAL ASPECXS Bessman and Baldwin (1962) reported that five children with juvenile amaurotic idiocy in three families excreted large amounts of carnosine, anserine, and l-methylhistidine in the urine. Normal children and adults in the same families but without neurological symptoms also excreted large amounts of the imidazole compounds. Since carnosine was not detected in blood of the patients, the aminoaciduria and imidazoluria, was thought to be due to a renal mechanism. Two years later, Levenson et al. (1964) reported four cases of juvenile amaurotic idiocy who excreted excessive amount of carnosine in urine. In 1966 Scriver et al. described an infant thought to be a case of inborn error of metabolism of p-alanine, which might be a deficiency of P-alanine-a-ketoglutarate transaminase. The baby had hyper-p-alaninemia associated with increased excretion of p-alanine, P-aminoisobutyric acid, taurine, and GABA. j3-Alanine and carnosine in skeltal muscle and brain were markedly elevated. More recently, Perry et al. (1967) described two cases of carnosinemia associated with severe mental defect and myoclonic seizures. The patients excreted carnosine in urine and had a high concentration of homocarnosine in the CSF. The latter pathological finding was thought not to be associated with the disease but with age, from other studies on developmental changes in the concentration of homocarnosine in CSF of children (Perry et al., 1968a). Oral administration of anserine to the patients caused an excretion of large amounts of anserine but no change in l-methylhistidine. Therefore, they were believed to have a deficiency of carnosinase activity, and this was confirmed by Perry et al. (196Sb) by determination of serum carnosinase activity. Ill.

-Acetyl Amino Acids and Peptides

A. N-ACXTYLASPARTIC ACID(NAA) This substance, though not a peptide, is mentioned here because it is also a N-substituted amino acid, and its presence in mammalian tissue is limited to nervous tissue ( Tallan et al., 1956). Tallan et d. (1954) were the first to observe the large amounts of aspartic acid

SIMPLE PEPTIDES IN BRAIN

253

released by hydrolysis of protein-free brain extract, and subsequently they identified its N-acetyl derivative ( Tallan et al., 1956) as its major source. It has been determined by column chromatography ( Tallan, 1957), gas-liquid chromatography ( Marcucci and Mussini, 1966), and enzyme assay (Fleming and Lowry, 1966). The high sensitivity of the enzyme method made it possible to measure its distribution in Ammon’s horn in fine detail (Fleming and Lowry, 1966). The distribution of NAA in brain has been extensively studied as illustrated in Table VII taken from Tallan. Gray matter has the highest concentration (Tallan, 1957), where it occurs mostly in the cell sap ( McIntosh and Cooper, 1!36S). Newborn rats and rabbits have a lower concentration of NAA than adults, and it rapidly increases to attain adult levels at 20 days of age (Jacobson, 1959). Increase with age was also observed in human brain (Jacobson, 1959). Because of a blood-brain barrier, peripherally administered NAA cannot enter the brain ( Berlinguet and Laliberte, 1966), and it is therefore generally accepted that it is synthesized in the brain. The biosynthesis of NAA has been studied in vivo during the time when its Concentration is rising. Jacobson ( 1959) observed an incorporation of radioactive carbon TABLE VII CONCENTRATION OF ACEITLASPARTIC ACID IN BOVINENERVOUSTISSUE^ Tissue Spinal roots, mixed cord Brain medulla pons cerebellum mesencephalon thalamus hypo thalamus cerebrum basal ganglia gray matter white matter Tallan et al., 1956.

Concentration (mg/gm) 0.48 0.54 0.55 0.66 0.86 0.88 0.92 0.72

1.12 1.24 0.68

254

ISAMU SANO

from glucose and acetate into NAA. Adult animals failed to show this incorporation (Berl et d.,1961). The biosynthesis of NAA in uitro has been studied by several investigators. The early work of Tallan (1957) failed to show the biosynthesis by brain preparations but Jacobson ( 1959) subsequently demonstrated the synthesis of NAA by crude brain homogenates. Goldstein (1959) prepared water-soluble enzyme systems from brain which acetylated L-aspartic acid. However, a recent study by Knizley (1967) clarified the mechanism much better. He showed that an acetone powder of cat brain synthesized NAA and citric acid in the presence of L-aspartic acid and acetyl-CoA. His improved methods of analyses of the products made it possible to separate NAA from citric acid, and it was shown that the enzyme system forming citric acid is present in a water extract of acetone powder, whereas NAA synthetase remains in the residue of the acetone powder after extensive extraction with water. This water-insoluble enzyme may be associated with the particulate fraction of brain cells. Other tissues than brain apparently cannot synthesize NAA, and the requirement, if any, by other tissues for NAA is assumed to be fulfilled by brain (Benuck and DAdamo, 1968). NAA has been shown to be relatively stable in the presence of brain extract. Some workers have even proposed that it is metabolically inert. Recently, however, the utilization of its acetyl moiety for lipid synthesis has been observed, and a brain deacylase has been reported by Buniatyan et al. (1964). Recently, McIntosh and Cooper (1%) described a “converting enzyme” in brain, which might be an enzyme complex rather than a single enzyme. The “converting enzyme” system requires Mg++, phosphate, and glucose, and is strongly inhibited by p-chloromercuribenzoate. The enzyme product of the “converting enzyme” system remains to be identified. Apparently the unknown product from the reaction with N-acetyl-L-aspartic acid is unstable and easily reverted back to NAA in its racemic form. Neither the acetyl nor the aspartyl moieties are lost by the enzyme action, and neither phosphate nor glucose, although they are indispensable for the reaction, are incorporated into the product, The role of NAA in brain metabolism has been extensively studied, but its exact role remains obscure. Tallan (1957) proposed that NAA in brain may serve to make up for any anion deficit in brain. This proposal was recently supported by the results of a study of NAA metabolism by McIntosh and

SIMPLE PEPTIDES IN BRAIN

255

Cooper ( 1 M ) . The concentration of NAA in brain under various experimental conditions was investigated. Male mice made aggressive by isolation had less concentration of NAA in brain (Marcucci et al., 1968). This was not the case with females. Buniatyan et al. (1965) showed that brain NAA decreased after forced swimming and increased during anesthesia. Administration of various kinds of drugs such as tranquillizers, convulsants, and hypnotics did not cause any significant change in cerebral NAA concentration (Jacobson, 1959; De Ropp and Snedeker, 1961; Tews et al., 1963). Ischemia, anoxia, insulin coma and convulsion, electric shock, and thiamine deficiency also had no effect (Fleming and Lowry, 1966; Jacobson, 1959; Curatolo et al., 1963; Curatolo and D’Arcangelo, 1964). Recently, McIntosh and Cooper (1964) showed that the administration of 5-hydroxytryptophan or a monoamine oxidase inhibitor caused a rise of NAA as well as serotonin in the brain. The mechanism is not clear. These observations suggest that NAA is metabolically rather inert. Further support for this contention is the fact that NAA in brain slices remains constant for an incubation period of 60 minutes. However the possibility that NAA might be used as a supplementary energy source has not been ruled out ( McIntosh and Cooper, 1965). The stability of NAA in adult brain is quite remarkable. Thus the administration of propiothiouracil caused a decrease in NAA in brain of young but not adult rats ( Mussini et al., 1967). Buniatyan et al. (1964) also suggested that NAA is used as a supplementary source of energy in brain. They observed that the oxygen consumption of rat brain cortex slices is increased by the addition of NAA. The utilization of NAA in lipid metabolism was demonstrated by D’Adamo and Yats (1966). They showed that the intracisternal administration of NAA in rats caused a rapid incorporation of I4C into lipids. These findings suggest a possible relationship between the metabolism of NAA and myelination. The acetyl moiety of NAA was recently shown to be used in the acetylation of glucosamine and choline by Buniatyan et al. (1965) and by Buniatyan and Oganesyan ( 1964). N-Acetylglutamic acid and N-acetylaspartic acid are known to be cofactors for carbamylphosphate synthesis. NAA failed to facilitate the production of carbamylaspartic acid, argininosuccinic acid, or purines when incubated with brain slices (Gebhard and Veldstra, 1964). Recently, NAA was found to be a constituent of G-actin, which

256

ISAMU SANO

has an N-acetylaspartyl-glutamyl-threonylgroup ( Gaetjens and Barany, 1966). Carboxyl activation of NAA and subsequent incorporation into soluble RNA is also known (Pearlman and Bloch, 1963).The presence of NAA in actomyosin implies some functional participation of this acid in muscle and other contraction mechanisms. A contractile protein in brain was recently studied by Puszkin et aE. (1968), and the presence of NAA in brain is provocative in this regard. In conclusion, as far as present knowledge permits, it seems that NAA may be functioning in the brain not only in its free form but also as a constituent of macromolecules.

B. N-ACETYLGLUTAMIC A m (NAG) Another N-acetylated amino acid, namely N-acetyl-L-ghtamic acid, recently was found in the brain. The release of glutamic acid from protein-free extracts of brain by acid hydrolysis led to a survey of compounds containing glutamic acid. In addition to many y-glutamyl peptides reported by us, another source of released glutamic acid was identified by Auditore et al. (1966a) as Nacetyl-L-glutamic acid. They identified the compound by paper chromatography and hydrazinolysis. The L-configuration of the glutamyl moiety was confirmed by the hog kidney acylase method. The concentration of NAG ( 5 1 0 pglgm) is much lower than that of NAA (1mg/gm) in brain, but its precise distribution is not yet known. NAG was shown to be present in mammalian liver and yeast (Hall et al., 1958) and was implicated as a cofactor in the formation of carbamylphosphate. NAG may well participate in this reaction in certain brain cells. There have been no reports on the biosynthesis and metabolism of NAG. A m (NAAG) C. N-ACETYL-~-ASPARTYLGLUTAMIC Auditore and Hendrickson (1964) reported the occurrence of NAAG in nervous tissue. This was confirmed by Auditore et al. (196613) and by Curatolo et al. (1965). Marchetti and Mattalia ( 1965) first proposed that it is N-acetyl-P-aspartyl-L-glutamic acid, but Miyamoto et al. (1966) have given conclusive evidences that it is N-acetyl-a-aspartylglutamic acid. This was later supported by Auditore et al. ( 1966b). Miyamoto and Tsujio (1967) also showed that it is specific for nervous tissue. Their results are shown in Table VIII. Curatolo et al. (1964) obtained similar results. The distribution does not parallel that of NAA.

257

SIMPLE PEPTIDES IN BRAEN

TABLE VIII CONCENTRATION (pg/gm) OF ACETYLASPARTYLGLUTAMIC ACIDIN VARIOUS PARTSOF NERVOUS TISSUEOF ANIMALS" Tissue Whole brain

Cerebrum Midbrain Cerebellum Brainstem

Spinal cord Sciatic nerve

Rabbit 192 167 160 120 119 242 293 351 333 403 353 283 279 319 376 274

Dog

Rat

Guinea pig

-

107 144 152

230 230

-

-

-

120 152 216 232

-

290

Miyamoto et al., 1966.

Reichelt and Kvamme (1967) showed that incubation of brain slices with 14C-labeled glutamine, glutamate, aspartate, glucose, or acetate, resulted in an incorporation of 14C into NAA as well as NAAG. The intraperitoneal administration of g l u c ~ s e - ~ also ~C resulted in an incorporation of radioactivity into NAA and NAAG into acetyl, aspartyl, and glutamyl residues (Miyamoto, 1967). It also has been shown that it has no acetylcholine-like activity (D. Morris and Staughan, 1965). Precise knowledge of the mechanisms of the biosynthesis of NAAG is still unknown, and there is almost no information on its physiological signscance. IV. Proteases and Peptides

The rapid turnover of cerebral protein, comparable to that of liver, suggests that there are high activities of proteases and peptidases in the brain. I n contrast to the synthesis of proteins in the brain, there is little information on the mechanisms of their breakdown. There is, of course, no doubt that various proteases and peptidases participate, but the sequential steps which lead to the formation of amino acids from a protein are unknown. A series of

258

ISAMU SANO

reactions may be expected to occur by the coupled reactions of hydrolyzing enzymes. It was rather surprising to us that in our previous attempts to detect small amounts of ninhydrin-positive compounds, we found no traces of oligopeptides which are considered to be the intermediates of protein breakdown. One mg of oligopeptides has been detected in our isolation experiments where 30 to 80 kg of the bovine brain were used as starting materials, and all fractions of the extracts were exhaustively examined. This means that concentrations of these peptides, even if present in the brain, were less than one ten-thousandth of the concentrations of free amino acids. The peptides we could detect were of structures such as y-glutamyl or 7-aminobutyryl peptides, which are not considered as fragments of protein structures. Synthesis of all the oligopeptides known to occur in the brain is not explained by the limited hydrolysis of the cerebral proteins. The presence of y-glutamyl bonds and P-alanyl or 7-aminobutyryl residues in the proteins has not been reported. The absence of detectable amcunts of peptides composed of protein amino acids suggests a tight coupling of the series of hydrolyzing steps of proteins. If proteases and peptidases are localized in different compartments of the cells, or loosely coupled, it would be probable that there are many peptides present in the brain in small amounts on a wet gram basis, but which may be highly concentrated ultrastructurally. In an experiment designed to isolate amino compounds, large amounts of bovine brain have been homogenized with dilute acetic acid and heated to CW'C to extract water-soluble compounds, and the basic fraction was further purified. Peptides such as phenylalanyllysine, serylarginine, arginylleucine, arginylphenylalanine, leucylarginine, threonyllysine, seryllysine, arginylalanine, and valylarginine ( Kakimoto et at., 1968) were isolated from its basic fraction. When the brain tissue was homogenized in the presence of trichloroacetic acid or alcohol, these peptides could not be detected. This result is explained by cell disintegration followed by the breakdown of cerebral protein by intracellular cathepsins which are active at acidic pH, where most peptidases are inactive. Although an increasing number of studies are being reported on the purification and characterization of proteases and peptidases (Lajtha, 1964;Guroff, 1964; Marks and Lajtha, 1965; Brecher and Sobel, 1967; Marks et al., 1968; Riekkinen and Rinne, 1968a,b), our knowledge is still very fragmental. This is mainly due to the

SIMPLE PEPTIDES IN BRAIN

259

use of different substrates, enzyme sources, conditions of enzyme assays, and classification of the cerebral enzymes used by every investigator. It is, therefore, impossible to identify an enzyme in any one report with that in other reports. In order to create order out of this chaos, it will be necessary to perform systematic experiments in investigations of the patterns of brain proteases and peptidases, using wide spectra of well-characterized substrates and enzyme sources, as well as purified enzymes from different subcellular fractions, and detailed studies of cofactors and enzyme kinetics. In considerations of the essential role of proteins in the brain, it will be necessary to describe precisely the sequence of the breakdown of the individual proteins. ACKNOWLEDGMENTS Grateful acknowledgment is made to Dr. W. G. Clark, Veterans Administration Hospital, Sepulveda, California and the University of California School of Medicine, Los Angeles, for critically reviewing the manuscript. REFEPENCES Abraham, D., Pisano, J. J., and Udenfriend, S. (1961). Biochim. Biophys. Acta 50, 570. Abraham, D., Pisano, J. J., and Udenfriend, S. (1962). Arch. Biochem. Bwphys. 99, 210. Ackermann, D., Timpe, O., and Poller, K. (1929). 2. Physiol. Chem. 183, 1. Aonuma, S., Hama, S., Aoki, M., and Tamaki, H. (1968). J. Phrmuceut. SOC. (Japan)$8, 1. Auditore, J. V., and Hendrickson, H. (1964). Intern. J. Neurophurmacol. 3, 1. Auditore, J. V., Wade, L., and Olson, E. J. (1966a). J. Neurochem. 13, 1149. Auditore, J. V., Olson, E. J., and Wade, L. (1966b). Arch. Biochem. Biophys. 114, 452. Awapara, J., Landua, A. J., Fuerst, R., and Soale, B. (1950). J. Biol. Chem. 187, 35. Baddiley, J. (1955). Aduun. Enzymol. 16, 1. Bazemore, A., Elliott, K. A. C., and Florey, E. (19156). Nature 178, 1052. Behrens, 0. K., and Du Vigneaud, V. (1937). J. Biol. C h . 120, 517. Benuck, M., and DAdamo, A. F. (1968). Biochim. Biophys. Acta 152, 611. Berl, S., and Purpura, D. P. (1963). J. Neurochem. 10, 237. Berl, S., and Purpura, D. P. (1966). J . Neurochem. 13, 293. Bed, S., and Waelsch, H. (1958). J . Neurochem. 3, 161. Berl, S., Lajtha, A., and Waelsch, H. (1961). J. Neurochem. 7, 186. Berlinguet, L., and Laliberte, M. (1966). Crmad. J. Biochem. 44, 783. Bessman, S. P., and Baldwin, R. (1962). Science 135, 789. Bizerte, G., Plaquet, R., and Bodanger, P. (1955). Bull. SOC. Chim. Biol. 37,831.

260

ISAMU SANO

Bocharnikova, I. M., and Petushkova, E. V. (1967a). Biokhimiya 33, 17. Bocharnikova, I. M., and Petushkova, E. V. (1967b). Biokhimiya 32, 119. Brecher, A. S., and Sobel, R. E. (1967). Biochem. J. 105, 641. Brockman, J. A., Jr., and Burson, S. L. (1957). Proc. SOC. Ezptl. Biol. Med. 94, 450. Buchanan, D. L., Haley, E. E., and Markiw, R. T. ( 1962). Biochemistry 1, 612. Buniatyan, G. Kh., and Oganesyan, V. S. (1964). Vopr. Biokhim. Mozga, Akad. Nauk Arm. SSR, I n s t . Biokhim. 1, 131. Buniatyan, G. Kh., Aprikyan, G. V., and Paronyan, Zh. A. (1964). Vopr. Biokhim. Mozga, Akad. Nauk Arm. SSR, Inst. Bwkhim. 1, 139‘. Buniatyan, H. Ch., Hovhannissian, V. S., and Aprikian, G. V. (1965). J. Neurochem. 12, 695. Cemoch, M., and Preininger, V. (1966). Casopis Lekaru Ceskych 105, 877. Cliffe, E. E., and Waley, S. G. (1958). Biochem. J. 69, 649. Cliffe, E. E., and Waley, S. G. (1961). Biochem. J. 79, 118. Connell, G. E., and Hanes, C. S. (1956). Nature 177, 377. Curatolo, A,, and D’Arcangelo, P. (1964). Boll. SOC. ltal. Biol. Sper. 40, 694. Curatolo, A., D’Arcangelo, P., and Brancati, A. (1963). Quart. Not. ( l t d y ) 23, 191. Curatolo, A., D’Arcangelo, P., Lino, A., and Bramcati, A. (1964). Boll. SOC. ltal. Biol. Sper. 40, 2136. Curatolo, A., D’Arcangelo, P., Lino, A., and Brancati, A. (1965). J. Neurochem. 12, 339. D’Adamo, A. F., and Yatsu, F. M. (1966). J. Neurochem. 13, 961. Davey, C. L. (1957). Nature 179, 209. Davey, C. L. ( 1960). Arch. Biochem. Biophys. 89, 303. De Ropp, R. S., and Snedeker, E. H. ( 1961). J. Neurochem. 7, 128. Dickinson, J. C., and Hamilton, P. B. (1966). J. Neurochem. 13, 1179. Engeland, R., and Kutscher, F. (1910). 2. Physiol. Chem. 69, 282. Fink, K., Henderson, R. B., and Fink, R. M. (1951). Proc. SOC. Erptl. B i d . Med. 78, 135. Fink, K., Henderson, R. B., and Fink, R. M. (1952). J. Biol. Chem. 197, 441. Fink, K., McGenghey, G., Henderson, R. B., and Fink, R. M. (1956).Federation Proc. 15, 251. Fleming, M. C., and Lowry, 0.H. (1966). J. Neurochem. 13, 779. Gaetjens, E., and BBrbny, M. (1966). Biochim. Biophys. Acta 117, 176. Gebhard, O., and Veldstra, I. H. (1964). J. Neurochem. 11, 613. Goldstein, F. B. (1959). J. Biol. Chem. 234, 2702. Gulewitsch, W., and Amiradzibi, S. (1900). Chem. Ber. 33, 1902. Guroff, G. (1964). J. Biol. Chem. 239, 149. Hall, L. M., Metzenberg, R. L., and Cohen, P. P. (19%). J. Biol. Chem. 230, 1013. Hamilton, P. B., and Dickinson, J. C. (1964). Proc. 6th Intern. Congr. Biochem., New York, 1964 Vol. V, p. 393. Hanes, C. S., Hird, F. J. R., and Isherwood, F. A. (1952). Biochem. J . 51, 25. Hanson, H. T., and Smith, E. L. (1949). J. Biol. Chem. 179, 789. Hotta, S. S., and Seventko, J. M. (1968). Arch. Biochem. Biophys. 123, 104.

SIMPLE PEPTIDES I N BRAIN

261

Hulme, A. C., and Arthington, W. (1950). Nature 165, 716. Irreverre, F., Evans, R. L., Hayden, A. R., and Silber, R. (1957). Nature 180, 704. Jacobson, K. B. (1959). J . Gen. Physiol. 43, 323. Jencks, W.P., and Hyatt, M. (1959). Biochim. Biophys. Acta 31, 262. Jones, N. R. (1955). Biochem. J . 60, 81. Jones, N. R. (1956). Biochem. J . 64,20P. Kakimoto, Y., and Armstrong, M. D. (1961). J. BWl. Chem. 236, 3280. Kakimoto, Y.,Nakajima, T., Kanazawa, A., Takesada, M., and Sano, I. (1964). Biochim. Biophys. Acta 93, 333. Kakimoto, Y., Kanazawa, A., Nakajima, T., and Sano, I. (1965). Biochim. Biophys. Acta 100, 426. Kakimoto, Y., Kumon, A., and Sano, I. (1966). Unpublished data. Kakimoto, Y., Kanazawa, A., and Sano, I. (1967). Biochim. Biophys. Acta 132, 472. Kakimoto, Y., Kanazawa, A,, and Kumon, A. (1968). Unpublished data. Kalyankar, G. D., and Meister, A. (1959). J . B i d . Chem. 234, 3210. Kalyankar, G. D., and Meister, A. (1965). Unpublished data [cited in Biochemistry of the Amino Acids ( A . Meister, Ed.), Vol. 1. Academic Press, New York, 1965.1 Kanazawa, A., and Sano, I. (1967). J. Neurochem. 14, 211. Kanazawa, A., Kakimoto, Y., Nakajima, T., Shimizu, H., Takesada, M., and Sano, I. (1965a). Biochim. Biophys. Acta 97, 460. Kanazawa, A., Kakimoto, Y., Nakajima, T., and Sano, I. (196%). Biochim. Biophys. Acta 111, 90. Kanazawa, A., Kakimoto, Y., Miyamoto, E., and Sano, I. ( 1 9 6 5 ~ ) J. . Neurochem. 12, 957. Kendo, D. (194.2). J . Biochem. (Tokyo) 36, 265. Kiriaki, K., Yakushiji, T., Noro, T., Shimizu, H., and Saji, S. (19158). Nature 181, 1336. Knizley, H., Jr. (1967). J. Biol. Chem. 242, 4619. Kumon, A. (1969). J. Japan. Biochem. SOC. 41, 256. Lajtha, A. (1964). Intern. Reo. Neurobiol. 6, 1. Levenson, J., Lindahl-Kiessling, K., and Rayner, S. (19.64). Lancet 11, 756. Maas, W. K. (1956). Proc. 3rd Intern. Congr. Biochem., Brussels, 1955, 8, p. 32. Maas, W. K., and Novelli, G. D. (1953). Arch. Biochem. Biophys. 43, 236. McIntosh, J. C., and Cooper, J. R. (1964). Nature 203, 658. McIntosh, J. C., and Cooper, J. R. (1965). J. Neurochem. 12, 825. McLennan, H. ( 1957). J. Physiol. (London) 139, 79. McManus, R. (1962). J. Biol. Chem. 237, 1207. Mandeles, S., and Bloch, K. (1955). J . B i d . Chem. 214, 639. Marchetti, E., and Mattalia, G. (1965). Erperientia 21, 687. Marcucci, F., and Mussini, E. (1966). J . Chromatog. 25, 11. Marcucci, F., Mussini, E., and Valzelli, L. (1968). J. Neurochem. 15, 53. Marks, N., and Lajtha, A. (1965). Biochern. J. 97, 74. Marks, N., Datta, R. K., and Lajtha, A. (1968). J. Biol. Chem. 243, 2882.

262

ISAMU SANO

Matsuoka, M., Nakajima, T., and Sano, I. (1W9). Biochim. Biophys. Acta 1‘77, 169. Matsuoka, Y., Kumon, A., Nakajima, T., Kakimoto, Y., Imaoka, N., and Sano, I. (1969). Bull. Japan. Neurochem. SOC. Suppl. S, 25. Meister, A., ed. (1%). “Biochemistry of the Amino Acids,” 2nd ed., Vol. 1, p. 456. Academic Press, New York. Miyamoto, E. (1967). J. Japan. Biochem. SOC. 39, 351. Miyamoto, E., and Tsujio, T. (1967). J . Neurochem. 14, 899. Miyamoto, E., Kakimoto, Y., and Sano, I. (1966). J . Neurochem. 13, 999. Morris, C. J., Thompson, J. F., Asen, S., and Irreverre, F. ( 1961). J . Biol. Chem. 236, 1181. Morris, C . J., Thompson, J. F., and Asen, S. ( 1964). I. Biol. Chem. 239, 1933. Morns, D., and Staughan, D. W. (1965). Biochem. Phurmacol. 14, 1679. Mussini, E., Marcucci, F., and Garattini, S. (1967). J . Neurochem. 14, 551. Nakajima, T., and Kakimoto, Y. ( 1968). Unpublished data. Nakajima, T., Wolfgram, F., and Clarks, W. G. (1967). J. Neurochem. 14, 1107. Nakajima, T., Kakimoto, Y., Kumon, A., Matsuoka, Y., and Sano, I. (1969). J. Neurochem. 16, 417. Neifakh, E. A. (1966). Dokl. &ad. Nauk SSSR 170, 1216. Oja, S. S., and Piha, R. S. (1966). Life Sci. 5, 865. 01-10, T., and Hirohata, R. (1956). Z. Physiol. Chem. 304, 77. Otsuka, M., Iverson, L. L., Hall, Z. W., and Kravitz, E. A. (1966). Proc. Natl. Acad. Sci. U.S . 56, 1110. Otsuka, M., Tanaka, Y., and Obata, K. (1968). Bull. Japan. Neurochem. SOC. 7, Suppl., 67. Pearlman, R., and Bloch, K. (1983). Proc. Natl. Acad. Sci. U. S . 50, 533. Perry, T. L., and Jones, R. T. (1961). J. Clin. Inoest. 40, 1363. Perry, T. L., Hansen, S., Tischler, B., Bunting, R., and Berry, K. (1967). New Engl. 1. Med. 277, 1219. Perry, T. L., Hansen, S., Stedman, D., and Love, D. (1968a). J. Neurocbem. 15, 1203. Perry, T. L., Hansen, S., and Love, D. (1968b). Lancet I , 1229. Pisano, J. J., Mitoma, C., and Udenfriend, S. (1957). Nature 180, 1125. Pisano, J. J., and Udenfriend, S. (1958). Federation Proc. 17, 403. Pisano, J. J., Wilson, J. D., Cohen, L., Abraham, D., and Udenfriend, S. (1961). J . Biol. Chem. 236, 499. Pocchian, F., Tentori, L., and Vivaldi, G. ( 1962). Sci. Rept. 1st Supar. Sanita 2, 188. Puszkin, S., Clarke, D. D., and Berl, S. (1968). Federation Proc. 27, 464. Reichelt, X. L., and Kvamme, E. (1967). J. Neurochem. 14, 987. Riekkinen, P. J., and Rinne, U. K. (1968a). J. Neurol. Sci. 7, 97. Riekkinen, P. J., and Rinne, U. K. (1968b). Brain Res. 9, 126. Roberts, E., and Frankel, S. (1950). J. Bwl. Chem. 187, 55. Roberts, E., Frankel, S., and Harman, P. J. (1950). Proc. SOC. Erptl. Bid. M e d . 74, 383. Rosenberg, A. (1960). Arch. Biochm. Biophys. 88, 83.

SIMPLE PEPTIDES IN BRAIN

263

Sano, I., Kakimoto, Y., Kanazawa, A., Nakajima, T., and Shimizu, H. (1966). J. Neurochem. 13, 711. Scriver, C. R., Pueschel, S., and Davies, E. (1966). New Engl. J . Med. 274,

635. Seo, S . (1957). Med. J . Osaka Univ. 7, 833. Severin, S. E., Liu, S. H., and Popova, I. A. (1966). Probl. Neirokhim., Akad. Nauk SSSR p. 169. Sifferd, R. H., and Du Vigneaud, V. (1935). J . Biol. Chem. 108, 753. Stenesh, J. J., and Winnick, T. (1960). Biochem. J. 77, 575. Strumeyer, D. H., and Bloch, K. (1960). J . B i d . Chcm. 235, PC27. Synge, R. L. M. (1951). Biochem. J. 48, 429. Takao, T. (1967). Med. J . Osaka Univ. 19, 201. Tallan, H. H. (1957). J . Biol. Chem. 224, 41. Tallan, H. H. (1962). In “Amino Acid Pools” (J. T. Holden, ed.), p. 466. Elsevier, Amsterdam. Tallan, H. H., Moore, S., and Stein, W. H. (1954). 1. Biol. Chem. 211, 927. Tallan, H. H., Moore, S., and Stein, W. H. (1956). J. Biol. C h m . 219, 257. Taniguchi, K. (1967). Unpublished data. Tews, J. K., Carter, S. H., Roa, P. D., and Stone, W. E. (1963).J. Neurochem. 10, 641. Tsunoo, S., Horisaka, K., Muto, M., Matsumoto, A., Takahashi, S., and Muto, T. (1966). Japan. J. Phurmacol. 16, 165. Udenfriend, S. (1950). J . Biol. Chem. 187, 65. Udenfriend, S., Pisano, J. J,, and Wilson, J. D. (1961). Proc. 4th Intern. Neurochem. Symp. Varenna, Italy, 1960 p. 348. Pergamon Press, Oxford. Varma, R. R., Khuteta, K. P., and Dandiya, P. C. (1968). Psychophamcobgiu 12, 170. Virtanen, A. I., and Berg, A. M. (1954). Acta Chem. S c u d . 8, 1089. Virtanen, A. I., and Laine, T. (1937). Enzymologiu 3, 266. Virtanen, A. I., and Mitikkala, E. J. (1960). Z. Physiol. Chem. 322, 8. Virtanen, A. I., Rintala, P., and Laine, T. (1938). Nature 142, 674. Waley, S. G. (1956). Biochm. J . 64, 715. Winnick, T., and Winnick, R. E. (1959a). Nature 183, 1466. Winnick, T., and Winnick, R. E. (195913). Biochim,. Biophys. Acta 31, 47. Wolff, J., Horisaka, K., and Fales, H. M . (1968). Biochemistry 7, 2455. Wood, T. ( 1957).Nature 180, 39.

This Page Intentionally Left Blank

THE ACTIVATING EFFECT OF HISTAMINE ON THE CENTRAL NERVOUS SYSTEM By M. Monnier,

R.

Sauer, and A. M. Hattl

Physiological Institute, Universily of Basel, Basel, Switzerland

I. Formation, Distribution, and Catabolism of Histamine in the Brain . . . . . . . . . (Bibliographic Report) A. Intravenous Uptake of Exogenous Histamine . . . . B. Intraventricular Uptake . . . . . . . . . C . Endogenous Histamine Formation (Synthesis, Breakdown) . D. Distribution of Exogenous and Endogenous Histamine . . E. Neurohumoral Nature of Histamine . . . . , . F. Cerebral Functions of Histamine . . . . . . . 11. Effects of Intravascular Administration of Histamine . . . A. Intra-arterial Administration (Bibliographic Report) . . B. Intravenous Administration . . . . . . . . 111. Effects of Intraventricular Administration of Histamine . . . A. Bibliographic Report . . . . . . . , . B. Personal Investigations . . . . . . . . . IV. Effects of Direct, Intracerebral Administration of Histamine (Bib. . . . . . . . . . liographic Report) V. Release of Histamine in the Hemodialysate of Aroused Animals . . . . . . . . . (Personal Investigations) A. Method; Assay of the Histamine-like Substance . . . . B. Results . . . . . . . , . . , . VI. Conclusions and Summary . . . . . . . . . References . . . . . . . . . . . .

266 266 267 267 268 268 269 270 271 271 288 288 290 293 294 294 297 298 303

A critical analysis of the action of histamine ( H ) on the central nervous system must consider the route of administration (intravascular, intraventricular, or intracerebral ) , the viscerul, somatic behauiora2, and electrographic effects of this substance on the central nervous system. The neurophysiological approach is rather recent and particularly important, since the neurohumoral nature ' W e are very much indebted to W. Mehlhose, B. Roesch and B. Herkert for their technical assistance and to Dr. M. Fallert for his help in bibliographic investigations. Furthermore, we wish to express our gratitude to the direction of Hoffmann-La Roche AG and of the Ciba-Stiftung in Basel who supported our research.

285

266

M. MONNIER, R. SAUER, AND A. M. HA'IT

of H becomes more and more evident. Scarcely investigated was histamine's involvement in the regulation of wakefulness and its relations to other neurohumors such as catecholamines or serotonin. Previous experimental data suggested that H has a definite waking effect (Bovet et al., 1958; Goldstein et aE., 1963; Monnier et al., 1967a), but the mechanisms of this effect need closer analysis. We will review, on one hand, the previous investigations devoted to the action of H on the central nervous system. On the other hand, we will report the results of our personal experiments, performed in order to differentiate the mechanisms of the agerent and central efects of H on the brain. It is indeed necessary to specify to what extent the waking effect is mediated reflexly by afferent systems and to what extent it is of a direct central nature. I. Formation, Distribution, and Catabolism of Histamine in the Brain (Bibliographic Report)

A brief review of the formation, regional distribution, and catabolism of H in the brain will help to understand the functional significance of this substance. A. INTRAVENOUS UPTAKE OF EXOGENOUS HISTAMINE Histamine-14C (H-"C) injected intravenously in the mouse is rapidly metabolized within 1 hour to methylimidazole acetic acid riboside. The latter is detectable in large amounts in the homogenate of the total organism. It disappears linearly within 48 hours (Snyder et al., 1964). In rats, 5 to 30 minutes after intravenous injection the greatest concentrations of H are found in kidney, heart, liver, and lung. One hour and up to 24 hours after the injection, the concentration increases in the spleen, part of the gastrointestinal musculature, and saliva glands. Some H was believed to be taken up from the blood by the brain (Green, 1964) in a sufficient concentration to evoke a pharmacological response ( Goldstein et aZ., 1963; Rosenberg and Savarie, 1964). However, recent whole body autoradiographs of H-"C in rats (partially depleted in mast cells with compound 48/80 and sacrificed 5 minutes after the intravenous injection) showed radiodense areas in salivary glands, myocardium, lung, liver, and kidney, but not in the central nervous system. Mast cells play a small role in the binding of blood-borne H (Gershwin et al., 1!369). Thus, the hypophysis exhibits a particu-

ACTIVATING EFFECT OF HISTAMINE ON THE CNS

267

larly high concentration of H, but since this structure has few mast cells, the cells cannot be responsible for the higher uptake of H. In dogs also, some H is extracted from the hypophyseal stalk and is found in mast cells, but the main part of cerebral H has a nonmast cell origin (Adam, 1961) .

B. INTRAVENTRICULAR UPTAKE In contrast, H injected into the cerebral ventricle of the cats is taken up by the brain and catabolized (White, 1960). It seems to be particularly accessible to the soluble portion of the cell. This results from the observation that exogenous H perfused through the ventricles is more readily methylated than endogenous H (White, 1960; Draskoci et al., 1960). C. ENDOGENOUS HISTAMINE FORMATION ( SYNTHESIS, BREAKDOWN) Perfusion of the brain with radioactive histidine leads to synthesis of H in this organ, since histidine, but not H, passes the blood-brain barrier. Indeed, histidine, considered as the amino acid precursor of H (White, 1960), leads to the formation of H under the influence of histidine decarboxylase in the presence of pyridoxal phosphate as cofactor. The action of histidine decarboxylase is blocked by a-methylhistidine. Catecholamines may facilitate the formation of H (Kahlson and Rosengren, 1968). In uitro, the H formation capacity (increase in growth and multiplication of cells) reaches a maximum in the hypothalamus of cat, dog, and pig (White, 1959, 1960). The preponderant distribution in the hypothalamus is similar to that of norepinephrine and 5-hydroxytryptamine. However, a greater amount of H than of norepinephrine enters the brain (Rose and Hrowne, 1938; Snyder et al., 1964; Whitby et al., 1961). I n dtro, the brain catabolizes H to 1,4-methylhistamine and 1,4methylimidazole acetic acid. This ring imidazole-N-methylation of H is catalyzed by imidazole-N-methyltransferase (Brown et al., 1959). Methylhistamine is the major metabolite of administered H in most species. The cat brain can synthetize methyl-H from intraventricularly injected H (White, 1960). The whole brain of cats and guinea pigs contains methylhistamine at a concentration of about 0.5 mpmole/gm, chiefly located in the initochondrial fraction ( Carlini and Green, 1963; Fram and Green, 1968).Methylhistamine

268

M. MONNIER, R.

SAUER,

AND A. M. HA‘IT

is found, like H, in highest concentration in the midbrain. Its subcellular distribution predominates in the crude mitochondria1 fraction, whereas that of H is preponderant in the microsomal fraction. AND ENDOGENOUS HISTAMINE D. DISTRI~UTIONOF EXOGENOUS The hypophysis takes up exogenous H, but cannot synthesize it. Conversely, the hypothalamus and area postrema can synthesize H and can bind its endogenous but not its exogenous form (Adam et a!., 1964). Reserpine reduces the concentration of H synthesized in the hypothalamus and the medial thalamus but not in the hypophysis. A particularly high concentration is detectable in the hypothalamus of swine, dog, and cat (Harris et aZ., 1952; Crossland, 1!360; White, 1966a,b). In cat and man, the brain and spinal cord contain the enzymes necessary for synthesis and breakdown of H. The highest concentration is found in the hypothalamus and a smaller amount in the caudate nucleus, midbrain, medulla, and some other brain stem structures (Ling, 1960). The mechanisms that determine the localization of endogenous H (histidine decarboxylase) are independent of the factors that determine the localization of exogenous H (high blood flow, presence of cells permeable to H, low activity of H-catabolizing enzymes). Therefore, endogenous and exogenous H are not necessarily found in the same cells. In the mast cells H is bound in granules denser than mitochondria. In brain elements the H-containing granules are less dense than mitochondria (Carlini and Green, 1963), and in adrenergic fibers H is not found in granules (von Euler, 1958).

NATUREOF HISTAMINE E. NEUROHUMORAL The conception of an intracellular localization of H is supported by detection of H (and 5-hydroxytryptamine) in the subcellular sediment containing synaptic vesicles ( Carlini and Green, 1963). Michaelson and Dowe (1963) demonstrated distribution of H in the subcellular fraction of brain tissue, and presumed that it might also be present in axon endings. Indeed, radioactive H injected inbaventricularly in the rat is recovered in the “nerve endings fraction” and to a lesser degree in the “microsome-myelin fraction” (Snyder et aZ., 1966). For instance, tritiated H injected into the lateral ventricle of the rat (0.25pg) is taken up and disappears

ACITVATING EFFEfX OF HISTAMINE ON THE CNS

269

with a half-life of 1.6 hours during the first 6 hours and a half-life of 11 hours from 6 to 24 hours. A great amount of this H is distributed in the subcellular fraction, except the microsomes, and in the synapse-soma portion or “pinched-off nerve ending fraction. The chief metabolite resulting from this H is tritiated imidazole acetic acid and to a minor extent tritiated methyl-H. In the guinea pig midbrain H was found both in a low-speed nuclear precipitate (containing also 47% spermidine) and in an intermediate-speed mitochondrial precipitate. However, it could not be related to subfractions of the mitochondrial fraction separated by centrifugation ( Michaelson and Coffman, 1967). Kataoka and de Robertis (1967) also found H in the synaptic vesicles and small axon endings of the rat cortex, along with acetylcholine, noradrenaline, dopamine, and serotonin; it was not detected in myelinated structures or free mitochondria. In the former structures, the concentration of H is higher than in the homogenate of the total cortex. This supports the assumption that H could be a neurohumor acting at the synaptic level.

F. CEREBRALFUNCTIONS OF HISTAMINE Investigations with a-methylhistidine, a specific nontoxic inhibitor of endogenous H formation in vitro and in viz)o, suggest that only the non-mast-cell H is concerned with physiological phenomena. The involvement of H in physical exercise, under the influence of catecholamines and in hypersensitivity reactions, is independent of the mast cell density of the tissues; these physiological properties are linked with an increased formation capacity of nonmast-cell H (Kahlson and Rosengren, 1968). The fact that monoamine oxidase (MAO) blockers increase the tritiated H content in the brain (Snyder et al., 1966) indicates some relation between the H metabolism and the catecholamine or serotonin metabolism. These data agree with the previous observations that H activates the adrenal cortex, but less than serotonin (Sayers, 1950; Sayers and Sayers, 1968; Van Cauwenberge et al., 1967). This activating effect of H is probably mediated by an increased secretion of adrenocorticotropic hormone ( ACTH ) from the hypothalamohypophyseal substrate. In the dog also, an intravenous injection of 0.1%H elicits within 1 minute a pituitary adrenocortical response expressed by an increased secretory rate of 17-hydroxycorticosteroidsin adrenal venous

270

M. MONNIER, R. SAUER, AND A. M. HATT

blood. This response is suppressed by lesions of the hypothalamus (Katsuki et al., 1967). The anterior hypothalamus participates in the regulation of the stress reaction induced by H. The posterior hypothalamus is concerned with the negative feedback mechanism of ACTH secretion, but the anterior hypothalamus also seems to participate in this mechanism as well as in the regulation of the stress reaction of H injection. The H content in the caudate nucleus and midbrain vanes like that of other biogenic amines according to the circadian rhythm. In normal or in bilaterally adrenalectomized rats, the rectal ternperature increases during the dark period (24 to 06 hours). During this period of maximal motor activity the H content (mglgm wet tissue) increases from 0.03 (at 18 hours) to 0.06 ( a t 6 hours) in the midbrain and from 1.17 to 2.3 in the caudate nucleus. Similarly, noradrenaline increases in the midbrain from 18 to 06 hours and also in the caudate nucleus. Conversely, serotonin tends to decrease in the midbrain during the dark, active period (but not significantly) and in the caudate nucleus. On the contrary, the serotonin concentration increases in the caudate nucleus homogenate during sleep, whereas the concentration of H and noradrenaline decreases ( Friedman and Walker, 1968). The analysis of the H effects on the central nervous system may be approached by three methods: ( 1 ) intra-arterial and intravenous injection or infusion, (2) intraventricular infusion or perfusion, and ( 3 ) direct application into the brain or its neurons. II. Effects of lntravascular Administration of Histamine

The action of H on the CNS has been investigated with various methods of administration ( intraperitoneal, intravenous, intraarterial, or intraventricular ) . All these methods must be critically appraised. The intraperitoneal administration is followed by a marked fall in blood pressure, due to the local dilatatory action of H on splanchnic vessels; this may impair the cerebral blood flow and, indirectly, the cerebral functions. The intra-arterial injection presents the advantage of a localized, direct action on definite brain areas, but the stress resulting from the surgical approach of the internal carotid artery is not negligible. In addition, this mode of administration may influence the pressoceptive and chemoceptive carotid systems involved in regulation of somatic and cerebral circulation. The intravenous injection evokes visceral side

ACTIVATING EFFECT OF HISTAMINE ON THE CNS

271

effects on blood pressure and on heart and respiration rate, depending on the dose injected. The question arises whether the cerebral changes observed primarily depend on these visceral effcts, or must they be considered as parallel effects of a common mechanism. Furthermore, all intravenous and intra-arterial (single or repeated) injections performed in a nonanesthetized animal elicit a waking reaction due, to a large extent, to the emotional stress produced by the injection itself and by the presence of the human operator (Gangloff and Monnier, 1956). In our opinion, only the experiments in which H is administered automatically into the vessel with an infusion pump, at a constant 00w during a sufficient time, are reliable. However, even in such conditions the conscious animal must be isolated from the operator in a soundproof cage. Finally, in all cases of intravascular administration, the problem of the blood-brain barrier and of the speed of: H breakdown must be faced. A. INTRA-ARTERIAL ADMINISTRATION( BLBLIOGR4PHIC REPORT) In midcollicular decerebrate rabbits and cats, injection of H (O.OP0.08 pg) in the carotid artery enhances, after 1 minute, the electrical activity of the cerebellum. This effect is blocked by chlorpromazine ( Crossland and Mitchell, 1956). A similar activating effect was confirmed by Ling (1960), who injected H-diHC1 (0.01 p g ) into the common carotid artery. The neurophysiological effect was most pronounced 9 minutes after rapid injection; it was accompanied by a decreased blood flow in the organs irrigated by the external and internal carotids. The blood pressure increased only slightly ( 5 mm Hg). This action on arterioles varies according to the animal’s species. In rabbits and cats H elicits cerebral vasoconstriction, but vasodilatation (Terasawa et al., 1965) at splanchnic level. B. INTRAVENOUS ADMINISTRATION

1. Cerebral and Visceral Efects (Bibliographic Report) a. Cerebml Sfects. Intravenous injection of H (30-300 mg/kg ) in the rat anesthesized with pentobarbital induces a prolonged “desynchronization” of the electrical brain activity, symptomatic of arousal, with a short blood pressure rise preceding a subsequent fall (Bovet et al., 1958). A similar brain activation after intravenous

272

M. MONNIER, R. SAUER, AND A. M. HATI:

injection has been described in the rabbit ( Goldstein et al., 1963). H-diHC1 ( 1 mglkg iv) in the anesthetized rabbit ( pentobarbital, 6 mglkg) produces an immediate and long-lasting arousal reaction, blocked by the antihistaminic promethazine-HCl. The stimulant effect of H is not due to a reflex release of catecholamines, since it persists after pretreatment with 1 mg/kg iv phenoxybenzamine which blocks the stimulating action of amphetamine and norepinephrine on the central nervous system. Conversely, quaternary antihistaminic drugs without specific effect on the EEG, such as promethazine methosulfate or diphenhydramine trimethylammonium, have a sedative effect; this agrees with the assumption that H may play a role in brain activation. In midbrain decerebrate cats pretreated with compound 48/80 intraperitoneally for H depletion (Tuttle, 1967), an antihistaminic substance (chlorpheniramin maleate, 5 mglkg) was administered for blocking H side effects. Forty-eight hours later, histidine-14C was injected intraperitoneally (23pCi/kg) after M A 0 blocking. In such a preparation, electrical stimulation of the sympathetic chain, of the ischiadic nerve (followed by blood pressure rise), or of the depressor nerve (followed by blood pressure fall) resulted in a transitory increase of H-14C in the venous blood. The author suggested that H could act as a neuromediator at the sympathetic nerve endings, but he did not consider the possibility that, pons and medulla being intact, electrical stimulation of visceral nerves might also act on brain stem centers, mediating the release of H. Furthermore, histidine passing the blood-brain barrier could also give rise to the formation of H in the brain itself. H infused intravenously or intra-arterially ( 1 5 0 pg/minute) facilitates the release of epinephrine. Low doses (1-15 pg/minute) produce an initial secretion which decreases during 3 minutes. Middle doses (20 pg/minute) induce a sustained secretion, whereas strong doses (50 pglminute) maintain secretion at higher rate, but only with small increase in blood pressure ( Staszewska-Barczak and Vane, 1965). This mechanism could explain why the depressing effect of chlorpromazine is reversed by H administered intravenously for a 30-minute period (Rosenberg and Savarie, 1964). b. Visceral Effects. In rabbits, injection of H often elicits a blood pressure rise, whereas in other animals it mostly produces an initial depressor reaction followed by a compensatory pressor reaction (Goodman and Gilman, 1965). With automatic perfusion,

AClTVATING EFFECT OF HISTAMINE ON THE CNS

273

however, Monnier et al. (1967a) observed in the conscious rabbit a slight blood pressure fall, with a compensatory increase in heart rate. The dilatation often observed in cutaneous, splanchnic, or muscle vessels under the infiuence of H is commonly attributed to afferent nerve fibers eliciting axon reflexes (Robinson et al., 1965). Such a stimulation of visceral afferent nerves could also explain to a great extent the activating effect of intravenous administration of H, when injected into a vein (rats), since it does not penetrate the brain tissue, but is metabolized within 30 minutes by other tissues (Halpern et al., 1959).

2. Action on Viscerosomutic Behavior, EEG, and Cortical Activities ( Personal Investigations) As criteria of the action of H on the brain, we chose on one hand the alterations of the spontaneous delta activities in the EEG and on the other hand those of the potentials evoked in the motor cortex by electrical stimulation of the mesencephalic reticular system, mediocentral intralaminary thalamus, and dorsal hippocampus. As regards the first criterion, i.e., the spontaneous slow (1.5 3.5 c/s ) and high-voltage delta activities, their increase means moderation of wakefulness or sleep, whereas their decrease is significant of arousal or alertness. The other criteria, i.e., the evoked potentials, give information about the cerebral systems involved in the first phase of the cortical response evoked by electrical stimulation of the midbrain reticular formation, of the intralaminary thalamus, and of the dorsal hippocampus (see Figs. lC, 2C, 3C). An increase of this first phase means activation, i.e., increased wakefulness or arousal (Tissot and Monnier, 1959; Okuma and Fujimori, 1963; Albe-Fessard et al., 1964; Allison et al., 1966). Conversely, the second phase of the evoked thalamocortical potential ( recruiting response of Morison and Dempsey, 1942) is usually less pronounced in increased wakefulness (Tissot and Monnier, 1959; see Fig. 2D). We performed our experiments on 30 alert rabbits of average weight 2.75 kg. After having, the day before, bored holes in the skull for the electrodes, we cannulated the superficial jugular vein under local anesthesia and introduced a polyvinyl catheter for H or control infusions.

274

M. MONNIER, R. SAUER, AND A. M. HAlT

The electrical activity of the cortex was recorded bipolarly on both sides by means of silver electrodes screwed in the skull. Needle electrodes were introduced bilaterally into the mesencephalic reticular formation, the mediocentral intralaminary thalamus, and the dorsal hippocampus (on one side, steel electrodes for electrical stimulation, on the other side platinum electrodes for recording). For precise adequate localization of the electrodes in these brain structures, we used the stereotaxic method and atlas of Monnier and Gangloff ( 1961). With a Schwarzer 16 channel electroencephalograph we recorded on both sides the electrical activities of the motor and sensory-motor cortex, occurring spontaneously or evoked by stimulation of the midbrain reticular formation, intralaminary thalamus, and dorsal hippocampus. The spontaneous activities of the motor cortex and hippocampus were quantsed by an automatic frequency analyzer ( Faraday Electrical Instrument Co. ). It was thus possible to determine quantitatively the increase or decrease of the cortical delta activities. The cardiac and respiratory activities (ECG and EPG) and, in some experiments, the blood pressure were also recorded electrically. The rectal temperature was measured before and after each experiment. Before each experiment opportunity was given to the conscious animal, hanging in a hammock, in a dark and soundproof room to adapt itself to its surroundings. We divided the rabbits into three groups of 10 each, in which either the mesencephalic reticular formation, the intralaminary thalamus, or the dorsal hippocampus were stimulated at regular intervals of about 5 minutes in order to evoke potentials in the motor cortex. The reticular formation was stimulated with 35-40 stimuli of 0.3 to 0.5 msec pulse duration and 0.7-2.5 V, and the intralaminary thalamus with 3035 stimuli of 0.5 msec pulse duration and 0.5-9 V. For the hippocampus, 2530 stimuli of 0.1-0.5msec and 0.652.0 V were adequate. The threshold voltage, once determined, was not changed during the experiment. The 25 to 40 evoked responses recorded from the motor cortex during one stimulation episode were averaged by a Mnemotron computer of average transients (CAT). The curve thus obtained was transferred to an automatic plotter (XY recorder) and considerably magnsed. In 5 rabbits out of a group of 10, a solution of 1 mg H (Hista-

ACTIVATING EFFECT OF HISTAMINE ON THE C N S

275

minum dehydrochloricum, Merck) in 30 ml (0.37 mg/kg) was infused, at a constant flow speed of 1 ml/minute in the jugular vein within 30 minutes whereas the other 5 received, during the same period, a control infusion of Tyrode-like solution. This solution had the composition of the dialyzing fluid ( D F ) used in our experiments on hemodialysis in the rabbit ( Monnier and Hosli, 1964; Hosli et al., 1965).A constant flow speed was maintained by means of a perfusion pump (Braun, Melsungen). The solution of H and the control infusion (DF) had a pH of 7.4. As reference value ( =0) for the changes in delta activity and in amplitude of the two first evoked potentials during H and control infusion, we took the arithmetic mean of all the values of a preexperiment performed during 30 minutes before infusion. This variation from zero was reckoned in percent and plotted on the ordinate of a diagram. After the end of the infusion, the experiment was carried on for 45 minutes more, in order to allow an eventual return of the altered values to zero. The amplitude of the first positive and second negative evoked potentials were measured from a zero line starting at the beginning of the stimulus artefact (see Fig. 3 ) . An increase of the first component was observed in alert animals, with decreased delta activities, whereas a diminution of the first component (with increase of the second) was symptomatic of drowsiness or sleep expressed by high voltage delta activities. a. Visceral and Somatic Behavioral Effects. The visceral effects of H are weak, under the experimental conditions described. The previously observed slight fall of blood pressure, increase of the heart rate, and mild increase of the respiratory rate were again present. These controls confirmed the fall of the systolic blood pressure (lower than 5%),and the increase of the heart rate (lower than 10%);the respiratory rate was mostly unchanged, or slightly raised. Such visceral and somatic alterations are too moderate for being primarily responsible for the electrical changes in brain activity (Monnier et al., 1967a). In longer-lasting experiments, a slight blood pressure fall starts 55 minutes after onset of the recording, accompanied by a heart rate increase. b. Effects on Spontaneous Cortical and Evoked Reticubcortical Activities. We called spontaneous activities those occurring in the EEG during the whole experiment (intermittent stimulation) and

STIMULAT. FORM. R E T I CU L. MESENCEPH. EEG

Stim. V

= 0.7-2.5

PD = 0.5 ms Fr

-

lnfu sion 1 rnl I rnin

+

204

-

20

= 3/sec.

3 .--,:;,

A Histamine di H f l l r n g in.30rnl Control i o l u t i o n

j

n=5 Y

100

80 60 40

+

20

0

-

20

40

60

-

20 40

80 30

20

10

1

"v I

,

,

10

20

30

I '

I

10

l

20

1

1

30

40

1

5 0 rnir

ACTIVATING EFFECT OF HISTAMINE ON THE CNS

277

expressed by variations in the delta rhythms. They were measured in control animals as well as in those having received H. The delta activities decrease under the influence of H (5 animals). The minimum of -50% is reached during minute 18. The effect persists with progressive decrease for 20 minutes after the end of the infusion (Fig. 1A). In 5 control animals, the infusion was followed, after 12 minutes, by a delta decrease of -35%. The activity then rose again and, from minute 18 on, reached the zero line as in the preinfusion period. The reticulocortical evoked response consists of a first positive component, with a latency of 3 5 msec, and a second negative component, with a latency between 7.5 and 10 msec (Fig. 1B). This negative component mostly shows a tendency to “recruiting,” that is, to increase progressively in amplitude on repetitive stimulations. These two components are followed by less constant alternating positive and negative waves of longer duration. During H infusion, the first positive component rises markedly, reaching +112% during minute 24. After cessation of the infusion, the amplitude decreases progressively and reaches, between minute 20 and minute 30, a lower value than the mean (100 = 0 level) of the preinfusion experiment. In the control animals, the first component does not increase and its amplitude oscillates around the zero line (Fig. 1C). The second negative component decreases in amplitude, under H, to a minimum of -74% during minute 24. After the end of the infusion, it rises again and reaches the zero line around minute 40. From this series of experiments, we may conclude that the waking effect of intravenous histamine, expressed by decreased delta activities in the cortex, seems to be mediated by the ascending activating reticular system, since these decreased activities are accompanied by increased reticulocortical evoked potentials ( first component), known to correspond to reticular activation.

FIG. 1. Changes in delta activity ( A ) , in the first positive downward ( C ) , and second negative upward ( D ) components of the reticulocortical evoked potentials during infusion of histamine-diHC1 (1 mg in 30 ml = 0.37 mg/kg). The delta activity decreases and the first evoked potential increases, whereas the second decreases. These waking effects of histamine do not occur in control animals. The vertical divisions of the curves correspond to a 30-minute preinfusion experiment, a 30-minute infusion experiment, and a 45-minute postinfusion experiment.

278

M. MONNIER, R. SAUER, AND A. M. H A l T

STIMULAT. THALAMUS M E D I A L I S Stim. V = 0.450.9 PD = 0.5 ms Fr = 3/sec.

EEG

-

I

Inf uston 1 mllmin.

A H i s t a m i n g di HCI l m g i n 30 m l Control solution

c-

, I

460 0

-\

!

Thalarno-

cort icai

evoked potentials

i

'

OlO

I

C

- 20 40

I

30

20

10

I

I

I

10

20

I "

30

I

,

I

I

10

20

30

40

,

50min.

FIG.2. Changes in delta activity ( A ) , in the first positive downward (C), and second negative upward ( D ) components of the thalamocortical evoked potentials due to histamine. The delta activity decreases and the first component increases, suggesting an activation of the excitatory reticulocortical projections

ACTIVATING EFFECT OF HISTAMINE ON THE CNS

279

c. Effects on Spontaneous Cortical and Evoked Thalumocortical Activities. The mediocentral intralaminary thalamic system is a mixed substrate, containing intermingled activating reticulocortical projections and a moderating hypnogenic mechanism. Stimulation of this system with pulses of high frequency and short duration activates the reticulocortical projections, whereas stimulation at low frequency enhances chiefly the moderating hypnogenic mechanism (Tissot and Monnier, 1959). Stimulation at 3/sec and 0.5 msec pulse duration, as for the reticular system, slightly activates both mechanisms, the excitatory (first component) and the inhibitory (second component) (Fig. 2). Immediately after the beginning of the H infusion (5 animals), the delta activities show a marked and continuous decrease (Fig. 2A). The lowest value is reached with -57% during minute 30 of infusion, i.e., at the end of the infusion. The delta activity increases then progressively to a value of -13% 48 minutes after the end of the infusion. In 5 control animals the delta activities increase 20 minutes after onset of the infusion, under the influence of the moderating mechanism. The thalamocortical evoked response consists of several phases (Fig. 2 B ) . A fkst positive component, with a latency of 3-6 msec (measured from the beginning of the stimulus to the peak), is followed by a second negative highly recruiting component with a latency of 9-14 msec. In the animals subjected to H infusion, the first component increases markedly and reaches a maximal amplitude of +80% during minute 30. It then decreases and returns to its initial value 30 minutes aftw the end of the infusion. No such change of the first component can be observed in the control animal; the values obtained remain around the zero line, as in the preexperiment. On the contrary, the second, negative recruiting component tends to decrease under the influence of H. Ten minutes after the beginning of the infusion, the decrease already amounts to -20% (Fig. 2D). At the end of the infusion, the initial value is restored and the amplitude returns to the mean value of the preexperiment. within the intralaminary thalamus, whereas the decrease of the second component indicates a depression of the moderating intralaminary mechanism. The vertical divisions of the curves correspond to a 30-minute preinfusion experiment, a 30-minute infusion experiment, and a 48minute postinfusion experiment.

280

M. MONNIER, R. SAUER, AND A. M. HATT

ST I M U L AT. H I P P O C AM P U S DOR SA L I s EEG

Delta

Stim.

V

= 0.6-2.0 PD = 0.1-0.5 ms Fr 3/sec.

O/O

Infusion 1 mllmin.

---

4

cortical evoked potentials 40 60 40

-+ 2 0 0

-

20

40 40

A Histamine di H C I l m g in 30ml Control solution

b,p

B

1: Component

C

11 Component

D

+

- 20 40

30

20

10

I

10

20

30

10

20

30 4 0

50min

FIG.3. Changes of the delta activities ( A ) and of the first positivc: (B,C) and second negative (B,D) components of the hippocampocortical evoked response to H. The delta activites decrease, whereas the first evoked potential

ACI'IVATDIG EFFECT OF HISTAMINE ON THE CNS

281

In the control animals, the amplitude of the recruiting component increases during the infusion, but returns to the zero line at the end of the infusion. This series of experiments indicates that the waking action of intravenous H, expressed by the decrease in delta activities and increase of the first thalamocortical evoked potential is mediated by an excitation of the activating reticulocortical projections running through the intralaminar thalamus and by a concurrent depression of the moderating, hynogenic thalamic mechanism. d . Effects on Spontamous Cortical and Evoked Hippommpocortical Activities. Here again, the cortical delta activities decrease during and after the infusion of H, as a sign of electrographic arousal reaction. The hippocampocortical evoked response consists of two chief initial components ( Fig. 3B ). The first component, surface positive, has a latency of 2 5 4 . 5 msec. The second component is surface negative with a latency of 10-20 msec, and a strong tendency to recruiting. In most cases, it is followed by a positive phase of about the same duration as the second component. The first component increases under H, with a maximum of +6% during minute 24 (Fig. 3C). There follows a decrease of amplitude (again interrupted by an intercurrent increase) until, 45 minutes after the end of the infusion, the initial value of the preexperiment is reached again. In the control animals, the &st component somewhat decreases. The second component is but slightly reduced by H (Fig. 3D). In the control animals, insignificant alterations occur during the infusion. It results from these observations that the decrease in delta activities, symptomatic of arousal, during intravenous infusion of H is coupled with an activation of the dorsal hippocampus (increased first component of the evoked hippocampocortical response). Summarizing, in the three groups of experiments, i.e., after reticular, thalamic, and hippocampal stimulation, the delta activity, considered as an adequate sleep parameter, decreased with intravenous H. This means that H enhances wakefulness. The highest decrease of the delta activity follows reticular and thalamic increases. These waking effects of H contrast with the curves of the control animals. The vertical divisions of the curves correspond to a 30-minute preinfusion experiment, a 30-minute infusion experiment, and a &-minute postinfusion experiment.

282

M. MONNIER, R. SAUER, AND A. M. H A l T

stimulation and the lowest hippocampal stimulation. In the three groups the effect of H lasts for about 30 minutes after the infusion. In none of the three control groups could a similar decrease of the delta activity be observed. These results confirm previous findings on the waking action of H (Monnier et al., 1967a). In the cortical evoked potentials, it was possible, by repetitive stimulation with short pulses (0.5 msec) and addition of the responses by the computer, to distinguish two initial components, one positive and the other negative. They often behave reciprocally and are differently influenced by the degree of wakefulness, the frequency of stimulation ( Tissot and Monnier, 1959), temperature, anoxia, pressure, curare, strychnine, and anesthesia ( Chang, 1959), as well as by the depth of the recording electrode (Amassian et al., 1964). Whereas the first component is not altered by repetitive stimulation, the second shows a tendency to recruitment. This tendency was particularly pronounced during thalamic and hippocampal stimulation (often variable in reticular stimulation). The first component is more conspicuous in waking states, the second in sleep (Tissot and Monnier, 1959; Goff et al., 1966; Schneiderman et al., 1966). These observations on the changes of the evoked potentials under conditions of sleep and wakefulness agree with the results of other authors, who chiefly considered the amplitude of the second, recruiting component (Favale et al., 1963, 1965; Yamaguchi et al., 1963, 1964; Albe-Fessard et al., 1964; Goff et al., 1966).

3. Histamine Effects Compared with Amphetamine and Chlorpromazine Effects ( Personal Investigations) In order to specify further the central activating mechanism of intravenous H we compared its effects to those of an excitatory drug like amphetamine ( Amphetaminum sulfuricum, Siegfried) (15 mg in 30 ml, i.e., 5.5 mg/kg) and of a moderating substance like chlorpromazine (Largactil, Specia) (17 mg in 30 ml, i.e., 6.2 mg/kg). These substances are known to act primarily on the brainstem ascending reticular system at the site where the reticular unspecific afferent systems receive collaterals from the specific afferent systems. Thus, according to Bradley and Key (1958), chlorpromazine elevates the threshold for the sensory induced arousal, but not that of the arousal induced by direct stimulation of the brain stem reticular formation (Bradley and Key, 195S), whereas amphetamine has the opposite effect.

283

ACTIVATING EFFECT OF HISTAMINE ON THE C N S

STIMUL. THALAMUS

MEDIALIS

Stim: 1.2 V PD = 0.5 ms Fr = 3/s 30-40stim.

EEG

-

DELTA

Int u s 1 o n

A

300

200

-.- C h I or pr o m a 2 i ne n=3 17mg130m1130min

100

+

--Amphetamine n.5 15mg130m1130min

- Hlmg/30m1130min istamine n.5

50

T

- 50 I COMPONENT

'/o

- 50

+

I \

I , 5

I

15

I

I

25

35

lo

20

30

10

20

30

40

50min

FIG. 4. Comparative effects of intravenous H, amphetamine, and chlorpromazine on spontaneous and reticularly evoked cortical activities in the rabbit. A: Intravenous H, like intravenous amphetamine elicits an electrographic arousal reaction with decreased delta activities, whereas intravenous chlorpromazine markedly increases these activities. B: Intravenous H, like intravenous amphetamine, enhances the first component of the reticulocortical evoked response, whereas chlorpromazine decreases it. These observations suggest that the waking effect of intravenous H is similar to that of amphetamine, which activates ( a t the site of specific afferents) the unspecific ascending activating reticular system.

284

M. MONNIER, R. SAUER, AND A. M. HATT

STIMUL.

FORM. RETICUL. MESENCEPH. Stim: 1.5 V PD = 0.5 ms Fr = 3/s 35-45stim.

EEG

'lo DELTA 300

Infusion

-

A

4

1

,

?

200. L

100

f

-

I

--Chlorpromazine n-3 1?mg/30m1/30min

*-

.I

--*Amphetamine n.5 15mg130m1130min -Histamine 17.5 lrng130m1130rnin

I t

1

I

\

Y-+-

I

'lo I COMPONENT

+

I

1

50-

0-

- 505

15

25

35

io

.

20

.

30

10

20

30

40

50min.

FIG.5. Comparative effects of intravenous H, amphetamine, and chlorpromazine on spontaneous and thalamically evoked cortical activities in the rabbit. A: Intravenous H, like intravenous amphetamine, elicits an arousal reaction, with decreased delta activities, whereas intravenous chlorpromazine markedly increases these activities. B: Intravenous H, like intravenous amphetamine, enhances the first component of the thalamocortical evoked response, whereas chlorpromazine depresses it. These observations suggest that the waking effect of intravenous H is similar to that of amphetamine, which activates (at the site of specific afferents) the unspecifx activating reticular

ACI'IVATING EFFECT OF HISTAMINE ON THE C N S

285

In our experiments, amphetamine elicits a stronger decrease of delta activities than H and induces arousal, whereas chlorpromazine, on the contrary, markedly increases the delta percentage and induces sleep (Fig. 4A). Concurrently, amphetamine increases the first component of the evoked reticulocortical response, but to a lesser extent than H (Fig. 4B). This potential is, on the contrary, reduced by chlorpromazine; it therefore expresses an activating reticular mechanism, enhanced by amphetamine or H and moderated by chlorpromazine. The same applies to the evoked thalamocortical response. H, like amphetamine, decreases the delta activities (Fig. 5A) and increases the first positive component of the response evoked in the motor cortex by stimulation of the mediocentral intralaminary thalamus, whereas chlorpromazine decreases it (Fig. 5B). This fact again confirms previous observations suggesting that the first component of the evoked thalamocortical potential corresponds to the waking action of the excitatory reticulo- and thalamocortical projections (Tissot and Monnier, 1959; Schneiderman et al., 1966). The similarity of the effects of H and amphetamine on the evoked reticulo- and thalamocortical potentials indicate that the waking effect of H (decreased delta activities) could be mediated by specific afferents acting on the reticular ascending system. Chlorpromazine, as an antagonist-like substance of amphetamine, acts at the site, where afferents from the specific sensory systems discharge their impulses into the unspecific ascending and activating reticular system. Chlorpromazine depresses this point of impact, whereas amphetamine activates it. The cortical response (first component) evoked by stimulation of the dorsal hippocampus is slightly increased by intravenous infusion of H, and to a lesser extent by intravenous amphetamine. Chlorpromazine decreases slightly this first component 20 minutes after infusion onset. Therefore, the indirect waking effect of H may be ascribed only to a small extent to a reflex stimulation (through specific afferents) of central ascending systems activating the hippocampal cortex. From all these observations it results that the waking effect of H system and its ascending projections through the intralaminary thalamus to the cortex.

286

M . MONNIER, R. SAUER, AND A. M. HA=

(decreased delta activities in the neocortex) is reflexly mediated through specific visceral afferents. These afferents stimulate chiefly the unspecific ascending reticular system, which activates the neocortex (reticulo- and thalamocortical projections), and to a smaller extent the paleocortex (hippocampus). The nature of the visceral afferent systems stimulated by H (nociceptive painful or chemoceptive) will be investigated by use of an analgesic.

4. Action of an Analgesic on the Waking Mechanisms of Intravenous Histamine ( Personal Investigation) We need to determine whether visceral afferent systems stimulated by intravenous infusion of H can reflexly activate the central reticular waking mechanisms. This question is pertinent since a direct action of intravenous H on this central mechanism is problematic, this substance being unable (in some species) to cross the blood-brain barrier. The waking effect of H could be mediated by nociceptive painful afferents from the vessel wall and the tissues, or by chemoceptive visceral afferents. Both these groups of afferents could discharge impulses into the activating reticular and perhaps also into the hypothalamic and hippocampal systems. To clarify this point, we administered orally a nonnarcotic analgesic (acetylsalicylic acid, ASA; 100 mglkg) . From previous investigations in the rabbit, we know that this substance produces (60 minutes later) a maximal analgesic effect on the dental pain threshold; it does not affect greatly the visceral and somatic behavior, nor the electrical brain activity (Monnier et al., 1963; Monnier and Nosal, 1968). The choice of this analgesic for the present research is appropriate as it does not markedly alter the spontaneous or evoked cortical activities in the rabbit. ASA slightly decreases the delta activities and lowers somewhat the threshold of the cortical arousal reaction elicited by stimulation of the reticular formation or of the posteroventral hypothalamus. Preparation of the rabbits was as described previously. The day after operation a cannula is introduced under local anesthesia into the jugular vein and 1 ml of a 3%heparin solution (Liquemine, Roche) is injected. Implanted on both sides are the bipolar needle electrodes for stimulating the midbrain reticular formation as well as the elecbodes for recording the activities of the motor and sensory-motor cortex ( stereotaxic method of Monnier and Gangloff, 1961). A small tube is then introduced into the stomach for adminis-

ACTIVATING EFFECT OF HISTAMINE ON THE CNS

287

tration of ASA (100 mg/kg) in 10 ml Tyrode-like dialyzing fluid. In control animals, the same amount of h i d without ASA is administered into the stomach. In an initial period of 30 minutes, the spontaneous motor and sensory-motor cortical activities are recorded bipolarly on both sides, as well as the ECG and EPG (electropneumogram). The brain activities are analyzed with the automatic frequency analyzer. Potentials are evoked in the motor cortex by stimulation of the midbrain reticular formation ( 20-30 stimuli; voltage 0.7-2.5; pulse duration 0.4 msec; frequency 3/sec); they are summated (Mnemotron CAT) and recorded with an XY plotter. This analysis of the reticular cortical potentials is repeated every 5 minutes. Concurrently, the delta activities are summed up with an automatic frequency analyzer every 5 minutes. After this preliminary period of 30 minutes following gastric administration of ASA, H is infused in the jugular vein of 5 rabbits for 30 minutes at a concentration of 1 mg in 30 ml dialyzing fluid. In a control group of 5 rabbits, Tyrode-like fluid is infused instead of H. ?;his infusion period of 30 minutes is followed by a terminal period of 30 minutes during which the spontaneous and evoked cortical activities are further recorded and analyzed. All results are plotted in the usual way: evoked potentials in percent on the ordinate; time in every fifth minute on the abscissa. The infusion of H (after oral administration of a control solution) strongly increases the first component of the reticulocortical evoked response (Fig. 6 ) . Since this first component is symptomatic of an activation of the reticulocortical projections, we must conclude that H, which has not passed the blood-brain barrier, has stimulated this central activating system through primary specific afferents. In order to control this hypothesis, we compared the effect of intravenous H (without previous ASA) with that of H after ASA. The curve now shows a decrease of the reticularly evoked cortical potentials starting 10 minutes after onset of the H infusion and reaching a maximum 25 minutes later. This strongly supports the conclusion that a part of the activating effect of intravenous H is due to a reflex activation of the reticulocortical projections through nociceptive or chemoceptive afferent systems. This activating mechanism is depressed by ASA, probably not only at peripheral levels (receptors) but also at the site of the brain stem where

288

M. MONNIER, R. SAUER, AND A. M. HATT

STIMUL. FORM. RETICUL. MESENCEPH. Stirn. V = 0.7-2.5 PD = 0.4 ms Fr 3/s

Ret iculo- c o r t i c a l evoked potentials

ASA

or DF

Ole

I

I

Component

,

r Infusion 1 mllmin

,...*Acetylsalicylic acid and histamine -Histamine lmg130ml Acetylsalicylic acid M A ) 1~OOmglkg

---

I

10

20

30min

FIG.6. Comparative effects of H on the reticularly evoked cortical activities, with and without previous administration of ASA. The reticulocortical evoked potentials (first component) enhanced by the intravenous infusion, of histamine are markedly depressed when this infusion is preceded by oral administration of ASA.

specific aff erents impinge on the unspecific ascending reticular system and its reticulocortical projections. Ill. Effects of lntraventricular Administration of Histamine

A. BIBLIOGRAPHIC REPORT Injection of H (100 pg, base neutralized in 0.3 ml Tyrode) into the lateral ventricle of conscious cats (4.5 kg) produces muscle hypotonia, sleepiness, slight tachypnea, salivation and licking with

A m V A T I N G EFFECT OF HISTAMINE ON THE CNS

289

depression of postural reflexes ( Feldberg and Sherwood, 1954).

A stronger dose (150-200 p g ) elicits muscle atonia and adynamia lasting 1 minute, followed by closure of the nictitating membrane, lacrimation, salivation, licking, swallowing, vomiting, defecation, tachypnea ( u p to 200/minute), and panting. In a later period, the animal falls lightly asleep. Similar reactions were reported by Bhave ( 1958). These effects were confirmed by Sawyer (1955) who injected H (0.35 mg in 0.05 ml saline within 1 minute) into the third ventricle of Nembutal anesthetized female nonpregnant rabbits of about 4 kg weight. He observed miosis, salivation, bradycardia with polypnea, and heat loss. The parasympathetic symptoms could be blocked with atropine. Electrographically, fast activities were detectable in the septal and preoptic region, probably induced from the olfactory bulb. They disappeared after lesions of the olfactory bulb and septal region. Sawyer (1955) concluded from these observations that H (under Nembutal anesthesia) activates a rhinencephalic mechanism (olfactory bulb, medial forebrain bundle, basal ganglia), which stimulates the adenohypophysis. Conversely, H in unanesthetized or etherized female rabbits evokes an arousal reaction (fast desynchronized activities and theta rhythm 5/ second) with parasympathetic symptoms, but does not activate the adenohypophysis and the release of gonadotropic hormones. The author presumed that H and its antagonistic enzymes might occur in great amount in the septal region and that interaction of both substances could be important for septal activities associated with behavior. In mice, intraventricular injections of histamine liberator 48/80 produces an excitation not prevented by mepyramine (Rocha e Silva, 1959). Injection of H in the lateral ventricle of cats (100 pg in 0.2 ml) anesthetized with ether and chloralose (75 mg/kg) induces deglutition, cough, vomiting, and defecation with tachypnea and blood pressure rise lasting 30-60 minutes. In waking animals the rcspiratory rate increases markedly (White, 1961). Injection of H (5-40 p g in 0.2 ml) into the second or fourth ventricle produces a blood pressure rise in cats anesthetized with chloralose (80 mg/kg). This effect is abolished by transection of the spinal cord as well as by mepyramine or hexamethonium (Trendelenburg, 1957). The author concludes that H activates not only the peripheral but also the central sympathetic. H adminis-

290

M.

MONNIER,

R.

SAUER,

AND A. M. HATT

tered into the ventricle antagonizes the circulatory effects of intraventricular injection of 5-HT ( serotonin) and noradrenaline (Kaneko et al., 1960). By contrast, methylhistamine, the catabolite of H, lacks the pharmacological effects of intraventricular H (White, 1961). H administered into cerebral ventricles does not infiltrate the brain tissue as a whole, but some does enter the gray matter (White, 1960; Draskoci et a,?., 1960). Radioactive H (5-200 pg in 0.01-0.02 ml) injected into the ventricle is eliminated up to 75%within 3 hours and the rest between 3 and 24 hours (Heath and de Balbian Venter, 1960-1961) . From these bibliographic data, it is apparent that H injected into the ventricular system of anesthetized animals produces muscular weakness, adynamia, and trophotropic parasympathetic symptoms such as sleepiness, relaxation of the nictitating membrane, lacrimation, salivation, swallowing, retching, vomiting, defecation, and strong panting (Feldberg and Sherwood, 1954). Conversely, in the unanesthetized rabbit an arousal reaction occurs with weaker parasympathetic effects.

B. PERSONAL INVESTIGATIONS A microcannula is introduced into the third ventricle according to a method recently developed with P. Sprungli and based on data of the atlas of Monnier and Gangloff (1961) for stereotaxic brain research in the conscious rabbit. Simultaneously bipolar recording silver electrodes are screwed through the skull in close contact with the motor and sensorymotor cortex on both sides. With two platinum needle electrodes the electrical activity of the dorsal hippocampus is recorded on the left side. Furthermore, cortical potentials are evoked by stimulation of the dorsal hippocampus (right side) and of the midbrain reticular formation by means of stimulation needle electrodes in these two structures. After an adaptation period in the dark soundproof shielded case the EEG is recorded and submitted to an automatic frequency analysis for 25 minutes. Then the tyrode-like control solution (the dialyzing fluid, D F ) is infused into the third ventricle at a rate of 0.05 ml within 25 minutes, that is, 2 pliter per minute. After a new adaptation period of 25 minutes a solution of H, at a concentration of 0.15 mg or 0.3 mg in 0.05 ml, is infused in 25 minutes. In a group of 7 rabbits the spontaneous cortical and hippo-

ACTIVATING EFFECT OF HISTAMINE ON THE CNS

291

campal activities (as well as the cortical potentials evoked from the midbrain reticular formation and dorsal hippocampus ) are recorded during the following period of 25 minutes, concurrently with the electrocardiogram and pneumogram. In another group of 5 rabbits the same procedure is performed, but after the initial recording period of 25 minutes, ASA is administered orally (100 mg/kg in 10 ml tyrode, by means of a tube introduced into the stomach). Immediately after this administration of the analgesic, D F is infused for 25 minutes and, after a pause of 25 minutes, the solution of H is given as previously. Finally, in a last group of 5 animals, the control ( D F ) is infused immediately after oral administration of the analgesic and again after a period of 25 minutes rest. The control ( D F ) has the composition of the dialyzing fluid previously described; it has an osmolarity of 333 mosmoles. When H was diluted in this DF, we reduced slightly the NaCl content (5.3 g instead of 6.6 g/liter) in order to avoid hyperosmolarity effects. The delta activities are summed up with a computer (Mnemotron CAT) each time during periods of 5 minutes. The average of all delta activities during the initial period of 25 minutes is taken as reference for the subsequent experiments and represents 100% =O line of the diagram. Variations from this basic value during the subsequent experiment are expressed in percent of this basic value. The intraventricular infusion of H (0.15 or 0.3 mg in 0.05 ml of D F during 25 minutes) elicits in the conscious rabbits of 2.75 kg weight a marked electrographic arousal reaction, with marked decrease of the delta activities (-80%) 25 minutes after onset of the infusion (Fig. 7A). This delta decrease still persists after the end of the infusion (-85%); it is highly significant ( p > 0.01) when compared to the effect of a control infusion. Analgesia with ASA (100 mg/kg in 10 ml D F orally) relaxes the animal's behavior and increases the amount of spontaneous delta activities. Under such conditions, however, the intraventricular infusion of H (0.15 mg) still elicits a strong arousal reaction, with a decrease of the delta activities (-858 25 minutes after the onset and -90% at the end of the infusion). This action is significant when compared to the effect on control animals having received the analgesic alone, and it demonstrates

292

M. M O N N W , R. SAUER, AND A. M. HATT

A. HISTAMINE WITHOUT ANALGESICS EEG '10 Delta Infusion DF Infusion

1

,

I

!

60

80 100

B HISTAMINE WITH ANALGESICS % Delta

+

20

-

20

acclylsalicyl acid lOOmgll (per as

I

40 60 80 100 5

15

25

5

15

25

5

15

25 min.

FIG.7 . Comparative effects of intraventricular H alone ( A ) and of H following oral administration of a nonnarcotic analgesic ( B ) . The strong waking effect of intraventricular H (decreased delta activities in A ) is not abolished by previous oral analgesia with ASA in B, though this substance slightly increases the delta activities. Therefore, the waking effect of intraventricular H must not be attributed to pain, but rather to a direct action on the activating mesodiencephalic centers.

that the waking effect of H infused into the third ventriclc of the rabbit is not mediated by a central nociceptive or pain-inducing mechanism, since it is not abolished by ASA. We may conclude, therefore, that H has also a direct central waking effect possibly on the activating midbrain reticular system and the hypothalamic and hippocampal systems.

ACTIVATING EFFECT OF HISTAMINE ON THE CNS

293

IV. Effects of Direct, lntracerebral Administration of Histamine (Bibliographic Report)

The effects of intracerebral administration of H are contradictory, since different species, modes of administration, anesthesia, and doses were used. In the isolated brain of the snail ( H e l i x aspersa) at a high level the effects of H and related compounds on the spontaneous electrical activity of single neurons (Kerkut et al., 1968) were investigated. Among the 90 cells tested, H stimulated the activity of 41 cells, inhibited the activity of 14, and had no effect in the remaining 35 neurons. Both the excitatory and the inhibitory effects of H were blocked by low concentration of mepyramine. Acetylhistamine, L-histidine, 1,4-methylhistamine, N-methylhistamine ( 4-[pmethylaminoethyl]imidazole-diHC1) and N,N-dimethylhistamine ( 4- [~-dimethylaminoethy1]imidazole-diHCl)tested on neurons which responded to H had no effects. In mammals, the activity of single neurons of the cortical pyramidal area (Betz cells), induced by antidromic stimulation of the bulbar pyramid, was recorded with micropipettes in anesthetized or decerebrate cats. It was slightly moderated by L-histidine and more rapidly by H phosphate iontophoretically applied. Surprisingly, however, the same depressing effect was also obtained with dopamine, catecholamines, ephedrine, hydroxytryptamine, or y-aminobutyric acid (Krnjevic and Phillis, 1964). Histamine administered iontophoretically to neurons in the precruciate cerebral cortex of anesthetized cats (without barbiturate ) had dual effects on many neurons (Phillis et al., 1968). Small amounts depressed the neuronal excitability (65.5%) as did histidine and imidazole acetic acid. Larger amounts caused a mixed depression and excitation. At the termination of the depressant effect an excitatory effect, perhaps under the influence of its metabolite l-methylhistamine may occur. Both effects, as well as the effects of noradrenaline, acetylcholine, and 5-hydroxytryptamine (5-HT), were antagonized by antihistamines (mepyramine maleate). At brainstem level, H applied iontophoretically to thalamic neurons had no effect, perhaps because of the barbiturate anesthesia (Curtis and Davis, 1%2). However, Bradley ( 1968) observed, in half of the neurons, depressant effects on brainstem

294

M. MONNIER, R. SAUER, AND A. M. H A l T

reticular neurons. Finally, at spinal level, H applied iontophoretically had no effects on spinal neurons (Curtis et al., 1961). Here also, the negative results could be due to the barbiturate anesthesia. Indeed, in cats anesthetized with nitrous oxide and methoxyflurane ( Penthrane) or anemically decerebrated, noradrenahe, 5HT, and H applied extracellularly to spinal neurons increased the membrane polarization, decreased the amplitude of the excitatory and inhibitory postsynaptic potentials, and blocked the synaptic or antidromic invasion of some motoneurons (Phillis et al., 1968). These effects were believed to be compatible with those of an inhibitory transmitter. However, strychnine ( iontophoretically applied) and intravenous picrotoxin failed to antagonize the depressant actions of these three biogenic amines on spinal interneurons. From the heterogeneous results of direct intracerebral application of H, we may conclude that this substance may have excitatory or inhibitory effects on single neurons, according to the species ( phylogenetical scale), the type of anesthesia, the mode of administration (iontophoresis), and the doses. The moderating effects are reported more often, but the fact that they were also obtained by catecholamines is puzzling. Therefore, we need more observations on functionally well-defined neuronal systems as well as studies under more physiological conditions. V. Release of Histamine in the Hemodialysate of Aroused Animals (Personal Investigations)

A. METHOD;ASSAYOF

HISTAMINE-LIKE SUBSTANCE We dialyzed rabbit blood according to the method of Fdonnier and Hosli (1964; Monnier et al., 1965) and investigated whether THE

the H concentration in this hemodialysate varies according to the degree of wakefulness. Blood from the cranial venous sinus is propelled by a roller pump through an artificial kidney (dialyzer) intercalated in the extracorporal blood circulation; it flows hetween two cellophane membranes with small pore size (25.50 angstroms), impermeable to proteins such as ovalbumin with a molecular weight of 44,OOO. Thus, smaller particles pass from the blood into the external dialyzing fluid. The blood flows back into the femoral vein (Fig. 8). The external dialyzing fluid (30 ml) now contains many substances extracted from the blood, among which arc: neurohumors such as acetylcholine, H, and serotonin.

295

ACTIVATING EFFECT OF HISTAMINE ON THE CNS

Confluens sinuum

Vena femoralis

-. Bubble Manometer Thermometer

Bubble trap

b

b I

Roller Pump

cl i I I

I

Constant temp. water- bath

I I I

I

I I I I

I

----

Dialyzing solution Blood

FIG. 8. Extracorporal hemodialysis in the aroused or sleeping rabbit. (According to Monnier and Hosli, 1964.)

296

M . MONNIER, R. SAUER, AND A. M. HAlT

If we stimulate intermittently during this hemodialysis (70 minutes ) the midbrain reticular activating system in order to evoke arousal, the dialysate shows specific changes in the concentration of H, provided that the same charge of total electrical stimulation was used. Experimental arousal is induced by stimulation of the midbrain reticular formation at a frequency of 150/sec, pulse duration 0.5 msec, and average voltage of 0.25. Each stimulation episode lasts 3 seconds; the total time of stimulation may reach 440 seconds and the total charge of electrical stimulation amounts to 2.7 mC. During dialysis, the electrical brain activities are recorded and automatically analyzed, as previously reported. A preliminary dialysis during 20 minutes is necessary for determining the iverage value of the delta activities and for providing a reference (base line) for evaluation of the subsequent changes in percent. During the next dialyzing period, prolonged for 75 minutes, the waking system is intermittently stimulated. Waking donors, kept aroused by stimulation of the midbrain reticular system (75 minutes), exhibit a decrease of delta activities (compared to the prestimulation period and to relaxed animals in which the intralaminary thalamus was stimulated during the hemodialysis with the same total electrical charge). The contrast between the increased delta activities in the sleeping donors and the decreased delta activities in the waking donors is particularly striking. The dialysates were classified according to the quality of the arousal or sleep experiment, defined by the amount of altered delta activities: Group I elite donors showed a marked decrease of delta activities in the waking experiments and an increase in the relaxed animals used as a reference. The dialysates were kept in small tubes in N, atmosphere in the icebox at -6°C after readjustment of the p H by slight lowering to 7.35. As far as possible, they were analyzed pharmacologically the same day or the next day. In order to determine the concentration of the H-like substance in the dialysates, we modified the original technique of Bucher (1951)) devised to assay very small amounts of H on the guinea pig ileum. Our modified method, as well as the corresponding apparatus, has been described previously (Monnier et al., 1968) and will be only summarized here. A short segment of intestine (2.5 cm length) is carefully in-

ACTIVATING EFFECT OF HISTAMINE ON THE CNS

297

verted, in order to bring the serous membrane inside, and kept for 3 hours in the icebox at 5°C. Thereafter it is put for 20 minutes in a Petri box containing Tyrode ( p H 7.36) at room temperature. Now one extremity of the intestine strip is attached to the myograph and the other end is tied to a U-shaped capillary tube. The organ and the capillary tube containing the solution to be tested are brought into Tyrode solution at 35°C and pH 7.36 (constant value in spite of the continuous perfusion of the bath with Carbogen (95%0, and 51%CO,). The lumen of the intestine is rinsed with Tyrode and kept at rest for 20 minutes. The reactivity of the intestine is first tested with acetylcholine and H ( 6 x The Tyrode bath is then replaced by ( 1X a combined solution of atropine (2.5 x 1O-l) for blocking acetylcholine and methysergide ( 5 x lo-*) for blocking serotonin. After a 10-minute adaptation period, the dose-activity curve 3 X lea, for H is worked out with concentrations of 1x 6 X 1c8, and so on until 1x 1e6. Each time only 0.2 ml of the respective H concentration is issued through the capillary tube into the intestinal lumen. Following an adequate increasing dose-activity curve, 0.2 ml of dialysate is infused into the organ. The atropine-methysergide bath is then replaced by an atropine (2.5 X )-methysergide (5 x 10-8)-mepyramine ( Neo-Antergan Specia) ( 5 x 1W6)bath, for additional blocking of the H effect. Fifteen minutes later, 0.2 ml dialysate is added and a control is performed in order to determine whether all formerly applied doses of H are fully blocked by mepyramine. Toward the end of the experiment, acetylcholine (1X lo-?) is added in order to control whether this substance is also fully blocked. The shortening of the ileum is measured by the decreased amplitude of its deflection on the record and plotted in millimeters on the ordinate of the diagram, whereas the various concentrations of H are plotted on the abscissa. The dose-activity curves thus obtained allow to determine the concentrations of H in the waking and sleep dialysates.

B. RESULTS Table I summarizes the results of the pharmacological dosage of H-like substance in the hemodialysates of one group of elite waking rabbits and one group of elite sleeping rabbits. In the experiments a greater release (about 5 times) of H-like

298

M. M O N N W , R. SAUER, AND A. M. HATI

TABLE I Animal No.

Quality

Total charge of electrical stimulation (mC)

OF AROUSEDANIMALS A. DIALYSATES 7 I 10 I 14 I 2 I 8 I 14 I

2.75 2.55 1.5 3.36 1.77 1.5

H-diCl (ng/ml) 3100

5400 3250 3670 6000 3250

-

-

2.72

4010

B. DIALYSATES OF RELAXED ANIMALS 6 I 2.75 5 I 2.5 11 I 2.45

710 600 1000

Average

Average

-

-

2.6

770

substance in the dialysate is associated with arousal (4010 ng/ml) than with relaxation (770 nglml). The release of the H-like substance is mostly independent of the charge (millicoulombs) of the total electrical stimulation during 70 minutes hemodialysis, since this charge is about the same in waking experiments (2.7 mC) and in the relaxed animals (2.6 mC). In the rabbits subjected to stress (by an increased frequency of the roller pump: 65 instead of 52 clmin), the release of €I in the dialysate increased considerably. This series of investigations, showing that an H-like substance is released in the blood to a greater extent in actioe states like arousal, alertness, or stress, confirms the preponderant activating effect of H on the central nervous system. As H is present under physiological conditions in the mesodiencephalon and reticular formations, its participation as neurohumoral factor in the regulation of wakefulness is probable. VI. Conclusions a n d Summary

1. Exogenous intrauenm H does not normally pass the bloodbrain barrier in contrast to its amino acid precursor histicline. In spite of this barrier, H administered into the vessels develops

ACTIVATING EFFECT OF HISTAMINE ON THE C N S

299

waking effects, the mechanism of which has to be elucidated. Exogenous intraventricular H is taken u p by the brain and found in the subcellular fraction ( axon endings, synapse-soma fraction rather than in the microsomal fraction) of the cerebral cortex. This suggests that H could act as a neurohumour at synaptic level. 2. Endogenous H , synthetized from histidine by histidine decarboxylase in presence of pyridoxal phosphate, is located mainly in non-mast-cells in the hypothalamus, medial thalamus, area postrema, and, to a lesser extent, in the midbrain and medulla. It has been detected also in the nongranular portion of adrenergic axons. Various facts suggest that the non-mast-cell brain H could have some relation with activating and defensive mechanisms, the regulation of which involves other neurohumours, such as catecholamines and serotonin; for instance, the cerebral distribution of H is similar to that of noradrenaline and serotonin. Furthermore, the synthesis of H is facilitated by catecholamines. The cerebral tritium content is increased by M A 0 blockers, whereas the H concentration in the hypothalamus and medial thalamus is reduced by reserpine. Finally, the H and noradrenaline contents of the caudate nucleus and midbrain increase during the waking period with maximal motor activity, whereas the serotonin content decreases (and vice versa in sleep). These data support our previous report on the activating and waking function of H (Monnier et al., 1967a). They justified the new investigations we performed with neurophysiological and electrophysiological methods. 3. Intravenous infusion of H (0.37 mg/kg in 30 ml during 30 minutes) in conscious rabbits elicits an electrographic arousal reaction with significant decrease of the spontaneous delta activities in the cortex. This outlasts the end of the infusion and is accompanied by a slight blood pressure fall with increased heart rate (Monnier et al., 1967a). Since H does not normally cross the blood-brain barrier and does not penetrate the brain tissue following intravenous injection, this waking effect could be reflexly elicited by afferents from the tissue and vessel wall acting on the brain centers. 4. In order to elucidate this mechanism, we analyzed the intravenous H effects on the potentials evoked in the motor cortex by stimulation of the midbrain reticular formation, the centromedial intralaminary thalamus, and the dorsal hippocampus. These potentials give information on the ascending activating reticular

300

M. MONNIER, R. SAUER, AND A. M. HA'IT

system and its projections through the intralaminary thalamus to the cortex. They also tell whether the hippocampal system (rhinencephalon) is involved in the H activating mechanism. The investigations showed that the waking effect of intravenous H, expressed by decreased cortical delta activities, is coupled with an increase of the reticulocortical evoked potentials ( first component), of the thalamocortical evoked potentials (first component), and of the hippocampocortical potentials ( first component), whereas the second component of all these potentials is reciprocally depressed. This suggests that the waking effect of H is mediated by the activating ascending reticular system and its projections to the cortex through the intralaminary thalamus. The hippocampal system is also partly involved in this waking mechanism. 5. Comparative intravenous infusion of amphetamine (5.5 mg/ kg) and chlorpromazine (5.2 mg/kg) showed that the activating effect of H is similar to that of amphetamine and opposed to that of chlorpromazine, known to act at the site where collaterals from the specgc sensory afFerents impinge on the unspecific ascending reticular system. Intravenously infused H seems to activate this system and the hippocampus through visceral afferent nerves. We have to demonstrate to what extent the afferents acting on the brain centers belong to the nociceptive pain system or to other (chemoceptive) systems. 6. The effects on the brain of iv H and of H infused after oral administration of a nonnarcotic analgesic, such as acetyl salicylic acid (37 mglkg), are different. The enhancing action (of H on the reticulocortical evoked potentials (first component symptomatic of reticulocortical activation) is markedly decreased by the previous analgesia. This suggests that the H activating effect is due partly to specific nociceptive pain afferents and perhaps to chemoceptive visceral afFerents stimulating the unspecific ascending activating reticular system and its reticulocortical projections. The present suggestion is supported by the observation that stimulation of dorsal roots by afferent nerves liberates a H-like factor not only peripherally (Kwiatkowski, 1943) but also centrally in the spinal cord, the reflexes of which are enhanced (Hausler and Sterz, 1952). The waking mechanism of intravenous H, however, may be mediated by specific afferents stimulating not only the reticulocortical projections, but also other cortex activating mechanisms (hypothalamus, hippocampus ) .

ACTIVATING EFFECT OF HISTAMINE ON THE CNS

301

7. In addition to its reflex activating effect, H has a direct central activating effect when infused into the third ventricle (0.05 mg/kg and 0.1 mglkg in 0.05 ml during 30 minutes). It induces a significant waking reaction of central origin, which is not abolished by analgesia with acetylsalicylic acid. This waking and activating effect of H agrees with the fact that antihistaminics (without effects on the EEG ) frequently develop hypnotic reactions with increased delta activities. 8. Finally, the central activating effect of H is demonstrated by the higher concentration of an H-like substance in the hemodialysate of rabbits kept aroused during 70 minutes by electrical stimulation of the activating midbrain reticular system. Conversely, the concentration of this substance is low in the hemodialysate from rabbits kept asleep by electrical stimulation of the hypnogenic thalamic center. As H is present under physiological conditions in the mesodiencephalon and reticular formations, its participation as neurohumoral factor in the regulation of wakefulness is not excluded. Several similarities with other biogenic amines ( catecholamines, 5-HT) as regards cerebral distribution and metabolism support this assumption. 9. Our present knowledge of the afferent (reflex) and central activating mechanisms of H are summarized in an anatomophysiological diagram (Fig. 9). It shows that the waking effect of intravascular H (which does not pass the blood-brain barrier) must be explained reflexly by visceral aff erents stimulating chiefly the reticular activating system and, to some extent also, the diencephalic (hypothalamus, thalamus) and rhinencephalic (hippocampus ) activating mechanisms. The afferents belong in part to nociceptive, pain-inducing systems and to chemoceptive systems, since the intravascular H effect is reduced by a nonnarcotic analgesic such as acetylsalicylic acid. The waking effect of intraventrkular H (third ventricle) is, no doubt, of central origin. It cannot be ascribed to pain, since it is not reduced by the nonnarcotic analgesic. It may be explained by direct stimulation of ascending reticular, hypothalamic, and hippocampal activating systems, running closely to the third ventricle. Some synergistic ergotropic effects of H may also be attributed to descending mesodiencephalic and reticular pathways which activate the preganglionic sympathetic system. This mechanism may explain many activating effects of H on endocrine systems

302

M. MONNIER, R. SAUER, AND A. M. HAlT

Central a c t i v a t i o n Intraven tricular

reticular--------e fFerents

-__-___ afferents

-----___ Specific visceral 1

Nc.n. X

afferents

4

Epinephrine

Ref I e x a c t i v a t i o n

\I/

FIG.9. Anatomophysiological diagram giving our present concept of the indirect, reflex activating mechanism of intravascular histamine and of the direct, central activating effect of intraventricular histamine.

(adrenal cortex) and defensive mechanisms, such as stress or hypersensitivity. Finally, definite oral and secretory activities seem to be due to a direct stimulation of the cranial nerves VII, IX, X by (intraventricular) H.

ACZIVATING EFFECT OF HISTAMINE ON THE CNS

303

REFERENCE^

Adam, H. M. (1961). In “Regional Neurochemistry” ( S . S. Kety and J. Elkes, eds.), p. 293. Pergamon Press, Oxford. Adam, H. M., Hye, H. K. A,, and Waton, N. G. ( 1964). J. Physiol. (London) 175, 70P. Albe-Fessard, D., Massion, J., Hall, R., and Rosenblith, W. (1964). Compt. Rend. 258, 353. Allison, T., Goff, W. R., and Sterman, M. B. (1966). Electroencephalog. Clin. Neurophysiol. 21, 461. Amassian, V. E., Waller, H. J., and Macy, J. J. (1964). Ann. N.Y. h o d . Sci. 112, 5. Bhave, W. B. (1958). J. Physiol. (London) 140, 169. Bovet, D., Kohn, R., Marotta, M., and Silvestrini, B. (1958). Brit. J. Pharmacol. 13, 74. Bradley, P. B. (1968). Intern. Rev. Neurobiol. 11, 1. Bradley, P. B., and Key, B. J. (1958). Electroencephabg. Clin. Neurophysiol. 10, 97. Brown, D. D., Tomchick, R., and Axelrod, J. (1959). 1. B i d . Chem. 234, 2948. Bucher, K. (1951). Experientia 7, 111. Carlini, E. A., and Green, J. P. (1963). Brit. J. Pharmacol. 20, 264. Chang, H. T. (1959). In “Handbook of Physiology” (Am. Physiol. Soc., J. Field, ed.), Sect. 1, Vol. I, p. 299. Williams & Wilkins, Baltimore, Maryland. Crossland, J. (1960). J. Pharm. Pharmucol. 12, 1. Crossland, J., and Mitchell, J. F. (1956). J. Physiol. (London) 132, 391. Curtis, D. R., and Davis, R. (1962). Brit. J. Pharmacol. 18, 217. Curtis, D. R., Phillis, J. W., and Watkins, J. C. (1961). J. Physiol. (London) 158, 296. Draskoci, M., Feldberg, W., Fleischhauer, K., and Haranath, P. S. R. K. (1960). J . Physiol. (London) 150, 50. Favale, E., Loeb, C., and Manfredi, M. (1963). Erperientia 19, 189. Favale, E., Loeb, C., Manfredi, M., and Sacco, G. (1965). Electroencephabg. Clin. Neurophysiol. 18, 354. Feldberg, W., and Shenvood, S. L. (1954). J . Physiol. (London) 123, 148. Fram, D. H., and Green, J. P. (1968). J. Neurochern. 15, 597. Friedman, A. H., and Walker, C. (1968). J. Physiol. (London) 197, 77. Gangloff, H., and Monnier, M. ( 1956). Electroencephabg. CZin. Neurophysiol. 8, 623. Gershwin, M. E., Smith, N. T., and Hood, N. (1969). Am. J . Physiol. 216, 46. Goff, W. R., Allison, T., Shapiro, A., and Rosner, B. S. (1966). Electroencephubg. Clin. Neurophysiol. 21, 1. Goldstein, L., PfeifFer, C. C., and Muiioz, C. (1963). Federation Proc. 2 5 424. Goodman, L. S., and Gilman, A., eds. (1965). “The Pharmacological Basis of Therapeutics,” 3rd ed. Macmillan, New York. Green, J. P. (1964). Federation Proc. 23, 1095. Hausler, H. F., and Sten, S. H. (1952). J. Mount Sinai Hosp. 19, 121.

304

M. M O N N W , R. SAUER, AhQ A. M. HATT

Hdpem, B. N., Neven, T., and Wilson, C. W. M. ( 1959). J. Physiol. (London) 147, 437. Harris, G. W., Jacobsohn, D. D., and Kahlson, G. (1952). Ciba F o u d . CoZZoq. Endocrinol. 4, 186. Heath, R. G., and d e Balbian Verster, F. (1960-1961). Am. J. Psychiut. 117, 980-990. Hosli, L., Monnier, M., and Koller, T. (1965). Arch. Ges. Physiol. 282, 54. Kahlson, G., and Rosengren, E. (1968). Physiol. Reu. 48, 155. Kataoka, K., and de Robertis, E. (1967). J . Phurmacol. Exptl. Therap. 156, 114. Kabuki, S., Ito, M., Watanabe, A., Iino, K., Yuji, S., and Kondo, S., (1967). Endocrinology 81, 941. Kerkut, G. A., Walker, R. J., and Woodruff, G. W. ( 1968). Brit. J. Phurmucol. 32, 241. Krnjevic, K., and Phillis, J. W. (1964). Brit. J . Pharmucol. 20, 471. Kwiatkowski, H. V. (1943). J . Physiol. (London) 102, 32. Ling, G. (1960). In “Comparative Neurochemktry” (E. Richter, ed.), p. 387. Pergamon Press, Oxford. Michaelson, A., and Coffman, P. Z. (1967). Biochem. P h a m o l . 16, 2085. Michaelson, A., and Dowe, G. (1963). Biochem. Phwmacol. 12, 949. Monnier, M., and Gangloff, H. (1961). “Rabbit Brain Research,” Vol. 1. Elsevier, Amsterdam. Monnier, M., and H&li, L. (1964). Science 146, 796. Monnier, M., and Nosal, G. (1968). I n “Pain” (A. Soulairac et aZ., eds.), p. 441. Academic Press, New York. Monnier, M., Nosal, G., and Radouco-Thomas, C. (1963). Intern. J . Neurophurmucol. 2, 217. Monnier, M., Hosli, L., and Koller, T. (1965). Arch. Ges. Physiol. 282, 54. Monnier, M., Fallert, M., and Bhattacharya, I. C. (1967a). Experientia 23, 21. Monnier, M., Radouco-Thomas, C., and Nosal, G. (1967b). Arch. Biol. Med. Exptl. 4, 5. Monnier, M., Romanowski, W., and Bhattacharya, I. C. (1968). Arch. Ges. Physiol. W2,279. Morison, R. S.,and Dempsey, E. W. (1942). Am. J . Physiol. 135, 281. Okuma, T., and Fujimori, M. (1963). Folia Psychiat. Neurol. Japon. 17, 25. Phillis, J. W., T e k i s , A. K., and York, D. H. (1968). European J. P h a m c o l . 4, 471. Robinson, J . D., Anderson, J. H., and Green, J. P. (1965). J. Phurmucol. Exptl. Therap. 147, 236. Rocha e Silva, M. (1959). Brit. J . Phurmucol. 14, 243. Rose, B., and Browne, J. S. L. (1938). Am. J . Physiol. 124,412. Rosenberg, F. J., and Savarie, P. J. (1964). J. Phurmacol. 146, 180. Sawyer, C. H. (1955). Am. J . Physiol. 180, 37. Sayers, G. ( 1950). Physiol. Reu. 30, 241. Sayers, G., and Sayers, M. A. (1968). Recent Progr. Homume Res. 2, 81. Schneiderman, N., Monnier, M., and Hosli, L. (1966). Arch. G e s . Physiol. 288, 65.

ACITVATING EFFECT OF HISTAMINE ON THE CNS

305

Snyder, S. H., Axelrod, J., and Bauer, H. (1964). J . Pharmucol. Exptl. Therap. 144, 373. Snyder, S . H., Glowisnki, J., and Axelrod, J. (1966). J. Phurmacol. Exptl. Therap. 153, 8. Staszewska-Barczak, J., and Vane, J. R. (1965). Brit. J . Pharmucol. 25, 728. Terasawa, F., Fujii, J., Murata, K., Hosoda, S., and Ikeda, M. (1965). Japon. Circulation J . 29, 577. Tissot, R., and Monnier, M. ( 1959). Electroencephubg. Clin. Neurophysiol. 11, 675. Trendelenburg, U. ( 1957). Circulation Res. 5, 105. Tuttle, R. S. (1967). Am. J. Physiol. 213, 620. Van Cauwenberge, H., Lecompte, J., and Palem-Vliers, M. (1967). Arch. Intern. Physiol. Biochim. 75, 614. van Euler, U. S. (1958). Acta Physiol. Scand. 43, 155. Whitby, L. G., Axelrod, J., and Weil-Malherbe, H. (1961). J. P h a m c o l . Exptl. Therap. 132, 193. White, T. (1959). J . Physiol. (London) 149, 34. White, T. (1960). J . Physiol. (London) 152, 299. White, T. ( 1961). J . Physiol. (London) 159, 191. White, T. (1964). Federation Proc. 23, 1103. White, T. (1966a). Brit. J . Pharmacol. 26, 494. White, T. ( 1966b ). In “Handbuch der experimentellen Pharmakologie” (0. Eichler and A. Farah, eds.), Vol. 18, p. 789. Springer, Berlin. Yamaguchi, N., Ling, G. M., and Marczynski, T. J. (1963). Nature 199, 186. Yamaguchi, N., Ling, G. M., and Marczynski, T. N. (1964). Ebctroencephalog. Clin. Neurophysiol. 17, 246.

This Page Intentionally Left Blank

MODE OF ACTION OF PSYCHOMOTOR STIMULANT DRUGS By Jacques M. van Rossum Department of Pharmacology, Catholic University, Nijmegen, The Netherlands

I. Psychomotor Stimulant Drugs . . . . . . . . A. History of Central Stimulant Drugs . . . . . . B. Survey of Psychomotor Stimulant Drugs and Anorectic Agents . C. Chemical Structure and Psychomotor Stimulant Action D. Absolute Configuration and Psychomotor Stimulation . . 11. Effects of Psychomotor Stimulant Drugs in Man . . . . A. Effects of Amphetamines in Therapeutic Doses . . . B. Toxic Effects of Psychomotor Stimulants in Man . . . C . Addiction to Psychomotor Stimulant Drugs . . . . D. Amphetamine Psychosis Induced by Psychomotor Stimulant . . . . . . . . . . . . Drugs 111. Effects of Psychomotor Stimulant Drugs in Animals . . . A. Locomotor Stimulant Effects . . . . . . . B. Psychomotor Stimulants on Operant Behavior . . . . C. Psychomotor Stimulants on Self-Stimulation . . . . D. Stereotyped Behavior through Psychomotor Stimulants . . E. Action of Psychomotor Stimulant Drugs on Social Behavior . IV. Kinetics of Absorption, Distribution, and Elimination of Amphetamines . . . . . . . . . . . A. Physicochemical Properties of Amphetamines . . . . B. Kinetics of Absorption of Amphetamines . . . . . C. Kinetics of Distribution of Amphetamines . . . . D. Kinetics of Metabolism of Amphetamines . . . . . E. Kinetics of Elimination of Amphetamines . . . . V. Antagonism of Amphetamine Action and Interaction with Other . . . . . . . . . . . . Drugs . A. Antagonism with Neuroleptics . . . . . . . B. Amphetamine Action in Reserpinized Animals . . . . C. Interaction of Monoamine Oxidase Inhibitors with Amphet. . . . . . . . . . . . amines D. Interaction of Thymoleptics with Amphetamines . . . E. Interaction of Amphetamines with Sympatholytic Drugs . F. Interaction of Amphetamines with Cholinolytic and Other . . . . . . . . . . . Drugs VI. Psychomotor Stimulant Action and Brain Catecholamines . . A. Effects of Amphetamine on Brain Monoamines . . . B. Inhibition of Synthesis of Catecholamines and Psychomotor . . . . . . . . . Stimulant Action . C. Inhibition of Synthesis of Noradrenaline and Psychomotor . . . . . . . . . Stimulant Action . 307

.

309 309 310 312 319 323 323 324 3% 325 327 327 329 331 334 335 337 337 339 341 343 346 347 348 351 353 354 355 355 356 356 357 360

308

JACQUES M. VAN ROSSUM

D. Inhibition of Synthesis of Brain Serotonin and Psychomotor Stimulant Action . . . . . . . . . . VII. Mechanism of Action of Psychomotor Stimulant Drugs . . . A. Significance of Brain Noradrenaline Receptors . . . B. Significance of Brain Dopamine Receptors . . . . C. The Midbrain Reticular Foxmation and Psychomotor Stimulant Action . . . . . . . . . . . . D. The Neostriatum and Psychomotor Stimulant Action . . References . . . . . . . . . . . .

360 361 361 365

369 370 373

Amphetamine or “Weckaminen” and a number of related drugs are classified as psychomotor stimulant drugs. Alertness, suppression of fatigue and sleepiness, stimulation of motor activity, and other symptoms of central excitation are a characteristic effect of the psychomotor stimulant drugs. The various members of this category in varying degrees have side effects in the peripheral and central nervous system (Welsh, 1962; Kalant, 1966). In the peripheral nervous system, the sympathomimetic effects of amphetamine and the local anesthetic effects of cocaine occur; in the central nervous system, the anorectic effects of amphepramon and phentermine predominate (Welsh, 1962). The amphetamines have little therapeutic value with regard to their central excitatory effects, but they are frequently used as appetite suppressants (Welsh, 1962; Editorial, 1968). Abuse of amphetamine and related drugs is widespread both as agents to increase performance and endurance in sport events (doping) and to induce excessive stimulation and euphoria in addicts (“Speed,” etc.), Kalant, 1966; Connell, 1958). The toxic effects of the amphetamines with respect to the sympathomimetic and central stimulant actions are severe. Cardiovascular collapse and drug-induced psychosis are often seen after chronic abuse of amphetamines ( Connell, 1958; Kalant, 1966). Amphetamine and related drugs have been used in many experimental pharmacological and neuropharmacological studies. AS a consequence an immense amount of data on these stimulants is now available and certain aspects of the mechanism of actnon have been elucidated. Evidence is available that the amphetamines act at neurons where catecholamines are transmitter substances. Braindopamine in particular appears to have a predominant role in psychomotor stimulant action.

ACTION OF PSYCHOMOTOR STIMULANT DRUGS

309

I. Psychomotor Stimulant Drugs

Dexamphetamine is the prototype of the category of psychomotor stimulant drugs. The pharmacology, the clinical application, and the toxic effects have been described in a number of monographs (Leake, 1958; Bett et al., 1955; Bonhoff and Lewrenz, 1954; Welsh, 1962; Kalant, 1966; Connell, 1958). A. HISTORY OF CENTRAL STIMULANTDRUGS Throughout the history of mankind, stimulants of plant origin have been used. Interesting surveys with many references are found in monographs by von Bibra ( 1855), Hartwich (1911), and Lewin ( 1927). The old Chinese herb Ma Huang (Ephedra vulgaris) has certain central stimulating properties. The active principle is the sympathomimetic drug levoephedrine, which has central stimulating properties and is now mainly used for the treatment of asthma and nasal congestion (Table I ) (Chen and Schmidt, 1924). The old Inca drug coca (Eythroxylon coca) is still used by the Indians living in the mountains of Peru (indos serranos), but the inhabitants at sea level seem to use this stimulant less. The Aymaras people who speak the Quecha language chew the leaves of the coca plant and thereby gain body strength and postpone fatigue and sleepiness (Hartwich, 1911; Mortimer and Golden, 1902; Lewin, 1927). The main alkaloid of the coca plant is ( - )-cocaine which has all the characteristics of a psychomotor stimulant ( Molina, 1946; Gutierrez-Noriega and Zapata-Ortiz, 1947). QAt, Cat, or Kafta, as it is variously called, are the fresh leaves and young shoots of the West African plant Catha edulis, grown in the highlands of Ethiopia (Hartwich, 1911; Beitter, 1901). The use of QAt provides strength to the people while it prevents the development of fatigue and hunger (Lewin, 1927; Beitter, 1901). The active principle of Catha is the alkaloid cathine or norpseudoephedrine (see Table I). Catha does contain other amines, but cathine alone is responsible for its psychomotor stimulant effects (Alles et al., 1961). Although deoxyephedrine, now known as methamphetamine and Pervitin, was synthesized early in this century, it was only after the discovery of amphetamine that the central stimulant properties became generally known. Amines of simple chemical structure were

310

JACQUES M. VAN ROSSUM

TABLE I STRUCTURES OF PSYCHOMOTOR STIMULANT ALHALO~DS

b'

ephedrine I-I,lR;ZS

mNH2 qHflN-c cocaine

-1 2R;3S

?H

\

4:

C'

OH

b

hyoscyamine 3a;l'S

CH3

nor pseudo ephedrine I+) Cathine 1s; 2 s

m3NH amphetamine (+I 2s

synthesized ( Alles, 1927), and amphetamine appeared to exert predominant sympathomimetic and central stimulant effects ( Haley, 1947). During World War 11, amphetamine and Pervitin were used extensively by army troups as energy tablets to combat fatigue and sleepiness and to improve endurance (Bett et a?., 1955; Bonhoff and Lewrenz, 1954; Hauschild, 1939). The structure of amphetamine is related to that of cathine and ephedrine; cocaine contains a piperidine base, but it is not related stereochemically to atropine or hyoscyamine ( see Table I ) .

B. SURVEY OF PSYCHOMOTOR STIMULANT DRUGS AND ANORECTIC AGENTS The amphetamines now available are of a variety of chemical structures. However, all stimulants except cocaine are phenylethylamine derivatives ( see Table 11). Some psychomotor stimulants have one or more centers of asymmetry, so optical antipodes and/or stereoisomers may exist. In general, one of the antipodes is more potent than the other. The dextrorotatory isomer of amphetamine or dexamphetamine is about three times as potent as the levorotatory antipode ( Alles, 1939; Jarowski and Hartung, 1943).

TABLE I1 STRUCTURES AND ABSOLUTE CONFIGURATION OF PSYCHOMOTOR S T l M U L A N T DRUGS

0”i””’ dexamphetamine * 1+12S; DexedrineR

alfetamine AletamineR

methamphetamine* I+)2s; PervitinR

*

dimet hamphetamine

I+)2 5 MetrotoninR

cypenamine

dextrofemine

1+I 2S

NH-C-C

p hen met razine three[+) 1 S ; Z S

**

fencam famine ReactivanR

zylofuramine threo 2S;35

l+l

PreludinA

C=O,n

F

Q

c-o\,/,o

0% \

facetoperan

methylphenidate Ritalin R 1+1threo 1R:ZR

I-)threo 1S;ZS LidepranR

pipradrol

1+1 Z R Meratran

cmNHz C

pernoline TradonR

*

**

xylopropamine Esantn

Prolintane Catovit R

-

also in use as anorectic drug to be considered as a dangerous addictive drug

312

JACQUES M. VAN ROSSUM

Dexamphetamine, methamphetamine, and phenmetrazine are the best-known psychomotor stimulant drugs ( Leake, 1958; Kalant, 1966). Sympathomimetic side effects are predominantly present in amphetamine and methamphetamine. Ephedrine is mainly a sympathomimetic, so the weak central stimulant effects are accompanied by strong peripheral side effects.Most of the other drugs have slight or no sympathomimetic properties. Some of the psychomotor stimulants are used therapeutically as appetite suppressants, eventually in combination with a sedative in order to reduce central stimulation or with a laxative to prevent an increase in dosage by the patient (Welsh, 1962). Others are mainly stimulants that lack anorectic properties. This is the case for zylofuramine and pipradrol (Harris d al., 1963; van Rossum and Simons, 1969) and also pemoline. A number of psychomotor stimulant drugs are predominantly appetite suppressants showing central stimulation in higher or slightly higher doses (see Table 111). This is the case for regenon ( Melander, 1960) and phentermine (van Rossum and Simons, 1969). Phenmetrazine, although used as an anorectic drug, is classified here as a stimulant since this drug has stimulant properties in doses that are lower than needed for the anorectic effect (van Rossum and Simons, 1969). In addition there are a number of anorectic drugs that are to a large extent completely devoid of stimulant properties (Gylys et al., 1962; le Douarec et aE., 1966) (see Table IV). The anorectic agents, although related to dexamphetamine, are not considered psychomotor stimulant drugs ( compare Tables 11, 111, and IV). Only the drugs presented in Table I1 are the psychomotor-stimulant drugs currently available. As pointed out before, a number of anorectic agents show psychomotor stimulation in doses higher than those employed for treatment of obesity. Since tolerance to anorectic action occurs, psychomotor stimulation may appear when doses are augmented during chronic treatment.

C. CHEMICAL STFKJCTURE AND PSYCHOMOTOR STIMULANTACITON Most amphetamine-like drugs except pemoline are relatively simple organic bases. They are primary, secondary, or occasionally tertiary amines. The phenylethylamine structure can be found in

TABLE I11 STRUCTURES OF ANORECTIC AGENTSWITH CENTRAL STIMULANT PROPERTIES

levamphetamine

1-1 R Levedrine

ethyl amphetamine [ d l ) Adiparthrol ApetinilR

furfenorex

mefenorex Anexate R

benzphetamine [+12S OidrexR

vNH phenlermine Mirapront R

mNH-c cYfHz \

\

levo -methamphetamine

diphemethoxidine CleofilR

&f::;

?H

norpseudo ephedrine I+)threo 1S;ZS CathineR

phendimetrazine [+) threo 1S;2S

Plegine R

pent orex Modatrop R

metarnfepramone

Id[)

fenbutrazate

0

amfepramone [ d l l Regenon R

aminorex

314

JACQUES M. VAN ROSSUM

TABLE IV STRUCTURES OF ANORECTIC AGENTSWITH LITTLEOR CENTRALSTIMULANT PROPERTIES

NO

mNHp ort etamine

f o r m e torex

amphecloral Acutran

mNH PNH‘ CF3wNH

CI

c F3

chlorphentermine Lucofen R

~ N H $ O - C - C

fenfluramine PonderalR

6 NH-C

trifluorex rri f lutamine R

mNH

n Pent

Cl

\

il cloforex

clominorex

fludorex

fluminorex

amfepentorex

fenmetramide

practically all drugs of this category (Tainter et al., 1939; Jacobsen, 1939; Schulte et a,?.,1939; Chen et a,?.,1929; Alles, 1933). Phenylethylamine is a very weak central stimulant since this drug is rapidly metabolized in most species. In rats pretreated with

ACTION OF PSYCHOMOTOR STIMULANT DRUGS

315

a monoamine oxidase inhibitor, phenylethylamine behaves as amphetamine (Stein, 19fXa,b; van der Schoot, 1961) (see Fig. 1 and Table V). The phenylethylamine structure is essential. Shortening or lengthening of the C-C bridge between the phenyl ring and the amino group results in a complete loss of activity (van der Schoot, 1961) ( see Table V ) . The phenyl ring that is separated by two carbon atoms from the

0265 82 b

FIG.1. Cumulative records of locomotor stimulant effects in mice ( m )or rats ( T ) following intraperitoneal (i.p.) administration of various central stimulant drugs by use of the light-beam method. ( a ) phenylethylamine simultaneously with the monoamine oxydase inhibitor pargyline; ( b ) the amino acid L-a-methylmetatyrosine in animals pretreated one hour before with the monoamine oxydase inhibitor nialamide; ( c ) the natural amino acid L-dioxyphenylalanine in an animal pretreated 30 min before with a periferal L-DOPAdecarboxylase inhibitor Ro4-4602; ( d ) dexamphetamine alone. Qualitative similar effects are obtained. The onset of action, the intensity, and the duration are different. The rather long onset of action of treatments ( b ) and ( c ) indicate accumulation of bioactive metabolites.

316

JACQUES M. VAN ROSSUM

TABLE V STRUCTURES AND ACTIVITIES~ OF AMPHETAMINE ANALOQS

0"j'"" WNH2 amp het amine 11001

phenylisobutylamine

101

C

@NH2 1-phenylet hylamine 10)

rNH2 2-phenylethylamine ( 2 ) [ a f t e r MA01 20)

a-methyldopamine

10)

methylene-a-methyldopamine (15)

WNH2 furyl-isopropylarnine

135)

propylhexedrine 115)

tetrahydro 8naphjylamine 1101

Locomotor activity in percent of activity of amphetamine.

amino group may be replaced by an isosteric planar aromatic nucleus such as thiophene and furane (Alles and Feigen, 1941). Substitution on the amino group of amphetamine results in a loss of activity except for substitution of one methyl group as in methamphetamine. Methamphetamine is the most potent drug in this category now available (Schulte et al., 1939). Since enzymatic dealkylation is easily carried out in most animal species, metabolites may be responsible for the psychomotor stimulant action of certain amphetamines. This is, for example, the case for dimethylamphetamine (Section IV,D). There is no correlation of the central stimulant effects with sympathomimetic effects ( Jacobsen and Wollstein, 1939). Although amphetamine may be regarded as a sympathomimetic amine most members of this category are less or not at all sympathomimetic (Jacobsen, 1939). By introduction of heavier substituents on the

ACITON OF PSYCHOMOTOR STIMULANT DRUGS

317

amino group the sympathomimetic action diminishes, so that benzphetamine is not sympathomimetic. Relatively potent peripheral vegetative effects are encountered when an OH group is present in the same position as in noradrenaline and ephedrine. Norpseudoephedrine having the OH group in the opposite position, is a much stronger central stimulant than ephedrine but a much weaker sympathomimetic. It may be concluded that there is a dissociation between the psychomotor and the peripheral sympathomimetic effects of the amphetamines. Amphetamine and a number of other psychomotor stimulant drugs such as phenmetrazine and phentennine are in use as anorectic agents. Through substitution of a chlorine group or a trifluoromethyl group in the 3 or 4 position in the phenyl ring of amphetamine, the central stimulant properties diminish whereas the anorectic properties remain. Thus chlorphentermine, f e d u r amine, and fluminorex are anorectic drugs that lack central stimulant effects. There appears to be a scale of related drugs that are pure stimulants, others are pure anorectic drugs, and a number are mixed in action (van Rossum and Simons, 1969). For instance, amphepramon is an anorectic in low doses and a central stimulant in higher doses. Amphetamine and certain analogs cause a rise in body temperature in animals and man. The fact that xylopropamine has less central stimulating properties than amphetamine, but, in contrast, has stronger pyretic effects suggests that the psychomotor stimulant action has no relation to the possible effects on temperature regulation (van der Schoot, 1961; Mantegazza et al., 1970). The typical psychomotor stimulant action in animals and man may be characterized by an alerting effect, stimulation of motor activity, and suppression of fatigue with adequate doses. Higher doses cause constant motor activity and psychotic behavior. Although most psychomotor stimulants are strongly related to amphetamine, it is possible to distinguish certain subcategories on the basis of the following structural features.

1. The amphetamines as presented in Table 11. Many phenylethylamine analogs have been synthesized that have effects similar to amphetamine. 2. Pemoline is not a basic amine but an acid. The Mg salt is well known for its well-publicizcd but dubious effects on learning

318

JACQUES M. VAN ROSSUM

(Plotnikoff, 1966a,b; Frey and Polidora, 1967; R. G. Smith, 1967). Although pemoline is similar in action to amphetamine, its cellular mechanism of action may be different. Naphtyridine is also an acidic stimulant ( Aceto et al., 1966). 3. The structure of cocaine is completely different from the phenylethylamines (see Table I ) . Cocaine presumably has a different mode of action than amphetamine (van Rossum et al., 1962) in that it potentiates the action of noradrenaline by inhibiting the re-uptake of noradrenaline in sympathetic neurons. The steroisomeric form pseudococaine has the same local anesthetic properties as cocaine but is devoid of central stimulant properties and the noradrenaline potentiating effects (Schmidt et al., 1961). Other local anesthetics do not have a psychomotor stimulant action. 4. Certain amino acids such as dopa and a-methyl-m-tyrosine in sufficient dosages may induce a psychomotor stimulant action in animals (see Fig. 1 and Table VI) (van der Wende and Spoerlein, 1962; Porter et al., 1961). These amino acids are converted into corresponding amines (Blaschko and Chrusciel, 1960; van Rossum, 1963; Carlton, 1963). The corresponding amines, e.g., dopamine, do not pass the blood-brain barrier and therefore can act only when they are formed in the brain. a-Methyl-m-tyrosine does not act as TABLE VI STRUCTURES OF APOMORPHINE AND AMINO ACIDSWITH CENTRAL STIMULATING PROPERTIES

r

HO OH

L-DOPA

1-1 2s

HO

apomorphine

1-1 R

ilH L-a-methylmetat yrosine

1-1 2 s

3 19

ACTION OF PSYCHOMOTOR STIMULANT DRUGS

such but by virtue of its amphetamine-like metabolites both with respect to increase of locomotor stimulation (van Rossum, 1963) and shock avoidance behavior ( Carlton and Furgiuele, 1965). 5. Apomorphine may be considered as a dopamine analog (Ernst, 1965). This drug induces constant motor activity in a variety of animals as does morphine and methadone ( see Table VI ) . D. AESOLUTE CONFIGURATION AND PSYCHOMOTOR STIMULATION Dextrorotatory amphetamine is 2 to 3 times as potent a stimulant as its optical antipode. Dexamphetamine has the same configuration as D-( +)-phenylalanine ( Fischer projection), The carboxyl group of phenylalanine corresponds to the a-methyl group of amphetamine (see Table VII ) . The Fischer projection is unambiguous for a-amino acids and sugars for which certain conventions have been adopted (Hartung and Andrako, 1961). For other classes of chemical substances new conventions have to be established. It is therefore advantageous to use the so-called priority sequence rule of Cahn et al. (1956). According to the sequence rule the absolute configuration of dexamphetamine is 2s. The Fischer projection and the absolute configuration of a number of analogous substances are given in Table VII. The priority sequence is based on atomic weight of the substituents of an asymmetric carbon atom I + Br -+ C1+ OH+ N H 2 + CO -, C N -+ C=C+

C-C-,

H

Replacement of a certain substituent without alteration of the absolute configuration therefore may result in a change of R + S or the reverse. So D-phenylalanine has the configuration 2R, whereas exchange of the carboxylic acid group by a methyl group results in dexamphetamine having the same configuration but indicated by the notation 2s. The absolute configuration thus can be indicated suitably by the R : S notation, whereas drawings of the formulas are necessary to indicate whether different drugs have the same configuration with respect to a given asymmetric center. In this paper the four groups of a center of asymmetry are arranged in such a way that the hydrogen atom is below or above the center (see Table VII). The formulas of the various amphetamine-like drugs insofar as the absolute configuration is known is given according to this convention (see Tables I-V). The various potent stereoisomers have identical configuration at

TABLE VII ABSOLUTE CONFIGWATIONB OF DEXAMPHETAMINE I N RELATION TO

THE

CONFIGURATION OF EPHEDRINE A N D NORADRENALINE~

COOH

CH3 I H-F-NHz

H2

H-C

6

It) phenylalanrne

DzFischer projection [ 2 R sequence rule )

0 I+)amphetamine

I+) 2s -amphetamine

I+)2 s amphetamine

sequence rule

CHNHZ I H-C-OH I

I-)eryt hro

1R:2S 0

a-methyl noradrenaline I-)erythro 1R ; 2 S

noradrenaline

I-)1R

norpseudo ephedrine I+)threo ; IS : 2s

norpseudo ephedrine I+)threo' 1 S : Z S

The notation of absolut'e configuration according to the Fischer projection and the priority-sequence rule is indicated.

ACTION OF PSYCHOMOTOR STIMULANT DRUGS

321

the carbon atom adjacent to the amino group as in dexamphetamine. For instance the active isomers of phenmetrazine and phendimetrazine are the threo 1S:2S form (Clarke, 1962; Dvornik and Schilling, 1965). (See Table 11.) Potent isomers do not always have identical configuration to that of dextroamphetamine. The levorotatory isomer of pipradrol is a potent stimulant whereas the dextro-rotatory isomer is inactive (Portoghese et aZ., 1968). It has been found that (-) pipradrol has the 2R configuration which therefore is not stereochemically superimposable upon 2s-dexamphetamine ( Portoghese et al., 1968; Shafiee and Hite, 1969). The configuration of methylphenidate is the more active CNS stimulant antipode is 1R:2R (Shafi'ee et al., 1967; Shafiee and Hite, 1969). (See also Table 11.) The configuration of the more active (+)threo isomer of phacetoperan is likely 1R:2R. For a number of amphetamines, for example, fencamfamin, the configuration of the most potent isomers have not been elucidated. Norephedrine has the same configuration with respect to the second carbon atom as dexamphetamine and the OH group in the 1R configuration as in L-noradrenaline; it is far less potent as a central stimulant than is norpseudoephedrine which also has the 2s configuration with respect to the second carbon atom but the OH group in the opposite position or 1s configuration (Fairchild and Alles, 1967). Norpseudoephedrine is, however, inferior to norephedrine as an alpha sympathomimetic drug. One must conclude that the central stimulant actions of the amphetamines have no relation to the peripheral alpha sympathomimetic action. L-Amphetamine is about 3 times less active as a locomotor stimulant in rats. Methamphetamine with a codguration identical to that of dexamphetamine is about twice as potent as dexamphetamine but about 10 times as potent as its optical antipode. L-Methamphetamine is therefore less potent than L-amphetamine, while the reverse holds true for the dextrorotatory isomers (van Rossum et al., 1970) (Table VIII). The relative central stimulant and anorectic activity of related stereoisomers deserve further study. The psychomotor stimulant ( - )-cocaine has two asymmetry centers with absolute configuration 2R:3s (Fodor, 1957) (Table I). Although cocaine like hyoscyamine is a tropanol derivative, the geometric configuration is quite different. The stereoisomer ( ) -pseudococaine, which is as potent as cocaine with regard to local anesthetic properties, lacks central stimulant anorectic and

+

322

JACQUES M. VAN ROSSUM

TABLE VIII RELATIONSHIP BETWEEN ABSOLUTE CONFIQURATION AND PSYCHOMOTOR STIMULANT ACTION’ Stimulant

Configuration

Dexamphetamine L-Amphetamine Methamphetamine (-)-Methamphetamine ( +)-Dimethamphetamine ( -)-Dimethamphetamine (+)-No rpseudoephedrine ( -)-Norephedrineb ( -)-Norpseudoephedrine ( +)-Norephedrineb

2s

2R 2s

2R 2s

2R 1s:2s

lR:2S 1R:2R 1S:2R

EDw (@mole/kg)

Relative potency (dexamph. = 100)

3.16 (9.0) 10.0 (38) 1.78 31.6 31.6 316 31.6 (93)

100 (100) 30 (24) 200 10 10 1 10 (‘10) (2) (2 .‘1) (1)

(500) (385) (1000)

a Unpublished results on locomotor stimulant action in rats. Data between brackets from experiments in mice by Fairchild and Alles, 1967. * Toxic in stimulatory doses.

peripheral sympathomimetic effects ( Gottlieb, 1923; Schmidt et aZ., 1961; Schmidt, 1965; Schmidt and Meisse, 1962). The amino acid L-a-methyl-m-tyrosine ( a-MMT) being enzymatically converted into 3-hydroxydexamphetamine produces the same effects as dexamphetamine (van Rossum, 1963). The central stimulant effect of a-MMT depends on the formation of metabolites 3-hydroxydexamphetamine and eventually also metaraminol. Metaramino1 is the erythro form with the 1R:2S configuration, thus having the hydroxy group in the side chain in the same position as natural L-noradrenaline and ephedrine. With regard to the differences in central stimulant activity of norephedrine ( 1R :2s ) and norpseudoephedrine ( 1s:2s ) , it is unlikely that metaraminol acts as an essential metabolite in the psychomotor stimulant action of a-MMT. Since aromatic decarboxylation is stereospecificfor ],-amino acids (Lovenberg et al., 1962), it can be visualized that dexamphetamine can be formed from a-methylphenylalanine. The lattc,-r compound, however, has no central stimulant activity, presumably because decarboxylation proceeds extremely slowly. Apomorphine has certain stimulant effects in common with dexamphetamine and L-dopa and has identical configuration (2R) at the asymmetry center as dexamphetamine ( 2 s ) (Corrodi and Hardegger, 1955); see Table VI.

ACITON OF PSYCHOMOTOR STIMULANT DRUGS

323

II. Effects of Psychomotor Stimulant Drugs in M a n

The alerting effects of amphetamine in man were soon observed following studies of the antinarcosis effect of amphetamine in animals ( Leake, 1958). A. EFFECTS OF AMPHETAMINESIN THERAPEUTIC DOSES The first clinical application of amphetamine was in the therapy of narcolepsy (Prinzmetal and Bloomberg, 1935). Except for certain applications in some forms of epilepsy, parkinsonism, and depressions, the treatment of narcolepsy is still the only unequivocal indication for psychomotor stimulant drug ( Editorial, 1968). For the treatment of obesity, certain amphetamines may be used, but the specific anorectics are preferable. The possibility exists that certain anorexigenic drugs such as aminorex and cloforex may induce pulmonary hypertension in man (Schwingshackl et al., 1969). Such toxic reactions are apparently unrelated to sympathomimetic effects. The alerting effect or the production of a state of wakefulness accompanied by an increase of all kinds of psychic and motor activity is the most characteristic effect of the psychomotor stimulant drugs in man. Diminished fatigue, a better adjustment toward work, as well as suppression of sleep (Bahnsen et at., 1938) and elevation of mood are the consequences (Nathanson, 1937; Leake, 1958; Connell, 1958; Kalant, 1966). The enhancement of human performance by amphetamines has been proved (see review by Weiss and Laties, 1962). The amphetamines may therefore be used in cases of severe emergency. During World War I1 methamphetamine and dexamphetamine were used extensively ( Bett et al., 1955; Ivy and Krasno, 1941; Ivy and Goetzl, 1943). The psychomotor stimulant effects of amphetamines in man are variable in intensity (Jacobson and Wollstein, 1939) and are experienced as a sense of energy and self-confidence and the occurrence of quicker mentation and decision making (Kalant, 1966). The euphoric effects experienced in a number of individuals is one of the reasons for its misuse by addicts. Following repeated administration of amphetamine, tolerance develops in man (Rosenberg et al., 1963). There is no cross tolerance to LSD in subjects tolerant to amphetamine and there is no cross tolerance to amphetamine in those tolerant to LSD (Rosenberg et al., 1963).Cross tolerance between the hallucinogens LSD,

324

J A C Q W M. VAN ROSSUM

psylocybine, and mescaline occurs (Isbell et d.,1961; Wolbach et al., 1962). Tolerance to the awakening effect of the amphetamines apparently does not occur, because in the treatment of narcolepsy the dose need not be increased (Leake, 1958). The peripheral side effects are of the sympathomimetic type: namely, increase in blood pressure and heart rate, pupillary dilation, and relaxation of the smooth muscle of the gastrointestinal tract ( Kalant, 1966). In higher doses excessive pupillary dilatation, hypertension, and tachycardia prevail. The various psychomotor stimulant drugs differ greatly in their relative potencies of central stimulant and peripheral sympathomimetic effects.

B. Toxrc EFFECTS OF PSYCHOMOTOR STIMULANTSIN MAN After injection of a dose of more than 20 mg dexamphetamine or intravenous administration of psychomotor stimulants, toxic effects of overstimulation of the central and peripheral sympathetic nervous system may be experienced. Kalant (1966) has given a thorough review of amphetamine intoxication with case reports. Toxic symptoms may include restlessness, hyperactivity, convulsions, tremor, tenseness, irritability, insomnia, confusion, delirium, and anxiety (Leake, 1958; Espelin and Done, 1968). Paranoid psychosis induced by amphetamine with suicidal or homicidal behavior is often encountered, especially after chronic use of intravenous administration ( Connell, 1958; Kalant, 1966). In addition, a number of toxic effects may be observed (due to sympathetic stimulation) such as profuse sweating, rapid breathing, headache, pallor, palpitation, tachycardia, hypertension, and cardiac arrhythmias or circulatory collapse (Leake, 1958). The treatment of amphetamine poisoning should not be symptomatic as with sedatives, but neuroleptic drugs such as chlorpromazine and haloperidol (Espelin and Done, 1968) should be used. The neuroleptics are specific amphetamine antagonists (see later). Chlorpromazine or thioridazine are the drugs of choice for intoxication due to dexamphetamine. Haloperidol is indicated for the treatment of poisoning with pure psychomotor stimulants.

C. ADDICITON TO PSYCHOMOTOR S m m m DRUGS Most adult male inhabitants of the Peruvian mountains are chronic coca chewers. They seldom develop psychotic reactions and they usually abandon their habit when they move to lower altitudes

ACTION OF PSYCHOMOTOR STIMULANT DRUGS

325

( Molina, 1946; Gutierrez-Noriega and Zapata-Ortiz, 1947). The use has spread to the highlands of other countries in South America such as Bolivia and the Salta area of Argentina. Addiction to cocaine with development of tolerance has been described vividly and extensively by Lewin (1893). Abuse of amphetamine, methamphetamine, phenmetrazine, and other drugs is well documented (Morimoto, 1957; Kramer et al., 1967; WHO, 1964). Addiction first occurred on a large scale in Japan and Germany shortly after World War I1 (Morimoto, 1957; Bonhoff and Lewrenz, 1954). In the last few years the abuse appears to be increasing in large cities, where adolescents administer the drugs by intravenous route. Extremely intensive but short-lasting euphoric effects are experienced by the intravenous users (Kramer et aZ., 1967). Many addicts initially use amphetamine orally but they rapidly change to the intravenous route and inject themselves with progressively larger doses of amphetamine. The amount injected generally ranges from 100300 mg per dose but may be as high as 1 gm methamphetamine every 2 hours (Kramer et al., 1967). The users get a sudden generalized overwhelming pleasurable feeling called a “rush.” The mood is euphoric with intense concentration of thoughts and activities which extends to flight of ideas with paranoia and anger (Kramer et aZ., 1967). A review on the abuse of methamphetamine has been given by Hawks et al. ( 1969). The rapid development of tolerance implies that only the very potent and lipophilic psychomotor stimulant drugs are dangerous addictive substances. These are dexamphetamine, methamphetamine, phenmetrazine, and cocaine. Other stimulants listed in Table I1 may be dangerous but presumably less so than the above mentioned. Also, the anorectic agents that have psychomotor stimulant “side” effects may to some extent be addictive stimulants. D. AMPHETAMINEPSYCHOSIS INDUCED BY PSYCHOMOTOR STIMULANT DRUGS Cases of paranoid psychosis due to chronic cocaine abuse have been reviewed as early as 1892 (Lewin, 1893). The cocaine addicts are paranoid, experience persecutory and auditory hallucinations, and eventually may become aggressive against suspected enemies (Lewin, 1893). The occurrence of paranoid psychosis and psychotic states with auditory hallucinations, delusions of persecution, anxiety,

326

JACQUES M. VAN ROSSUM

and hostility elicited by chronic or sometimes even by an occasional use of amphetamine is well documented and resembles paranoid schizophrenia ( Connell, 1958; Kalant, 1966; Kosman and Unna, 1968). Amphetamine intoxication in humans is accompanied frequently with stereotyped movements, as, for instance, repeating the same sentences constantly (Tatetsu et aZ., 1956; Bonhoff and Lewrenz, 1954; Connell, 1958; Munkvad et al., 1968). The first indication of amphetamine-induced psychosis has been noticed in the treatment of patients for narcolepsy with amphetamine (Young and Scoville, 1938). Although a number of individuals who show psychotic reactions associated with amphetamine intoxication are abnormal personalities, normal persons may become psychotic following amphetamine abuse (Connell, 1958; GrifEth et d.,1W; Kalant, 1966). Paranoid psychotic reactions are a real danger to society since there is a causal relationship between amphetamine use and crimes of violence and sex offenses ( Walsh, 1964). Drug-induced psychosis is predominantly the case following use of dexamphetamine, methamphetamine, and phenmetrazine ( Askevold, 1959; M. Herman and Nagler, 1954; Bonhoff and L,ewrenz, 1954; Connell, 1958; Marley, 1960).Chronic use of methylphenidate also causes paranoid psychosis ( McCormick and McNeel, 1963), as does amphepramon (Kuenssberg, 1963; Baumer, 1966) and ephedrine (Herridge and a’Brook, 1968). Other stimulants and anorexogenic drugs with stimulating properties may be the cause of psychoses in the future. It has been shown that methyldopa in combination with a monoamine oxidase inhibitor elicits a strong amphetamine-like stimulation in mice (van Rossum and Hurkmans, 1963).These findings suggest that amines related to dopamine might cause psychotic reactions in man. A patient with Huntington’s chorea reacted well to a treatment with methyldopa but became psychotic when the therapy was extended with isocarboxazide ( Korten and Pelckmans, 1968). In male schizophrenics there is apparently an excessive brain stimulation as detected by quantitative EEG (Goldstein and Beck, 1965). Since amphetamine addicts experience a florid paranoid psychosis but do not exhibit the typical dissociated and autistic disorganization of thinking associated with schizophrenia ( Kramer d al., 1967), an amphetamine-like stimulation might be one but not the only etiologic factor in schizophrenia.

327

AClTON OF PSYCHOMOTOR STIMULANT DRUGS

Ill. Effects of Psychomotor Stimulant Drugs in Animals

A. LOCOMOTOR STIMULANTE F F E ~ The amphetamines exert an awakening effect in animals anesthetized by various hypnotics (Tainter et al., 1939). It has been observed that, although certain analogs of amphetamine had little power to hasten awakening from deep narcosis, they exhibited excitation after the animals were aroused (Tainter et al., 1939). This central excitation, characterized by restless coordinated movements, differs from the convulsive seizures that occur from stimulation of medullary centers and the spinal cord, suggesting stimulation of higher centers in the brain (Tainter et al., 1939). Motor stimulation in small laboratory animals may be measured through observation or recorded in a number of ways (Kinnard and Watzman, 1966; Riley and Spinks, 1958). Observations of cats before and after treatment with amphetamine have been recorded (Norton, 1967). Some behavioral patterns, for example, ‘%head down simultaneously with stretching” (which is not seen significantly in controls), occurred with amphetamine. Amphetamine (P. W. Dews, 1953) and other psychomotor stimulant drugs such as pipradrol and cocaine cause a strong increase in locomotion in mice and rats (Schulte et al., 1941; Tainter, 1943; van der Schoot, 1961; van Rossum et al., 1962; C. B. Smith, 1965).

w-1 0964 852 b m 23

0964 852c

if&& I5 min

t

Cocaine 56.2 p l e / k g i.p.

0964 852 d

Cocaine 56.2 pmole/kg oral

0964 852 e

k f Cocaine

10 pmole/kg i.v

FIG. 2. Cumulative records of locomotor activity induced by 1-cocaine administered in the mouse via various routes. The onset of action is obviously the most rapid for i.v. injection. The intraperitoneal route is faster than the subcutaneous route. Cocaine is practically devoid of activity when given orally.

328

JACQUES M. VAN ROSSUM

Figure 2 shows cumulative records of locomotor stimulation in mice following administration of cocaine via various routes. A comparison of the action of various psychomotor stimulants in rats is found in Fig. 3. The onset of action and the intensity and duration of locomotor stimulation differs for the various stimulants. The stimulant effects of cocaine and phenmetrazine are fast, intense, and short. The effect of norpseudoephedrine is long lasting. It is probable that cocaine and phenmetrazine have a high addiction liability because of the fast and intense actions, which would be intensified after intravenous administration. Increase of locomotor activity is also obtained with a-methylInterruptiondmin

I00

R2o dexamphetamine 562pnole/kg I P

] A

p

w

ppseudoephedrine.

( t ) nor-

Interruptions/min nterruotions/min

1

R2a cocaine 31.6pmole/kg i p.

I00 (threo)

H 20 (t) methomphetamine

3 16 pmole/kg i.p.

50

Interrupti o n s h i n

R3m 3.16 p o l e / kg i.v.

1

6b

'

( t ) methamphetamine

'

j0

90 min

1

R3m (t) methamphetamine

3.16 p o l e / k g i.p.

30

60

90

R3m

( t ) methamphetomine

3.16 p o l e / k e p.0.

min

FIG. 3. Records of average locomotor activity per interval, following i.p. administration of various psychomotor stimulant drugs in one rat as well as administration of methamphetamine via various routes in another rat. Cocaine is short acting with a fast onset and high intensity. Norpseudoephedrine is long acting. Methamphetamine is the most potent drug presently available.

ACTION OF PSYCHOMOTOR STIMULANT DRUGS

329

m-tyrosine (van Rossum, 1963) and with dopa in rats pretreated with a monamine oxidase (MAO) inhibitor or a peripheral decarboxylase inhibitor. Even methyldopa in rats pretreated with a MA0 inhibitor exerts after a few hours a strong amphetamine-like effect on locomotor activity. Apomorphine stimulates to some extent, but stereotyped movements predominate over locomotor stimulation. Locomotor stimulation may be considered as a characteristic effect of psychomotor stimulants in mice and rats.

B. PSYCHOMOTOR STIMULANTS ON OPERANT BEHAVIOR Facilitating actions of amphetamine in behavior can be objectively measured by the operant behavioral techniques; rats, cats, monkeys, etc., can be trained to press a lever to obtain food or water (positive reinforcement) or to avoid an electric shock (negative reinforcement). The training can be done on the basis of different schedules of reinforcement ( Ferster and Skinner, 1957). The effect of the amphetamines on operant behavior has been reviewed by P. B. Dews and Morse (1961). In rats working for food on a fixed ratio schedule (FR), amphetamine and related drugs cause cessation of response similar to the behavior of satiated rats (Fig. 4 ) . Eventually they still press the lever but do not eat the food presented. The best example is fencanfamine in rats. Long pauses in food reinforcement in rats on a fixed ratio performance occurs after cocaine administration (Pickens and Thompson, 1968). In rats working for food on a fixed interval schedule (FI) amphetamine increases the rate of responding (Cook and Kelleher, 1962; see also Ray and Bivens, 1966). On a schedule in which animals must wait a certain period in order to obtain the next reinforcement ( differential reinforcement of low rates schedule DRL), amphetamine causes disruption of efficient behavior by shortening of the waiting times (Sidman, 1955; Kelleher et al., 1961; Segal, 1962). The temporal pattern of behavior influenced by amphetamine has been studied by Weiss and Laties (1964). Performance of rats under influence of amphetamine on a multiple schedule has been studied by Clark and Steele (1966). Changes in instrument responding elicited by amphetamine are disturbed to some extent by the inhibitory effects of amphetamine on hunger drives (see Poschel, 1963). Amphetamine and methamphetamine increase the rate of responding of rats conditioned

330

JACQUES M. VAN ROSSUM

0

8

15min

FIG.4. Cumulative records of lever pressing for food of a trained rat on a fixed ratio (FR 30) schedule following injection of saline and increasing doses of dexamphetamine. Pausing normally occurs after lever pressing for Iof an hour but occurs within 10 minutes when dexamphetamine has been given. A 1 mg/kg (5.62 pmole/kg) dose causes response to cease in all rats tested.

to press a lever in order to avoid shocks. Consequently the number of shocks received is decreased ( Verhave, 1958). Amphetamine differentially increases responding on a variable interval schedule and decreases the numbers of shocks accepted by rats in the conflict situation (Geller and Seifter, 1960) (Fig. 5 ) . Behavioral tolerance in response to chronic administration of dexamphetamine has been observed in schedules of reinforcement (FR, DRL, FI) when responses led to a decrease of food reinforcement ( Schuster et d.,1966). However, chronic administration of dexamphetamine led to a uniform increase in rate of responding on scheduled negative reinforcement; as a consequence the rate of shock reinforcement was decreased ( Schuster, 1966). In a test situation in which rats were trained to regulate their environmental temperature by pressing a lever that turned on a heat lamp, it could be shown that amphetamine increased the frequency even though the skin temperature was driven above normal (Weiss and Laties, 1963). Amphetamine reduces freezing behavior and consequently causes improvement of shock avoidance (Krieckhaus et al., 1965). Suppression of freezing may be the reason that psychomotor stimu-

ACTION OF PSYCHOMOTOR STIMULANT DRUGS

331

No drug

J d -Amphetamine 0.5 mg/kg immediately

prior

FIG. 5. Cumulative records of lever pressing of a rat conditioned on a variable interval schedule (VI 2 min) alternating with continuous reinforcement and simultaneous shock (conflict situation). In the absence of a drug the rat does respond to the conflict situation when the shocks are of low intensity (number indicated). After injection of dexamphetamine response in the conflict situation is drastically suppressed. Reproduced after Geller and Seifter (1960) with permission of the authors and Psychophamcologiu.

lants (including pemoline) seem to facilitate learning of instrument responding in an avoidance situation but not in case of a positive reinforcement situation. Amphetamine has a positive influence on performances ( “learning”) in rats which showed consistently low rates of avoidance in a shuttle-box in spite of extensive training ( Rech, 1966). It may be concluded that amphetamine facilitates ongoing operant as well as spontaneous behavior.

C. PSYCHOMOTOR STIMULANTS ON SELF-STIMULATION Electrical stimulation in certain areas of the brain are experienced by cats and rats as a reward, and animals learn to press a lever in order to receive an electric stimulus to their brains (Olds, 1956, 1962). Amphetamine increases self-stimulation in cats with electrodes in the lateral hypothalamus or caudate nucleus (Horo-

332

JACQUES M. VAN ROSSUM

Responsesl

I

H

10min

FIG.6. Cumulative records of lever pressing for self-stimulation in a cat with electrodes in the lateral hypothalamus. Amphetamine facilitates selfstimulation, whereas chlorpromazine exerts the opposite effect. Reproduced after Horovitz et a2. (1962a,b) with permission of the authors and Psychopharmacologia. vitz et al., 1962a) and in rats with electrodes in the posterior hypothalamus and midbrain segmentum (Stein and Ray, 1960; Stein, 1964a,b, 1967) (Fig. 6). The facilitating effect is not due to nonspecific augmentation of motor activity since the effect abruptly stops if the electric current is turned off (Horovitz et al., 1962a; Stein, 1964a). Amphetamine facilitates the hypothalamic reward system which may be its mode of action as a stimulant (Stein and Seifter,,1961). The facilitation of self-stimulation is probably due to an amphetamine-induced release of catecholamines in the rewarding system (Stein, 1962a, 1964b). Other psychomotor stimulant drugs do the same. For instance, facilitation of hypothalamic self-stimulation has been observed for methamphetamine and methylphenidate (Stein, 1964b,c). Phenylethylamine, as such, is without effect [Fig. 7( a ) ]

ACTION OF PSYCHOMOTOR STIMULANT DRUGS

333

Phenethylomine hydrochloride ( I mg/kg*) Responses

Phenethylamine hydrochlorlde ( I rng/kg*) 3 hours after iproniazid phosphate (100m g / k g )

///)(/)////// AA/+$vl/lA/

A-

Equimolar with I mg/kg d-Amphetamine sulfate

200 Self stimulations

L

10 min

Q

-methyl-m- tyrosine 300rng/kg

A-

Cont'D

FIG. 7. Cumulative records of lever pressing for self-stimulation on a variable schedule interval in rats with electrodes in the posterior hypothalamus. Dexamphetamine alone and phenylethylamine in combination with a monoamine-oxidase inhibitor both induce a facilitation of self-stimulation. The amino acid a-methyl-rn-tyrosine also facilitates self-stimulation but only after a latency period of 45 minutes. Reproduced after Stein (1964b) with permission of the author and Fed. Proc. Compare with the locomotor stimulant effects of the same treatments in Fig. 1.

334

JACQUES M. VAN ROSSUM

but simulates amphetamine after M A 0 inhibition (Stein, 1964b) [Fig. 7(b)]. The effect of cocaine needs clarification. Low doses that have no facilitating effect of amphetamine on self-stimulation (Stein, 1 9 6 4 ~ )a-Methyl-rn-tyrosine . exerts an amphetamine effect after a latency period of about 45 minutes (Fig. 7) (Stein, 1964b). The time pattern is similar as that for locomotor stimulation (van Rossum, 1963). The facilitating action on hypothalamic self-stimulation may be due to an action on a different brain area such as the reticular formation (Horovitz et al., 1962b). Since the lateral hypothalamus is involved in feeding behavior and since amphetamine is an anorectic drug, it might be expected that amphetamine would inhibit self-stimulation ( Umemoto and Kido, 1967). In relatively high doses Umemoto and Kid0 observed such an inhibition in cats. However, this may have been due to overstimulation. Facilitation of self-stimulation parallels locomotor stimulation to a large extent except for cocaine, which may indicate a different mechanism for cocaine.

D. STEREOTYPED BEHAVIORTHROUGH PSYCHOMOTOR STIMULANTS Amphetamine (0.5-2 mg/kg) hyperactivity has a stereotyped character in mice and rats, such as biting the wires of the cage, backward walking, and self-mutilation (Hohn and Lasagna, 1960; Schulte d al., 1941; Irwin et nl., 1958; Randrup et al., 1963; Janssen, 1961).Furthermore, excessive sniffig over a restricted area of the cage (Randrup et al., 1963), compulsory gnawing ( Janssen, 1961; Janssen et al., 1965), and purposeless searching head movements (Emele et al., 1961) have been observed. Stereotyped movements other than increased locomotion generally occurs with higher doses in rats and mice (Lht, 1965; van Nueten, 1962; Quinton and Halliwell, 1963; Randrup and Munkvad, 1965, 1966a,b, 1967a). Other psychomotor stimulant drugs cause similar stereotype movements. In a number of other animal species this activity appears to be absent, and only stereotyped behavior has been observed. For instance, in cats and monkeys coiitinuous staring and certain head movements occur (Randrup and Munkvad, 1967a) . Apomorphine causes stereotyped movements such as constant snifEng and compulsory gnawing (Harnack, 1874; Amsler, 1923; Ther and Schramm, 1962; Ernst, 1965, 1967; Janssen et al., 1965),

whereas locomotor stimulation is minimal. Apomorphine causes excessive hoarding of ‘‘nonsense food” by golden hamsters (van Rossum and Simons, 1970). Stereotyped behavior is predominant following apomorphine treatment. whrreas with amphetamine increased locomotion suppressrs stcrcotype behavior to some extent.

E. ACTION OF PSYCHOMOTOR STIMULANT DRUGS ON SOCIALBEHAVIOR Rats kept together in a cage. exhibit a number of social interactions such as gnawing, crouching, rearing, crowding in a corner of the cage, agression, and mating. Time spent in these activities decreases under the influence of psychomotor stimulants (Irwin et al., 1958). Mating activity in rats was found to be increased to some extent by low doses of amphetamine, whereas high doses which produce stereotype movements have adverse effects on mating activity ( Bignami, 1966). Amphetamine-treated rats separate from each other and no longer stay together in a corner of the cage. Aggression decreases and exploration drives are exaggerated so that they run around aimlessly (Chance and Silverman, 1964). Behavioral changes resembling aggression and fighting have been observed (Chance, 1947, 1948; Moore, 1963). Fighting between rats after repeated doses, but not after a single dose, has been reported ( Ehrich and Krumbhaar, 1937; Randrup and Munkvad, 1967a,b). Amphetamine in high doses facilitates the attack behavior in cats, which is elicited by electrical stimulation of the lateral hypothalamus as well as the enhancement of facilitation of such attacks by simultaneous stimulation of the midbrain reticular formation ( Sheard, 1967). The psychomotor stimulant methylphenidate but not amphetamine shows inhibition of aggressive behavior in mice (Valzelli et d.,1967). Social interaction in rats is augmented into bizarre forms when amphetamine is given after pretreatment with reserpine ( Morpurgo and Theobald, 1966) or with diethyldithiocarbaniate ( van Rossum and Lammers, 1970). Rats assume positions in pairs standing on their hind legs with their noses and forcyaws in close contact (see Fig. 8a ) . This bizarre behavior resembles aggressive postures but the rats do not fight (Morpurgo and Theobald, 1966). Similar bizarre behavior can be elicited with apomorphine (Schneider, 1968; van Rossum and Simons, 1970).

FIG. 8. Bizarre social behavior of male Wistar rats induced by various treatments. ( a ) Dexamphetamine (5.62 pmole/kg=l mg/kg) in 4 rats pre-

ACTION OF PSYCHOMOTOR STIMULANT DRUGS

337

Apomorphine-injected Wistar rats show excessive sniffing within 5 minutes after injection and bizarre behavior with sound. The rats arrange themselves in pairs when they meet a partner and make rhythmic movements with thcir forepaws (Fig. 8b). It is interesting to note that through a selective increase of dopamine in the brain after injections of L-dopa and a peripheral dopa decarboxylase inhibitor a bizarre social behavior has been observed (Lammers and van Rossum, 1968) (Fig. 8c). This type of behavior is reminiscent of psychosis in humans. It may be concluded that apomorphine as such and amphetamine under certain situations may cause “psychotic” behavior in animals. IV. Kinetics of Absorption, Distribution, a n d Elimination of Amphetamines

The intensity of psychomotor stimulant action depends on the activity of the stimulant and on the concentration at the locus of action, i.e., the concentration in neurons in certain areas of thc brain. The concentration in the brain in turn depends on the plasma concentration of the amphetamines. For the psychomotor stimulant drugs the relationship between brain concentration and plasma concentration has not been studied. In analogy with other drugs, this relation will depend strongly on the physicochemical properties of the stimulants. The rate of absorption, distribution, and elimination also is largely determined by the same physicochemical properties. A. PHYSICOCHEMICAL PROPERTIES OF AMPHETAMINES The amphetamines are weak to moderate-strong organic bases with a pK, value of roughly between 8 and 11.The pK, of amphetamine is 9.8 (Leffler et al., 1951), so at the physiological pH this drug is largely in the ionized form while only a small fraction of treated 15 hours before with reserpine ( 2 mg/kg). The onset of bizarre behavior occurs within 10 minutes after dexamphetamine. ( b ) Apomorphine (3.16 @mole/kg = 1 mg/kg) in 4 rats. The onset of action is within 5 nunutes. ( c ) L-Dopa (316 amole/kg = 63 mg/kg) in 6 rats treated with the peripheral Dopa decarboxylase inhibitor Ro4-4602 ( 316 amole/kg = 52 mg/kg). The onset of bizarre behavior occurs after about 30 minutes, while maximum effect is obtained between 1 and 2 hours after the Dopa injection. Reproduced after Lammers and van Rossum (1968) with permission of the authors of the E. J. P. The effects of the three treatments are qualitatively similar. The bizarre social behavior in treated animals always can be elicited by sounds.

338

JACQUFS M. VAN ROSSUM

this drug is in the neutral form. Passage of drugs through the lipid cell barriers and the blood-brain barrier occurs predominantly in the neutral form. This would imply that amphetamine would act slowly. Cocaine, on the other hand, is a weak base with a pK, of 7.6 with about 40%in the neutral form. The onset of action of cocaine is therefore predictably fast. The pK, value and the fraction of drug in the neutral form of a number of psychomotor stimulant drugs is given in Table IX (Vree et al., 1969). The capacity of the neutral form to pass lipid barriers depends on the lipid solubility of the drug and its affinity for the tissues in the brain that bind the drug. One measure of lipid solubility is the distribution coefficient TABLE I X D~SSOCIATION COXSTANTS AND PARTITIOS COEFPICIENTS OF AMPHETAMINESO Apparent partition True partition neiitral coefficient coefficient at pH 7.4 at pH 7.4 CHCl,/H,O Hept,/H,O CHCla/H20

Percent

Ihig

Phenylet,hylamine Dexamphetamine Methamphetamine Et.hylamphetamine Iaopropylamphetamirie Benzylamphet.amine L)imet,hylamphetamine Rlethylethylamphetamine Methylisopropylaniphetamine Benzphetamine Phentermine Mephentermine Chlorphentermine Norephedrine Ephedrine Methylephedrine Norpseudoephedrine Pseudoephedrine Propylhexedrine Phenmetrazine Phendimetrazine

pK.

9.88 0.33 9.90 0.31 10.11 0.19 10.23 0.15 10.14 0.18 7.50 44.1 9.80 0.39 9.80 0.39

0.078 0.48 1.11 2.67 8.09 1000 11.5 19.0

9.4.7 0.88 6.55 87.2 10.11 0.19 10.25 0.13 9.60 0.62 9.5.i 2.80 9.60 0.62 9.30 1.25 9.40 1.00 9.86 0.33 10.74 0.043 8.45 16.8 7 . . 3 44.0

100 1000 1 .oo 1.22 4.00 0.0010 0.0152 1.00 0.001

Aft,er Vree et al. (1969).

0.070 1.11 15.6 1000

0.28 1.88 5.14 38.6 117 110 108 166

20.8 146 565 17!)0 4460 2250

28!)0 47fiO

11300 200 74.8 1400 63.2 514 110.6 806 17.5 797 levels alone are sufficient to guarantee amphetamine-indiicctl stereotypy but that for locomotor stimulation noradrenaline is necessary in addition. The apomorphine stereotyped movcmcnts and bizarre social behavior is not iiifluenced by (U-MPT(Ernst, 1967) h i t is to some extent by reserpine plus wMPT (van Rossum and Simons, 1970). Apomorphine therefore may directly interact with dopamine receptors (Ernst, 1967). It has been shown that the combination of apomorphine ( a dopamine receptor activator in thc brain) and catapresan ( a noradrenaline receptor activator i n thr~brain) can counteract the reserpine-induced behavioral dcprcssioii ( motor activity) in a similar way to dopa or aniphetaminc: ( A n d i ~ i1970). , Neither catapresan nor apomorphinc does this. Tlicse data suggest the importance of both noradrenalinc and dopaininc receptors in the psychomotor stimulant action of amphetaminc ( AndPn, 1970). The neuroleptic drugs are potent antagonists of dopamine ( van Rossum, 1967) and of apomorphine and various components of the amphetamine effect (Jaiissen et al., 1965, 1967). From the data presented above it may be concluded that dopamine receptors play an important role in various aspects of amphetamine-induced bchavioral cxcitation. In stereotyped and bizarre social behavior dopamine has the predominant role. In locomotor activity dopamine rcccptors are involved, but also noradrenaline receptors have some function.

c. THE MIDBRAINKETICULAR

F o R n i A T I o h - AKD

PSYCHOMOTOR

STIMULANTACTIOX Amphetamine induccs activation of the EEG similar to stimulation of the reticular formation ( Imigo and Silvestrini, 1957; Hiebel et al., 1957; Bradley and Elkes, 1957; van Meter and Ayala, 1961). With low doses ( u p to 0.5 mg/kg) a disappearance of slow waves

370

JACQUES M. VAN ROSSUM

and spindles is observed; furthermore, waves of the subconvulsive type occurred in the occipital recording. With larger doses ( u p to 3 mg/ kg ) there is activation, especially in the frontal recordings ( Longo and Silvestrini, 1957). In addition, amphetamine lowers the threshold for behavioral and EEG arousal following stimulation of the midbrain reticular formation (Bradley and Elkes, 1957; Bradley and Key, 1958). Amphetamine-induced EEG activation is obtained in cats with a section at the c1 level (encbphalc isolk) but not in cats with a section at a higher level (cerveau isolk). From such data it has been concluded that the locus of action of amphetamine is in the midbrain reticular formation (Bradley and Key, 1958). The acetylcholine output in the cerebral cortex is increased by amphetamine as well as by electrical stimulation of the reticular formation ( Pepeu and Bartolini, 1967). The increase caused by amphetamine was found to be blocked by chlorpromazine, not influenced by pentobarbital, and facilitated by imipraminc, whereas the increasc caused by reticular formation stimulation was blocked by pentobarbital, reduced by chlorpromazine, and not influenced by pentobarbital ( Pepeu and Bartolini, 1968). Amphetamine affects the firing rate of cortical neurons in different ways. The majority show increased discharge frequency, whereas a few are inhibited or not affected by amphetamine (Herz and Fuster, 1964). Neurons with relatively high spontaneous activity ( >3/sec) are further activated by amphetamine, while slower units ( < 3 / s e c ) are not affected or inhibited (Herz and Fuster, 1964). Under influence of methamphetamine, resting alpha motoneurons of the tonic extensors start to fire while already active neurons show an increase in firing frequency. As a consequence, the monosynaptic reflex amplitude increases, and activation of the tonic alphamotoneurons occurs (Haase and Tan, 1965). It has heen concluded that the interneurons of the tonic systems are activated by methamphetamine. It may be concluded that, as far as the awakening effect had increased locomotion is concerned, amphetamine may act directly in the midbrain reticular formation as well as in the cerebral cortex and, at a lower lever, in the spinal cord.

D. THENEOSTRIATUM AND PSYCHOMOTOR STIMULANT ACTION Convincing evidence has accumulated that brain dopamine is involved in the psychomotor stimulant action of amphetamine and related drugs (van Rossum and Hurkmans, 1964).

ACTION OF PSYCHOMOTOR STIMULANT DRUGS

371

In the central nervous system dopamine is found mainly in the nucleus caudatus, putamen, tuberculum olfactorium, nucleus accumbens, nuclei interstitialis striae terminalis and eminentia mediana (Fuxe, 1965). Of these areas the neostriatum and nucleus accumbens at first seem to be of primary importance. Furthermore only certain components in the central effects such as stereotyped and bizarre social behavior are related to brain dopamine. When the locomotor stimulant effect of amphetamine is studied in rats with chronic brain lesions, it is observed that the sensitivity to amphetamine is increased in rats with lesions in the frontal or posterior cortex but not in rats with lesions in the caudate nucleus (Adler, 1961). These data suggest that the caudate nucleus is not an important area for amphetamine-induced locomotor stimulation. Apomorphine exerts only the stereotyped and bizarre social behavioral effects of amphetamines. After ablation of the striatum in guinea pig and rats the apomorphine-induced gnawing was abolished ( Amsler, 1923) . Also amphetamine-induced stereotyped behavior is prevented in rats with lesions in the caudate nucleus. Microinjections of apomorphine or dopamine in the striatum induces stereotyped behavior (Ernst and Smelik, 1966; Fog et al., 1967; Fog, 1967). In cats with cannulae in the head of the caudate nucleus, dopamine and dexamphetamine produce similar effects in behavior registered on ethograms, whereas noradrenaline injections produce totally different effects ( Cools and van Rossum, 1970). Furthermore, the behavioral effects of stereotactically administered dopamine and amphetamine were abolished by low doses of a neuroleptic drug such as haloperidol while the blockade could be surmounted by increasing the dose of amphetamine or dopamine (Cools and van Rossum, 1970). The neuroleptic drugs are selective dopamine antagonists (van Rossum, 1967). Rats with unilateral lesions in the nigrostriatal system turned to the homolateral side following systemic dopa injections and to the heterolateral side to the lesion following administration of a neuroleptic drug (And6n et al., 1966). Microinjection of quaternized neuroleptics directly into the neostriatum antagonized also the amphetamine-induced stereotyped behavior (Fog et al., 1968). The neuroleptic drugs that are selective amphetamine antagonists are also dopamine antagonists presumably at the level of the caudate nucleus. They increase turnover of striatal dopamine as evidenced by increased levels of metabolites of dopamine such

372

JACQUES M. VAN ROSSUM

as homovanillic acid and dioxyphenylacetic acid (Roos, 1965; Sharman, 1966). Dexamphetaminc acts in the striatum and other structures containing dopamine nerve terminals by release of dopamine. Perfusion through a push-pull cannula with dexamphetamine has been found to result in a significant increase in dopamine output from the caudate nucleus ( McKenzic and Szerb, 1968). Evidence was obtained that the aniphetamine-induced dopamine [Fig. 20( a ) ] release occurred from a nondiffusible bound form. The metabolite dioxyphenylacetic acid [Fig. 20 ( b ) ] was found to be increased by amphetamine and, as anticipated, reduced by an M A 0 inhibitor (McKenzic and Szerb, 1968) with an elegant in vitro preparation, it has been shown that low doses of amphetamine 0 Locke ( 4 ) 0 Pheniprazine ( 3 )

I

Q Dextroamphetamine ( 8 ) Dextroamphetamine (4)

83 Locke ( 3 ) Pheniprazine (3) 0 Dextroamphetamine ( 8 )

Caudate nucleus 300

-

Cerebellum

250 -

I -

40 -

R

E

s

5 30E

8 \

m

= 20&

6

R

0" l o 050

100

150

200 rnin

FIG.20. Release of dopainine and its metabolite dioxyphenylacetic acid by perfusion of the caudate nucleus with dexarnphetamine and other dnigs through a push-pull cannula system. Dexamphetamine causes a strong increase of dopamine (left) and dopac (right) in the effluent in untreated cats and even inore in cats treated with d o p a . The monoamine oxiclase inhibitor pheniprazine has little effect on the release of dopainine but as anticipated suppresses the oxydation of dopamine to dopac. Reproduced after McKcnzie and Szerb (1968) with permission of the authors and J. Phamacol uncl Exptl. Therap.

ACTION OF PSYCHOMOTOR STIMULANT DRUGS

373

induced a release of newly synthcsizcd dopamine (Besson et al., 1969) . The effect of amphetaiiiinc at the level of the striatal neurons is probably indirect by rclease of dopamine which in turn acts as an inhibitory transmitter. It has heen found that firing of neurons of the caudate nucleus and putamen cither spontaneously or following stimulation of the mcdial thalnmus is inhibited by microelectrophoretically applied dopaminc (Herz and Zieglggnsberger, 1966; Bloom et al., 1965). From experiments with a-methyldopa on histochemical fluorescence in rat brain, evidence is provided which suggests that dexamphetaminc cmi also release a-methyldopamine present in dopamine nervc’ terminals after administration of amethyldopa ( Carlsson et NL., 1967). I t is questionable whetlicsr thc dcxainphctamine effects in man at therapeutic doses 1ia1.e much in common with a dopainine release in the basal ganglia. It is more likely that chronic toxic effects, especially amphetanrinc-induccd psychosis, is due to effects in the dopamine containing titiclei. The relationship between amphetamine-induced psychosc~s and certain forins of schizophrenic psychosis in the light of thc prcwnt knowledge of amphetamine action on dopamine neurons m a y suggest the importance of dopamine in the pathogenesis of schizophrenia (see also the review by Munkvad et al., 1968).

Aceto, M. D., Harris, L. S., atid L e s h ~ r ,G . Y. (1966). Pharmacologist 8, 222. Adler, M. W. (1961). J. Pliurmuco/. Erptl. Therap. 134, 214-221. Alles, G. A. ( 1927). J. Phannucd. Erptl. Therup. 32, 121-133. Alles, G. A. ( 193.3). J . Phannoco/. E q d . Therap. 47, 339-354. Alles, G. A. ( 1939). Unio. Coliforriiu P i t h l . Pharmacol. 1, 129-150. Alles, G . A., and Feigen, G. A. ( 1941 ) . Ant. J. Physiol. 133, 194-195. Alles, C. A,, Fairchild, hl. D., and Jensen, h l . (1961). J. Med. Pharm. Chem. 3, 323-351. Amsler, C. ( 1923). Arch. Exptl. Putliol. Phurmcikol. 97, 1-14. AndBn, N. E. ( 1970). Proc. Ziiterii. Symp. Amphetamines Related Compounds, Milan, 1969. Haven Press, N ( w York ( i n press). And&, N. E., Carlsson, A,, Dahlstriitn, A,, Fuxe, K., Hillarp, N. A,, and Larsson, K. ( 1964). Life Sci. 3, S25-530. And&, N. E., Corrodi, H., Dahlsti-iiin, A., Fuxe, K., and I-Iiikfelt, T. ( 1966). Life Sci. 5, 561-568. Aprison, M. H., and Ferster, C. B. (1961). J. Neurochem. 16, 350-357. Amfred, T., and Ranclrup, A. ( 1968). Actu Phnrmcicol. Toxicol. 26, 384-394.

374

JACQUES M. VAN ROSSUM

Asatoor, A. M., Galman, B. R., Johnson, J. R., and Milne, M. D. (1965). &it. J. Pharmucol. 24, 293300. Askevold, F. (1959). A d a Psychiat. Neurol. Scand. 34, 145-164. Axelrod, J. (1954). J. Phurmucol. Exptl. T h r a p . 110, 315-326. Axelrod, J. (1955). J. Biol. Chem. 214, 753-763. Bahnsen, P., Jacobsen, E., and Thesleff, H. (1938). Acda Med. Scand. 97, 89131. Baird, J. R. C., and Lewis, J. J. (1964). Biochem. Phumcol. 13, 1475-1482. Baumer, L. (1966). Amneimittel-Forsch. 16, 249. Beauvallet, M. (1968 ), Acta Pharmucol. Toxicol. 21, 15-39. Beauvdet, M., Fugazza, J., and Solier, M. (1963). Compt. Rend. 257, 32513253. Beckett, A. H. (1966). Dansk Tidsskr. Farm. 40, 197-223. Beckett, A. H., and Brooks, L. G. (1967). J. Pharm. Phurma~ol.19, Suppl., 41S-52S. Beckett, A. H., and Rowland, M. (1965a). J. P h r m . P h m c o l . 17, 59-60. Beckett, A. H., and Rowland, M. (1965b). J. Pharm. Phamacol. 17, 628-639. Beckett, A. H., and Rowland, M. ( 1 9 6 5 ~ )J. . Phurm. Phumcol. 17, Suppl., 109s-114s. Beckett, A. H., and Triggs, E. J. (1967). J. Pharm. Phumcol. 19, Suppl, 31S41S. Beckett, A. H., Tucker, G. T., and Moffat, A. C. (1967). J. Pharm. Pharmacol. 19, 273-294. Beckett, A. H., Boyes, R. N., and Triggs, E. J. ( 1968a). J. Pharm. Phurmacol. 20, 92-97. Beckett, A. H., Boyes, R. N., and Tucker, G. T. (196813). J. P h r m . PhuTmaco~. 20, 277-282. Beckett, A. H., Salmon, J. A., and Mitchard, M. (1969). J. Pha~m.Phamacol. 21, 251-258. Beitter, A, (1901). Arch. Pharm. 239, 17-34. Bertler, A. (1961). Actu Physiol. Scand. 51, 75-83. Besson, M. J., Cheramy, A., Feltz, P., and Glowinski, J. (1969). PTOC.Notl. A c d Sci. U. S . 02, 741-748. Bett, W. R., Howells, L. H., and MacDonald, A. D. (1955). “Amphetamine in Clinical Medicine-Action and Uses.’’ Livingstone, London. Beyer, K. H., and Skinner, J. T. (1940). J. Phumcol. Exptl. Therap. 68, 419-432. Bignami, G. (1966). Psychopharmacologia 10, 44-58. Bhschko, H. (1952). Phurmucol. Reu. 4, 415-458. Blaschko, H., and Chrusciel, T. L. ( 1960). J . Phydol. (London) 151,272-284. Bloom, F. E., Costa, E., and Salmoiraghi, G. C. (1965). J. P ~ ~ T V U Z C Ezptl. O~. Therap. 150, 244-252. Bogdanove, L. H., and Nir, I. (1965). Phamcolog* 7, 174. Bonhoff, C., and Lewrenz, H. (1954). “Wber Weckamine (Pervitin und Benzedrin) .” Springer, Berlin. Bradley, P. B., and Elkes, J. (1957). Brain 80, 77-117. Bradley, P. B., and Key, B. J. (1958). Ekctroencephalog. Clin. Neurophyssiol. 10,97-111.

ACTION OF PSYCHOMOTOR STIMULANT DRUGS

375

Brink, J. J., and Stein, D. G. (1968). Biochem. Pharmucol. 17, 2365-2366. Bruce, R. B., and Maynard, W. R. (1968). J. Phurm. Sci. 57, 1173-1176. Burn, J. H., and Rand, M. J. (1958). J. Physiol. (London) 144, 314-336. Calm, R. S., Ingold, C. K., and Prelog, V. (1956). Elcperientia 12, 81-94. Campbell, D. B. (1969). J. Pharm. Pharmacol. 21, 129-131. Carlsson, A., Lindqvist, M., and Magnusson, T. (1957). Nature 180, 1200. Carlsson, A., Lindqvist, M., Dahlstrdm, A., Fuxe, K., and Masuoka, D. (1965). J. Pharm. Pharmucol. 17, 521-524. Carlsson, A., Lindqvist, M., Fuxe, K., and Hamberger, B. (1966a). J. Pharm. Pharmacol. 18, 128-131. Carlsson, A., Fuxe, K., Hamberger, B., and Lindqvist, M. (196613). Acta Physiol. Scand. 67, 481497. Carlsson, A., Fuxe, K., and Hokfelt, T. (1967). European J. Pharmacol. 2, 196-201. Carlton, P. L. ( 1961). Psychopharmacologia 2, 364-376. Carlton, P. L. (1963). Nature 200, 586-587. Carlton, P. L., and Didamo, P. (1961). J. Phurmacol. Exptl. Therap. 132, 91-96. Carlton, P. L., and Furgiuele, A. R. (1965). Life Sci. 4, 1099-1106. Chance, M. R. A. (1947). J. Pliarmacol. Exptl. Therap. 89, 289-296. Chance, M. R. A. (1948). Behavior 1, 64-69. Chance, M. R. A., and Silverman, A. P. (1964). Ciha Found. Symp., Animal Behavior Drug Action pp. 65-79. Chen, K. K., and Schmidt, C. F. (1924). J. Pharmucol. Exptl. Therap. 24, 339-357. Chen, K. K., Wu, Ch.-K., and Hendriksen, E. (1929). J. Pharmacol. Expti. Therap. 36, 363400. Clark, F. G., and Steele, B. J. (1966). PPrJchopharmucologia 9, 157-169. Clarke, F. H. (1962). J. Org. Chem. 27, 3251-3253. Connell, P. H. ( 1958). “Amphetamine Psychosis.” Chapman & Hall, London. Consolo, S., Dolfini, E., Garattini, S., and Valzelli, L. (1967). J. Pharm. Pharmacol. 19, 253-256. Cook, L., and Kelleher, R. T. (1962). Ann. N . Y. Acad Sci. 96, 315-335. Cools, A. R., and van Rossum, J. M. (1970). Brain Res. (to be published). Corrodi, H., and Hardegger, E. (1955). Helu. Chim. Acta 38, 2038-2043. Costa, E., and Neff, N. H. (1966). In “Biochemistry and Pharmacology of the Basal Ganglia” ( E. Costa, L. J. C&, and M. D. Yahr, eds.), pp. 141156. Raven Press, New York. Dagirmanjian, R., Laverty, R., Mantegazzini, P., Sharman, D. F., and Vogt, M. (1963). J. Neurochem. 10, 177-182. Davis, J. L., (1965). J. Exptl. Anal. Behauior. 8, 325-327. Dews, P. B., and Morse, W. H. ( 1961). Ann. Rev. Pharmacol. 1, 145-174. Dews, P. B. (1953). Brit. J. Pharniucol. 8, 4649. Dingell, J. V., Langlois, M., Robinson Norvich, M., and Sulser, F. (1967). Life Sci 6, 1155-1162. Dring, L. G., Smith, R. L., and Williams, R. T. (1966). J. Pharm. Pharmucol. 18, 402405. Dvornik, D., and Schilling, G. (1965). J. Med. Chem. 8, 466469.

376

JACQUES M. VAN ROSSUM

Editorial. (1968). D m g Therap. Bull. 6, 33-34. Ehrich, W., and Krumbhaar, E. (1937). Ann. Internal Med. [N.S.] 10, 18741882. Ellison, T., Gulzail, L., and van Loon, E. Y. (1965). J. Pharmacol. Exptl. Therap. 152, 383-387. Emele, J., Shanaman, J., and Warren, M. R. (1961). Federation Proc. 20, 328 Ernst, A. M. (1965). Psychopharmucologia 7, 391399. Ernst, A. M. (1967). Psychopharmacologiu 10, 316-323. Ernst, A. M., and Smelik, P. G. (1966). Experientia 22, 837-838. Espelin, D. E., and Done, A. K. (1968). New Engl. J. Med. 278, 1361-1366. Everett, G. M. ( 1961). Proc. 2nd Intern. Meeting CINP, Bade, 1960 Vol. 2, pp. 479484. Elsevier, Amsterdam. Everett, G. M., and Toman, J. E. P. (1959). “Biological Psychiatry,” pp. 7581. Grune & Stratton, New York. Everett, G. M., and Wiegand, R. G. (1962). Proc. 1st Intern. Pharmacol. Meeting, Stockholm, 1961 Vol. 8 pp. 85-93. Pergamon Press, Oxford. Everett, G. M., Toman, J. E. P., and Smith, A. H., Jr. (1957). Federation Proc. 16, 295. Fairchild, M. D., and Alles, G. A. (1967). J. Pharmacol. Exptl. Therap. 158, 135-139. Ferster, C. B., and Skinner, B. F. (1957). “Schedules of Reinforcement.” Appleton, New York. Fodor, G. (1957). Tetrahedron 1, 86-102. Fog, R. L. (1967). Acta Pharmacol. Toxicol. 25, Suppl. 4, 59. Fog, R. L. (19G9). Psychopharmacologia 14, 299-305. Fog, R. L., Randrup, A., and Pakkenberg, H. (1967). Psychopharmacologia 11, 179-183. Fog, R. L., Randrup, A., and Pakkenberg, H. ( 1968). Psychopharmacologia 12, 428-432. Frey, P. W., and Polidora, V. J. (1967). Science 155, 1281-1282. Fuxe, K. (1965). Acta Physiol. Scand. 64, Suppl. 247, pp. 37-85. Fuxe, K., and Hanson, L. C. F. (1967). Psychopharmacologia 11, 439447. Geller, I., and Seifter, J. (1960). Psychopharmucologia 1, 482-492. Gershon, S., Hekimian, L. J., Floyd, A., and Hollister, L. E. (1967). Psychopharmacologia 11, 189-194. Glowinski, J., and Baldessarini, R. J. (1966). Pharmucol. Rev. 18, 1201-1238. Goldstein, L., and Beck, R. A. (1965). Intern. Rev. Neurobiol. 8, 265-312. Goldstein, M., and Nakajima, K. (1966). Life Sci. 5, 175-179. Gottlieb, R. (1923). Arch. Exptl. Pathol. Phnrmakol. 97, 113-147. GrifEth, J., Cavanaugh, J., and Oates, J. (1968). Pharmacologist 10, 180. Gunne, L. M., and Galland, L. (1967). Biochem. Pharmacol. IS, 1374-1377. Gutierrez-Noriega, C., and Zapata-Ortiz, V. ( 1947). Estudios Coca Cocaina, Lima, Peru, 1947 Min. Educ. Publ., p. 140. Gylys, J. A., Hart, J. J. D., and Warren, M. R. (1962). J. Pharmacol. Exptl. Therap. 137, 365-373. Haase, J., and Tan, U. (1965). Arch. Exptl. Pathol. Phurmakol. 252, 20-31. Haley, T. J. (1947). J. Am. Pharm. As.roc., Sci. Ed. 36, 161-169.

ACTION OF PSYCHOMOTOR STIMULANT DRUGS

377

Hanson, L. C . F. (1965). Psycliopharlnucologia 8, 100-110. Hanson, L. C. F. (1967a). Psyc/rophnrtnucologia 11, 8-17. Hanson, L. C. F. ( 196711). Psychophnrniucologia 10, 289-297. Harnack, E. ( 1874). Arch. Exptl. Puthol. Pharmakol. 2, 255306. Harris, L. S., Clarke, R. L., ;und Deml)inski, J. R. (1963). Arch. Intern. Pharmacodyn. 146, 392405. Harking, W., and Andrako, J. ( 1961). J. Phurm. Sci. 50, 805818. Hartwich, C. ( 1911 ) , “Die nienschlichen Gennssmittel. Tauchnitz, Leipzig. Hauschild, F. ( 1939). Arch. Exptl. Pathol. Pharmakol. 191, 465-481. Havlicek, V. ( 1967 ). Interir. J. Nerrropl~urnaacol. 6, 83-88. Hawks, D., Mitcheson, M., Ogborne, A., and Edwards, G. (1969). Brit. Med. J. 11, 715721. Herman, M., and Nagler, S. €I. ( 1 9 5 4 ) . J . Nervous Mental Diseuse 120, 268272. Herman, Z. S. ( 1967). PsychophnrnLncologia 11, 136-142. Herridge, C. F., and a’Brook, h l . F. ( 1968). Brit. Med. J. 11, 160. Herz, A., and Fuster, J. (1964). Arch. E x l d Pathol. Plzarmukol. 249, 146-161. Herz, A., and Zieglgiinsberger, W. (1966). Experientia 22, 839-840. Hiebel, C., Bonvallet, h4., IIrivC, l’., and Dell, P. ( 1957). Seniaine Hop. Pari.c 30, 1880. Hohn, R., and Lasagna, L. ( 1960). Psychopharmacologia 1, 210-220. Holtz, P., Balzer, H., Westermarin, E., and Wezler, E. (1957). Arch. Exptl. Puthol. Pharmak. 231, 333-348. Hornykiewicz, 0. ( 1966). Pl~urt~iucol. Rev. 18, 925-964. Horovitz, Z. P., Chow, M. I., ant1 Carlton, 1’. L. ( 1962a). Psychopharmacolt,gin 3, 449454. Horovitz, Z. l’., Chow, h4. I., and Carlton, 1’. L. ( 1962b). Psychophnrmacologiu 3, 455-462. Irwin, S., Slalmk, M., and l h o i n a s , G. ( 1958). J. Pharmacol. Exptl. Therap. 123, 206-211. Isbell, H., Wolbach, A. R . , Wikker, A , , and hliner, E. J. (1961). Psychophurmacologiu 2, 147-150. Iversen, L. L. (1965). J. PIiurvi. Phurmcicol. 17, 62-64. Ivy, A. C., and Krasno, L. H . ( 1841 ). W a r Med., Chicago 1, 15-42. Ivy, A. C., and Coetzl, F. R . ( 194:3). W a r Med., Chicago 3, 60-77. Jacobsen, E. ( 1939 ). Actu A l e t l . Scurid. 100, 188-202. Jacobsen, E., and Wollsteiii, A. (1939). Actu Med. Scand. 100, 159-187. Janssen, P. A. J. ( 1961 ). I~,z)ic.iniittcZ-Forsck. 11, 819-824 and 932-938. Janssen, P. A. J., Kieinegecrs, (:. J. E., and Jagenean, A . (1960). ArzneimittelForsch. 10, 1003-1005. Janssen, P. A. J., Niernegwrs, C. J. E., and Schellekens, K. H. J. (1965). Arzneimittel-Forsch. 15, 104-117. Janssen, P. A . J., Niemegeers,